Dynamic time division inter-cell interference coordination and resource allocation in heterogeneous networks -

Abstract

Heterogeneous Networks are considered a promising solution to meet the exponentially increasing data demand. To make the best use of the small base stations added near the existing macro base stations, each user should be associated to the most appropriate base station, the base stations should cooperate to reduce the inter-cell interference, and each base station should allocate its resources (power and subcarriers) efficiently to its users. In this thesis we studied the time domain inter-cell interference coordination technique for the downlink of Long Term Evolution (LTE), which was proposed in 3GPP release 10. It consists of using the almost blank sub-frames to reduce the interference on the victim users. We proposed an efficient method to optimize the total network performance, and it is divided into two stages: the resource allocation and the cell association. The resource allocation problem for Orthogonal Frequency Division Multiplexing (OFDM) systems is a mixed integer non-linear programming problem and it is NP hard problem. We proposed to use the K-best branch and bound as a sub-optimal solution. The K-best branch and bound proved to have very low complexity at the cost of a slight reduction in the total data throughput. The cell association problem is also a mixed integer non-linear programming; we used the exhaustive search to find the optimal solution because it has a low complexity. We divided the users into three categories: pico cell center users, pico cell edge users, and macro cell users. We used the dynamic cell range expansion to determine the three groups. We proved by simulations that the muting rate of the almost blank sub-frames has an optimal solution, and we calculated the bias of the pico cell center users and the pico cell edge users using the equal average throughput and the maximum minimum rate methods respectively.