Logo: to the web site of Uppsala University

uu.sePublications from Uppsala University
Change search
Refine search result
12345 1 - 50 of 225
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1.
    Abdalmoaty, Mohamed
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Systems and Control. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Automatic control.
    Coimbatore Anand, Sribalaji
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Teixeira, André
    Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Systems and Control. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Automatic control.
    Privacy and Security in Network Controlled Systems via Dynamic Masking2023In: IFAC-PapersOnLine, E-ISSN 2405-8963, Vol. 56, no 2, p. 991-996Article in journal (Refereed)
    Abstract [en]

    In this paper, we propose a new architecture to enhance the privacy and security of networked control systems against malicious adversaries. We consider an adversary which first learns the system using system identification techniques (privacy), and then performs a data injection attack (security). In particular, we consider an adversary conducting zero-dynamics attacks (ZDA) which maximizes the performance cost of the system whilst staying undetected. Using the proposed architecture, we show that it is possible to (i) introduce significant bias in the system estimates obtained by the adversary: thus providing privacy, and (ii) efficiently detect attacks when the adversary performs a ZDA using the identified system: thus providing security. Through numerical simulations, we illustrate the efficacy of the proposed architecture

  • 2.
    Absalyamov, Artur
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Gladh, Jimmy
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Navigation Robot2020Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
    Abstract [en]

    Journeying into the information era the need for new technologies used forsending vast amounts of data eciently has risen, as has the possibilities and needfor dierent types of AI-controlled robots. In this project a RC-car was modiedand equipped with a Raspberry PI and laser radar to let it automatically navigatearound a room using UWB transmitters and receivers. A framework for roboticapplications called ROS, Robot Operating System, was used with a large numberof open source packages to ll dierent functions. Custom scripts was created totie everything together, allowing all dierent components in the system to work inunison.

    Download full text (pdf)
    fulltext
  • 3.
    Aitman, Victor
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Conductive interferences from multiple EC-motor installation: To measure and mitigate harmonics2022Independent thesis Basic level (professional degree), 10 credits / 15 HE creditsStudent thesis
    Abstract [en]

    Products produced by a Swedish company are requested to be investigated regarding their harmonic and inter-harmonic currents injected into the public supply system to comply with the Swedish standard SS-EN IEC 61000-3-2, Electromagnetic compatibility (EMC).   

    On behalf of Systemair Sweden AB, this bachelor thesis aims firstly to investigate if their product, PAFEC 4225 WH (Air Curtain), complies with standard SS-EN IEC 61000-3-2, and if not, what measures should be taken; and secondly to develop a low-cost instrument for the measurement of harmonics and inter-harmonics according to standard IEC 61000-4-7 related to the requirements on equipment used in standard SS-EN IEC 61000-3-2.

    To this purpose, the standards have been looked into thoroughly, the preconditions for measurements have been studied in detail, and external meetings with a consultant at Delta Development Technology AB have been performed. Measurements of the harmonic spectrum generated by PAFEC 4225 WH have been performed first at Systemair Sweden AB’s Technical center in Skinnskatteberg, and later at Delta Development in Västerås. After this, a low-cost instrument was developed, including hardware and software design and implementation. The hardware implementation consists of a circuit board designed using EasyEDA (an online PCB Design Tool), a NI myDAQ (a data acquisition device made by National Instruments), and an enclosure designed with Solidworks and made with a 3D-printer. The software implementation was conducted using LabVIEW – a graphical programming language. 

    A few measurements were performed using instruments complying IEC 61000-4-7 at Delta Development, and later with the low-cost instrument. Different line chokes were measured. The results showed that a 15 mH line choke connected in series with each motor would make the PAFEC 4225 WH comply with SS-EN IEC 61000-3-2. The results from the low-cost instrument did not match Delta Developments results regarding harmonic and inter-harmonic content. The difference could be caused by unfinished algorithm, different measurements conditions, and missing anti-aliasing-filter.         

    For the future work it is recommended that Systemair Sweden AB can either develop the low-cost instrument or buy an existing instrument that complies with IEC 61000-4-7, to enable to do measurements that comply with SS-EN IEC 61000-3-2. One does also need to investigate the grid during low activity or consider buying a signal generator for the purpose of fulfilling the preconditions to enable measurements. It is also recommended that further measurements are performed with the proposed line choke installed to check for any change in performance of the product. 

    Download full text (pdf)
    fulltext
  • 4.
    Al-Barghouthi, Mohammad
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Control system for a radio car2020Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
    Abstract [en]

    In this project, an Arduino Uno car was built and tuned with a PID controller to avoid obstacles in the reverse stage. The distance to the obstacles was shown on an LCD display. The result was as expected where the steady state error had an average around 1-2 cm. The rise rime was around 2.4 seconds which is excellent considering the motor limitations. There was no overshoot indicating that no collision occurred, which fulfills the whole point with the modified parking sensor.  

  • 5.
    Al-Barghouthi, Mohammad
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Extended and Unscented Kalman Filtering for Estimating Friction and Clamping Force in Threaded Fasteners2021Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Threaded fasteners tend to break and loosen when exposed to cyclic loads or potent temperature variations. Additionally, if the joint is held tightly to the structure, distortion will occur under thermal expansion issues. These complications can be prevented by identifying and regulating the clamping force to an appropriate degree – adapted to the properties of the joint. Torque-controlled tightening is a way of monitoring the clamping force, but it assumes constant friction and therefore has low accuracy, with an error of around 17% - 43%.This thesis investigates if the friction and clamping force can be estimated using the Extended and Unscented Kalman filters to increase the precision of the torque-controlled methodology. Before the investigation, data were collected for two widely used tightening strategies. The first tightening strategy is called Continuous Drive, where the angular velocity is kept at a constant speed while torque is increased. The second strategy is TurboTight, where the angular velocity starts at a very high speed and decreases with increased torque. The collected data were noisy and had to be filtered. A hybrid between a Butterworth lowpass filter and a Sliding Window was developed and exploited for noise cancellation.The investigations revealed that it was possible to use both the Extended and Unscented Kalman filers to estimate friction and clamping force in threaded fasteners. In Continuous Drive tightening, both the EKF and UKF performed well - with an averagequality factor of 81.87% and 88.38%, and with an average error (at max torque) of 3.54% and 4.09%, respectively. However, the TurboTight strategy was much more complex and had a higher order of statistical moments to account for. Thus, the UKF outperformed the EKF with an average quality factor of 93.02% relative to 24.49%, and with an average error (at max torque) of 3.50% compared to 4.19% 

    Download full text (pdf)
    fulltext
  • 6.
    Andersson, Tim
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Automatic Crack Detection in Sand Molds Using Image Processing and Convolutional Neural Networks2022Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Sand casting is used to manufacture large metal workpieces. The processing is executed by pouring molten metal into a sand mold. During the process, the mold is subjected to mechanical and thermal stress. It is of economic interest to inspect the molds for defects that can affect casting results, in the worst case leading to discarded products.

    This thesis investigates and proposes an automated solution for inspecting surface cracks in sand molds. A hybrid solution using image processing and convolutional neural networks has been developed. The first is to find and implement a crack detection method that can perform equally well or better than a human. The second objective is to investigate the amount of training data needed. Twenty-one machine learning models have been trained to evaluate the impact training data size along with transfer learning, fine-tuning, data augmentation, and image processing have on performance.

    As a result, it was found that the image processing part of the method is not effective in finding cracks in its current form. However, the convolutional neural network still achieves good performance. The method has been trained and tested on sand mold core images captured with a test workbench along with images of concrete walls and pavement acquired from the SDNET2018 data set. Sand mold images achieve 82% accuracy and 79% recall when training on 90 images while testing on 28 images separate from training. A maximal performance of 97.9% accuracy and 99.7% recall is achieved when training on 5400 SDNET2018 images and then testing on 608 images. When training on 100 SDNET2018 images and tested on the same 608 images, a performance of 86.0% and 96.7% recall is achieved.

    It is concluded that the proposed solution is feasible. Transfer learning and data augmentation are essential techniques to achieve good performance if a small amount of data is available, while fine-tuning may give a slight performance boost. Further work should be performed considering the impact of curved geometry on performance. Investigating alternative structures of the convolutional neural network and testing alternative hyperparameters may improve generalization performance. The image processing performance may be improved if the manufacturing process is more precisely defined, as parameters can be more optimally tuned.

    Download full text (pdf)
    Automatic Crack Detection in Sand Molds Using Image Processing and Convolutional Neural Networks
  • 7.
    Antonsson, Tobias
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Automation of printed circuit board testing on a bed-of-nails testbench2022Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Syntronic Research & Development AB both designs and performs tests of PCBs of differentkinds. At lower volumes, big parts of the test process are manual. This thesis examines oneway to automate the loading and unloading process with a cheap and simple solution so that itcan be profitable even at low volumes. If all parts of the process are automated manpower canbe freed, hence the cost of testing can be lowered, and tedious monotonous work can beavoided. Human error can also be removed from the equation, which could result in lessstochastic errors.

    The approach to automate the loading and unloading of test objects in this thesis is to design atwo-axis linear robot. This way the PCBs can be picked from incoming plates, placed in thetestbench, and then be placed in outgoing plates, as long as they are all in a straight line.

    To find weakness in the design a prototype was constructed on which tests were performed.These tests showed some areas on which improvements are needed before it can beconsidered a finished product. These improvements are discussed.

    The tests also showed that this approach can be made profitable, with some limitations. The cost savings are greatly dependable on how the other process are automated.

    There are both limitations on the setup of the test fixture used and, perhaps mainly, on how theother process are automated. This is also discussed.

    One challenge yet to overcome is how to make it easy to adapt and implement for a specifictest. When and if a robot based on this concept is implemented the setup time will by far be thelargest contributor to the cost.

    Download full text (pdf)
    fulltext
  • 8.
    Arghavani, Abbas
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Ahlén, Anders
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Teixeira, André
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    A Game-theoretic Approach to Covert Communications in the Presence of Multiple Colluding Wardens2021In: 2021 IEEE WIRELESS COMMUNICATIONS AND NETWORKING CONFERENCE (WCNC), IEEE Institute of Electrical and Electronics Engineers (IEEE), 2021Conference paper (Refereed)
    Abstract [en]

    In this paper, we address the problem of covert communication under the presence of multiple wardens with a finite blocklength. The system consists of Alice, who aims to covertly transmit to Bob with the help of a jammer. The system also consists of a Fusion Center (FC), which combines all the wardens' information and decides on the presence or absence of Alice. Both Alice and jammer vary their signal power randomly to confuse the FC. In contrast, the FC randomly changes its threshold to confuse Alice. The main focus of the paper is to study the impact of employing multiple wardens on the trade-off between the probability of error at the FC and the outage probability at Bob. Hence, we formulate the probability of error and the outage probability under the assumption that the channels from Alice and jammer to Bob are subject to Rayleigh fading, while we assume that the channels from Alice and jammer to the wardens are not subject to fading. Then, we utilize a two-player zero-sum game approach to model the interaction between joint Alice and jammer as one player and the FC as the second player. We derive the pay-off function that can be efficiently computed using linear programming to find the optimal distributions of transmitting and jamming powers as well as thresholds used by the FC. The benefit of using a cooperative jammer is shown by means of analytical results and numerical simulations to neutralize the advantage of using multiple wardens at the FC.

  • 9.
    Arghavani, Abbas
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Ahlén, Anders
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Covert Outage Minimization in the Presence of Multiple Wardens2023In: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 71, p. 686-700Article in journal (Refereed)
    Abstract [en]

    The idea of covert communication is to conceal the presence of a transmission from an illegitimate receiver, known as a warden. This paper tackles the problem of finite blocklength covert communication in the presence of multiple colluding wardens. The system consists of Alice, who aims to covertly transmit to Bob with the help of a cooperative jammer (henceforth known as Jammer), and a Fusion Center (FC) in charge of combining the wardens' information and deciding on the presence of Alice's transmission accordingly. In our proposed approach, we utilize a two-player zero-sum game to model the interaction between Alice and Jammer jointly as one player and FC as the second player. In this game, Alice and Jammer cooperatively randomize over a range of transmitting and jamming powers to confuse FC. In contrast, FC randomly changes the detection threshold to confuse Alice. The main focus of the paper is to study the impact of employing multiple wardens on the trade-off between the probability of error at FC and the outage probability at Bob. We derive a pay-off function that can be efficiently computed using linear programming to find the optimal distributions of transmitting and jamming powers as well as thresholds used by FC. The benefit of using a cooperative jammer in neutralizing the advantage of employing multiple wardens is shown by analytical results and numerical simulations.

