Dynamic optimization of radio resource management in LTE-based high-speed railway wireless networks

High-speed railway (HSR) has been deployed worldwide as the dominant fast public transportation system. HSR infrastructure features an increasing amount of control and multimedia transmissions to ensure the safety and efficiency of the transportation, and to support the increasing demand from passengers for the Internet services. Since GSM-R offers limited data transmission capabilities and inefficient radio resource usage, alternative communication systems should be investigated to support the increasing traffic demand. Therefore, the deployment of wideband wireless communication technologies is an urgent need for the operation of HSR to provide reliable and robust connection with the central control systems, and to offer seamless Internet access for passengers. LTE is envisioned as the leading candidate to replace GSM-R for the future railway infrastructures due to its packet-switched network architecture with high data transmission capacities and low latency. The increasing amount of wireless traffic in HSR necessitates the development of radio resource management (RRM) techniques to enable the efficient resource utilization. However, the design of such techniques is quite challenging due to the very dynamic channel conditions between the fast moving train and base stations, and the operational and passenger services with heterogeneous QoS requirements. RRM techniques for HSR wireless networks should jointly optimize the different aspects of resource allocation, admission and power control while observing different network constraints to enhance the system performance. This optimization should be performed in a cross-layer manner by considering the distinctive parameters and constraints stemmed from different network layers. Furthermore, it should manage the inherent trade-offs between these layers in a comprehensive close-to-optimal manner. In the light of the above, this work presents a delay-aware optimization framework which exploits the cross-layer design methodology and optimizes the resource allocation, admission and power control in LTE-based HSR wireless networks. The proposed framework extends previous works in literature by introducing an additional delay constraint to the dynamic network management of HSR infrastructure, and, hence, maximizing the throughput performance of the network while jointly optimizing the aforementioned aspects. The proposed scheme maximizes system utility under the constraints of average delay, queue stability and average power by formulating a joint stochastic problem which is transformed into a queue stability problem. Using Lyapunov drift theory, the queue stability problem is decomposed into separate convex subproblems. Golden section search method is used to achieve a globally optimal solution for the resource allocation subproblem. Then, a distributed dynamic resource management technique is proposed, and its performance is evaluated under realistic conditions for HSR wireless networks. Numerical simulations is performed on MATLAB and the results show that the proposed approach provides an improved throughput performance and enhanced link capacity, while satisfying the average delay requirements.