Intelligent Channel Assignment in Cognitive Wireless Sensor Networks: Learning from Virtual Channel Environment

Abstract

One of the crucial challenges experienced by Wireless Sensor Networks (WSNs) is the increasing number of devices sharing the limited spectrum of the ISM band. As a result, a new concept of Cognitive Wireless Sensor Networks (CWSNs) has been proposed in order to achieve reliable and efficient communication via spectrum awareness and intelligent adaption. Based on such concept, this paper proposes novel machine-learning technique to provide an intelligent channel management for WSNs. The proposed method is based on the learning and prediction technique so called finite-state Partially Observable Markov Decision Process (POMDP) together with virtual channel environment classification. In comparison with traditional learning techniques, our proposed method offers reduced reaction time in respond to environmental changes. Simulation result shows that the proposed method provides a quick recovery to the interference that affects the system, hence maintaining better system throughput.