IEEE 802.16 無線系統上行鏈路之以基因演算法與用戶分組資源分配

Abstract

隨著無線系統傳輸在現今通訊系統的重要性日益增加，要提供多媒體傳輸服務品質保證(quality of service, QoS)成為一個很重要的議題。為了達到最佳的系統資源利用並提供服務品質保證，我們在IEEE802.16 虛擬多發多收(virtual multiple-input-multiple-output (MIMO))上行鏈路系統提出了一個採用基因演算法暨用戶分組的資源分配(resource allocation with genetic algorithm and SS grouping, GGRA)機制。在機制中使用基因演算法來決定排程順序，並設計了速率分配策略(rate assignment strategy)來動態調整各種服務的優先次序及分配資源大小。我們提出的GGRA 機制會在考慮系統限制條件如能量、接收天線數等限制下，找到最佳的排程結果，來使得系統傳輸速率最大化，並滿足各種服務的品質保證。在計算時間上，本方法也能利用平行處理的方式達到足夠快的處理速度，因此GGRA 可應用於實際系統中，成為一個有效並可以實現的排程機制。

In this thesis, a resource allocation with genetic algorithm and SS grouping (GGRA) scheme is proposed for IEEE 802.16 uplink communication systems. The GGRA scheme designs a rate assignment strategy applied with a predefined residual lifetime to dynamically allocate resource to each service type. It also aggregates high correlation subscriber stations (SSs) into the same group to avoid mutual user interference and adopts genetic algorithm to find an optimal assignment vector. It can maximize the system throughput while satisfying the QoS requirements. Simulation results show that the proposed GGRA scheme has better performance than the EFS algorithm and the MLWDF algorithm in system throughput, voice/video packet dropping rate, unsatisfied ratio of HTTP users/packets, and FTP throughput. It can also be found that the GGRA scheme has a feasible computation complexity so that it is feasible in real applications.