Research On Resource Allocation Algorithm In LTE-A Systems With Carrier Aggregation

With the development of mobile communication technology, the demand of the bandwidth is growing. In order to meet the needs of users, ITU proposed several key technologies to expand the bandwidth, such as carrier aggregation, but that its application makes resource allocation become complicated. Based on that, this article focuses on resource allocation optimization problems under the carrier polymerization in the LTE-A system.In this paper, the research on resource allocation is done in two situations, one is for narrow band users; the other one is for narrow band users and broad band users coexisted. Firstly, for narrow band users, research on resource allocation with carrier aggregation in the LTE-A systems is of great importance. In the traditional algorithms, for example, one is aiming at optimal system throughput, and the best throughput can be get, but users suffer serious unfairness; the other is aiming at optimal the fairness of users, and the equitable distribution about resource between users can be get, but the system throughput decreases, in which resources are wasted, because users with poor channel quality also occupy a lot of resources. In the resource allocation for narrow band users, a new algorithm is designed based on utility function with cross-entropy in this article, in which users obtain resources according to their different needs and maximize the satisfaction of users. The algorithm can balance the system throughput and the fairness of users. Secondly, under the narrow band users and broad band users coexisting conditions, BF algorithm and MSU algorithm are two traditional ones. BF algorithm gets the best solution by combining all of the resource allocations, comparing and analyzing the performances. But it is with great time complexity and system overhead; MSU algorithm gets the maximum e satisfaction of customers based on utility function, which is to maximize the system throughput while ensures the fairness of users, but its calculating time is large. In this paper an improved resource allocation method based on cross-entropy is proposed, which can ensure that its performance is close to BF algorithm and MSU algorithm while the calculating time is shortened. And the simulation results show the effectiveness and superiority of the algorithms.