標題: 針對具載波聚合LTE-A網路所設計且以提升能源效率為目標之下鏈路資源分配技術
Energy-Efficient Downlink Resource Allocation with Carrier Aggregation in LTE-Advanced Network
作者: 郭治緯
Kuo, Chih-Wei
方凱田
Feng, Kai-Ten
電信工程研究所
關鍵字: 能源效率;資源分配技術;Energy-Efficient;Resource Allocation
公開日期: 2014
摘要: 隨著使用者對資料傳輸的需求快速地成長,以及全球暖化的嚴峻形勢,資源分配和能源消耗的問題近年來逐漸受到相當大的重視。為了增進載波聚合系統的能源效率,我們首先制定一個為非線性規劃問題的最佳化問題。然而,要透過竭盡式搜尋解決該問題所耗費的計算量是極度複雜且不切實際的。因此,在本論文中我們提出幾個低複雜演算法來解決這個問題。在基於能源效率權重分量載波的選擇和功率分配(WEECP) 的方法和基於可取性分量載波的選擇和功率分配(DCP) 的方法,我們介紹了啟發式的方法來解決服務基地台、分量載波、資源區塊的配置,以及藉由假設基地台間的干擾為一個期望常數以降低複雜度,解決功率分配。另一方面,在基於交叉熵分量載波的選擇和功率分配(CECP) 的方法中,我們透過反覆地更新機率同時解決了服務基地台、分量載波、資源區塊和功率的配置問題,並且可以因此獲得接近最佳解。模擬結果明顯地呈現提出的方法的能源效率比相較的方法好得多。
With the rapidly growing demands of data transmission from users and the serious situation of global warming, the problems about resource allocation and energy consumption attract significant attention in recent years. In order to enhance the energy efficiency (EE) for LTE-Advanced systems with carrier aggregation (CA), we first formulate an optimization problem as non-linear programming problem. However, the computation of solving this problem by exhausted search is extremely complex and impractical. Therefore, in this thesis we propose several low-complex algorithms to solve this problem. In weighted energy efficiency based component carrier selection and power allocation (WEECP) method and desirability based component carrier selection and power allocation (DCP) method, we present heuristic approach to solve the serving base station (BS) assignment, component carrier (CC) selection, and resource block (RB) configuration and solve the power allocation where the inter-cell interference term is assumed to be a expected constant to reduce the computational complexity. On the other hand, in cross-entropy based component carrier selection and power allocation (CECP) method, we jointly solve the allocation of serving BS of user equipment (UE), CC, RB, and power through updating probability iteratively and can thus obtain near-optimum solution. Simulation results show that EE of proposed methods are obviously much better than that of compared methods.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT070060219
http://hdl.handle.net/11536/76365
Appears in Collections:Thesis