  • 10.
    Arghavani, Mahdi
    et al.
    Univ Otago, Dept Comp Sci, Dunedin 9016, New Zealand.
    Zhang, Haibo
    Univ Otago, Dept Comp Sci, Dunedin 9016, New Zealand.
    Eyers, David
    Univ Otago, Dept Comp Sci, Dunedin 9016, New Zealand.
    Arghavani, Abbas
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    StopEG: Detecting when to stop exponential growth in TCP slow-start2020In: Proceedings of the 2020 IEEE 45th Conference on Local Computer Networks (LCN 2020) / [ed] Tan, H.P., Khoukhi, L. & Oteafy, S., 2020, p. 77-87Conference paper (Refereed)
    Abstract [en]

    TCP slow-start grows the congestion window exponentially, aims to quickly probe the throughput of the network path. Stopping this growth at the wrong time can affect the overall network performance. In this paper, we introduce StopEG, an efficient mechanism to accurately and quickly detect when to stop this exponential growth. StopEG reacts to the changes on congestion window size rather than traditional congestion signals such as packet loss. We show that theoretically the number of inflight packets in the forward path is no more than 56.8% of all the inflight packets when the bottleneck link is unsaturated, and use this value as the threshold to stop the exponential growth. StopEG is evaluated through simulations in ns-3 by incorporating it into Google's BBR congestion control algorithm. Simulation results demonstrate its effectiveness in BBR, with a reduction of ≈68% in the length of the bottleneck queue when new connections are initiated.

  • 11.
    Aso Abbas, Ismail
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Isaksson Sandberg, Mats
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Low-Cost Electrical Resistance Tomography2023Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    ​​Electrical resistance tomography (ERT) and electrical impedance tomography (EIT) are imaging techniques reconstructing the internal conductivity distribution image of an object based on voltage measurements at the periphery of the object with a given applied current. ERT uses a direct current (DC), while EIT uses an alternating current (AC). However, for low frequencies both ERT and EIT have the same governing equation, which is often referred to as a non-linear and ill-posed inverse problem. Both methods have diverse applications in biology, biomedicine, and industry. ​This master’s degree project aims to create a low-cost imaging system for the ERT, which is the main focus, as well as for the EIT. The project includes three main components: 1) Simulations and reconstructions using EIDORS (Electrical Impedance Tomography and Diffuse Optical Tomography Reconstruction Software), 2) Developing an experimental workbench (a measurement system), and 3) developing a machine learning model for the ERT. ​EIDORS was used to simulate and reconstruct ERT and EIT images. It was also used to generate training data for the machine learning model to be developed. ​The measurement system includes a circular water tank with electrodes, power supplies, and measurement units. Tanks with 8 and 16 electrodes were designed using 3D printers. Initially, aluminium electrodes provided inconsistent measurements due to magnetization and electrolysis, later replaced by graphite electrodes, offering better but not yet accurate enough results. ​After implementing reconstruction algorithms in EIDORS, a machine learning model was developed for ERT. It involved: 1) generating a training set, containing over 5000 simulated data points, 2) preprocessing the generated data set which included PCA dimensionality reduction, 3) and lastly a linear regression model developed. The model struggled with small object detection and occasional inconclusive results, likely due to limited training dataset diversity. Additionally, images of two cases were reconstructed using EIT and comparing it to ERT it can be concluded that EIT performs better than ERT. ​

    Download full text (pdf)
    ERT_Thesis
  • 12.
    Badran, Rasha
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    LoRa Based Moisture Sensing System2023Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Water is an important parameter for crop growth, and the information about the moisture content in soil at different depths is very useful for farmers to determine the best time to water the soil and to irrigate farmland so as to maximize their yield.

    This thesis project aims to develop a prototype of a multi-depth moisture sensor probe that is part of a large sensing system used in agriculture. The sensor probe has three sets and is required to last for 6-12 months of usage and to be reproduced at a low cost.

    The sensor probe consists of three sensor boards, on each of which has two different capacitive based sensors and one analog temperature sensor. The three boards are placed approximately 20 cm from each other in the probe. During this project, the two capacitive based sensors were developed, one with arc-shaped plates operating at a frequency less than 1 MHz, and one with electrodes in the form of annular rings operating at a high frequency, approximately 100 MHz. The moisture content in the soil is calculated based on the measurement of the frequency, which depends on the dielectric constant of the soil.

    For the implementation of the sensor probe, three printed circuit boards (PCBs) for the sensor boards were designed using Altium Designer and then ordered; an STM32 Nucleo board with low power microcontroller was used and the software was implemented in STM32CubeIDE. The lifetime of the sensor probe was calculated for different duty-cycles. With a duty-cycle of 15 minutes, where the sensor probe is active for 1 minute and in sleep mode for 14 minutes, the lifetime of the sensor probe would only be 16 days. With a duty-cycle of 120 minutes instead, with the sensor probe being active for 1 minute, the lifetime is increased to 130 days (less than4,5 months).

    Due to challenges with the high frequency capacitive sensor, the multi-depth sensor probe does not fully work, and thus cannot be tested with a large testbed. Further work needs to be conducted on the high frequency capacitive sensor and the communication with the gateway.

    Download full text (pdf)
    fulltext
  • 13. Barros, M
    et al.
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Feed-forward and feedback control in astrocytes for Ca2+-based molecular communications nanonetworks2020In: IEEE/ACM Transactions on Computational Biology & Bioinformatics, ISSN 1545-5963, E-ISSN 1557-9964, Vol. 17, no 4, p. 1174-1186Article in journal (Refereed)
  • 14.
    Batti, Parwand
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Implementering av datadriven digital tvilling för förbättrande underhåll på ett torkcylindersystem2023Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    A digital twin is a digital representation of a real-world physical object (a physical twin), and useful for product lifecycle management and maintenance. The aim of this project is to implement a data-driven digital twin for a system consisting of an electric motor, a gearbox and a drying cylinder, in order to improve the maintenance plan and reduce maintenance costs for a drying cylinder system. 

    To achieve this, the data on the electric motor and the gearbox was first measured with ABB smart sensors and thereafter stored on ABBHubb and Billerudhubb via wireless communication. The data on the drying cylinder was measured with the measuring frames and, via wireless communication as well, stored on BillerudHubb. Then all the data were further transferred to the ThingSpeak cloud by means of an application developed in C# programming in the .Net developer platform. 

    Since the ThingSpeak is an IoT analytics platform service from MathWorks (the makers of MATLAB and Simulink®), it can directly execute the MATLAB codes developed for data visualization and analysis. The data-driven digital twin for the drying cylinder system is able to be realized on the ThingSpeak. The data shown on the digital twin include the data on the electric motor (speed, bearing condition, overall vibration, skin temperature and output power),  the gearbox (overall vibration, overall acceleration and skin temperature)  and the cylinder (speed, load and elongation/current).

    However, ThingSpeak has limitations related to reading and writing data, which make transferring data difficult. Other IoT platforms such as AWS IoT TwinMaker and Azure Digital Twins can be better alternatives for realizing a digital twin.

  • 15.
    Beas Petersson, Patric
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Characterisation and Modeling of RF environment2022Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Devices that transmit and receive electromagnetic signals are today central to the way people communicate and acquire information from the surroundings. WiDAR, Wideband Digital Array Receiver, is the name of a new system of digital antenna array receivers, which has recently been developed at Saab in Järfälla and is the focus of this thesis. Digital antenna arrays have the benefit over conventional antennas of having their lobes steered electrically, allowing for beamforming, both in reception and transmission. The goal of the thesis has been to characterise certain signal technologies operating in the spectrum of WiDAR, and during the process learn about the limitations of digital antenna array receivers and the system. After studying the system and telecommunication technologies present in the wideband, several measurements were conducted using WiDAR in the field to gather raw data for processing in Matlab. With WiDAR having numerous channels, as well as high sampling rate, large amounts of data was received, leading to difficulties in the processing thereafter. Concluding that three types of signal technologies are certain to be found in WiDAR's spectrum, UMTS/3G, LTE/4G, and DVB-T, their respective narrowbands were studied further through the production of spectrograms of the signal data. Within each band, probability distributions were fit to the histograms of the data. Each of the signal technologies were then characterised by their respective fit to the probability distributions. This resulted in a way of identifying unknown signals from new measurement data from WiDAR. While this method could prove useful as a first step in characterisation, weaknesses such as its lack of depth in the narrowbands are discussed. For further work and the future of the system, it is suggested to e.g. explore the concepts of the multipath problem, or TDM/TDMA in the data. Ultimately, the characterisation of the found signal technologies was moderately successful, however with a sizeable list of limitations and area of improvements.

    Download full text (pdf)
    MasterThesis_PatricBeasPetersson
  • 16.
    Bergström, Joakim
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Transfer Learning on Ultrasound Spectrograms of Weld Joints for Predictive Maintenance2020Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    A big hurdle for many companies to start using machine learning is that trending techniques need a huge amount of structured data. One potential way to reduce the need for data is taking advantage of previous knowledge from a related task. This is so called transfer learning. A basic description of it would be when you take a model trained on existing data and reuse that for another problem. The purpose of this master thesis is to investigate if transfer learning can reduce the need for data when faced with a new machine learning task which is, in particular, to use transfer learning on ultrasound spectrograms of weld joints for predictive maintenance. The base for transfer learning is VGGish, a convolutional neural network model trained on audio samples collected from YouTube videos. The pre-trained weights are kept, and the prediction layer is replaced with a new prediction layer consisting of two neurons. The whole model is re-trained on the ultrasound spectrograms. The dataset is restricted to a minimum of ten and a maximum of 100 training samples. The results are evaluated and compared to a regular convolutional neural network trained on the same data. The results show that transfer learning improves the test accuracy compared to the regular convolutional neural network when the dataset is small. This thesis project concludes that transfer learning can reduce the need for data when faced with a new machine learning task. The results indicate that transfer learning could be useful in the industry.

    Download full text (pdf)
    fulltext
  • 17.
    Bethdavid, Simon
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Zero-Knowledge Agent Trained for the Game of Risk2020Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Recent developments in deep reinforcement learning applied to abstract strategy games such as Go, chess and Hex have sparked an interest within military planning. This Master thesis explores if it is possible to implement an algorithm similar to Expert Iteration and AlphaZero to wargames. The studied wargame is Risk, which is a turn-based multiplayer game played on a simplified political map of the world. The algorithms consist of an expert, in the form of a Monte Carlo tree search algorithm, and an apprentice, implemented through a neural network. The neural network is trained by imitation learning, trained to mimic expert decisions generated from self-play reinforcement learning. The apprentice is then used as heuristics in forthcoming tree searches. The results demonstrated that a Monte Carlo tree search algorithm could, to some degree, be employed on a strategy game as Risk, dominating a random playing agent. The neural network, fed with a state representation in the form of a vector, had difficulty in learning expert decisions and could not beat a random playing agent. This led to a halt in the expert/apprentice learning process. However, possible solutions are provided as future work.

    Download full text (pdf)
    fulltext
  • 18. Birk, Wolfgang
    et al.
    Hostettler, Roland
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Razi, Maryam
    Atta, Khalid
    Tammia, Rasmus
    Automatic generation and updating of process industrial digital twins for estimation and control: A review2022In: Frontiers in Control Engineering, E-ISSN 2673-6268, Vol. 3, article id 954858Article in journal (Refereed)
    Abstract [en]

    This review aims at assessing the opportunities and challenges of creating and using digital twins for process industrial systems over their life-cycle in the context of estimation and control. The scope is, therefore, to provide a survey on mechanisms to generate models for process industrial systems using machine learning (purely data-driven) and automated equation-based modeling. In particular, we consider learning, validation, and updating of large-scale (i.e., plant-wide or plant-stage but not component-wide) equation-based process models. These aspects are discussed in relation to typical application cases for the digital twins creating value for users both on the operational and planning level for process industrial systems. These application cases are also connected to the needed technologies and the maturity of those as given by the state of the art. Combining all aspects, a way forward to enable the automatic generation and updating of digital twins is proposed, outlining the required research and development activities. 

    Download full text (pdf)
    fulltext
  • 19.
    Biswas, Sinchan
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Distributed Detection and Its Applications with Energy Harvesting Wireless Networks2020Doctoral thesis, comprehensive summary (Other academic)
    Abstract [en]

    With the advent and widespread applications of high data-rate wireless services and devices, two of the fundamental resources in wireless communication have become extremely important and are scarce. These two resources are bandwidth and energy respectively. To tackle the problem of the ever-growing requirements of bandwidth, the paradigm of cognitive radio has been proposed in the literature where the users without any license are capable of utilizing the wireless radio spectrum allocated to the licensed user when it is idle. The performance of such a dynamic spectrum allocation policy depends heavily on the unlicensed users' ability to detect the vacancy in a licensed user's radio spectrum. Different types of detection algorithms have been investigated in the literature for this purpose. Some classic detection techniques like energy detection, matched filter detection, cyclo-stationary detector, generalized likelihood ratio test detector have found significant applications in sensing the licensed spectrum. These group of detection techniques focuses on collecting samples and performing the detection in a non-sequential fashion. The sequential counterpart of such techniques which are implemented sequentially at every time instant has also been studied extensively.

    Tackling the evergrowing energy requirement has been the other major challenge for wireless communication. To address this issue, a significant amount of research has been dedicated to the idea of incorporating the capability of energy harvesting in wireless devices. Further research in this domain has also introduced the idea of wireless energy sharing, where individual users are additionally capable of sharing energy with each other. The problem with such systems is the inherently stochastic nature of the energy harvesting process. Furthermore, there are practical limitations of the size of the battery for each user, which limits the amount of energy that can be stored at a particular time instant.

    Motivated by these two factors, the work presented in this thesis has its focus on cognitive radio networks with energy harvesting capability. In the aforementioned network, unlicensed users are concerned with achieving two fundamental goals. Firstly, they want to efficiently utilize the radio spectrum when the licensed user is not active, which results in the sum-throughput maximization problem with energy harvesting constraint. We have also investigated this problem where individual unlicensed users are capable of sharing energy with each other. Secondly, they want to detect the change in the activity in the licensed user spectrum as soon as possible. Motivated by this goal, we have investigated the problem of change point detection delay minimization in wireless sensor networks with energy harvesting constraints in a decentralized setting. Furthermore, we have explored the detection delay parameter for the decentralized settings with local decisions with similar constraints of energy availability due to energy harvesting.

