In IEEE 802.11 standard, the backoff parameters of its collision avoidance mechanism can be very inefficient and hence, the network becomes far from its optimal behavior. There have been several mechanisms to tune the Contention Window (CW) with the aim to achieve the optimal throughput in the IEEE 802.11 WLAN, however, the mechanisms do not specifically address a proper setting of the backoff parameters under non-saturated conditions. Noting that typical 802.11 networks are usually non-saturated, in this paper, we analytically derive the CW sizes that maximize the WLAN system throughput under both saturated and non-saturated conditions. Then, using the CW sizes derived, we propose a distributed algorithm that enables each station to dynamically adapt its CW according to the channel congestion status. The performance of the proposed algorithm is investigated through simulation. Simulation results indicate that our proposed backoff algorithm provides a remarkable performance improvement in terms of the delay experienced by a packet in the MAC layer, while maintaining an optimal throughput close to the theoretical throughput limit of the IEEE 802.11 Distributed Coordination Function (DCF) access scheme.

In an integrated WLAN and cellular network, if all mobile users whose connections originate in the cellular network migrate to the WLAN whenever they enter the double coverage area, the WLAN will be severely congested and its users will suffer from performance degradation. Therefore, we propose a Call Admission Control (CAC) algorithm that allows the WLAN to limit downward Vertical Handovers (VHOs) from the cellular network to reduce unnecessary VHO processing. Numerical and simulation results demonstrate that our CAC scheme reduces the unnecessary VHO processing while keeping the DVHO blocking rate within acceptable limits and maintaining reasonable throughput in the WLAN.

As the number of WiFi-enabled devices in IEEE 802.11 networks becomes large, existing paging support over link or IP layer becomes insufficient to support mobility across these networks. For more efficient support, we propose an IP paging scheme in this paper. The proposed scheme is an extension of existing mobile IP paging, adopting a two-step overlapping paging mechanism. In addition, the size of paging areas is designed to be determined based on the speed of a mobile node. Thorough analysis and simulation demonstrate that our proposal can reduce the registration cost for paging area configuration, overall signaling cost in the network and hence the power consumption of mobile nodes significantly.