Almost Blank Subframe (ABS) is a technique which allows an interfering cell to mute some of its transmissions in selected subframes to reduce interference in these resources to other cells. The set of ABS is called ABS pattern. This way, ABS pattern properly configured can increase the system capacity, while ABS pattern unsuitable can generate significant interference and loss of capacity. In this work, we investigated the effects of the Bernoulli distribution in ABS pattern to Heterogeneous Network (HetNet). The Bernoulli distribution was chosen due to its simplicity of implementation, which allows an intuitive analysis of scenarios with and without ABS pattern. Also, it can be used as baseline for complex algorithms. Results indicate that the ABS pattern through Bernoulli distribution has different performances, depending on the parameterization of the distribution, which can provide capacity gains between 3.7 % and 10.4 % to the system.

The interference suffered by users at the CRE area of pico cells is one of the main factors that affects performance of a HetNet. In order to assist pico downlink transmissions, the Macro-eNB can mute all downlink transmissions to its users in certain subframes termed ABS. In this paper, we propose a genetic algorithm to find optimal ABS pattern that maximizes the network capacity and provides better fairness in resource allocation between all users. Results indicate that the ABS pattern provided by GA achieve a Fainess Jain Index 59,63% better compared with use of random pattern.