Power Efficiency for Device-to-Device Communications

TL;DR

This paper models device-to-Device (D2D) communication networks using stochastic geometry, proposing power control schemes to minimize power consumption while maintaining performance, with dependent control significantly reducing power use.

Contribution

It introduces a stochastic geometry model for D2D networks and develops both independent and dependent power control schemes, including a closed-form solution for the former and an efficient method for the latter.

Findings

Dependent power control saves about 50% of power compared to independent control.

Closed-form expression derived for optimal independent power control.

Proposed dependent control adapts to channel conditions for improved efficiency.

Abstract

The concept of device-to-Device (D2D) communication as an underlay coexistence with cellular networks gains many advantages of improving system performance. In this paper, we model such a two-layer heterogenous network based on stochastic geometry approach. We aim at minimizing the expected power consumption of the D2D layer while satisfying the outage performance of both D2D layer and cellular layer. We consider two kinds of power control schemes. The first one is referred as to independent power control where the transmit powers are statistically independent of the networks and all channel conditions. The second is named as dependent power control where the transmit power of each user is dependent on its own channel condition. A closed-form expression of optimal independent power control is derived, and we point out that the optimal power control for this case is fixed and not relevant to the randomness of the network. For the dependent power control case, we propose an efficient way to find the close-to-optimal solution for the power-efficiency optimization problem. Numerical results show that dependent power control scheme saves about half of power that the independent power control scheme demands.