RBSRA: RSSI based Statistical Rate Adaptation for IEEE 802.11 WLANs

Abstract

In recent research, estimations for condition of channels are based on either (i) physical layer parameters (e.g. SNR, RSSI etc.) or (ii) statistics of transmitted data (e.g. no of retransmissions, throughput etc.). Our research evaluates that depending upon one approach entirely while ignoring other can lead to loss of valuable information about adaptation of rate to send data. This in turn can highly degrade throughput. In view of this, the proposed scheme, RBSRA, defines channel assessment as the function of combining PHY layer parameter with statistical metrics. RBSRA estimates accurate assessment of fluctuating channel conditions, so it leads to the substitution of data rates to yield higher throughput. RBSRA uses, the PHY layer parameter RSSI in cooperation with the statistical metric that is number of received ACK, in such a way that it avoids unnecessary fluctuations in data rates and provides more realistic selection amongst the available link speeds. Performance of RBSRA depicts that the data rate selection has been precisely synchronized with ups and downs of the channel conditions along with statistics of transmitted stream.