[Paper Review] Energy-Efficient Communication over Gaussian Interference Networks with Processing Energy Cost

Xi Liu, Elza Erkip|arXiv (Cornell University)|Jan 8, 2013

Wireless Communication Security Techniques13 references3 citations

TL;DR

This paper proposes energy-efficient transmission schemes for Gaussian interference networks by accounting for processing energy costs at transmitters, which invalidates the conventional 'always-on' transmission strategy. It identifies a larger very strong interference regime than previously known and extends analysis to a three-user cascade Z-channel with interference-aware scheduling, showing that optimal sum-rates require dynamic on-off control based on interference geometry.

ABSTRACT

This work considers communication over Gaussian interference networks 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. Also, extensions to a three-user cascade Gaussian Z interference channel with processing energy cost are provided, where scheduling of user transmissions based on interference geometry is investigated.

Motivation & Objective

• To address the inefficiency of always-on transmission in interference networks when processing energy is non-negligible.
• To model and analyze the impact of transmitter processing energy cost on achievable sum-rates in Gaussian interference channels.
• To identify conditions under which interference does not incur rate penalties, particularly in the very strong interference regime.
• To extend the analysis to multi-user scenarios, specifically a three-user cascade Gaussian Z interference channel.
• To develop scheduling strategies based on interference geometry to optimize energy efficiency and sum-rate.

Proposed method

• Proposes multiple transmission schemes with varying complexity for a two-user Gaussian interference channel, including constant-power and on-off strategies.
• Introduces a processing energy cost model where energy is consumed whenever a transmitter is active, not just during transmission.
• Derives an interference-free upper bound to evaluate the performance of proposed schemes.
• Identifies the very strong interference regime by analyzing conditions under which interference does not reduce sum-rate.
• Extends the framework to a three-user cascade Gaussian Z-channel, modeling interference geometry to guide user scheduling.
• Uses interference alignment and power control principles to optimize sum-rate under energy constraints.

Experimental results

Research questions

• RQ1 How does processing energy cost affect the optimality of always-on transmission in Gaussian interference networks?
• RQ2 What is the size of the very strong interference regime when processing energy cost is considered, and how does it compare to the conventional case?
• RQ3 Can dynamic on-off scheduling based on interference geometry improve sum-rate and energy efficiency in multi-user interference channels?
• RQ4 How does the presence of processing energy cost alter the fundamental trade-off between spectral efficiency and energy consumption?
• RQ5 What scheduling policies maximize sum-rate in a three-user cascade Gaussian Z-channel under processing energy constraints?

Key findings

• The very strong interference regime, where interference causes no rate penalty, is larger than in the conventional no-processing-cost model.
• Transmitting all the time is no longer optimal when processing energy cost is considered, necessitating dynamic on-off strategies.
• The proposed schemes achieve sum-rates significantly closer to the interference-free upper bound, especially in high-interference regimes.
• In the three-user cascade Z-channel, scheduling based on interference geometry leads to improved sum-rate performance.
• The processing energy cost fundamentally changes the optimal transmission strategy, favoring intermittent transmission over continuous transmission.
• The sum-rate gains from interference-aware scheduling are most pronounced in high-interference, high-processing-cost scenarios.