基於 Wi-Fi Direct 之行動網路內容共享框架

Abstract

隨著網路與無線網路技術的發展，人們能夠透過智慧手機、平板電腦和其他移動設備連上Wi-Fi、4G或LTE輕鬆地瀏覽網頁並擷取網路資訊，帶給我們方便使我們能在任何時間、任何地點連接到網路。相反的，現今仍然存在一些挑戰需要克服。首先，網頁的文件內容大小越來越大因此產生大量的網路流量；其次，越來越多的移動設備在一區域同時間連接到有限頻寬的網路存取點，引起設備間相互競爭網路頻寬導致各個設備的吞吐量降低，造成高度延遲及不良的用戶體驗。為了優化移動網路的流量與提供良好的用戶體驗，本研究為移動網路環境提供了一個新的點對點內容共享框架，此設計的框架使用兩層主從式架構來管理和處理終端設備之間的請求，除此之外並使用了Wi-Fi Direct，一種Device-to-Device通信技術來幫助終端設備將緩存的網路資源共享到附近的設備，從而減少通過網路接入點的流量。為了評估結果，此研究設置了實驗，在基於Android 智慧手機與作為終端設備和Wi-Fi存取點的 Raspberry Pi之下，測試及探討不同大小的網頁內容在不同網路頻寬下的網路延遲及流量。實驗結果顯示，該框架有效降低了網絡延遲，並減少了通過接入點的流量。

With the advance of the Internet and wireless network technologies, people can easily browse website and retrieve information on the internet with their smartphones, tablets, and other mobile devices via Wi-Fi and 4G LTE. It brings them to the convenience connecting to the Internet at any place and any time. On the opposite of benefits, there are a few challenges are made and needed to be overcome. Firstly, file contents of websites have become much larger that generates large amounts of network traffics. Secondly, more and more mobile devices are joining in the local area and connecting to bandwidth-limited network access points. They compete network bandwidth and reduce the throughput for each device. It causes high response latencies and bad user experiences. For optimizing mobile network traffics and generate good user experiences, this study provided a newly Peer-to-Peer Content Sharing Framework on the mobile network environment. The designed framework used two layers’ client-server model proxies to manage and process requests among end devices. Moreover, it used Wi-Fi Direct, a mobile D2D content distribution technology, to help end devices share their cached network resources to the nearby devices that reduces network traffics via the Access Point. For evaluation, this study set up experiments to test network latency and traffics in different size of content and in different bandwidth under android-based smartphones and Raspberry Pi which acting as end devices and the Wi-Fi access point. The result showed that the framework efficiently reduced the network latency and also reduced the traffic through access point.