    List of papers
    1. Sensing throughput optimization in fading cognitive multiple access channels with energy harvesting secondary transmitters.: Extended version of conference paper
    Open this publication in new window or tab >>Sensing throughput optimization in fading cognitive multiple access channels with energy harvesting secondary transmitters.: Extended version of conference paper
    2016 (English)Report (Other academic)
    Place, publisher, year, edition, pages
    Uppsala universitet, 2016
    National Category
    Engineering and Technology
    Identifiers
    urn:nbn:se:uu:diva-302637 (URN)
    Available from: 2016-09-07 Created: 2016-09-07 Last updated: 2020-04-22
    2. Sensing Throughput Optimization in Cognitive Fading Multiple Access Channels With Energy Harvesting Secondary Transmitters
    Open this publication in new window or tab >>Sensing Throughput Optimization in Cognitive Fading Multiple Access Channels With Energy Harvesting Secondary Transmitters
    2016 (English)In: 2016 24Th European Signal Processing Conference (EUSIPCO), 2016, p. 577-581Conference paper, Published paper (Refereed)
    Abstract [en]

    The paper investigates the problem of maximizing the expected achievable sum rate in a fading multiple access cognitive radio network when secondary user (SU) transmitters have energy harvesting capability, and perform cooperative spectrum sensing. We formulate the problem as maximization of throughput of the cognitive multiple access network over a finite time horizon subject to a time averaged interference constraint at the primary user (PU) and almost sure energy causality constraints at the SUs. The problem is a mixed integer non-linear program with respect to two decision variables, namely, spectrum access decision and spectrum sensing decision, and the continuous variables sensing time and transmission power. In general, this problem is known to be NP hard. For optimization over these two decision variables, we use an exhaustive search policy when the length of the time horizon is small, and a heuristic policy for longer horizons. For given values of the decision variables, the problem simplifies into a joint optimization on SU transmission power and sensing time, which is non-convex in nature. We present an analytic solution for the resulting optimization problem using an alternating convex optimization problem for non-causal channel state information and harvested energy information patterns at the SU base station (SBS) or fusion center (FC) and infinite battery capacity at the SU transmitters. We formulate the problem with causal information and finite battery capacity as a stochastic control problem and solve it using the technique of dynamic programming. Numerical results are presented to illustrate the performance of the various algorithms.

    Series
    European Signal Processing Conference, ISSN 2076-1465
    National Category
    Electrical Engineering, Electronic Engineering, Information Engineering
    Identifiers
    urn:nbn:se:uu:diva-316230 (URN)10.1109/EUSIPCO.2016.7760314 (DOI)000391891900469 ()9780992862657 (ISBN)
    Conference
    24th European Signal Processing Conference (EUSIPCO), AUG 28-SEP 02, 2016, Budapest, HUNGARY
    Available from: 2017-02-27 Created: 2017-02-27 Last updated: 2020-04-22Bibliographically approved
    3. Quantized non-Bayesian quickest change detection with energy harvesting
    Open this publication in new window or tab >>Quantized non-Bayesian quickest change detection with energy harvesting
    2018 (English)Conference paper, Published paper (Refereed)
    Abstract [en]

    This paper focuses on the analysis of an optimal sensing and quantization strategy in a multi-sensor network where each individual sensor sends its quantized log-likelihood information to the fusion center (FC) for non-Bayesian quickest change detection. It is assumed that the sensors are equipped with a battery/energy storage device of finite capacity, capable of harvesting energy from the environment. The FC is assumed to have access to either non-causal or causal channel state information (CSI) and energy state information (ESI) from all the sensors while performing the quickest change detection. The primary observations are assumed to be generated from a sequence of random variables whose probability distribution function changes at an unknown time point. The objective of the detection problem is to minimize the average detection delay of the change point with respect to a lower bound on the rate of false alarm. In this framework, the optimal sensing decision and number of quantization bits for information transmission can be determined with the constraint of limited available energy due to finite battery capacity. This optimization is formulated as a stochastic control problem and is solved using dynamic programming algorithms for both non-causal and causal CSI and ESI scenario. A set of non-linear equations is also derived to determine the optimal quantization thresholds for the sensor log-likelihood ratios, by maximizing an appropriate Kullback-Leibler (KL) divergence measure between the distributions before and after the change. A uniform threshold quantization strategy is also proposed as a simple sub-optimal policy. The simulation results indicate that the optimal quantization is preferable when the number of quantization bits is low as its performance is significantly better compared to its uniform counterpart in terms of average detection delay. For the case of a large number of quantization bits, the performance benefits of using the optimal quantization as compared to its uniform counterpart diminish, as expected.

    Place, publisher, year, edition, pages
    IEEE, 2018
    Series
    IEEE Global Communications Conference, E-ISSN 2576-6813
    National Category
    Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-363299 (URN)10.1109/GLOCOM.2018.8647715 (DOI)000465774303093 ()978-1-5386-4727-1 (ISBN)
    Conference
    IEEE Global Communications Conference (GLOBECOM), 9-13 December 2018, Abu Dhabi, United Arab Emirates
    Available from: 2018-10-16 Created: 2018-10-16 Last updated: 2020-04-22Bibliographically approved
    4. Sum Throughput Maximization in a Cognitive Multiple Access Channel With Cooperative Spectrum Sensing and Energy Harvesting
    Open this publication in new window or tab >>Sum Throughput Maximization in a Cognitive Multiple Access Channel With Cooperative Spectrum Sensing and Energy Harvesting
    2019 (English)In: IEEE Transactions on Cognitive Communications and Networking, E-ISSN 2332-7731, Vol. 5, no 2, p. 382-399Article in journal (Refereed) Published
    Abstract [en]

    This paper focuses on the problem of sensing throughput optimization in a fading multiple access cognitive radio (CR) network, where the secondary user (SU) transmitters participate in cooperative spectrum sensing and are capable of harvesting energy and sharing energy with each other. We formulate the optimization problem as a maximization of the expected achievable sum-rate over a finite horizon, subject to an average interference constraint at the primary receiver, peak power constraints, and energy causality constraints at the SU transmitters. The optimization problem is a non-convex, mixed integer non-linear program (MINLP) involving the binary action to sense the spectrum or not, and the continuous variables, such as the transmission power, shared energy, and sensing time. The problem is analyzed under two different assumptions on the available information pattern: 1) non-causal channel state information (CSI), energy state information (ESI), and infinite battery capacity and 2) the more realistic scenario of the causal CSI/ESI and finite battery. In the non-casual case, this problem can be solved by an exhaustive search over the decision variable or an MINLP solver for smaller problem dimensions, and a novel heuristic policy for larger problems, combined with an iterative alternative optimization method for the continuous variables. The causal case with finite battery is optimally solved using a dynamic programming (DP) methodology, whereas a number of sub-optimal algorithms are proposed to reduce the computational complexity of DP. Extensive numerical simulations are carried out to illustrate the performance of the proposed algorithms. One of the main findings indicates that the energy sharing is more beneficial when there is a significant asymmetry between average harvested energy levels/channel gains of different SUs.

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2019
    Keywords
    Energy harvesting, cognitive radio, multiple access channel, spectrum sensing, fading channel
    National Category
    Communication Systems Telecommunications Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-389809 (URN)10.1109/TCCN.2019.2908860 (DOI)000471115000016 ()
    Funder
    Swedish Research Council, 2017-04053
    Available from: 2019-07-30 Created: 2019-07-30 Last updated: 2023-01-25Bibliographically approved
    5. Asymptotic Performance Analysis of Distributed Non-Bayesian Quickest Change Detection with Energy Harvesting Sensors
    Open this publication in new window or tab >>Asymptotic Performance Analysis of Distributed Non-Bayesian Quickest Change Detection with Energy Harvesting Sensors
    2022 (English)In: IEEE Transactions on Aerospace and Electronic Systems, ISSN 0018-9251, E-ISSN 1557-9603, Vol. 58, no 4, p. 3697-3707Article in journal (Refereed) Published
    Abstract [en]

    This paper focuses on the distributed non-Bayesian quickest change detection of the probability distribution of a random process in a wireless senor network (WSN), where the distributions before and after the change point is assumed to be known. The individual sensors are capable of harvesting energy from their surroundings. Each sensor decides to sense the observation signal depending on the available energy at its disposal. Once, a sensor decides to sense, it takes a sample of the observation signal and computes the log-likelihood ratio (LLR) of the aforementioned two distributions, if enough energy is available in its battery for sensing and processing the sample. On the other hand, if enough energy is not available, the sensor decides to abstain from the sensing process during that time slot and waits until a future time slot when it accumulates enough energy to perform the sensing and processing. Once a sensor computes the LLR, it uses that information to calculate the Cumulative Sum (CUSUM) test statsistic to arrive at a local decision about the change point. When a change is detected, these decisions are then sent to the FC (provided the transmitting sensor has enough energy to transmit the decision to the fusion centre successfully), where they are collated to form a single decision about the detection of the change point based on some pre-decided fusion rule. In this work, using asymptotic results on the detection delay for CUSUM tests for a single sensor, we have derived asymptotic results for the expected detection delay (when the change occurs) for three common fusion rules, namely, OR, AND and $r$ out of $N$ rule respectively. These results are analyzed for the scenario when the average harvested energy ($\mathit{\bar{H}}$) at each sensor is greater than or equal to the amount of energy required for sensing ($E_{s}$). We show that in such cases, the standard existing asymptotic results for CUSUM test holds for the local decisions. Consequently, we have determined corresponding results for the detection delay for decisions taken at the FC with the three aforementioned fusion rules by using the theory of order statistics. Numerical results are provided to support the theoretical claims.

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2022
    Keywords
    Energy harvesting, Sensor Networks, Distributed Change-Point Detection
    National Category
    Signal Processing
    Research subject
    Electrical Engineering with specialization in Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-409526 (URN)10.1109/TAES.2022.3156109 (DOI)000838710200081 ()
    Funder
    Swedish Research Council, 2017-04053
    Available from: 2020-04-22 Created: 2020-04-22 Last updated: 2022-10-31Bibliographically approved
    6. On Optimal Quantized Non-Bayesian Quickest Change Detection with Energy Harvesting
    Open this publication in new window or tab >>On Optimal Quantized Non-Bayesian Quickest Change Detection with Energy Harvesting
    2020 (English)In: IEEE Transactions on Green Communications and Networking, ISSN 2473-2400, Vol. 4, no 2, p. 433-447Article in journal (Other academic) Published
    Abstract [en]

    In this paper, we consider a problem of decentralized non-Bayesian quickest change detection using a wireless sensor network where the sensor nodes are powered by harvested energy from the environment. The underlying random process being monitored by the sensors is subject to change in its distribution at an unknown but deterministic time point and the sensors take samples (sensing) periodically, compute the likelihood ratio based on the distributions before and after the change, quantize it and send it to a remote fusion centre (FC) over fading channels for performing a sequential test to detect the change. Due to the unpredictable and intermittent nature of harvested energy arrivals, the sensors need to decide whether they want to sense, and at what rate they want to quantize their information before sending them to the FC, since higher quantization rates result in higher accuracy and better detection performance, at the cost of higher energy consumption. We formulate an optimal sensing and quantization rate allocation problem (in order to minimize the expected detection delay subject to false alarm rate constraint) based on the availability (at the FC) of non-causal and causal information of sensors’ energy state information, and channel state information between the sensors and the FC. Motivated by the asymptotically inverse relationship between the expected detection delay (under a vanishingly small probability of false alarm) and the Kullback-Leibler (KL) divergence measure at the FC, we maximize an expected sum of the KL divergence measure over a finite horizon to obtain the optimal sensing and quantization rate allocation policy, subject to energy causality constraints at each sensor. The optimal solution is obtained using a typical dynamic programming based technique, and based on the optimal quantization rate, the optimal quantization thresholds are found by maximizing the KL information measure per slot. We also provide suboptimal threshold design policies using uniform quantization and an asymptotically optimal quantization policy for higher number of quantization bits. We provide an asymptotic approximation for the loss due to quantization of the KL measure, and also consider an alternative optimization problem with minimizing the expected sum of the inverse the KL divergence measure as the cost per time slot. Numerical results are provided comparing the various optimal and suboptimal quantization strategies for both optimization problem formulations, illustrating the comparative performance of these strategies at different regimes of quantization rates.

