Energy-Efficient Relaying over Multiple Slots with Causal CSI

Chin Keong Ho, Peng Hui Tan, Sumei SunarXiv 2012 · DOI · arXiv

Abstract

In many communication scenarios, such as in cellular systems, the energy cost is substantial and should be conserved, yet there is a growing need to support many real-time applications that require timely data delivery. To model such a scenario, in this paper we consider the problem of minimizing the expected sum energy of delivering a message of a given size from a source to a destination subject to a deadline constraint. A relay is present and can assist after it has decoded the message. Causal channel state information (CSI), in the form of present and past SNRs of all links, is available for determining the optimal power allocation for the source and relay. We obtain the optimal power allocation policy by dynamic programming and explore its structure. We also obtain conditions for which the minimum expected sum energy is bounded given a general channel distribution. In particular, we show that for Rayleigh and Rician fading channels, relaying is necessary for the minimum expected sum energy to be bounded. This illustrates the fundamental advantage of relaying from the perspective of energy efficient communications when only causal CSI is available. Numerical results are obtained which show the reduction in the expected sum energy under different communication scenarios.

Introduction

Real-time wireless applications, such as industrial control or video streaming, demand reliable data delivery within strict deadlines. However, the inherent unpredictability of wireless fading channels poses a significant challenge. If a channel is currently in a deep fade, a transmitter could either expend massive energy to overcome the noise or wait for a better channel state in a future slot.

The difficulty is compounded in practical systems where only Causal Channel State Information (CSI) is available. While past and present channel conditions are known, future states remain unknown. This paper addresses the challenge of minimizing the total expected energy used by a source and a relay to deliver a message within a fixed number of time slots ($K$), operating without the benefit of foresight.

Methodology

The authors frame the energy minimization problem as a finite-horizon dynamic program. The transmission process is divided into two distinct phases based on the relay's status:

  1. Phase 1: The source (S) transmits. Both the destination (D) and the relay (R) listen and accumulate mutual information.
  2. Phase 2: Once the relay successfully decodes the message, it joins the source in transmitting to the destination, providing spatial diversity.
Transmission structure over multiple slots. The relay (R) begins assisting the source (S) in delivering the message to the destination (D) only after it has successfully decoded the source's signal.

The optimal power allocation is derived using Bellman’s Equation. Since receivers "collect" information over multiple slots—a concept known as Mutual Information Accumulation (MIA)—the state space tracks the remaining information required for decoding. To manage the continuous nature of this state space, the authors employ discretization and Monte Carlo simulations.

State transition diagram for the dynamic programming model. It tracks the system's progress through slots and phases, where transitions depend on whether the relay or destination has accumulated enough information to decode.

The performance of the proposed policy is evaluated using standard statistical channel models, including Rayleigh fading (no line-of-sight) and Rician fading (strong direct path).

Key Insights and Results

The Necessity of Relaying

A primary finding of this research is the "Infinite Energy Trap." In pure Rayleigh or Rician fading environments, if a system relies solely on a direct source-to-destination link with causal CSI, the expected energy required to guarantee delivery by a strict deadline is mathematically infinite. This occurs because there is a non-zero probability that the channel remains poor for the entire duration.

Relaying acts as a critical safeguard by introducing spatial diversity. Even if the direct link is failing, the source-to-relay or relay-to-destination links may be healthy. This diversity bounds the expected energy, making reliable communication feasible and efficient.

Performance Gains

Numerical results demonstrate that relaying significantly reduces the Normalized Minimum Expected Sum Energy (NMESE). For instance, at a transmission rate of 1 nat/slot, relaying can provide over 1 dB of power savings compared to non-relaying schemes. Furthermore, the study shows that most energy-saving gains are realized with just two slots ($K=2$), with diminishing returns as more slots are added.

Heuristic vs. Optimal

The authors also propose a simplified heuristic policy. Interestingly, this heuristic performs remarkably close to the complex, mathematically optimal dynamic programming solution, particularly at lower data rates, offering a practical alternative for real-time implementation.

Conclusion and Future Work

This work highlights that in wireless networks with strict deadlines, a relay is not merely an enhancement but a mathematical necessity for energy-efficient delivery under causal CSI.

Future research directions include scaling to multi-relay networks, developing more computationally efficient algorithms to overcome the "curse of dimensionality" in dynamic programming, and applying these strategies to multi-carrier systems like OFDM.

Summary updated

Abstract

In many communication scenarios, such as in cellular systems, the energy cost is substantial and should be conserved, yet there is a growing need to support many real-time applications that require timely data delivery. To model such a scenario, in this paper we consider the problem of minimizing the expected sum energy of delivering a message of a given size from a source to a destination subject to a deadline constraint. A relay is present and can assist after it has decoded the message. Causal channel state information (CSI), in the form of present and past SNRs of all links, is available for determining the optimal power allocation for the source and relay. We obtain the optimal power allocation policy by dynamic programming and explore its structure. We also obtain conditions for which the minimum expected sum energy is bounded given a general channel distribution. In particular, we show that for Rayleigh and Rician fading channels, relaying is necessary for the minimum expected sum energy to be bounded. This illustrates the fundamental advantage of relaying from the perspective of energy efficient communications when only causal CSI is available. Numerical results are obtained which show the reduction in the expected sum energy under different communication scenarios.
BibTeX
@misc{ho2012energy,
  author = {Chin Keong Ho and Peng Hui Tan and Sumei Sun},
  title = {{Energy-Efficient Relaying over Multiple Slots with Causal CSI}},
  year = {2012},
  eprint = {1205.4785},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1205.4785},
}
Figure