Adaptive MAC Scheduling Using Channel State Diversity for Wireless Networks

Head-of-line blocking problem compromises the throughput of multi-hop wireless networks. FIFO scheduling in the current IEEE 802.11 MAC causes this problem when the network is highly loaded. One solution is to increase the RTS success rate by extending the RTS frame to MRTS (multicast RTS), so that multiple receivers could be checked simultaneously. There is a tradeoff for the length of the MRTS-frame receiver list, since longer lists increase transmission success rate, but shorter lists impose lower control overhead. We present an adaptive learning process that observes the dynamic channel-state diversity among the candidate receivers. By maximizing the receiver diversity, we can achieve high transmission success rate using short receiver lists. This is supported by our simulation results