A Multi-Service Train-to-Ground Bandwidth Allocation Strategy Based on Game Theory and Particle Swarm Optimization

The rapid development of high-speed trains renders more information exchange between the train and ground. However, the bandwidth resources for various service communications are limited. The bandwidth should be allocated effectively to ensure that the vital information related to safety and efficiency can be transmitted in prior with better use of the residual bandwidth. This paper presents a multi-task train-to-ground bandwidth allocation strategy for wireless communication systems of high-speed trains. First, by analyzing the bandwidth demand and data characteristics of each service, the utility and weighting functions are formulated. The optimization model is established by using the asymmetric cooperate Nash game theory. Second, the Particle Swarm Optimization (PSO) is adopted to realize the nonlinear bandwidth allocation optimization described as a Non-Deterministic Polynomial (NP) problem. Finally, the novel bandwidth allocation scheme is verified with comparative simulations. It is shown that the scheme allocates the bandwidth dynamically according to the weighting factor of each service, and maximizes the bandwidth utility with satisfying the real-time demands of the running train. It optimizes the train-to-ground bandwidth allocation, and thus improves the operation safety, efficiency, and bandwidth usage of high speed trains.