On the use of learning automata in medium access control of single-hop lightwave networks

A Self-Adaptive Protocol for WDM Passive Star Networks, which is capable of operating efficiently under bursty and correlated traffic, is introduced. According to the proposed protocol, the stations that grant permission to transmit at each time slot are selected by means of learning automata. The choice probabilities of the selected stations are updated by taking into account the network feedback information. The system which consists of the automata and the network is analyzed and it is proved that the probability that a station grants permission to transmit on a specific wavelength, asymptotically tends to be proportional to the probability that this station has one or more packets to transmit on this wavelength. Therefore, although the traffic parameters are unknown and time-variable, the bandwidth of each wavelength is allocated to the stations according to their needs. In this way, the number of idle slots is reduced, resulting in a significant increase of the network throughput. Furthermore, extensive simulation results are presented which indicate that the proposed protocol achieves a significantly higher performance than other well-known protocols.