Behavioral learning of exposed terminals in IEEE 802.11 wireless networks

In this paper, we tackle issues on the exposed terminal problem in IEEE 802.11 based wireless networks from a simulative perspective by investigating the internal relationship of node transmission behaviors under various network scenarios. The proposed solution is designed to learn node behaviors according to incoming traffic patterns and to assist wireless nodes make accurate transmission judgment to avoid exposed terminal problems as much as possible. Extensive simulations using ns-2 have been performed to validate the proposed scheme by varying system parameters including carrier sense range and hop counters. Our experimental outcomes show that the proposed solution yields performance gains in terms of aggregate throughput in excess of 10% to 15% compared to the standard.