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2020
    Keywords
    Energy harvesting, Sensor Networks, Decentralized Change-Point Detection, Quantization
    National Category
    Signal Processing
    Research subject
    Electrical Engineering with specialization in Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-409392 (URN)10.1109/TGCN.2019.2961113 (DOI)000722240800010 ()
    Funder
    Swedish Research Council, 621-2013-5395
    Available from: 2020-04-20 Created: 2020-04-20 Last updated: 2023-07-19Bibliographically approved
    Download full text (pdf)
    fulltext
    Download (jpg)
    presentationsbild
  • 20.
    Biswas, Sinchan
    et al.
    Department of Electrical and Information Technology, Lund University, Lund, Sweden..
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Asymptotic Performance Analysis of Distributed Non-Bayesian Quickest Change Detection with Energy Harvesting Sensors2022In: IEEE Transactions on Aerospace and Electronic Systems, ISSN 0018-9251, E-ISSN 1557-9603, Vol. 58, no 4, p. 3697-3707Article in journal (Refereed)
    Abstract [en]

    This paper focuses on the distributed non-Bayesian quickest change detection of the probability distribution of a random process in a wireless senor network (WSN), where the distributions before and after the change point is assumed to be known. The individual sensors are capable of harvesting energy from their surroundings. Each sensor decides to sense the observation signal depending on the available energy at its disposal. Once, a sensor decides to sense, it takes a sample of the observation signal and computes the log-likelihood ratio (LLR) of the aforementioned two distributions, if enough energy is available in its battery for sensing and processing the sample. On the other hand, if enough energy is not available, the sensor decides to abstain from the sensing process during that time slot and waits until a future time slot when it accumulates enough energy to perform the sensing and processing. Once a sensor computes the LLR, it uses that information to calculate the Cumulative Sum (CUSUM) test statsistic to arrive at a local decision about the change point. When a change is detected, these decisions are then sent to the FC (provided the transmitting sensor has enough energy to transmit the decision to the fusion centre successfully), where they are collated to form a single decision about the detection of the change point based on some pre-decided fusion rule. In this work, using asymptotic results on the detection delay for CUSUM tests for a single sensor, we have derived asymptotic results for the expected detection delay (when the change occurs) for three common fusion rules, namely, OR, AND and $r$ out of $N$ rule respectively. These results are analyzed for the scenario when the average harvested energy ($\mathit{\bar{H}}$) at each sensor is greater than or equal to the amount of energy required for sensing ($E_{s}$). We show that in such cases, the standard existing asymptotic results for CUSUM test holds for the local decisions. Consequently, we have determined corresponding results for the detection delay for decisions taken at the FC with the three aforementioned fusion rules by using the theory of order statistics. Numerical results are provided to support the theoretical claims.

  • 21.
    Biswas, Sinchan
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Knorn, Steffi
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Ahlén, Anders
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    On Optimal Quantized Non-Bayesian Quickest Change Detection with Energy Harvesting2020In: IEEE Transactions on Green Communications and Networking, ISSN 2473-2400, Vol. 4, no 2, p. 433-447Article in journal (Other academic)
    Abstract [en]

    In this paper, we consider a problem of decentralized non-Bayesian quickest change detection using a wireless sensor network where the sensor nodes are powered by harvested energy from the environment. The underlying random process being monitored by the sensors is subject to change in its distribution at an unknown but deterministic time point and the sensors take samples (sensing) periodically, compute the likelihood ratio based on the distributions before and after the change, quantize it and send it to a remote fusion centre (FC) over fading channels for performing a sequential test to detect the change. Due to the unpredictable and intermittent nature of harvested energy arrivals, the sensors need to decide whether they want to sense, and at what rate they want to quantize their information before sending them to the FC, since higher quantization rates result in higher accuracy and better detection performance, at the cost of higher energy consumption. We formulate an optimal sensing and quantization rate allocation problem (in order to minimize the expected detection delay subject to false alarm rate constraint) based on the availability (at the FC) of non-causal and causal information of sensors’ energy state information, and channel state information between the sensors and the FC. Motivated by the asymptotically inverse relationship between the expected detection delay (under a vanishingly small probability of false alarm) and the Kullback-Leibler (KL) divergence measure at the FC, we maximize an expected sum of the KL divergence measure over a finite horizon to obtain the optimal sensing and quantization rate allocation policy, subject to energy causality constraints at each sensor. The optimal solution is obtained using a typical dynamic programming based technique, and based on the optimal quantization rate, the optimal quantization thresholds are found by maximizing the KL information measure per slot. We also provide suboptimal threshold design policies using uniform quantization and an asymptotically optimal quantization policy for higher number of quantization bits. We provide an asymptotic approximation for the loss due to quantization of the KL measure, and also consider an alternative optimization problem with minimizing the expected sum of the inverse the KL divergence measure as the cost per time slot. Numerical results are provided comparing the various optimal and suboptimal quantization strategies for both optimization problem formulations, illustrating the comparative performance of these strategies at different regimes of quantization rates.

  • 22.
    Björsell, Joachim
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Predictor Antennas: Enabling channel prediction for fast-moving vehicles in wireless broadband systems2022Doctoral thesis, comprehensive summary (Other academic)
    Abstract [en]

    Many advanced transmission techniques utilize channel state information (CSI) at the transmitter (CSIT) to improve throughput, spectral efficiency, power efficiency, and other performance metrics. Estimating CSI accurately is important to fully benefit from many of these techniques. In situations where users travel at high speed, the channel can change rapidly, especially in small-scale fading environments. In many systems, there is also a delay between measuring CSI and using it for transmission. If the channel changes significantly during this delay, CSI becomes outdated and the benefits of advanced transmission techniques are typically negatively affected. Long-range channel prediction can be used to counteract this delay and enable advanced transmission to vehicles that travel at high velocity. Conventional prediction methods use channel extrapolation and have a limited prediction horizon that does not support high vehicular velocities for the current size of these delays. The predictor antenna concept has been shown to increase the prediction horizon by at least an order-of-magnitude. It does so by placing an antenna array on the exterior of a vehicle, in the direction of travel. The first antenna can then measure the channel at positions that the following antennas will visit later.

    This thesis uses channel measurements to investigate how practical aspects affect the prediction performance of predictions based on predictor antennas. It also develops a general framework that can be used to calculate the predictions in a real system. This includes addressing the causality of all the processing methods involved and adapting these methods to the design of the system and the radio environment. In a massive multiple-inputmultiple-output (MIMO) system, multi-user transmission is enabled by channel prediction and increases the sum capacity by 100% compared to 1 ms old channel estimates at a velocity of 150 km/h. This is achieved with relatively dense pilots in time. The prediction performance of the proposed framework is shown to degrade if pilots are spread further than 0.3–0.5 wavelengths in space, if spline interpolation is used to interpolate between the channel estimates.

    List of papers
    1. Using predictor antennas for the prediction of small-scale fading provides an order-of-magnitude improvement of prediction horizons
    Open this publication in new window or tab >>Using predictor antennas for the prediction of small-scale fading provides an order-of-magnitude improvement of prediction horizons
    2017 (English)Report (Other academic)
    Place, publisher, year, edition, pages
    Uppsala: Signals and Systems, Uppsala university, 2017. p. Report r161
    National Category
    Engineering and Technology
    Identifiers
    urn:nbn:se:uu:diva-330706 (URN)
    Available from: 2017-10-03 Created: 2017-10-03 Last updated: 2022-03-27
    2. A Framework for Predictor Antennas in Practice
    Open this publication in new window or tab >>A Framework for Predictor Antennas in Practice
    2022 (English)In: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359, Vol. 71, no 7, p. 7503-7518Article in journal (Refereed) Published
    Abstract [en]

    Channel predictions are important to achieve high spectral efficiency for high-mobility vehicles. Channel extrapolation, used by many prediction methods, suffers from a limited prediction horizon in difficult radio environments. The predictor antenna (PA) concept provides the prediction horizons required for efficient transmission to fast-moving vehicles by measuring the channel ahead of time with an extra antenna placed on the vehicle. This paper presents a general framework that addresses the practical signal processing challenges of the PA concept. It is adaptable to a vast variety of vehicular deployment, mobility, and communication scenarios. A new theoretical prediction normalized mean-squared error (NMSE) expression is derived based on the presented framework. The framework is demonstrated by applying it to extensive channel measurements and comparing the PA predictions to Kalman-based channel predictions and outdated channel estimates. By studying the impact of vehicular velocity and radio environment on the prediction performance, it is shown that PA prediction is weaker at low velocities, where Kalman prediction methods are sufficient, but is uncontested at high velocities in environments without a dominating path. At high velocities in dominating path environments, the Kalman predictor provides usable predictions, but it is still outperformed by the PA predictions. 

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2022
    Keywords
    Channel state information (CSI), High-mobility, Predictor antenna
    National Category
    Telecommunications Signal Processing
    Research subject
    Electrical Engineering with specialization in Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-470620 (URN)10.1109/TVT.2022.3168225 (DOI)000876768200052 ()
    Available from: 2022-03-27 Created: 2022-03-27 Last updated: 2022-11-25Bibliographically approved
    3. Enabling Multi-User M-MIMO for High-Mobility Users with Predictor Antennas: A Deep Analysis Based on Experimental NLOS Measurements
    Open this publication in new window or tab >>Enabling Multi-User M-MIMO for High-Mobility Users with Predictor Antennas: A Deep Analysis Based on Experimental NLOS Measurements
    2022 (English)In: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359, Vol. 71, no 7, p. 7456-7471Article in journal (Refereed) Published
    Abstract [en]

    Predictor antenna channel measurements from a 64-antenna base station are used to investigate the effects of channel aging on the precoder design for high-mobility users in a non-line-of-sight (NLOS) massive multiple-input multiple-output (M-MIMO) environment. The effects are evaluated in terms of received downlink (DL) signal-to-interference-and-noise ratio (SINR) and the corresponding ergodic capacity bound. A simulated velocity of 150 km/h is used with a carrier frequency of 2.18 GHz. Maximum ratio (MR) and a codebook-based precoders are used to evaluate single-user transmission and zero-forcing (ZF), regularized zero-forcing (RZF), and minimum mean-squared error (MMSE) precoders are used to evaluate multi-user transmission with up to nine active users in a cell. Furthermore, predictor antenna predictions are evaluated as a mean of combating channel aging. It is also investigated how the predictor antenna can be used during data reception. Simulations show that outdated channel estimates significantly reduce the SINR and consequently the capacity for all investigated transmission techniques. Basic predictor antenna predictions outperform the use of outdated channel estimates for delays larger than 0.6 ms. In single-user transmission, channel prediction can improve the capacity by 6–14%. The gain from multi-user transmission typically disappears when using outdated channel estimates older than 1 ms. In contrast, the use of predictor antennas enables multi-user MIMO for these high-mobility users, which is demonstrated to increase the capacity bound by 100% compared to 1 ms old channel estimates. 

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2022
    Keywords
    Channel prediction, High-mobility, M-MIMO, Outdated channel state information (CSI), Predictor antenna
    National Category
    Telecommunications Signal Processing
    Research subject
    Electrical Engineering with specialization in Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-470622 (URN)10.1109/TVT.2022.3167630 (DOI)000876768200049 ()
    Funder
    Uppsala University
    Available from: 2022-03-27 Created: 2022-03-27 Last updated: 2022-11-24Bibliographically approved
    4. Channel Interpolation of Fading Channels and the Pilot Density Required for Predictor Antennas
    Open this publication in new window or tab >>Channel Interpolation of Fading Channels and the Pilot Density Required for Predictor Antennas
    (English)In: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359Article in journal (Refereed) Submitted
    Abstract [en]

    Predictor antennas (PAs) are a potential solution to severe channel aging that can occur at high vehicular velocities in non line-of-sight (NLOS) environments. Channel aging reduces the performance of many advanced communication schemes based on channel state information at the transmitter (CSIT). Although PAs have been shown to work in combination with dense pilots in time and space, prediction performance can be reduced when channel estimates are sparse. This paper answers how densely pilots must be placed for PAs to be feasible when performing basic interpolation between channel estimates. This is important, especially for establishing upper limits on the length of the downlink (DL) frames required in a time-division duplex (TDD) system with PAs. First, nearest-neighbor, linear, and spline interpolation are analyzed when applied to stochastic radio channels. A theoretical expression is derived for the power of the expected interpolation error for any interpolation method that can be expressed as a linear function of a set of measured values. The interpolation methods are evaluated on three theoretical channels with Rayleigh, flat, and Rician fading, and on two sets of channel measurements. For a given requirement on interpolation normalized mean-squared error (NMSE), a minimum pilot density is extracted, both for theoretical channels and measured channels. The interpolation techniques are used to evaluate the PA channel prediction for different distances between pilots. The results indicate that linear and spline interpolation can be used with down to five and three samples per wavelength, respectively, without affecting the PA-based prediction NMSE. At two samples per wavelength, the prediction NMSE is still at a level that can be useful for precoding design in massive multiple-input multiple-output (M-MIMO) systems. These results can be used to adapt the uplink/downlink frame rate in TDD to the speed of the terminals, if technically possible. 

