Swarm-Intelligent Multi-Hop Relay Selection for Throughput Maximization in Vehicular Ad Hoc Networks (VANETs)

Abstract

Vehicular ad hoc networks are emerging as an essential communication substrate for cooperative driving, traffic safety, and infotainment services along road infrastructures. In these networks, moving vehicles form spontaneously connected topologies in which wireless links are short lived and subject to rapid fluctuations. Multi-hop relaying becomes necessary when direct connectivity to roadside infrastructure is intermittent, yet the dynamic nature of vehicular mobility makes relay selection and route construction challenging. Throughput maximization over such volatile paths requires relay decisions that balance link quality, contention, and path length while reacting to time-varying channel and topology conditions. Conventional deterministic optimization methods often rely on simplified assumptions that make them less adaptable to realistic vehicular environments. Swarm intelligence, which coordinates a large population of candidate solutions through low-complexity interactions, provides a flexible framework for exploring high-dimensional relay selection spaces under uncertainty. This work studies a swarm-intelligent multi-hop relay selection strategy for throughput maximization in vehicular ad hoc networks, formulating relay decision making as a constrained linear optimization problem embedded inside a population-based metaheuristic. The approach maps candidate relay sets to particles that traverse the solution space according to locally evaluated throughput and connectivity metrics. A linear capacity model and a compact representation of path feasibility guide the swarm search while remaining compatible with the stringent timing constraints of vehicular communications. The study examines how swarm parameters, mobility patterns, and channel models influence the achievable throughput and outlines several implementation considerations for deployment in realistic vehicular scenarios.