Cognitive relaying in wireless sensor networks: Performance analysis and optimization

Abstract

The anticipated increase in the density of the deployed wireless sensor networks calls for spectrum sharing through unlicensed access to licensed spectrum. The key technology for spectrum sharing in this scenario is cognitive radio networks. Cognitive relaying scenarios, where a cognitive (unlicensed) user provides relaying services to a licensed (primary) user, have been proposed before as a method to increase chances of spectrum white spaces. In this paper, we develop a wireless sensor network framework containing a cognitive user (sensor node), with delay sensitive data and limited power budget. The cognitive user offers relaying capability to the primary traffic when the primary connection fails to deliver. The cognitive user utilizes a scheduling mechanism that grants priority to relayed traffic over its own traffic. In our framework, the cognitive user is allowed to control the volume of the relayed traffic through an admission control parameter. The objective of the sensor node (cognitive user) is to minimize its traffic delay, subject to certain power budget allowed for relaying the primary traffic. Our key contributions in this work are the development of the aforementioned framework, the establishment of a mathematical formalization for this problem, the derivation of mathematical expressions for average power consumption and average packet delay, and the solution for the developed optimization problem by finding a value for the admission control parameter that minimizes delay while satisfying power budget constraints.