    Keywords
    Channel interpolation, Channel prediction, Channel state information (CSI), Predictor antenna, Spline interpolation
    National Category
    Signal Processing Telecommunications
    Research subject
    Electrical Engineering with specialization in Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-470623 (URN)
    Available from: 2022-03-27 Created: 2022-03-27 Last updated: 2022-04-03
    Download full text (pdf)
    UUThesis_Björsell,J_2020
    Download (jpg)
    presentationsbild
  • 23.
    Björsell, Joachim
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Sternad, Mikael
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Grieger, Michael
    Xilinx, Dresden, Germany.
    A Framework for Predictor Antennas in Practice2022In: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359, Vol. 71, no 7, p. 7503-7518Article in journal (Refereed)
    Abstract [en]

    Channel predictions are important to achieve high spectral efficiency for high-mobility vehicles. Channel extrapolation, used by many prediction methods, suffers from a limited prediction horizon in difficult radio environments. The predictor antenna (PA) concept provides the prediction horizons required for efficient transmission to fast-moving vehicles by measuring the channel ahead of time with an extra antenna placed on the vehicle. This paper presents a general framework that addresses the practical signal processing challenges of the PA concept. It is adaptable to a vast variety of vehicular deployment, mobility, and communication scenarios. A new theoretical prediction normalized mean-squared error (NMSE) expression is derived based on the presented framework. The framework is demonstrated by applying it to extensive channel measurements and comparing the PA predictions to Kalman-based channel predictions and outdated channel estimates. By studying the impact of vehicular velocity and radio environment on the prediction performance, it is shown that PA prediction is weaker at low velocities, where Kalman prediction methods are sufficient, but is uncontested at high velocities in environments without a dominating path. At high velocities in dominating path environments, the Kalman predictor provides usable predictions, but it is still outperformed by the PA predictions. 

  • 24.
    Björsell, Joachim
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Sternad, Mikael
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Phan-Huy, Dinh-Thuy
    Orange Labs, Châtillon, France.
    Enabling Multi-User M-MIMO for High-Mobility Users with Predictor Antennas: A Deep Analysis Based on Experimental NLOS Measurements2022In: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359, Vol. 71, no 7, p. 7456-7471Article in journal (Refereed)
    Abstract [en]

    Predictor antenna channel measurements from a 64-antenna base station are used to investigate the effects of channel aging on the precoder design for high-mobility users in a non-line-of-sight (NLOS) massive multiple-input multiple-output (M-MIMO) environment. The effects are evaluated in terms of received downlink (DL) signal-to-interference-and-noise ratio (SINR) and the corresponding ergodic capacity bound. A simulated velocity of 150 km/h is used with a carrier frequency of 2.18 GHz. Maximum ratio (MR) and a codebook-based precoders are used to evaluate single-user transmission and zero-forcing (ZF), regularized zero-forcing (RZF), and minimum mean-squared error (MMSE) precoders are used to evaluate multi-user transmission with up to nine active users in a cell. Furthermore, predictor antenna predictions are evaluated as a mean of combating channel aging. It is also investigated how the predictor antenna can be used during data reception. Simulations show that outdated channel estimates significantly reduce the SINR and consequently the capacity for all investigated transmission techniques. Basic predictor antenna predictions outperform the use of outdated channel estimates for delays larger than 0.6 ms. In single-user transmission, channel prediction can improve the capacity by 6–14%. The gain from multi-user transmission typically disappears when using outdated channel estimates older than 1 ms. In contrast, the use of predictor antennas enables multi-user MIMO for these high-mobility users, which is demonstrated to increase the capacity bound by 100% compared to 1 ms old channel estimates. 

  • 25.
    Björsell, Joachim
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Sternad, Mikael
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Phan-Huy, Dinh-Thuy
    Orange Labs, Châttillon, France.
    Grieger, Michael
    Xilinx, Dresden, Germany.
    Channel Interpolation of Fading Channels and the Pilot Density Required for Predictor AntennasIn: IEEE Transactions on Vehicular Technology, ISSN 0018-9545, E-ISSN 1939-9359Article in journal (Refereed)
    Abstract [en]

    Predictor antennas (PAs) are a potential solution to severe channel aging that can occur at high vehicular velocities in non line-of-sight (NLOS) environments. Channel aging reduces the performance of many advanced communication schemes based on channel state information at the transmitter (CSIT). Although PAs have been shown to work in combination with dense pilots in time and space, prediction performance can be reduced when channel estimates are sparse. This paper answers how densely pilots must be placed for PAs to be feasible when performing basic interpolation between channel estimates. This is important, especially for establishing upper limits on the length of the downlink (DL) frames required in a time-division duplex (TDD) system with PAs. First, nearest-neighbor, linear, and spline interpolation are analyzed when applied to stochastic radio channels. A theoretical expression is derived for the power of the expected interpolation error for any interpolation method that can be expressed as a linear function of a set of measured values. The interpolation methods are evaluated on three theoretical channels with Rayleigh, flat, and Rician fading, and on two sets of channel measurements. For a given requirement on interpolation normalized mean-squared error (NMSE), a minimum pilot density is extracted, both for theoretical channels and measured channels. The interpolation techniques are used to evaluate the PA channel prediction for different distances between pilots. The results indicate that linear and spline interpolation can be used with down to five and three samples per wavelength, respectively, without affecting the PA-based prediction NMSE. At two samples per wavelength, the prediction NMSE is still at a level that can be useful for precoding design in massive multiple-input multiple-output (M-MIMO) systems. These results can be used to adapt the uplink/downlink frame rate in TDD to the speed of the terminals, if technically possible. 

  • 26.
    Bodin, Emanuel
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Furniture swap: Segmentation and 3D rotation of natural images using deep learning2021Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Learning to perceive scenes and objects from 2D images as 3D models is atrivial task for a human but very challenging for a computer. Being ableto retrieve a 3D model from a scene just by taking a picture of it canbe of great use in many fields, for example when making 3D blueprintsfor buildings or working with animations in the game or film industry.Novel view synthesis is a field within deep learning where generativemodels are trained to construct 3D models of scenes or objects from 2Dimages.

    In this work, the generative model HoloGAN is combined together with aU-net segmentation network. The solution is able to, given an imagecontaining a single object as input, swap that object to another oneand then perform a rotation of the scene, generating new images fromunobserved view points. The segmentation network is trained with pairedsegmentation masks while HoloGAN is able to in an unsupervised mannerlearn 3D metrics of scenes from unlabeled 2D images. The system as awhole is trained on one dataset containing images of cars while theperformance of HoloGAN was evaluated on four additionaldatasets. The chosen method proved to be successful but came with somedrawbacks such as requiring large dataset sizes and being computationalexpensive to train.

    Download full text (pdf)
    fulltext
  • 27.
    Bonnedahl, Marcus
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Using artificial intelligence to improvetime estimation for project management2024Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Time estimation is an important aspect in project management. Failure to make accurateestimates can lead to large consequences. Despite this, humans tend to make fairly inaccurateestimates when tasked to, often underestimating the time something will take substantially. Thisthesis explores using artificial intelligence and machine learning to produce time estimates forthe life science company Biotage. A predictive model can be trained using previous projects assamples, including time reporting data for employees as the output variable.A total of 12 completed projects were found that had both sufficient time reporting data andsome project information. Previous projects took on average 55.1% longer to complete thanestimated at the start of the project. Every project had one or more of the following: projectdescription, work breakdown structure and/or Gantt chart. However, the level of detail in almostall of the projects was very low, making it difficult to extract useful features. A constant-timemodel (predicting that every project takes the same amount of time), had a Root Mean SquaredError (RMSE) of 5058 hours and a Mean Absolute Percentage Error (MAPE) of 282%. Anothermodel that took into account whether the project was a software only, hardware only or both hada RMSE of 4269 hours and MAPE of 320%. Due to the scarcity of data, no furtherimprovements were made. It was determined that in order to develop a predictive model thatcan match human estimates, at least one of the following had to be true: Better level of detail inthe data, bigger sample size of previous projects, or projects being more similar so that theyshare common features more often.

    Download full text (pdf)
    fulltext
  • 28. Bosch, David
    et al.
    Panahi, Ashkan
    Özcelikkale, Ayca
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Double Descent in Feature Selection: Revisiting LASSO and Basis Pursuit2021Conference paper (Refereed)
    Abstract [en]

    We present a novel analysis of feature selection in linear models by the convex framework of leastabsolute shrinkage operator (LASSO) and basis pursuit (BP). Our analysis pertains to a generaloverparametrized scenario. When the numbers of the features and the data samples grow proportionally, we obtain precise expressions for the asymptotic generalization error of LASSO and BP. Considering a mixture of strong and weak features, we provide insights into regularization trade-offs for double descent for l1 norm minimization. We validate these results with numerical experiments.

  • 29.
    Bossér, Daniel
    et al.
    Automatic Control, Linköping University .
    Forsling, Robin
    Automatic Control, Linköping University .
    Skog, Isaac
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems. Swedish Defense Research Agency..
    Hendeby, Gustaf
    Automatic Control, Linköping University .
    Nordenvaad, Magnus
    Swedish Defense Research Agency, FOI, Sweden.
    Underwater Environment Modeling for Passive Sonar Track-Before-Detect2023In: OCEANS 2023 - LIMERICK, Institute of Electrical and Electronics Engineers (IEEE), 2023Conference paper (Refereed)
    Abstract [en]

    Underwater surveillance using passive sonar and track-before-detect technology requires accurate models of the tracked signal and the background noise. However, in an underwater environment, the signal channel is time-varying and prior knowledge about the spatial distribution of the background noise is unavailable. In this paper, an autoregressive model that captures a time-varying signal level caused by multi-path propagation is presented. In addition, a multi-source model is proposed to describe spatially distributed background noise. The models are used in a Bernoulli filter track-before-detect framework and evaluated using both simulated and sea trial data. The simulations demonstrate clear improvements in terms of target loss and improved ability to discern the target from the noisy background. An evaluation of the track-before-detect algorithm on the sea trial data indicates a performance gain when incorporating the proposed models in underwater surveillance and tracking problems.

  • 30.
    Boström, Viktor
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Positioning and tracking using image recognition and triangulation2021Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Triangulation is used in a wide range of applications of position estimation. Usually it is donebymeasuring angles by hand to estimate positions in land surveying, navigation and astronomy.Withthe rise of image recognition arises the possibility to triangulate automatically. The aim ofthis thesisis to use the image recognition camera Pixy2 to triangulate a target i threedimensions. It is basedon previous projects on the topic to extend the system to estimatepositions over a larger spaceusing more Pixy2s. The setup used five Pixy2s with pan-tilt kitsand one Raspberry Pi 4 B. Somelimitations to the hardware was discovered, limiting the extentof the space in which triangulationcould be successfully performed. Furthermore, there weresome issues with the image recognitionalgorithm in the environment positioning was performed.The thesis was successful in that it managesto triangulate positions over a larger area thanprevious projects and in all three dimensions. Thesystem could also follow a target’s trajectory,albeit, there were some gaps in the measurements.

    Download full text (pdf)
    fulltext
  • 31.
    Brenner, Elvira
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Hultmar, Oscar
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Robust Drone Mission in the Arctic2020Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    During environmental research projects in the Arctic region AFRY has come across an unproportionally high number of cases where the navigation of drones have not worked as intended, compared to other regions. The main objective of this thesis is to investigate the cause of these navigational problems and determine the main cause. A second objective is to design a solution that can mitigate these observed errors and improve the navigation. To establish the main error sources flight logs from flight tests performed at Svalbard are analyzed. The drone considered in this project is a quadcopter with a Pixhawk Cube flight controller and the Ardupilot software. A Pixhawk Here+ module is used for external sensors. The data logs show several cases of drones having troubles flying along a straight line. Analyzing the sensor data for the flights show that many of the flights suffer from the gyroscope drifting around the z-axis. The data show that the varying temperature on the IMU board is the cause of the drifting gyroscope. The on-board heaters did not manage to keep the temperature constant due to a too high target temperature and low outside temperatures. The system is aided with information from the magnetometer to estimate the drift around the z-axis. Results show that the estimation system is having trouble correctly estimating large drifts. To investigate why the magnetometer cannot properly compensate for the gyroscope, simulations of the magnetometer and estimation system are made. The results show that an increasing angle of inclination increases the gyroscope bias estimation errors. The large angle of inclination causes the horizontal components of the magnetic field to become too small for the magnetometer to measure correctly. The solution consists of instructions on how to operate the drone to properly use the on-board heaters, as well as an external module consisting of multiple magnetometers. Multiple magnetometers reduced the variance in the readings, but did not reach the accuracy needed to replace the external magnetometers on the drone. A better calibration method could be explored in the future, or another solution such as an improved magnetometer, a gyro compass, a GPS compass or dual GNSS antennas.

    Download full text (pdf)
    fulltext
  • 32.
    Brunell, David
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Deep neural networks for food waste analysis and classification: Subtraction-based methods for the case of data scarcity2022Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Machine learning generally requires large amounts of data, however data is often limited. On the whole the amount of data needed grows with the complexity of the problem to be solved. Utilising transfer learning, data augmentation and problem reduction, acceptable performance can be achieved with limited data for a multitude of tasks.

    The goal of this master project is to develop an artificial neural network-based model for food waste analysis, an area in which large quantities of data is not yet readily available. Given two images an algorithm is expected to identify what has changed in the image, ignore the uncharged areas even though they might contain objects which can be classified and finally classify the change. The approach chosen in this project was to attempt to reduce the problem the machine learning algorithm has to solve by subtracting the images before they are handled by the neural network. In theory this should resolve both object localisation and filtering of uninteresting objects, which only leaves classification to the neural network. Such a procedure significantly simplifies the task to be resolved by the neural network, which results in reduced need for training data as well as keeping the process of gathering data relatively simple and fast.

    Several models were assessed and theories of adaptation of the neural network to this particular task were evaluated. Test accuracy of at best 78.9% was achieved with a limited dataset of about 1000 images with 10 different classes. This performance was accomplished by a siamese neural network based on VGG19 utilising triplet loss and training data using subtraction as a basis for ground truth mask creation, which was multiplied with the image containing the changed object.

