Probabilistic estimation of achievable maximum throughput from wireless interface

Abstract

A maximum achievable bit rate estimation for the expected bit rate between a wireless client and an access point is calculated in two phases. In the first phase, the maximum achievable bit rate is estimated using signal quality information and optionally congestion level information of the access point(s) of interest. In a second phase of the bit rate estimation, the first phase estimator output is corrected by using historical information relating the estimated bit rate values to the bit rates actually experienced. In the second phase of the estimation, a repository stores known signal quality behavior of the access points so as to provide an Access Point History (APH). The APH can be implemented as a simple memory (database) entity where a record is maintained on how accurate the previous bit-rate estimations turned out to be in practice. The actual physical embodiment of APH memory is implemented using any adequate memory technology such as FLASH, local hard disk, and the like. The APH is an ordered list where Access Point ID (SSID) and MAC address are combined with the data in an experimentally known relation between the SNR and the stable throughput. This two phase technique thus provides the implicit capability to include a rough bias to throughput estimation by congestion level of the access point as well as by historical bit rate averages.