Self-adaptive polling protocols for wireless LANs: a learning-automata-based approach

A learning automata-based polling (LEAP) protocol for infrastructure wireless LANs (WLANs), which is capable of operating efficiently under bursty traffic conditions, is proposed. In LEAP, the mobile station that grants permission to transmit is selected by the base station by means of a learning automaton. It is proved that the learning algorithm asymptotically tends to assign to each station a portion of the bandwidth proportional to the station's needs. Simulation results reveal superiority of LEAP over more complex to implement polling protocols for WLANs (RAP, GRAP) under bursty traffic conditions.