Issue No. 03 - March (2008 vol. 7)
The limited number of orthogonal channels andautonomous installations of hotspots and home wireless networksoften leave neighboring 802.11 basic service sets (BSSs) operatingon the same or overlapping channels, therefore interferingwith each other. However, the 802.11 MAC does not workwell in resolving inter-BSS interference due to the well-knownhidden/exposed receiver problem, which has been haunting inthe research community for more than a decade. In this paperwe propose SELECT, an effective and efficient self-learning collisionavoidance strategy to address the hidden/exposed receiverproblem in 802.11 wireless networks. SELECT is based on theobservation that carrier sense with received signal strength (RSS)measurements at the sender and the receiver can be stronglycorrelated. A SELECT-enabled sender exploits such correlationusing automated on-line learning algorithm, and makes informedjudgment of the channel availability at the intended receiver. SELECTachieves collision avoidance at packet-level time granularity,involves zero communication overhead, and easily integrateswith the 802.11 DCF. Our evaluation in analysis, simulations,and prototype experiments show that SELECT addresses thehidden/exposed receiver problem well. In typical hidden/exposedreceiver scenarios SELECT improves the throughput by up to140% and channel access success ratio by up to 302%, whilealmost completely eliminating contention-induced data packetdrops.
Wireless communication, Access schemes, Algorithm/protocol design and analysis
Haiyun Luo, Hwangnam Kim, Chun-cheng Chen, Eunsoo Seo, "SELECT: Self-Learning Collision Avoidance for Wireless Networks", IEEE Transactions on Mobile Computing, vol. 7, no. , pp. 305-321, March 2008, doi:10.1109/TMC.2007.70723