Playing ‘games’ with human health the role of game theory in optimizing reliability in wireless health networks

Abstract

Current advances in wireless sensor networks have contributed to the development of multi-hop wireless body sensor networks (WBSNs). These networks, which are primarily comprised of low-power sensor nodes, provide long term health and physiological monitoring of patients without obstructing their normal activities. However, the time-varying nature of network resources poses a significant challenge to the high reliability of operations required of a body sensor network. In this paper, we address the problem of reliability optimization in multi-hop WBSNs, where the nodes co-operate with each other in maximizing system performance. We model our problem as a co-operative game which inherently induces a distributed mechanism that optimizes the reliability of network operations in a multi-hop WBSN supporting concurrent applications. By the term ‘reliability’, we imply the system utility generated when each sensor node (user) services its specific application. We denote the system utility as an instance of the vector of utilities of individual users, where each user's utility is a function of its allocated bandwidth. Our mechanism jointly accounts for the dynamistic nature of the wireless medium, application quality of service, user fairness in achieving the resources to service applications, and results in an optimal operating point (utility vector) at which the WBSN should operate. We arrive at the result that the Nash bargaining solution of the co-operative game gives us the optimal system operating point, i.e., the Nash bargaining solution optimizes the overall system reliability.