Intelligent reflecting surface aided wireless networks-Harris Hawks optimization for beamforming design

Huaqiang Xu, Guodong Zhang, Jun Zhao, Quoc-Viet PhamarXiv 2020 · arXiv

Introduction

As the telecommunications industry pivots toward 6G, reducing energy consumption and hardware costs remains a paramount challenge. Intelligent Reflecting Surfaces (IRS) have emerged as a promising "green" technology. These surfaces consist of numerous passive reflecting elements that can steer wireless signals toward users by adjusting their phase shifts. However, maximizing the Signal-to-Noise Ratio (SNR) requires the joint optimization of active beamforming at the Access Point (AP) and passive beamforming at the IRS. This non-convex optimization problem is traditionally solved using complex, mathematically intensive methods like Semi-Definite Relaxation (SDR). This paper explores a nature-inspired alternative: the Harris Hawks Optimization (HHO) algorithm.

Methodology

The Harris Hawks Optimizer (HHO) is a population-based, gradient-free meta-heuristic algorithm inspired by the cooperative hunting behavior of Harris hawks. The algorithm operates through three distinct phases:

  1. Exploration: Hawks search for prey (the optimal solution) across the search space.
  2. Transition: The search narrows based on the "escaping energy" of the prey.
  3. Exploitation: Hawks execute various chasing maneuvers to capture the prey.

To apply HHO to beamforming, the authors transformed complex-valued beamforming vectors into real-valued search spaces. Additionally, a penalty method was integrated into the fitness function to enforce hardware constraints, such as maximum power limits at the AP.

Performance Evaluation

The proposed HHO-based design was evaluated through computer simulations using standard wireless channel models. The AP-to-IRS link was modeled as Line-of-Sight (LoS), while user links incorporated Rayleigh fading with a 10 dB penetration loss.

SNR performance of HHO compared to centralized and distributed benchmarks across varying distances.

As shown in Figure 1, the HHO-based joint design achieves SNR levels nearly identical to established centralized and distributed benchmark algorithms. Notably, the IRS-aided system provides a significant SNR boost compared to a system without IRS, particularly as the distance between the AP and the user increases. At a distance of 50 meters, the benefit of the IRS becomes most pronounced.

Convergence and Scalability

A key advantage of meta-heuristic approaches is their robustness in non-convex landscapes. The HHO algorithm demonstrated rapid convergence, typically reaching a near-optimal solution within the first few hundred iterations.

Convergence behavior of the HHO algorithm for a user distance of 50m.

Figure 3 illustrates this convergence trend for a user located 50 meters from the AP. The "best fitness" (SNR) stabilizes quickly, validating the algorithm's efficiency. Furthermore, while increasing the population size improves the optimality of the solution, it does so with a linear increase in computational time, offering a predictable trade-off for system designers.

Conclusion and Future Work

The application of Harris Hawks Optimization to IRS beamforming design proves that nature-inspired algorithms can match the performance of traditional mathematical solvers while being significantly easier to implement. Future research should address the assumption of perfect Channel State Information (CSI) and explore the impact of discrete phase shifts, which are more common in practical IRS hardware.

Summary updated

Abstract

Intelligent Reflecting Surface (IRS) is envisioned to be a promising green and cost-effective solution to enhance wireless network performance by smartly reconfiguring the signal propagation. In this paper, we study an IRS-aided multiple-input single-output wireless network where a multi-antenna Access Point (AP) services a single-antenna user assisted by an IRS. The goal is to maximize the received signal power by jointly optimizing the transmit beamforming at the AP and the reflection coefficient at the IRS. The formulated optimization problem is non-convex and subject to constraints. We adopt a novel nature-inspired optimization technique named Harris Hawks Optimizer (HHO) to tackle the problem. After transforming the constrained problem into an unconstrained problem using penalty method, the formulated problem is optimized by the HHO. To the best of our knowledge, this is the first time to use a meta-heuristic algorithm to solve the IRS-aided network optimization problem. Simulation is conducted to verify the feasibility of the HHO-based scheme. The results show that the HHO-based scheme could provide similar or even better optimization results compared with other optimization algorithms.
BibTeX
@misc{xu2020intelligent,
  author = {Huaqiang Xu and Guodong Zhang and Jun Zhao and Quoc-Viet Pham},
  title = {{Intelligent reflecting surface aided wireless networks-Harris Hawks optimization for beamforming design}},
  year = {2020},
  eprint = {2010.01900},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2010.01900},
}
Figure