LTE-A在靜態頻率重用的情形下，雖能有效減少干擾，但在UE (User Equipment)較為集中情形下，會有頻譜使用不足的情況。在傳統的傳輸模式中，若UE之間欲進行傳輸，通常需由傳輸端發送訊號給基地台，基地台再發送訊號給接收者，需要進行兩次的無線傳輸。若UE在彼此距離相近的環境中，D2D (Device-to-Device)傳輸可讓UE之間直接利用LTE-A頻譜進行傳輸，進而達到節省頻譜資源的效果。 本研究探討靜態頻率重用的缺點與位置相近的D2D傳輸模式，提出DFRDD (Dynamic Frequency Reuse for D2D transmission)方法，使用動態頻率重用與D2D傳輸技術。我們將一個細胞(cell)劃分為中央區域及外圍區域，外圍區域又劃分為三個扇形區，使用動態頻率重用的方法調整頻譜，在頻譜不足時，中央區域可使用外圍區域的頻譜，外圍區域最多可使用中央區域三分之一的頻譜。在使用D2D技術時，利用D2D UE與BS/RS (Base Station/Relay Station)的距離，計算出對基地台UE干擾較少的頻譜，進而提升傳輸效率。 實驗結果顯示，DFRDD利用動態頻率重用與D2D選擇頻譜的方法、在吞吐量方面表現得較H. S.Chae [18]、Bao [19]、Zhang [20]所提出的方法為佳。 / In the case of LTE-A static frequency reuse, although it can effectively reduce the interference, however, in the case of more dense UEs (User Equipment) environment, there will be problem of insufficient spectrum. In the traditional transmission method, if a pair of UEs want to communicate with each other, the transmitter sends a signal to the base station, the base station then sends a signal to the receiver, the signal need to to be wirelessly transmitted twice. If a pair of UEs are within a close distance, D2D (device-to-device) transmission allows users to communicate with each other directly using the same LTE-A spectrum to save spectrum resource. In order to improve LTE-A system performance, this paper proposes a DFRDD (Dynamic Frequency Reuse for D2D transmission) method. Using dynamic frequency reuse and D2D transmission, we divide a cell into center region and outer area, where the outer area is divided into three sectors. We use dynamic frequency reuse method to allocate spectrum. When the spectrum is insufficient, the center region can use the spectrum of the outer region. On the other hand, the outer area can use up to one third of the spectrum of the center region. When using D2D technic, we calculate the distance between D2D UE and the BS / RS (Base Station / Relay Station), choose the frequency that may reduce the interference of cellular UE and improve transmission efficiency. The experimental results show that DFRDD uses the method of dynamic frequency reuse and D2D to select the spectrum, which has better performance than those methods proposed by Chae [18], Bao [19] and Zhang [20].