Improve the problem of insufficient bandwidth and path congestion in the cloud computing environment by using self-adjustment scheduler

Abstract

Users can effectively achieve the purpose of sharing hardware and software resources through the cloud computing. However, when the network load is too heavy or the network bandwidth is insufficient, the cloud computing often occupies a lot of network bandwidth. The operation delay makes the transmission performance decline rapidly. Therefore, the theme of this study focuses on the effective use of network resources to enhance the efficiency of the cloud computing environment. This study proposes a self-adjustment scheduler (SAS) that regulates the operation of the switch by detecting the traffic during the packet transmission. SAS is included by two mechanisms, the flow assignment (FA) and the flow weighted round robin (FWRR). With FA, when the delivering node is congested, the path table is quickly modified so that the next packet can be selected for alternate paths. Through this method, the path can be changed dynamically to deal with packet congestion. FWRR, taking into account the priority of data and the problem of preventing starvation, allows the bandwidth configuration to be dynamically adjusted to achieve fairness. Also, an experimental simulation established by using EstiNet network simulator to verify the SAS mechanism.