Modeling And Performance Analysis Of Multi-Layer/Multi-Cell Networks

Multi-Layer/Multi-Cell networks are multiple-tier cellular networks, which introduce small base stations (BS) into the traditional cellular networks to improve the coverage of macro BS. Before the deployment of Multi-Layer/Multi-Cell network, cell splitting method is adopted to enhance the spectrum efficiency. Compared with this technique, Multi-Layer/Multi-Cell network is much more flexible in network construction, and also needs less cost in terms of capital expenditures (CAPEX) and operating expenses (OPEX). Therefore, research on Multi-Layer/Multi-Cell network is of great significance and has attracted attention from both academia and industry. Compared with macro BS, small BS has its own advantages, but also introduces more interference. In this paper, modeling of Multi-Layer/Multi-Cell networks is studied, along with the performance analysis and improvement.In this paper, our investigation is carried out from two aspects. Performance analysis of Multi-Layer/Multi-Cell network is studied in both the static interference model and the dynamic interference model. In the static interference model, Signal to Interference plus Noise Ratio (SINR) is employed as the major indicator. Firstly, the system simulation platform is set up, together with the channel model and the interference model. Then, the deployment of small BS is researched to analyze the system performance of Multi-Layer/Multi-Cell network. Furthermore, techniques like partial frequency reuse scheme are employed to enhance the cell edge user performance.In the dynamic interference model, performance analysis of user association in multiple cells is studied. Most studies in this field have used static interference models for interfering base stations. However, the transmit power of a small base station is lower. Thus the number of users associated with a small base station is small. The small base station may stay in the state of idle and thus the interference situations of each related user may dynamically change. In this paper, we study a dynamic interference model first, which considers both the fast fading effects of the wireless channel and the dynamic characteristic of data packets. The two-dimension Discrete Time Markov Chain (DTMC) is employed to solve the model. Since the state space of DTMC increases exponentially with the user number, model decomposition and iteration method in Stochastic Petri Nets is utilized to obtain the approximate steady state probability. Then, the system performances such as average throughput, queue length, delay and dropping probability are derived using the steady state probability. Finally, the analytical results are verified through simulations.