毫米波多蜂巢式系統之效能建模與佈建策略

Abstract

本論文旨在針對兩層的毫米波異質性網路進行效能建模與分析框架的開發。首先，在取得精確的地理資訊下，本文提出真實環境下的毫米波小細胞佈建策略以及建立光跡追踪(Ray Tracing)三維通道模型，模擬結果證實所提出的佈建策略在毫米波蜂巢式網路中可以比傳統方法節省更多的發射功率。然而，在缺乏精確地理資訊的情況，本文建立了隨機幾何框架以表徵毫米波蜂巢式網路下的阻塞效應，同時推導出其共存的覆蓋概率，其中，大規模異質性網路的細胞模型包含了毫米波基站和傳統特高頻大型基站。從模擬發現，該隨機基站佈建（隨機幾何模型）與真實環境相似之決定性基站佈建（光跡追踪）提供覆蓋概率的下限與上限。我們所提隨機幾何模型可以提供覆蓋率的保守預測值。

This thesis aims at developing a performance modeling and analysis framework for the two-tier millimeter wave (mmWave) heterogeneous networks. First of all, with accurate geographic information, we propose a LOS-paths-determined (LPD) strategy for small cell deployment in real built-up environments based on a ray tracing 3D channel model. Simulation results show that the proposed deployment strategy can save more transmit power than the traditional approach in mmWave cellular network. However, with inaccurate geographic information, a stochastic geometrical (SG) framework for mmWave heterogeneous networks is established to characterize the blockage effect. A cellular model for a large-scale heterogeneous network consisting of mmWave base stations (BSs) and conventional UHF macro BSs is proposed and its coexisting coverage probability is derived. It is shown that the random BS deployment (SG model) and deterministic BS deployment (ray tracing) provide the lower and upper bound values of the coverage probability, respectively. Therefore, the proposed model can provide conservative predictions, as well as being analytically tractable.