Sharpe ratio for user association design in downlink heterogeneous cellular networks

This paper considers the downlink heterogeneous cellular network, which contains multiple types of access nodes, such as macrocells and picocells. A novel utility function maximization technique is proposed based on the Sharpe ratio (introduced by the Nobel laureate in Economic Sciences William F. Sharpe in 1966) for the design of user association to the base stations, and the corresponding resource allocation, e.g. a base station allocates different number of subcarriers to different users associated to it. The proposed user association method evaluates all the users' achievable rates under certain association patterns, and chooses the pattern as the optimal one which maximizes the Sharpe ratio of the mean to the deviation of the achievable rates. Simulation results show that the Sharpe ratio obtained by the proposed method is more than twice as large as that of the previous user association schemes.