具廣義基地台連結異質性小細胞網路之覆蓋率與傳輸量分析

Abstract

近年來採用空間點程序去建構細胞網路模型，已經逐漸變成研究覆蓋率、平均干擾等之閉合型式解的有效方法。此種網路模型之所以在數學上具有容易處理的特性是基於同質性泊松點過程（Poisson point process, 以下簡稱PPP）的 Slivnyank 理論。基於PPP模型，我們可以研究來自基地台和使用者間產生細胞連結以及對覆蓋率和傳輸量造成衝擊的資料流問題。我們推導出細胞連結的通用數學表示式，並且驗證此式在遵守特定條件下，多基地台選擇之演算法可以充分且有效地測定反應通道之統計特性，因而具有實用上的重要性。在分析網路覆蓋率和傳輸量的過程中，我們可以獲得一個重要的觀察：首先，連結最近基地台之基地台選擇演算法相對於最大接受訊號強度之基地台選擇演算法，具有更高的能量效率，然而卻會造成較小之傳輸量與覆蓋率的代價。這個觀察結果促使我們提出據能量效率之基地台選擇演算法，這個演算法可以在提高連結最近基地台之基地台選擇演算法的效能時，同時也減少能量效率的損失。最後，我們提出一個嶄新的架構來研究在多層網路內使用多重基地台選擇演算法之網路效能。基於這個架構我們得到第二個重要的觀察：最大接收能量排程並不能永遠得到較好的能量效率及效能。

The recent adoption of spatial point processes in modeling cellular networks has been instrumental in discovering some closed-form expressions of network performance metrics, such as the coverage and mean interference. The tractability of this network model is due to the Slivnyak theorem of a homogeneous \textit{Poisson point process} (PPP). Based on a PPP model, we study the traffic flow problem through cell association between mobile users and base stations (BS) and its impact on the coverage and throughput performance. We derive a general mathematical expression for cell association function in which we demonstrate its practical importance in determining useful statistical properties of the network under various specific cell association schemes that obey certain rules. The network coverage and throughput performance subsequently studied can provide a crucial insight. Namely, the nearest base station association (NBA) scheme is more energy-efficient in terms of the bits-per-joule energy efficiency if compared to the maximum received power association (MRPA) scheme, but at the expense of less throughput and smaller coverage. This motivates us to propose an energy-efficient cell association scheme aminig to increase the performance of NBA without degrading much of its achievable energy efficiency. In the practical context of cell association, we consider the situation that multiple cell association (MCA) schemes are used in a multi-tier heterogeneous network (HCN). In the multi-tier HCN, users are able to use MCA schemes consisting of a few different cell association schemes. The insight we can obtain from the MCA scheme is that small cell users do not benefit much from using MRPA in terms of the coverage and throughput performance.