    Download full text (pdf)
    fulltext
  • 33.
    Brynolf, Max
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Towards Real-Time Patient Monitoring: A Neural-Network based Decision Support System for Heart-Failure Prediction2021Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    Heart diseases are a significant challenge within the medical field, being the primary cause for premature death among adults in Sweden. Therefore, finding ways to diagnose patients as early as possible is of considerable interest. Accordingly, a decision support system has been developed which includes software functionality to upload data to a server for the patients as well as reading said patient data for the doctors. The decision support system also includes a machine learning (ML) model which predicts whether the patients have a heart disease.

    A decision support system for doctors and patients, based on earlier development in previous projects, consisting of an Android application, webserver and machine learning API was expanded with multiple features. The system is meant to implement the trained model and allow doctors and patients to use it for heart disease prediction. A machine learning model which predicted whether patients would get heart failure within the next 15 years based on multiple medical tests, was trained and evaluated. These medical tests were electrocardiography (ECG), different risk factors and variables extracted from x-ray images of the heart.

    Several functions were added to the system. These were an interface for doctors to handle their patients, the ability to upload ECG data from a Holter Monitor C3+ device, the ability for patients to make predictions from the Android application, local storage of measurement data, support for downloadable models along with a corresponding updating system, session-based user accounts and support for ECG data sent from Samsung Galaxy Watch 3 devices. All functions were observed to work as intended with minor issues left to future developers.

    A multilayer perceptron (MLP) network was trained to predict heart failure and its performance was compared to that of simpler logistic regression models. Different techniques against dataset imbalance such as random oversampling (ROS), synthetic minority oversampling technique (SMOTE) and weighted cost functions were used to see if they improved the classification performance. The performance of the MLP network was not significantly better than that of logistic regression, which is believed to be caused by the small number of samples along with the imbalance in the dataset. In spite of this, it was seen that including more medical tests improved the performance, even though a complex neural network was not required given the dataset at hand. Different oversampling techniques such as ROS and SMOTE did not improve the performance significantly for the neural network but did make a difference for logistic regression models.

    Overall, the system has been further developed to include features for both doctors and patients for real-time operation and interesting insight has been gained in regards to heart disease prediction using different machine learning methods. Future improvements should aim at making the system ready for production by finishing the account system and improving minor issues, as well as improving the classification performance. Examples are using different target diseases or exploring other types of machine learning models.

  • 34.
    Bushara, Adam
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Clyne, Sebastian Daniel
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Electrical Impedance Profiling of Plants, Coffee and TREO2022Independent thesis Basic level (professional degree), 10 credits / 15 HE creditsStudent thesis
    Abstract [en]

     The aim of the project was to investigate the possibility of using low-cost hardware to monitor plant health, measure the impedance of water containing dissolved substances at different concentrations, and observe how the plant impedance is affected compared to auxiliary measurements of soil moisture, ambient light, humidity, and temperature.To measure impedance, we used electrical impedance spectroscopy which gives the impedanceacross a spectrum of frequencies. This gives a bigger data set than just measuring resistanceand a broader information on the thing being measured.Studying the electrical impedance characteristics of the plant Rosemary was one of the work's specific application-related objectives. This includes the calibration of the impedance measurement system and how the parameters change impedance measurement. We saw a need for a minimum voltage of at least 200mV for input voltage for measuring impedance. The averaging parameter was insignificant and so was the sampling amount. As the measurement sample did not have a resonance tank the settling could be as low as 10ms.Open compensation gave significantly better measurement at higher frequencies over 100kHz as this accounts for the shunt admittance in the circuit. Short compensation gave no significant improvements.For the measurement of coffee and TREO at different concentrations, we found that the impedance profile changes over the frequency sweep.Finally, the measurement of the plant impedance as it changes over time did not produce useful data because of connection problem between the Impedance Analyzer and the plant probes. 

    Download full text (pdf)
    fulltext
  • 35.
    Carlgren, Jonathan
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Oskarsson, Per William
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    State Machine Model-To-Code Transformation In C2023Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    A state machine model can turn a complex behavioural system into a more accessible graphical model, and can improve the way people work with system design by making it easier to communicate and understand the system. The clear structure of a state machine model enables automatic generation of well structured, and consequently readable, and maintainable code. There are many known implementations and even plenty of commercial software available for working with state machines, and the goal of this project is to compare and discuss some of these implementations in the context of Ericsson's demands for run time, memory usage, scalability, readability, and maintainability.

    More specifically, the project focuses on the state machine models specified in the UML (the unified modeling language [15]), utilizing UML's associated markup language, XMI, to go from graphical model to generated C code. The resulting C code is primarily a code skeleton which only provides the basic behaviour of transitioning between states given a specific event, it is expected that the developers manually implement additional features themselves. The examined implementations are: Nested Switch, Array of Structs, Function Pointers, Basic State Pattern, State-Table Pattern, and Hierarchical State Pattern. Additionally, the project investigates how multiple state machines can communicate and work together as interacting state machines. And finally, to showcase how a state machine implementation can be maintainable, we develop an iterative code editor that can edit already operating and manually modified state machine implementations.

    The implementations are tested on a case study example provided by Ericsson, aimed to represent a sort of typical state machine design when it comes to number of states and events. The implementations are further tested with randomly generated state machines, to examine their scalability properties.

    In our opinion the results favour the Array of Structs and Basic State Pattern implementations and the choice depends on the optimisation used and the priority between run time and memory.

    Download full text (pdf)
    fulltext
  • 36.
    Cengiz, Heja
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Quadcopter Modeling and Linear Quadratic Regulator Design Using Simulink2024Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
    Abstract [en]

    This thesis project aims to model a quadcopter and design a linear quadratic regulator (LQR) by means of Matlab/Simulink. To this purpose, the LQR-based optimal control theory for controllinga quadcopter is first studied which includes state-space representation (SSR) of a dynamicprocess or system, cost function, LQR, quadcopter flight dynamics and system linearization. A quadcopter model is developed in Matlab/Simulink, followed by the implementation of a LQR-based control system. The LQR parameters are tuned and the system is tested under various flight conditions (wind disturbance, in the simulation, specific/simplified model, etc.).

    The simulation results show that the LQR is an effective controller for maintaining stable hover at a height straight up and compensating for wind disturbances. However, when the quadcopter moves to a new position, oscillations occur, highlighting the limitations of the LQR due to its reliance on a simplified and linearized model. Additionally, modifications to the model parameters, such as mass and inertia, impact the system performance, indicating potential robustness issues with the controller. It can be concluded that Matlab/Simulink is an effective tool for quadcopter modeling, LQR designing and LQR performance analyzing.

    In this thesis project, only the LQR method is used for controlling a quadcopter and the LQR tuning process is not efficient. In future work other techniques such as regional linearization and alternative non-linear controllers, like model predictive control (MPC) or sliding mode control (SMC), can be explored. Development of optimization algorithms for LQR tuning in the LQR method is highly recommended.

    Download full text (pdf)
    fulltext
  • 37.
    Chaki, Soumi
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Routray, Aurobinda
    Indian Inst Technol Kharagpur, Dept Elect Engn, Kharagpur 721302, West Bengal, India..
    Mohanty, William K.
    Indian Inst Technol Kharagpur, Dept Geol & Geophys, Kharagpur 721302, West Bengal, India..
    A probabilistic neural network (PNN) based framework for lithology classification using seismic attributes2022In: Journal of Applied Geophysics, ISSN 0926-9851, E-ISSN 1879-1859, Vol. 199, article id 104578Article in journal (Refereed)
    Abstract [en]

    This paper proposes a Probabilistic Neural Network (PNN) based framework for classification of lithology from a number of seismic attributes. The PNN has been the natural choice for classification in several research areas due to its insensitivities towards outliers and higher computational speed compared to multilayer perceptron (MLP) networks. Initially, the lithology is labelled into four classes such as sand, shaly sand, sandy shale, and shale through thorough analysis of multiple well logs by a proficient geologist. The seismic attributes and well logs pertaining to twelve closely spaced boreholes from a western onshore hydrocarbon field in India are used in this study. The performance of the designed framework consisting of preprocessing, classification, and lithological maps generation stages is compared with existing supervised classifiers in terms of classification accuracy, sensitivity, and specificity and the results are reported. The selection of appropriate parameters associated with individual classifier and importance of individual seismic predictors are also investigated. Finally, lithology maps indicating the different classes are produced using the tuned parameters of PNN over the study area. This framework would be of immense help to geologists along with other geological measures to estimate the probability of the presence of hydrocarbon in a large study area.

  • 38.
    Chantzi, Efthymia
    et al.
    Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Medical Sciences, Cancer Pharmacology and Computational Medicine. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Hammerling, Ulf
    Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Medical Sciences, Cancer Pharmacology and Computational Medicine.
    Gustafsson, Mats G
    Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Medical Sciences, Cancer Pharmacology and Computational Medicine. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Exhaustive in vitro evaluation of the 9-drug cocktail CUSP9 for treatment of glioblastoma using COMBImageDLIn: Molecular Cancer Therapeutics, ISSN 1535-7163, E-ISSN 1538-8514Article in journal (Refereed)
    Abstract [en]

    The CUSP9 protocol (aprepitant, auranofin, captopril, celecoxib, disulfiram, itraconazole, minocycline, quetiapine, sertraline) is currently undergoing a clinical trial as add-on treatment to standard-of-care temozolomide for recurrent glioblastoma. Although the theoretical repurposing rationale of this 9-drug cocktail is well defined, there is no in vitro experimental data yet supporting its superiority over all its plausible subsets. Such an exhaustive in vitro evaluation may provide preliminary evidence of whether only a fraction of all 9 drugs is needed to achieve an equivalent or even higher effect. Such information could be further used to guide and optimize individualized glioblastoma therapy selection both in terms of efficacy and adverse effects.

    Here, we employed COMBImageDL, a deep learning improved version of our recently developed COMBImage2 framework, to design, perform and analyze an exhaustive in vitro experiment of the CUSP9 protocol. More specifically, all 511 plausible subsets were evaluated as add-on treatment to temozolomide on a drug resistant glioblastoma cell line (M059K), by combining endpoint cell viability analysis and quantitative live-cell imaging. The experiment was performed in quadruplicate (eight 384-well plates, > 100GB of image data). Fixed clinically achievable concentrations were used for all drugs.

    Our results suggest that only disulfiram from the CUSP9 cocktail is required, together with temozolomide, in order to induce major changes in cell viability, confluence and morphology. Only slightly increased effects were observed by a few unique higher-order subsets of the CUSP9 protocol, which also contained disulfiram. This finding indicates that for the particular glioblastoma cell line used, the whole CUSP9 protocol could in principle be replaced solely with disulfiram. Notably, it may be worth testing in vitro the few slightly more potent higher-order subsets on primary patient derived glioblastoma cells. 

    This work demonstrates the feasibility and potential of performing exhaustive in vitro evaluation of higher-order drug cocktails prior to subsequent assessment for clinical use. Although the experimental in vitro disease models are not optimal, they can still pinpoint which among all plausible subsets should be further considered. From a personalized therapy selection perspective, in vitro sensitivity testing of primary patient derived tumor cells could thereby advance from the current practice based on single drugs and only cytotoxicity readouts to also include higher-order drug cocktails and quantitative live-cell imaging.

  • 39.
    Chantzi, Efthymia
    et al.
    Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Medical Sciences, Cancer Pharmacology and Computational Medicine. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Neidlin, Michael
    Technical University of Athens, Department of Mechanical Engineering.
    Macheras, George A.
    4th Orthopedic Department, General Hospital KAT, Athens, Greece.
    Alexopoulos, Leonidas G.
    Technical University of Athens, Department of Mechanical Engineering.
    Gustafsson, Mats G.
    Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Medical Sciences, Cancer Pharmacology and Computational Medicine. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    COMBSecretomics: a pragmatic methodological framework for higher-order drug combination analysis using secretomics2020In: PLOS ONE, E-ISSN 1932-6203, Vol. 15, no 5, article id e0232989Article in journal (Refereed)
    Abstract [en]

    Multi drug treatments are increasingly used in the clinic to combat complex and co-occurring diseases. However, most drug combination discovery efforts today are mainly focused on anticancer therapy and rarely examine the potential of using more than two drugs simultaneously. Moreover, there is currently no reported methodology for performing second- and higher-order drug combination analysis of secretomic patterns, meaning protein concentration profiles released by the cells.

    Here, we introduce COMBSecretomics (https://github.com/EffieChantzi/COMBSecretomics.git), the first pragmatic methodological framework designed to search exhaustively for second- and higher-order mixtures of candidate treatments that can modify, or even reverse malfunctioning secretomic patterns of human cells. This framework comes with two novel model-free combination analysis methods; a tailor-made generalization of the highest single agent principle and a data mining approach based on top-down hierarchical clustering. Quality control procedures to eliminate outliers and non-parametric statistics to quantify uncertainty in the results obtained are also included. COMBSecretomics is based on a standardized reproducible format and could be employed with any experimental platform that provides the required protein release data. Its practical use and functionality are demonstrated by means of a proof-of-principle pharmacological study related to cartilage degradation.

    COMBSecretomics is the first methodological framework reported to enable secretome-related second- and higher-order drug combination analysis. It could be used in drug discovery and development projects, clinical practice, as well as basic biological understanding of the largely unexplored changes in cell-cell communication that occurs due to disease and/or associated pharmacological treatment conditions.

