Modeling and Resource Allocation for E-Businesses

In this paper, we consider a quantitative approach in modeling customer access pattern and traffic load of the E-Business website and present a scheme of resource allocation based on the customer visits. We model the session of the customer visit as a discrete Markov process, the states of which are e-business functions that serve the users. The traffic load can be calculated based on customer behavior model. Given the models of customer behavior and predicted traffic load, the resource allocation schemes are then presented by using stochastic design and optimization methods. A case study is demonstrated showing the usefulness of the customer behavior model and the effectiveness of the resource allocation scheme.