Topology design and cross-layer optimization for wireless body sensor networks

This article proposes a mathematical optimization problem that jointly considers network topology design and cross-layer optimization in WBSNs, and introduces multilevel primal and dual decomposition methods to solve the proposed non-convex mixed-integer optimization problem.

Abstract

Wireless Body Sensor Networks play a crucial role in digital health care nowadays. Due to the size limitation on the sensor nodes and the life critical characteristics of the signals, there are stringent requirements on network's reliability and energy efficiency. In this article, we propose a mathematical optimization problem that jointly considers network topology design and cross-layer optimization in WBSNs. We introduce multilevel primal and dual decomposition methods and manage to solve the proposed non-convex mixed-integer optimization problem. A solution with fast convergence rate based on binary search is provided. Simulation results have been supplemented to show that our proposed method yields much better performance than existing solutions.