Utility Function-based Optimal Resource Allocation with a Mixture of Reallocation-tolerant and Reallocation-intolerant Users

摘要

In this work, we consider the problem of utility function-based resource allocation when a mixture of reallocation-tolerant and reallocation-intolerant users are present. Unlike reallocation-intolerant users, reallocation-tolerant users can be reallocated a different amount of resource during the course of their call. We develop a resource allocation mechanism that maximizes the average aggregate utility per unit time. By formulating the resource allocation problem as a Markov decision process (MDP), we determine the optimal quantity of resource to be allocated to newly arriving and the optimal reallocation of resources to reallocation-tolerant calls whenever there is a change in the state of the system. We present numerical results that show that our resource allocation scheme performs better than the greedy resource allocation scheme. To reduce the computational complexity involved in determining the optimal policy, we identify problem-specific model reduction techniques that do not compromise the optimality of the solution.