結合使用端與網路端資訊的行動通訊與區域網路之無線卸載

Abstract

LTE已成為新一代最重要的通訊技術之一，但是在近代移動裝置的快速成長之下，產生大量的移動資料流量，使得蜂巢網路面臨過載的問題。為此，營運商與標準組織制訂了HetNets的解決方案，原理是使得原本存在於蜂巢網路上的流量可以卸載到不同種類的網路通訊技術，藉此舒緩過載問題。為了最小化服務阻斷率並滿足服務品質，需要更有效的分配異質網路環境下的通道資源。但異質網路之間存在著協定上的差異，因此本文基於3GPP提出的標準架構，提出Combined UE and BS Information (CUBI)的方法，分別討論使用者設備端可以獲得的資訊，和網路端所能獲得的資訊，比較其做出的決策結果差異，並且進一步的提出兩輪式的結合兩種資訊的決策方法。使用者設備端可以獲得單一使用者和多個網路之間的通道品質，藉此選擇有最佳通道品質的目標網路。網路端則是可以獲得單一網路端與多個使用者之間的通道品質和網路本身的負載狀態，藉此能同時分配多個使用者端與網路進行連線。結合兩種資訊的決策目的就是要獲得此兩種資訊的優點，使得所有的使用者端都可以與其最適合的網路進行連線。實驗結果可以發現，全域資訊的決策可以降低使用者端資訊決策的服務阻斷率24.93%和網路端資訊決策10.53%。

LTE becomes one of the most important communication technologies in the next generation. With the increasing popularity of mobile devices, cellular networks suffer the dilemmas of mobile data traffic explosion. To resolve this problem, operators and standard development organizations have developed HetNet solutions, which offload traffic from cellular networks to other radio access technologies to reduce the overloading situation. To minimize the service blocking ratio and satisfy the service demand, HetNets need to design an efficient approach to allocate the channel resource. However, HetNets have to deal with the heterogeneous protocol problem. Thus, this work proposes the Combined UE and BS Information scheme (CUBI), which is based on the standard architecture which is proposed by 3GPP. First, we discuss the UE and BS information, respectively. We compare the impacts of using the two types of information to make the offloading decision. Afterwards, this work proposes the 2-round decision to obtain the advantage of UE and BS information. The decision with UE information is based on the channel qualities of all the available networks, hence, it can choose the target network which has the highest channel quality. The decision with BS information is based on the channel resource utilization of all the UEs to allocate the channel resource to multiple UEs at the same time. On the other hand, the main idea of the decision with combined UE and BS information is to obtain the advantage of these two type information and allocate the channel resource properly for UEs. The simulation results demonstrate that decision with combined UE and BS information can decrease the service blocking ratio down to 24.93% and 10.53% compared to the decision with UE and BS information schemes, respectively.