    Download full text (pdf)
    COMBSecretomics - A pragmatic methodological framework for higher-order drug combination analysis using secretomics
  • 40.
    Chen, Libo
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Solid-State Electronics.
    Karilanova, Sanja
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Wen, Chenyu
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Solid-State Electronics.
    Wang, Lisha
    Karolinska Institutet.
    Winblad, Bengt
    Karolinska Institutet.
    Zhang, Shi-Li
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Solid-State Electronics.
    Özcelikkale, Ayca
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Zhang, Zhibin
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Solid-State Electronics.
    Spike timing–based coding in neuromimetic tactile system enables dynamic object classification2024In: Science, ISSN 0036-8075, Vol. 384, p. 660-665Article in journal (Refereed)
    Abstract [en]

    Rapid processing of tactile information is essential to human haptic exploration and dexterous object manipulation. Conventional electronic skins generate frames of tactile signals upon interaction with objects. Unfortunately, they are generally ill-suited for efficient coding of temporal information and rapid feature extraction. In this work, we report a neuromorphic tactile system that uses spike timing, especially the first-spike timing, to code dynamic tactile information about touch and grasp. This strategy enables the system to seamlessly code highly dynamic information with millisecond temporal resolution on par with the biological nervous system, yielding dynamic extraction of tactile features. Upon interaction with objects, the system rapidly classifies them in the initial phase of touch and grasp, thus paving the way to fast tactile feedback desired for neuro-robotics and neuro-prosthetics. 

  • 41.
    Chen, Xiaomeng
    et al.
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Kowloon, Hong Kong, Peoples R China..
    Huang, Lingying
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Kowloon, Hong Kong, Peoples R China..
    Ding, Kemi
    Southern Univ Sci & Technol, Sch Syst Design & Intelligent Mfg, Shenzhen Key Lab Biomimet Robot & Intelligent Syst, Shenzhen 518055, Peoples R China.;Southern Univ Sci & Technol, Guangdong Prov Key Lab Human Augmentat & Rehabil R, Shenzhen 518055, Peoples R China..
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Shi, Ling
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Kowloon, Hong Kong, Peoples R China..
    Privacy-Preserving Push-Sum Average Consensus via State Decomposition2023In: IEEE Transactions on Automatic Control, ISSN 0018-9286, E-ISSN 1558-2523, Vol. 68, no 12, p. 7974-7981Article in journal (Refereed)
    Abstract [en]

    Average consensus is extensively used in distributed networks for computation and control, where all the agents constantly communicate with each other and update their states in order to reach an agreement. Under a general average consensus algorithm, information exchanged through wireless or wired communication networks could lead to the disclosure of sensitive and private information. In this article, we propose a privacy-preserving push-sum approach for directed networks that can protect the privacy of all agents while achieving average consensus simultaneously. Each node decomposes its initial state arbitrarily into two substates, and their average equals to the initial state, guaranteeing that the agent's state will converge to the accurate average consensus. Only one substate is exchanged by the node with its neighbors over time, and the other one is reserved. That is to say, only the exchanged substate would be visible to an adversary, preventing the initial state information from leakage. Different from the existing state-decomposition approach, which only applies to undirected graphs, our proposed approach is applicable to strongly connected digraphs. In addition, in direct contrast to offset-adding-based privacy-preserving push-sum algorithm, which is vulnerable to an external eavesdropper, our proposed approach can ensure privacy against both an honest-but-curious node and an external eavesdropper. A numerical simulation is provided to illustrate the effectiveness of the proposed approach.

  • 42.
    Chen, Xiaomeng
    et al.
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Hong Kong, Peoples R China..
    Huang, Lingying
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Hong Kong, Peoples R China.;Nanyang Technol Univ, Sch Elect & Elect Engn, Singapore 639798, Singapore..
    He, Lidong
    Nanjing Univ Sci & Technol, Sch Automat, Nanjing, Peoples R China..
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Shi, Ling
    Hong Kong Univ Sci & Technol, Dept Elect & Comp Engn, Hong Kong, Peoples R China..
    A Differentially Private Method for Distributed Optimization in Directed Networks via State Decomposition2023In: IEEE Transactions on Control of Network Systems, E-ISSN 2325-5870, Vol. 10, no 4, p. 2165-2177Article in journal (Refereed)
    Abstract [en]

    In this article, we study the problem of consensus-based distributed optimization, where a network of agents, abstracted as a directed graph, aims to minimize the sum of all agents' cost functions collaboratively. In the existing distributed optimization approaches (Push-Pull/AB) for directed graphs, all the agents exchange their states with neighbors to achieve the optimal solution with a constant step size, which may lead to the disclosure of sensitive and private information. For privacy preservation, we propose a novel state-decomposition-based gradient tracking approach (SD-Push-Pull) for distributed optimization over directed networks that preserves differential privacy, which is a strong notion that protects agents' privacy against an adversary with arbitrary auxiliary information. The main idea of the proposed approach is to decompose the gradient state of each agent into two substates. Only one substate is exchanged by the agent with its neighbors over time, and the other one is not shared. That is to say, only one substate is visible to an adversary, protecting the sensitive information from being leaked. It is proved that under certain decomposition principles, a bound for the suboptimality of the proposed algorithm can be derived, and the differential privacy is achieved simultaneously. Moreover, the tradeoff between differential privacy and the optimization accuracy is also characterized. Finally, a numerical simulation is provided to illustrate the effectiveness of the proposed approach.

  • 43.
    Cherubini, Giovanni
    et al.
    IBM Res Zurich, CH-8803 Ruschlikon, Switzerland..
    Guay, Martin
    Queens Univ, Kingston, ON K7L 3N6, Canada..
    Tarbouriech, Sophie
    Univ Toulouse, CNRS, LAAS, F-31000 Toulouse, France..
    Ariyur, Kartik
    Purdue Univ, W Lafayette, IN 47907 USA..
    Broucke, Mireille E.
    Univ Toronto, Toronto, ON M5S, Canada..
    Dey, Subhrakanti
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Ebenbauer, Christian
    Univ Stuttgart, D-70174 Stuttgart, Germany..
    Frasca, Paolo
    CNRS, GIPSA Lab, F-38400 Grenoble, France..
    Gharesifard, Bahman
    Queens Univ, Kingston, ON K7L 3N6, Canada..
    Girard, Antoine
    CNRS, L2S Cent Supelec, F-91190 Gif Sur Yvette, France..
    Gomes da Silva, Joao Manoel
    Univ Fed Rio Grande do Sul, BR-90040060 Porto Alegre, RS, Brazil..
    Grune, Lars
    Univ Bayreuth, D-95447 Bayreuth, Germany..
    Kellett, Christopher M.
    Univ Newcastle, Callaghan, NSW 2308, Australia..
    Khan, Usman
    Tufts Univ, Medford, MA 02155 USA..
    Notarstefano, Giuseppe
    Univ Bologna, I-40126 Bologna, Italy..
    Scardovi, Luca
    Univ Toronto, Toronto, ON M5S, Canada..
    Vamvoudakis, Kyriakos G.
    Georgia Inst Technol, Atlanta, GA 30332 USA..
    Guest Editorial Introduction to the Special Issue of the IEEE L-CSS on Learning and Control2020In: IEEE Control Systems Letters, E-ISSN 2475-1456, Vol. 4, no 3, p. 710-712Article in journal (Refereed)
  • 44.
    Chockalingam, Sabarathinam
    et al.
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands.;Inst Energy Technol, Dept Risk Safety & Secur, Halden, Norway..
    Pieters, Wolter
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands.;Radboud Univ Nijmegen, Behav Sci Inst, Nijmegen, Netherlands..
    Teixeira, André
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    van Gelder, Pieter
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands..
    Bayesian network model to distinguish between intentional attacks and accidental technical failures: a case study of floodgates2021In: Cybersecurity, E-ISSN 2523-3246, Vol. 4, no 1, article id 29Article in journal (Refereed)
    Abstract [en]

    Water management infrastructures such as floodgates are critical and increasingly operated by Industrial Control Systems (ICS). These systems are becoming more connected to the internet, either directly or through the corporate networks. This makes them vulnerable to cyber-attacks. Abnormal behaviour in floodgates operated by ICS could be caused by both (intentional) attacks and (accidental) technical failures. When operators notice abnormal behaviour, they should be able to distinguish between those two causes to take appropriate measures, because for example replacing a sensor in case of intentional incorrect sensor measurements would be ineffective and would not block corresponding the attack vector. In the previous work, we developed the attack-failure distinguisher framework for constructing Bayesian Network (BN) models to enable operators to distinguish between those two causes, including the knowledge elicitation method to construct the directed acyclic graph and conditional probability tables of BN models. As a full case study of the attack-failure distinguisher framework, this paper presents a BN model constructed to distinguish between attacks and technical failures for the problem of incorrect sensor measurements in floodgates, addressing the problem of floodgate operators. We utilised experts who associate themselves with the safety and/or security community to construct the BN model and validate the qualitative part of constructed BN model. The constructed BN model is usable in water management infrastructures to distinguish between intentional attacks and accidental technical failures in case of incorrect sensor measurements. This could help to decide on appropriate response strategies and avoid further complications in case of incorrect sensor measurements.

    Download full text (pdf)
    FULLTEXT01
  • 45.
    Chockalingam, Sabarathinam
    et al.
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands.;Inst Energy Technol, Dept Risk & Secur, Halden, Norway..
    Pieters, Wolter
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands.;Radboud Univ Nijmegen, Behav Sci Inst, Nijmegen, Netherlands..
    Teixeira, André
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Automatic control. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Systems and Control.
    van Gelder, Pieter
    Delft Univ Technol, Fac Technol Policy & Management, Delft, Netherlands..
    Probability elicitation for Bayesian networks to distinguish between intentional attacks and accidental technical failures2023In: Journal of Information Security and Applications, ISSN 2214-2134, E-ISSN 2214-2126, Vol. 75, article id 103497Article in journal (Refereed)
    Abstract [en]

    Both intentional attacks and accidental technical failures can lead to abnormal behaviour in components of industrial control systems. In our previous work, we developed a framework for constructing Bayesian Network (BN) models to enable operators to distinguish between those two classes, including knowledge elicitation to construct the directed acyclic graph of BN models. In this paper, we add a systematic method for knowledge elicitation to construct the Conditional Probability Tables (CPTs) of BN models, thereby completing a holistic framework to distinguish between attacks and technical failures. In order to elicit reliable probabilities from experts, we need to reduce the workload of experts in probability elicitation by reducing the number of conditional probabilities to elicit and facilitating individual probability entry. We utilise DeMorgan models to reduce the number of conditional probabilities to elicit as they are suitable for modelling opposing influences i.e., combinations of influences that promote and inhibit the child event. To facilitate individual probability entry, we use probability scales with numerical and verbal anchors. We demonstrate the proposed approach using an example from the water management domain.

    Download full text (pdf)
    fulltext
  • 46.
    Coimbatore Anand, Sribalaji
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Risk-Based Analysis and Design of Secure Control Systems2024Doctoral thesis, comprehensive summary (Other academic)
    Abstract [en]

    Networked Control Systems (NCSs) are integral to many critical infrastructures such as power grids, transportation, and production systems. The resilient operation of such NCS against cyber-attacks is essential for society, and risk management presents an effective framework for addressing these security challenges. The risk management framework encompasses two steps: risk assessment and risk mitigation. The risk assessment step aims to quantify the risk, whereas the risk mitigation step focuses on designing mitigation strategies. This thesis leverages the risk management framework to analyze and design NCSs that are resilient to cyber-attacks. In particular, this thesis aims to address the following research challenges. 

    Firstly, we aim to assess the risk of attack scenarios that are realistic (risk assessment step). In particular, we consider adversaries and operators with different levels of knowledge about the NCS. For instance, an adversary or operator may possess complete knowledge of the system dynamics or have only partial knowledge with varying degrees of uncertainty. Hence, we describe a systematic approach to assess the risk considering the interplay between the knowledge levels of adversaries and operators.

    Secondly, we aim to design the NCS to minimize the risk of attacks (risk mitigation step). We explore three different strategies to minimize the risk: (a) controller/detector design, (b) security measure allocation, and (c) system architecture design. In the first strategy, we design the controller and detector gains to minimize the risk of attacks. Here, risk is characterized by the performance loss caused by stealthy attacks on the NCS. In the second strategy, we consider a distributed NCS where certain distributed devices can be secured from attacks by deploying secure sensors and actuators. Then, we aim to strategically determine the devices to secure and mitigate the risk of attacks effectively. Finally, inspired by digital watermarking, we explore the idea of introducing watermarks in NCS to detect attacks efficiently. Throughout the thesis, we provide various numerical examples to depict the efficacy of risk assessment and risk mitigation algorithms. We also provide numerous discussions and avenues for future research directions.

