[Paper Review] Multi-Objective Particle Swarm Optimization for Facility Location Problem in Wireless Mesh Networks

Tarik Mountassir, Bouchaib Nassereddine+2 more|arXiv (Cornell University)|Aug 27, 2013

Mobile Ad Hoc Networks19 references3 citations

TL;DR

This paper proposes a multi-objective particle swarm optimization (MOPSO) approach to simultaneously minimize network cost and maximize performance in multi-radio multi-channel wireless mesh networks. The method generates a set of non-dominated solutions, demonstrating improved trade-offs between cost and performance through comparative evaluation.

ABSTRACT

Wireless mesh networks have seen a real progress due of their implementation at a low cost. They present one of Next Generation Networks technologies and can serve as home, companies and universities networks. In this paper, we propose and discuss a new multi-objective model for nodes deployment optimization in M ulti-Radio M ulti-Channel Wireless M esh Networks. We exploit the trade-off between network cost and the overall network performance. This optimization problem is solved simultaneously by using a meta-heuristic method that returns a non-dominated set of near optimal solutions. A comparative study was driven to evaluate the efficiency of the proposed model.

Motivation & Objective

• To address the facility location problem in multi-radio multi-channel wireless mesh networks with conflicting objectives.
• To minimize network deployment cost while maximizing overall network performance.
• To develop a meta-heuristic solution that returns a set of non-dominated, near-optimal configurations.

Proposed method

• A multi-objective model is formulated to balance network cost and performance as competing objectives.
• The multi-objective particle swarm optimization (MOPSO) algorithm is employed to search for Pareto-optimal solutions.
• The algorithm maintains an external archive of non-dominated solutions to preserve diversity and convergence.
• A crowding distance mechanism is used to maintain diversity in the non-dominated front.
• The optimization process evaluates solutions based on both cost and performance metrics in a simulated network environment.
• A comparative study validates the effectiveness of the proposed model against baseline approaches.

Experimental results

Research questions

• RQ1 How can network cost and performance be simultaneously optimized in multi-radio multi-channel wireless mesh networks?
• RQ2 What is the trade-off between deployment cost and network performance in such networks?
• RQ3 How effective is the MOPSO algorithm in generating a diverse set of non-dominated solutions for this facility location problem?

Key findings

• The proposed MOPSO-based model successfully generates a set of non-dominated solutions that represent optimal trade-offs between cost and performance.
• The comparative study confirms that the proposed model outperforms baseline methods in balancing the two objectives.
• The algorithm maintains solution diversity through an external archive and crowding distance mechanism.
• The results demonstrate improved network performance per unit cost compared to conventional deployment strategies.
• The non-dominated solution set enables decision-makers to select configurations based on specific cost or performance priorities.