Research On The Coverage Of Base Station In Massive MIMO System Based On Stochastic Geometric Model

Recently, with the widespread smart devices, customers’demand for data increases exponentially. In order to satisfy the increasing demand for data, scholars start the research on the next generation mobile communication system, which is faster and more efficient. Due to its high spectral efficiency and energy efficiency, massive multiple-input multiple-output (MIMO) is regarded as one of the most promising technologies in next generation wireless communication system (5G).At first, this paper introduces the characteristics and benefits of massive MIMO and states the recent development of massive MIMO. Then this paper presents the advantages of stochastic geometric modeling of wireless networks and leads to the conclusion that it is necessary to use stochastic geometry to model massive MIMO system. In the following, this paper describes the basic knowledge of massive MIMO and stochastic geometry. The part of massive MIMO includes point-to-point, multi-user, and multi-cell massive MIMO, where the causes and effects of pilot contamination in multi-cell massive MIMO system is analyzed in depth. The part of stochastic geometry includes point process theory and commonly used theorem and method, such as Campbell’s theorem and probability generating functional (PGFL).Then, this paper models the massive MIMO system employing stochastic geometry. The base station (BS) is distributed according to Poisson point process (PPP). User equipment (UE) is uniformly distributed in the circle centered at each BS. The average pilot reuse probability is calculated to estimate the effect of pilot contamination on system performance. At last, we present the closed-form expressions for downlink coverage probability and average achievable rate per user are derived. Numerical simulation results show the relationship among downlink coverage probability, average achievable rate per user, BS density, and frequency reuse factor.Finally, this paper studies the BS density bounded by maximum outage probability based on stochastic geometric model of massive MIMO system. In order to satisfy the maximum outage probability, BS density has to be below the upper bound. Using Lambert W function, this paper derives the closed-form expressions for the upper bound of BS density and the upper bound of successful area spectral efficiency (ASE). Numerical simulation results show the relationship among maximum outage probability, upper bound of BS density, and the upper bound of successful area spectral efficiency (ASE).