    List of papers
    1. Joint controller and detector design against data injection attacks on actuators
    Open this publication in new window or tab >>Joint controller and detector design against data injection attacks on actuators
    2020 (English)In: IFAC PapersOnline, Elsevier BV , 2020, Vol. 53, no 2, p. 7439-7445Conference paper, Published paper (Refereed)
    Abstract [en]

    This paper addresses the issue of data injection attacks on actuators in control systems. Considering attacks that aim at maximizing impact while remaining undetected, the paper revisits the recently proposed output-to-output gain, which is compared to classical sensitivity metrics such as H-infinity and H_. In its original formulation, the output-to-output gain is unbounded for strictly proper systems. This limitation is further investigated and addressed by modifying the performance output of the system and ensuring that the system from attack signal to performance output is also strictly proper. With this system description, and by using the theory of dissipative systems, a Bi-linear Matrix Inequality (BMI) is formulated for system design. Using this BMI, a design algorithm is proposed based on the heuristic of alternating minimization. Through numerical simulations of the proposed algorithm, it is found that the output-to-output gain presents advantages over the other metrics: the effect of the attack is reduced in the performance output and increased in the detection output in a relatively large spectrum of frequencies.

    Place, publisher, year, edition, pages
    Elsevier BV, 2020
    Keywords
    System security, Quadratic performance indices, Fault detection, H-infinity control, Optimization
    National Category
    Control Engineering Signal Processing
    Identifiers
    urn:nbn:se:uu:diva-447676 (URN)10.1016/j.ifacol.2020.12.1291 (DOI)000652593000483 ()
    Conference
    21st IFAC World Congress on Automatic Control - Meeting Societal Challenges, JUL 11-17, 2020, ELECTR NETWORK
    Funder
    Swedish Research Council, 2018-04396
    Available from: 2021-06-29 Created: 2021-06-29 Last updated: 2024-04-07Bibliographically approved
    2. Risk-averse controller design against data injection attacks on actuators for uncertain control systems
    Open this publication in new window or tab >>Risk-averse controller design against data injection attacks on actuators for uncertain control systems
    2022 (English)In: 2022 AMERICAN CONTROL CONFERENCE (ACC), IEEE, 2022, p. 5037-5042Conference paper, Published paper (Refereed)
    Abstract [en]

    In this paper, we consider the optimal controller design problem against data injection attacks on actuators for an uncertain control system. We consider attacks that aim at maximizing the attack impact while remaining stealthy in the finite horizon. To this end, we use the Conditional Value-at-Risk to characterize the risk associated with the impact of attacks. The worst-case attack impact is characterized using the recently proposed output-to-output l(2)-gain (OOG). We formulate the design problem and observe that it is non-convex and hard to solve. Using the framework of scenariobased optimization and a convex proxy for the OOG, we propose a convex optimization problem that approximately solves the design problem with probabilistic certificates. Finally, we illustrate the results through a numerical example.

    Place, publisher, year, edition, pages
    IEEE, 2022
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-488231 (URN)10.23919/ACC53348.2022.9867257 (DOI)000865458704099 ()978-1-6654-5196-3 (ISBN)
    Conference
    American Control Conference (ACC), JUN 08-10, 2022, Atlanta, GA
    Funder
    Swedish Research Council, 2018-04396Swedish Foundation for Strategic Research
    Available from: 2022-11-14 Created: 2022-11-14 Last updated: 2024-04-07Bibliographically approved
    3. Risk Assessment of Stealthy Attacks on Uncertain Control Systems
    Open this publication in new window or tab >>Risk Assessment of Stealthy Attacks on Uncertain Control Systems
    2023 (English)In: IEEE Transactions on Automatic Control, ISSN 0018-9286, E-ISSN 1558-2523, p. 1-8Article in journal (Refereed) Published
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-512904 (URN)10.1109/tac.2023.3318194 (DOI)
    Funder
    Swedish Research Council, 2018-04396
    Available from: 2023-09-29 Created: 2023-09-29 Last updated: 2024-04-07
    4. Risk assessment and optimal allocation of security measures under stealthy false data injection attacks
    Open this publication in new window or tab >>Risk assessment and optimal allocation of security measures under stealthy false data injection attacks
    2022 (English)In: 2022 IEEE Conference on Control Technology and Applications (CCTA), Institute of Electrical and Electronics Engineers (IEEE), 2022, p. 1347-1353Conference paper, Published paper (Refereed)
    Abstract [en]

    This paper firstly addresses the problem of risk assessment under false data injection attacks on uncertain control systems. We consider an adversary with complete system knowledge, injecting stealthy false data into an uncertain control system. We then use the Value-at-Risk to characterize the risk associated with the attack impact caused by the adversary. The worst-case attack impact is characterized by the recently proposed output-to-output gain. We observe that the risk assessment problem corresponds to an infinite non-convex robust optimization problem. To this end, we use dissipative system theory and the scenario approach to approximate the risk-assessment problem into a convex problem and also provide probabilistic certificates on approximation. Secondly, we con-sider the problem of security measure allocation. We consider an operator with a constraint on the security budget. Under this constraint, we propose an algorithm to optimally allocate the security measures using the calculated risk such that the resulting Value-at-risk is minimized. Finally, we illustrate the results through a numerical example. The numerical example also illustrates that the security allocation using the Value-at-risk, and the impact on the nominal system may have different outcomes: thereby depicting the benefit of using risk metrics.

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2022
    Series
    Control Technology and Applications, E-ISSN 2768-0770
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-510181 (URN)10.1109/CCTA49430.2022.9966025 (DOI)978-1-6654-7339-2 (ISBN)978-1-6654-7338-5 (ISBN)
    Conference
    IEEE Conference on Control Technology and Applications (CCTA), 23-25 August 2022, Trieste, Italy
    Available from: 2023-08-25 Created: 2023-08-25 Last updated: 2024-04-07Bibliographically approved
    5. Risk-based Security Measure Allocation Against Actuator Attacks
    Open this publication in new window or tab >>Risk-based Security Measure Allocation Against Actuator Attacks
    2023 (English)In: IEEE Open Journal of Control Systems, E-ISSN 2694-085X, p. 1-12Article in journal (Refereed) Published
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-510331 (URN)10.1109/ojcsys.2023.3305831 (DOI)
    Funder
    Swedish Research Council, 2018-04396
    Available from: 2023-08-28 Created: 2023-08-28 Last updated: 2024-04-07Bibliographically approved
    6. Privacy and Security in Network Controlled Systems via Dynamic Masking
    Open this publication in new window or tab >>Privacy and Security in Network Controlled Systems via Dynamic Masking
    2023 (English)In: IFAC-PapersOnLine, E-ISSN 2405-8963, Vol. 56, no 2, p. 991-996Article in journal (Refereed) Published
    Abstract [en]

    In this paper, we propose a new architecture to enhance the privacy and security of networked control systems against malicious adversaries. We consider an adversary which first learns the system using system identification techniques (privacy), and then performs a data injection attack (security). In particular, we consider an adversary conducting zero-dynamics attacks (ZDA) which maximizes the performance cost of the system whilst staying undetected. Using the proposed architecture, we show that it is possible to (i) introduce significant bias in the system estimates obtained by the adversary: thus providing privacy, and (ii) efficiently detect attacks when the adversary performs a ZDA using the identified system: thus providing security. Through numerical simulations, we illustrate the efficacy of the proposed architecture

    Place, publisher, year, edition, pages
    Elsevier, 2023
    Keywords
    Networked systems, Secure networked control systems, Linear systems, Privacy
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-525721 (URN)10.1016/j.ifacol.2023.10.1694 (DOI)001196708400157 ()
    Conference
    22nd IFAC World Congress, Yokohama, Japan, July 9-14, 2023
    Funder
    Swedish Research Council, 2018-04396Swedish Foundation for Strategic Research
    Available from: 2024-03-27 Created: 2024-03-27 Last updated: 2024-04-18Bibliographically approved
    7. Design of multiplicative watermarking against covert attacks
    Open this publication in new window or tab >>Design of multiplicative watermarking against covert attacks
    2021 (English)In: 2021 60th IEEE CConference On Decision And Control (CDC), Institute of Electrical and Electronics Engineers (IEEE), 2021, p. 4176-4181Conference paper, Published paper (Refereed)
    Abstract [en]

    This paper addresses the design of an active cyber-attack detection architecture based on multiplicative watermarking, allowing for detection of covert attacks. We propose an optimal design problem, relying on the so-called output-to-output l(2)-gain, which characterizes the maximum gain between the residual output of a detection scheme and some performance output. Although optimal, this control problem is non-convex. Hence, we propose an algorithm to design the watermarking filters by solving the problem suboptimally via LMIs. We show that, against covert attacks, the output-to-output l(2)-gain is unbounded without watermarking, and we provide a sufficient condition for boundedness in the presence of watermarks.

    Place, publisher, year, edition, pages
    Institute of Electrical and Electronics Engineers (IEEE), 2021
    Series
    IEEE Conference on Decision and Control, ISSN 0743-1546
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-475324 (URN)10.1109/CDC45484.2021.9683075 (DOI)000781990303112 ()978-1-6654-3659-5 (ISBN)
    Conference
    60th IEEE Conference on Decision and Control (CDC), DEC 13-17, 2021, ELECTR NETWORK
    Funder
    Swedish Research Council, 2018-04396Swedish Foundation for Strategic Research
    Available from: 2022-06-02 Created: 2022-06-02 Last updated: 2024-04-07Bibliographically approved
    8. Switching Multiplicative Watermark Design against Covert Attacks
    Open this publication in new window or tab >>Switching Multiplicative Watermark Design against Covert Attacks
    (English)In: Article in journal (Refereed) Submitted
    Place, publisher, year, edition, pages
    N/A:
    National Category
    Control Engineering
    Identifiers
    urn:nbn:se:uu:diva-525724 (URN)
    Available from: 2024-03-27 Created: 2024-03-27 Last updated: 2024-04-15Bibliographically approved
    Download full text (pdf)
    UUThesis_S-Coimbatore-Anand-2024
    Download (jpg)
    preview image
  • 47.
    Coimbatore Anand, Sribalaji
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems. Delft Center for Systems and Control, Delft University of Technology (TU Delft), Delft, Netherlands.
    Baldi, Simone
    School of Mathematics, Southeast University, Nanjing, China; Delft Center for Systems and Control, TU Delft, Delft, Netherlands.
    Optimal tracking strategies for uncertain ensembles of thermostatically controlled loads2020In: 2020 IEEE 16th International Conference on Control & Automation (ICCA), Institute of Electrical and Electronics Engineers (IEEE), 2020, p. 901-906Conference paper (Refereed)
    Abstract [en]

    Demand side energy management (DSEM) promises to regulate ensembles of loads to track desired power levels, in response to grid events (demand peaks, emergencies, variable renewable power generation, etc). A large fraction of such loads are Thermostatically Controlled Loads (TCLs) such as refrigerators, electric water heaters, and air conditioners. Such loads exhibit parametric uncertainty and heterogeneity which make power tracking difficult. Adaptive control strategies are explored in this work as a way to achieve power tracking. Effectiveness of such strategies are studied via numerical simulations.

  • 48.
    Coimbatore Anand, Sribalaji
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Teixeira, André
    Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Systems and Control. Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Automatic control.
    Risk-based Security Measure Allocation Against Actuator Attacks2023In: IEEE Open Journal of Control Systems, E-ISSN 2694-085X, p. 1-12Article in journal (Refereed)
  • 49.
    Coimbatore Anand, Sribalaji
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Teixeira, André M. H.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Joint controller and detector design against data injection attacks on actuators2020In: IFAC PapersOnline, Elsevier BV , 2020, Vol. 53, no 2, p. 7439-7445Conference paper (Refereed)
    Abstract [en]

    This paper addresses the issue of data injection attacks on actuators in control systems. Considering attacks that aim at maximizing impact while remaining undetected, the paper revisits the recently proposed output-to-output gain, which is compared to classical sensitivity metrics such as H-infinity and H_. In its original formulation, the output-to-output gain is unbounded for strictly proper systems. This limitation is further investigated and addressed by modifying the performance output of the system and ensuring that the system from attack signal to performance output is also strictly proper. With this system description, and by using the theory of dissipative systems, a Bi-linear Matrix Inequality (BMI) is formulated for system design. Using this BMI, a design algorithm is proposed based on the heuristic of alternating minimization. Through numerical simulations of the proposed algorithm, it is found that the output-to-output gain presents advantages over the other metrics: the effect of the attack is reduced in the performance output and increased in the detection output in a relatively large spectrum of frequencies.

    Download full text (pdf)
    fulltext
  • 50.
    Coimbatore Anand, Sribalaji
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Technology, Department of Electrical Engineering, Signals and Systems.
    Teixeira, André M. H.
    Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Systems and Control. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Automatic control.
    Risk-averse controller design against data injection attacks on actuators for uncertain control systems2022In: 2022 AMERICAN CONTROL CONFERENCE (ACC), IEEE, 2022, p. 5037-5042Conference paper (Refereed)
    Abstract [en]

    In this paper, we consider the optimal controller design problem against data injection attacks on actuators for an uncertain control system. We consider attacks that aim at maximizing the attack impact while remaining stealthy in the finite horizon. To this end, we use the Conditional Value-at-Risk to characterize the risk associated with the impact of attacks. The worst-case attack impact is characterized using the recently proposed output-to-output l(2)-gain (OOG). We formulate the design problem and observe that it is non-convex and hard to solve. Using the framework of scenariobased optimization and a convex proxy for the OOG, we propose a convex optimization problem that approximately solves the design problem with probabilistic certificates. Finally, we illustrate the results through a numerical example.

12345 1 - 50 of 225
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf