近年來，隨著行動通訊技術的發展，人們對於行動通訊的需求日益增加，行動裝置的數量也呈現爆炸性的成長。過多的裝置用戶進行通訊也會造成中心網路壅塞的情形產生，因此不必透過基地台幫忙的裝置間通訊(Device-to-Device, D2D)被視為下世代通訊發展的重要技術。 在本篇論文當中，我們探討裝置間通訊資源分配問題。為了能夠使資源頻譜能夠有效地被利用，我們讓裝置間通訊用戶和蜂巢式通訊用戶進行非正交資源共享，將裝置間通訊用戶所需要的服務品質視為取得資源的順序。此演算法讓一個蜂巢式通訊用戶能夠和兩個裝置間通訊用戶共享資源，由於在同一個資源塊上有多個通訊用戶存在，會造成訊號互相干擾問題，此演算法亦透過功率的調配處理此問題，保障裝置間用戶的通訊品質，同時確保蜂巢式通訊用戶最低的訊號傳輸功率。本研究論文的目標為將裝置間通訊用戶的系統傳輸吞吐量優化，並且提高裝置間通訊用戶的服務品質滿意度。 在模擬過程中，我們也將此演算法和隨機演算法及貪婪演算法做比較。研究結果顯示，此演算法能夠提高裝置間通訊用戶的系統傳輸吞吐量，並且能讓用戶端有較高的服務品質滿意度，其表現結果皆優於隨機演算法及貪婪演算法。;In recent years, due to the revolution of communication technology, Requirements for mobile communication has been increasing day after day. The amount of user equipment will be growing explosively, which cause congestion problems for the central network when there are abundant users trying to communicate at the same time. Therefore, D2D that can transmit data directly without assistance of eNodeB is becoming the critical part of communication in next generation. In this thesis, we discuss the issue of resource allocation in D2D communication underlying cellular networks. To make resource utilized effectively, non-orthogonal sharing mode is adopted. The eNodeB allocates resources according to the quality of service of each DUE. The proposed algorithm has a CUE share resources with two pairs of DUE, dealing with the severe mutual interference among the different users via power coordination. It can sustain the communication quality of DUE and ensure the basic transmit power of CUE simultaneously. The purpose of this thesis is to optimize the throughput as well as to enhance the service satisfaction degree of D2D. We compare the proposed method to random and greedy algorithms, and numerical results show that the proposed algorithm can increase the throughput and obtain higher user satisfaction degree in D2D communication.
