Cross-layer resource allocation for wireless distributed computing networks

Abstract

The allocation of communication power consumption and computing rate is inherently one of the cross-layer problems in wireless distributed computing networks (WDCNs). This paper exploits a subgradient approach to choose the optimal power-rate pair for maximizing the network computing capability and minimizing the power consumption. The impact of heterogeneous channel conditions for different nodes are modeled as a penalty term in the objective function to balance the performance and power consumption, and increase the amount of the power for the computing task. Simulation results show that both energy efficiency and the distributed computing power ratio (DCPR) can be increased by considering the wireless channel conditions. These results provide some guidelines for designing the protocol dedicated to WDCNs.