Efficient multipolling and link adaptation for WLAN

Abstract

Throughput and QoS are important issues in WLAN. In this paper, we present the Efficient Multipolling Strategy(EMS) for throughput and QoS enhancement. The key idea is that Hybrid Coordinator(HC) gathers the stations`` information such as traffic type, pending data size during the contention period(CP) and schedules the polling order. The HC transmits multipoll frame which informs the stations regarding the polling schedule during the CFP uplink phase. The EMS is compared against basic Single Polling(SP) and Two Step Multipolling(TS-MP). The EMS shows a maximum 47% improvement against SP and a 5% performance improvement against TS-MP under normal ack mode, and an improvement of 4.8% and 10% as compared to SP and TS-MP under block ack mode in given simulation environment. Drop rate of the EMS is only 1% higher than TS-MP. Energy efficiency on wireless communication is an important issue of the battery power based wireless communication devices. This chapter presents the energy efficiency model for MIMO OFDM WLAN system with the block ack mechanism and also presents the analysis of the system. Based on the analysis, optimal PHY mode table and optimal transmit power table are provided. Several parameters are considered such as number of antennas, modulation, path loss, payload size and number of burst transmission in block ack mechanism. Our experiment results show that energy efficiency is higher at the lower number of antennas, higher modulation, larger data payload size and burst transmission environment. Comparison between using optimal PHY mode table and transmit power table and using 15dBm fixed transmit power case shows that optimal table case is 59% higher energy efficiency than fixed transmit power case. Goodput analysis considers the PHY mode and MAC procedure. We analyze goodput performance of MIMO OFDM WLAN system which uses block ack mechanism and provides optimal PHY mode table based on goodput analysis. Naturally, goodput performance shows hig...