Transmission Schemes for Gaussian Interference Channels with Transmitter Processing Energy

TL;DR

This paper investigates how processing energy costs influence transmission strategies in Gaussian interference channels, proposing schemes that optimize sum rates and identifying conditions where interference becomes non-penalizing, with extensions to multi-user scenarios.

Contribution

It introduces transmission schemes considering processing energy costs and analyzes their impact on interference regimes and multi-user extensions.

Findings

Transmitting all the time is suboptimal with processing energy costs.

The very strong interference regime expands under processing energy considerations.

Scheduling based on channel setup improves performance in multi-user channels.

Abstract

This work considers communication over Gaussian interference channels with processing energy cost, which explicitly takes into account the energy expended for processing when transmitters are on. In the presence of processing energy cost, transmitting all the time as in the conventional no-cost case is no longer optimal. For a two-user Gaussian interference channel with processing energy cost, assuming that the on-off states of transmitters are not utilized for signaling, several transmission schemes with varying complexities are proposed and their sum rates are compared with an interference-free upper bound. Moreover, the very strong interference regime, under which interference does not incur any rate penalty, is identified and shown to be larger than the case of no processing energy cost for certain scenarios of interest. Also, extensions to a three-user cascade Gaussian Z interference channel with processing energy cost are provided, where scheduling of user transmissions based on the channel set-up is investigated.