Cloud Based Architectural Framework For Implementation Of Erp In Small And Medium Enterprises

Abstract

: Cloud computing has become one of the latest application used in Information Technology industry. Cloud based Enterprise Resource Planning systems is one of the optimal substitute for those organizations that are using traditional ERP system. Organization using cloud ERP requires low cost. This thesis unravels cloud ERP adoption on the basis of company size by identifying and classifying the opportunities. The huge research is performed to integrate the ERP with cloud. ERP deployment and its factors, problems of performance and implementation have been studied in the past, but findings are not up to date with regard to having the best cloud service provider. As list of CSPs are available and it is very difficult from the customer that which CSP provide the best services. So to solve these problems, Quality of Service attributes and Service Level Agreement is used. By using SLA and QoS attributes, selection from the set of the comparable services available in the global market become easier. Cloud brokers are taken into the account while choosing the best CSP. Multiple organizations are available that provide the cloud services such as Amazon, HP, and IBM. However, no technique is available to measure the real-world usage and estimate the ranking of the CSPs. This thesis proposes a ranking prediction framework which meets the QoS. To validate the performance of framework three approaches are proposed named Cloud Rank3, BOOST, and MOEPO. Experiments are performed on the real world data by using EC2 services of Amazon. In Cloud Rank3, both Cloud Rank1 and Cloud Rank2 are hybridized which compares the results with well known approaches to filter the results and to rank the CSPs. Next approach named BOOST used the modified preference function and the weight factor to predict the ranking. The various experiments are performed on real life QoS dataset. Such experimental results show that Cloud Rank3, BOOST provides optimal results. The MOEPO approach is used to tuning the parameters of Cloud Rank3 and BOOST algorithm