Adaptive wireless networks using learning automata

Abstract

Wireless networks operate in environments with unknown and time-varying characteristics. The changing nature of many of these characteristics will significantly affect network performance. This fact has a profound impact on the design of efficient protocols for wireless networks and as a result adaptivity arises as one of the most important properties of these protocols. Learning automata are artificial intelligence tools that have been used in many areas where adaptivity to the characteristics of the wireless environment can result in a significant increase in network performance. This article reviews state of the art approaches in using learning automata to provide adaptivity to wireless networking.