Study on Transmission Performance Enhancement Using Adaptive Path Selection in Software Defined Networks

• Main Authors: Shih-Chang Wu, 吳石昌
• Other Authors: Yeong-Sheng Chen
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/55101876021332089278

碩士 === 國立臺北教育大學 === 資訊科學系碩士班 === 103 === Routing Control is a useful feature in Software Defined Networks (SDNs). A routing path with high bandwidth and low packet loss rate can enhance the transmission throughput. So, identifying the best transmission routing path is an important issue in the SDN environments. However, because of the dynamic nature of the traffic in a network, the best transmission path should be dynamically and adaptively determined. In addition, many network factors such as packet delay and packet loss will also affect the transmission performance. Thus, in this study, an adaptive path selection scheme based on considering the bandwidth, packet delay, and packet loss of all the links in the network is proposed. Moreover, because of the different transmission mechanisms in TCP and UDP traffics, different schemes are proposed for these two traffic types.This study proposes two schemes for determining the routing path to maximize the throughput. In Scheme 1, at first, all the paths between source and destination are derived, and then the TCP transmission mechanism and the UDP throughput formula are utilized to develop the metric formulation for path selection. However, finding all paths is a NP problem, which incurs high computational complexity. To reduce the computation cost, formulation for determine the link weights is developed in Scheme 2. In our design, a low link weight implies the link can achieve high transmission performance. Thus, the Bellman-Ford algorithm is used to derive the minimum weight path, which should have the best transmission performance. Our simulation results show that the transmission paths derived by Scheme 1 and Scheme 2 are similar, and they are either the best or the second best transmission paths that achieve the highest throughput.