Estimator Goore Game based quality of service control with incomplete information for wireless sensor networks

Quality of Service (QoS) control is one of the key mechanisms to ensure the effective operation of wireless sensor networks (WSNs). In many specific applications such as battlefield communications, the optimum number of sensors sending information at any given time is an important QoS requirement. For such requirement, some Goore Game based QoS control approaches have been exploited to dynamically adjust the number of active sensors to the optimal one in the literatures, but they assume that the optimum number is known to the base station explicitly, and each method has its own shortcomings. In this paper, a novel approach is proposed to solve the QoS control problem. A framework for solving such problem is firstly proposed, in which the optimum sensor number is not known in advance but learnt by interacting with upper level applications in a reinforcement manner. On this basis, an algorithm named Estimator Goore Game (EGG) is designed to adjust the number of active sensors effectively. The proposed algorithm cannot only relax the above assumption, but also has a competitive performance. Furthermore, the algorithm is also applicable for both stationary and non-stationary environments. Simulation results on the convergence time, the network life and the ability of tracking dynamic QoS have demonstrated the effectiveness of the proposed algorithm. (C) 2015 Elsevier B.V. All rights reserved.