LTE-A下行鏈路內螞蟻群聚最佳化解多點式協調資源分配

Abstract

LTE-A 系統被視為新一代的4G技術，可服務大量的資訊流量，每個用戶使用行動上網的時間將會大幅增加。為了盡可能的服務多個用戶且達到可容忍的干擾，多點協調(CoMP) 的資源分配變成一個重要的議題。不同資源與不同用戶間有多種不同的可能配對，在資源與用戶變多的情況下，這種配對問題會變得相對複雜難解，此問題在不同領域都可能發生，一般被稱做NP-hard的問題。 在此類問題下，螞蟻演算法(Ant colony optimization)有相當良好的解。螞蟻演算法會根據內存的費洛蒙，將好的解記錄下來，因為好的解會持續增加費洛蒙與且差的解費洛蒙會揮發，而螞蟻會傾向選擇較高費洛蒙的解，最後得到的解就會逼近最佳解。在此論文中利用費洛蒙把有好的通道品質的用戶記錄下來，因此整個系統就會得到較好的系統傳輸量。同時在這裡也有考慮了不同資料封包的服務品質，使用螞蟻演算法找到服務品質跟系統傳輸量的平衡點。

Inter-cell interference of multi-cell in LTE-A downlink systems is a key concept for improving throughput. Coordinated multi-point (CoMP) transmission could avoid interference, enlarge the coverage of base station and enhance data rate of cell-edge user. How to allocate proper resource with different CoMP transmission type for distinct users is a typical tough NP-hard problem. In this paper, we used ant colony optimization (ACO) to solve this resource allocation problem then proposed a CoMP-based radio resource allocation by ant colony optimization (CRACO) scheme. The CRACO scheme can assign an appropriate CoMP transmission type, modulation order and proper resource to users. And the CRACO scheme improves the system throughput and quality of service (QoS) for different traffic types. The simulation results shows that the CRACO scheme enhances the system throughput by about 8.8% compared to other scheme without violating the QoS requirement.