Real-time signal strength prediction and load balance algorithm in high-speed environment
|關鍵字:||高速移動;WiMAX訊號;訊號預測;軟體定義網路;系統平衡;high-speed mobility;WiMAX signal;signal prediction;SDN;load balancing|
|摘要:||隨著高速鐵路在全球普及，高速移動下行動裝置上網的需求也快速上升。目前台灣高鐵車內無線網路節點是以線狀布建且只有一個WiMAX CPE (Customer Premises Equipment)對外閘道。不僅斷線風險高，也造成頭尾車廂的使用者上網延遲比中間的使用者高。本篇論文針對以上描述的問題提出多個對外閘道的解決方法。第一部分我們量測高速行駛下CPE與WiMAX基地台之間的通道品質，並用四個低時間複雜度的即時訊號強度預測方法計算出預測值。預測的結果不僅可以用於規劃與排程網路流量，也可以應用在節省電力與調整多媒體傳輸的編碼方式。第二部分是利用訊號強度預測的結果與多個閘道的即時負載狀況設計出以軟體定義網路(Software-defined networking, SDN)為基礎的系統平衡與成本函數演算法。|
As high-speed trains are getting more common worldwide, the demand of Internet under high-speed is increasing rapidly. Currently, the wireless network nodes in Taiwan high-speed railway were deposed linearly and all of the devices were connected by wireless signal. There is only one outbound CPE gateway to transmit data to ground base station. It will result in a high disconnection rate and will also cause the delay time of the two ends of the train to be higher than the middle part of the train. In this paper, we propose a multi outbound gateway method to solve this problem. In the first section, we measure the wireless channel quality between the CPE and the base station under high-speed, and predict the signal strength by four methods. The results could be the foundation for future research, such as power saving, network traffic scheduling and multimedia encoding algorithm adjustment. In the second section, we design a SDN based load balance algorithm by using the prediction results and the loading of different gateways.