Range estimation and performance optimization for IEEE 802.11 based on filter
Hui Ma,Xing Li,Hewu Li,Peiyun Zhang,Shixin Luo,Cong Yuan
DOI: https://doi.org/10.1109/WCNC.2004.1311660
2004-01-01
Abstract:The dynamic character of complex and variable network environment, the key problem, puzzles the corresponding protocols design for wireless LANs. However, the distributed coordination function (DCF) of IEEE 802.11 MAC protocol, which is carrier sense multiple access with collision avoidance (CSMA/CA) using constant parameters, could not perform well when network environment changes. Thus many researchers try to optimize IEEE 802.11 DCF. However early dynamic optimization mechanisms for the protocol mostly depend on measuring the number of "competing" stations accurately. The problem of them is that the algorithms are too complex to apply in reality. In our research, we find that system performance approaches optimization with the same protocol parameters, when the number of competing stations changes within a certain range dynamically. Therefore, we propose a self-adaptive optimization mechanism, DOOR (dynamic optimization on range), for the IEEE 802.11 DCF. DOOR uses filter to estimate the range of competing station number and adjusts the protocol parameters to optimize system performance effectively. The detailed analytical model and performance evaluation for the new mechanism are given. Moreover, the measurement method and parameters of filter are introduced. At last, the elaborate numerical results show that our mechanism could not only achieve much higher throughput and lower delay than the standard IEEE 802.11 DCF along with the change of competing stations, but also improve the stability of system performance based on reasonably partitioned ranges.