利用快取資料之協調多點傳輸於次世代蜂巢系統骨幹流量最小化

Abstract

近年來隨者智慧行動裝置的出現，以及行動網路的普及，使得無線頻寬的需求越來越大。在頻譜有限及標金昂貴的狀況下，提升頻譜使用效率為一個有效的方法。因使從第三代行動網路開始，第三代合作夥伴計畫決定採用頻率重複使用提升頻譜效率，亦及相鄰基地台都使用相同頻寬。但是相對的也帶來在基地台邊界同頻率干擾的問題，因此在第四代行動網路進階版的時候，基地台間干擾協調被提出來解決這問題。此機制為相鄰基地台錯開對基地台邊界使用者的資源方塊，缺點為犧牲基地台邊界的頻譜效率，而多點協調同時傳輸就是一個解決此問題的方案。多點協調同時傳輸利用原先為解決干擾而保留起來的資源方塊，傳輸相同的資料給基地台邊界的使用者，藉此提高訊號雜訊比以增加速度。為了讓多個同時傳輸的基地台都有相同資料，此方法會帶來骨幹網路傳輸的流量負擔。相較於頻譜效率普提升帶來好處，因此我們需要解決骨幹網路流量的問題。由於目前網路流量主要是以影音串流為主，加上思科預測未來更會高達七成。使用基地台搭配快取是一個未來趨勢，在考量快取的情況下，一個不好的多點協調同時傳輸分配機制會造成骨幹網路流量的浪費。因此在這篇論文中，我們基於資源方塊的限制及快取影片，提出了一種使網路流量最小化的分配機制。並提出一個實作範例，利用類似內容傳遞網路的機制，使配合的內容提供者可以放快取影片到基地台。我們的貢獻在於，配合提出的實作架構並搭配分配機制，達到我們想要的目標。

Operators consume more spectrums to provide service because mobile devices become more and more popular. Spectrum is the scarce and expensive resource. 3rd Generation Partnership Project (3GPP) decides to improve spectrum efficiency by frequency reuse starting from Wideband Code Division Multiple Access (WCDMA), which means every base station operates on the same frequency. But the side effect is inter-cell interference at cell edge. Therefore, Inter-cell interference coordination (ICIC) was proposed on Long Term Evolution Advanced (LTE-A) to solve this issue. This mechanism reserves the resource blocks (RBs) used by neighboring cell at cell edge. However, it reduces the spectrum efficiency at cell edge. Coordinated Multi Point Joint Transmission (CoMP JT) is a solution to this problem. To achieve higher SNR and throughput, which transmits data to cell edge user by reserved neighboring RBs. But delivering data to those cells pose the traffic burden on backhaul. Video streaming is major traffic nowadays, and Cisco predicts it will reach 70% of total traffic in the future. Cache on cell is a trend in the near future. By taking cache into consideration, bad CoMP JT cluster decision could waste resource on backhaul. In this thesis, we propose a method to select CoMP JT set with RB constraint and cached data to minimize the backhaul traffic consumption. We also provide an implementation example, a content delivery network like (CDN-like) mechanism, to content provider for putting data in cell cache. Our contribution is we proposed a new cache mechanism which can cooperate with our heuristic algorithm to reach our goal.