Low-Power Listening Goes Multi-Channel
2014 (English)In: 2014 IEEE INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTING IN SENSOR SYSTEMS (IEEE DCOSS 2014), 2014, 2-9 p.Conference paper (Refereed)
Exploiting multiple radio channels for communicationhas been long known as a practical way to mitigateinterference in wireless settings. In Wireless Sensor Networks,however, multichannel solutions have not reached their fullpotential: the MAC layers included in TinyOS or the ContikiOS for example are mostly single-channel. The literature offersa number of interesting solutions, but experimental results wereoften too few to build confidence. We propose a practical extensionof low-power listening, MiCMAC, that performs channel hopping,operates in a distributed way, and is independent of upper layersof the protocol stack. The above properties make it easy todeploy in a variety of scenarios, without any extra configuration/scheduling/channelselection hassle. We implement our solutionin Contiki and evaluate it in a 97-node testbed while runninga complete, out-of-the-box low-power IPv6 communication stack(UDP/RPL/6LoWPAN). Our experimental results demonstrateincreased resilience to emulated WiFi interference (e.g., data yieldkept above 90% when ContikiMAC drops in the 40% range). In noiseless environments, MiCMAC keeps the overhead low incomparison to ContikiMAC, achieving performance as high as 99% data yield along with sub-percent duty cycle and sub-secondlatency for a 1-minute inter-packet interval data collection.
Place, publisher, year, edition, pages
2014. 2-9 p.
, IEEE International Conference on Distributed Computing in Sensor Systems
Wireless Sensor Networks
Research subject Computer Science with specialization in Computer Communication; Computer Science with specialization in Embedded Systems; Computer Science with specialization in Real Time Systems
IdentifiersURN: urn:nbn:se:uu:diva-237685DOI: 10.1109/DCOSS.2014.33ISI: 000361020100002OAI: oai:DiVA.org:uu-237685DiVA: diva2:768451
9th IEEE International Conference on Distributed Computing in Sensor Systems (DCOSS), MAY 26-28, 2014, Marina Del Rey, CA