Concurrent MAC protocols can improve channel usage of wireless sensor networks (WSNs), and provide a high-performance infrastructure for data intensive applications. Most of the existing concurrent MAC protocols are based on proactively constructed physical interference models, i.e., PRR-SINR models (PSM). However, it incurs relatively high bandwidth and energy overheads to construct PSM for WSNs. In this paper, we propose NoPSM, which does not take PSM as base to determine transmission concurrency. Instead, the base of NoPSM is reactively constructed interference relationships by passively analyzing overlapping relationships among time logs of block data transmissions and corresponding reception status of each packet in blocks. In this way, NoPSM has two salient features. First, NoPSM is able to construct interference relationships among nodes quickly and accurately along with block data transmissions without needs of network downtime. Second, based on the constructed interference relationships, NoPSM can make decisions of transmission concurrency with a comprehensive criterion, which not only estimates quality of any active links after initiating a new link, but also estimates throughput improvement gained from concurrent transmissions. NoPSM has been implemented in Tinyos-2.1 and extensively evaluated in TOSSIM. Experimental results show that NoPSM improves system throughput by up to 60 percent compared with a traditional CSMA protocol, which cannot exploit potential transmission concurrency. Moreover, NoPSM can gain up to 55 percent throughput improvement as compared to an existing reactive concurrent MAC.

Additional Metadata
Keywords concurrent transmission, data intensive application, interference relationship, Wireless sensor network
Persistent URL dx.doi.org/10.1109/TMC.2016.2547867
Journal IEEE Transactions on Mobile Computing
Citation
Chen, H. (Haiming), Zhang, Z. (Zhaoliang), Cui, L. (Li), & Huang, C. (2017). NoPSM: A Concurrent MAC Protocol over Low-Data-Rate Low-Power Wireless Channel without PRR-SINR Model. IEEE Transactions on Mobile Computing, 16(2), 435–452. doi:10.1109/TMC.2016.2547867