Load balancing in self-organized heterogeneous LTE networks: A statistical learning approach

The continuous evolution of cellular communication networks into dense, dynamic and heterogeneous networks has posed new challenges for system configuration as well as coverage and capacity optimization, especially in areas with unequal user traffic distribution. In a mixed macro/small (or heterogeneous) cell scenario, load balance is one of those challenges since users typically select the base station with the highest received signal power. Hence, the higher transmit power of macro-cells causes difficulties in offloading a sufficient number of users to small cells. This paper propose a Self-Optimizing Cell Range Expansion Scheme based on a statistical learning approach for an LTE heterogeneous network. System level simulations show the effectiveness of this approach in dynamically expanding the small cell coverage according to traffic conditions, balancing traffic load, reducing cell congestion, and diminishing packet losses.