In CSMA-based wireless multi-hop networks, consecutive packets on a single multi-hop path interfere with each other, causing hidden/exposed terminal problems and frequent contention. This phenomenon is called <i>intra-flow interference</i> and is considered as one of the principal factors degrading the throughput performance, since it increases the number of collisions as well as backoff intervals. Here, we propose a probabilistic packet dropping algorithm that mitigates the intra-flow interference on wireless multi-hop networks. In the proposed algorithm, a wireless node classifies all flows into several <i>1-hop forwarding domains</i>, which are determined by the next relay node. For this classification, each node regards the flows that will pass the same 1-hop neighbor as ones belonging to the same forwarding domain. Each node probabilistically drops incoming packets if the <i>inter-arrival time</i> of incoming packets belonging to the same forwarding domain is so short as to make intra-flow interference. Through extensive simulation studies, we show that the proposed algorithm effectively reduce the intra-flow interference, and increases the overall throughput as well as fairness on wireless multi-hop networks.
Unfortunately, ACM prohibits us from displaying non-influential references for this paper.
To see the full reference list, please visit http://dl.acm.org/citation.cfm?id=1853087.