Performance analysis for high-speed railway communication network using stochastic network calculus

High-speed railway (HSR) wireless communication network has won a great development in recent years. However, due to its unique characteristics, such as high speed, fixed mobility track, special traffic, frequent handover and bandwidth constraints, it is quite a challenge to carry out the performance analysis on HSR. In this paper, we derive the backlog and delay bounds for HSR communication network by using stochastic network calculus (SNC) with moment generating functions (MGFs). SNC is a highly useful tool in packet switched network performance analysis since 1990s. Different from the previous studies about SNC which mostly concentrate on general low-speed, traditional public communication network, we take the varying small-scale fading and large-scale fading into consideration by means of a novel multi-dimension finite-state Markov channel (FSMC) model. We elaborate the differences in packet-level channel model, traffic model of train control information and performance metrics in HSR communication network. Particularly, the backlog and delay bounds of HSR with different packet arrival rates and violation probabilities are derived when train travels from the cell center to the cell edge. What is more, the numerical results can help to provide a guideline for designing the proper violation probabilities and arrival rates to different types of traffic in HSR communication network.