Comparison of Abstract Resource Management Model for SON Algorithm of eICIC with Real Radio Resource Management

Mobility load balancing alone does not improve the performance of intra-frequency heterogeneous networks because at higher offloading, users in the range extended region suffer from low throughput and radio link failures. Enhanced Inter-Cell Interference Coordination (eICIC) improves mobility load balancing by protecting users in the range extended region. However, setting optimal blanking pattern configuration and range extension is not an easy task since the algorithm needs to take into account among other things user measurement capabilities, traffic variation due to mobility, mobility-associated characteristics such as radio link failures, etc. Moreover, eICIC is used in HetNets along with other SON features such as mobility robustness optimization and carrier aggregation. Coexistence studies cannot be undertaken by using a radio resource management model that have simulation step equivalent to a transmission time interval because of unfeasible computation time. Thus, an abstract radio resource management model is needed. This paper contributes the comparison of an abstract radio resource management with a real radio resource management based on proportional-fair scheduler. The abstract model uses ideal proportional fair scheduler based on utility maximization whereas the real radio resource management model use metric based proportional fair in the presence of link adaptation, packet error and retransmission. Despite the differences between the two models due to approximations and resource sharing behavior, the comparison shows that both models result in the same trend and order of gain in terms of 5-% user throughput.