Research on SaaS Service Performance Prediction Method in Dynamic Resource Environment

Users behavior´s uncertainty and service resource dynamic´s change make the prediction of SaaS performance trend more complex and difficult. This paper proposes a set SaaS performance prediction approach based on time series, defines three key indexes including SaaS Service Transactions Index (STI), Service Resource Occupancy Index (SROI) and Service Performance Index (SPI), and describes the computing methods of the three indexes´ time series. The paper also presents a fuzzy matching algorithm of SaaS performance prediction, and designs experiments to identify the effectiveness of the approach.