Performance Evaluation of Service Broker Policies in Cloud Computing Environment Using Round Robin

With its ability to provide scalable and practical solutions for various applications, the cloud computing platform has emerged as a pillar of Information Technology. The Resource Allocation Algorithms (RAA) of the cloud computing platform and its Service Broker Policies (SBP) plays an impactful significance for its performance. Hence, it becomes essential to examine these SBPs used by the RAA. The primary aim of this research paper is experimentally examining the SBPs of the cloud, namely Closest Data Centre (CDC), Optimized Response Time (OptiResTime), and Dynamically Reconfigured (RD) using the RAA Round – Robin (RR). To do so, this paper includes experimenting with a cloud simulation platform and computing tasks in the cloud DCs in various scenarios. The performance parameters used for the study are Overall Response Time (OvrallResTime) and Data Centre Processing Time (DCPT) measured in milliseconds (ms). The experimental results convey that the SBPs CDC, OptiResTime, and RD take an average OvrallResTime of 1310.68 ms, 1310.92 ms, and 8226.72 ms, respectively. Concerning DCPT, the SBPs CDC, OptiResTime, and RD take an average of 1010.71 ms, 1010.66 ms, and 7925.28 ms, respectively. Hence, the SBP CDC outperforms other SBPs in terms of OvrallResTime, and the SBP OptiResTime outperforms other SBPs in terms of DCPT. To enhance SBPs and maximize cloud results, the cloud needs to be embedded with an external intelligence mechanism. Therefore, this paper also presents a Reinforcement Learning model to enhance these SBPs and provide Quality of Service (QoS).