Interference Management for Heterogeneous Networks in 3GPP LTE Advanced and Beyond

Sammanfattning: The primary goals of the thesis study were to evaluate the Closed Subscriber Group (CSG) based Heterogeneous Network (HetNet) deployments and identify the impact of CSG LPN cells on various system performance metrics (cell throughput, UE throughput and interference levels etc.) in contrast to the case when Open Access (OA) LPN cells are deployed in the macro cell area. Moreover, we aimed to analyze the system performance (metrics) in the presence of Inter-cell Interference Coordination (ICIC) techniques based on Time Domain Resource Partitioning (TDRP) and Frequency Domain Resource Partitioning (FDRP). As part of the initial objectives, we aimed to characterize the new interference scenarios consequent from the presence of CSG cells and contrast the performance of the system with the case when we have Open Access (OA) LPNs in the system. Furthermore, we simulated cases where the User Equipment (UEs) had the capability to perform Cell-specific Reference Signal Interference Cancellation (CRS-IC) in contrast to the studies where the UEs were not capable of performing CRS-IC. The simulations were performed in a heavily loaded traffic scenario (15Mbps & 10 LPNs per macro cell area) with 20% of the total users clustered around CSG LPNs. It was observed that in the presence of UEs capable of performing CRS-IC, incorporation of the CSG cells brings about 20% increase in the overall macro cell area T.P for DL and about 47% increase in the same for UL. Furthermore, the system capacity (i.e. the no. of UEs supported) was also seen to increase (17% for DL and 15% for UL respectively). However for the same scenario, when the system contained UEs not capable of performing CRS-IC, CSG cells tend to degrade the system capacity (10% degradation), followed by 18% decrease in the overall macro cell area throughput in DL. This is mainly due to high interference (due to CRS transmission) experienced by the un-authorized macro UEs in the DL. With regard to the case when ICIC techniques are employed in the system, we simulate the Almost Blank Sub-frame (ABS) technique with CSG LPN cells muting their transmission in alternate sub-frames and contrast it to the FDRP scenario where the transmission from the LPN cells is limited to 50% of the total bandwidth (with no muting employed). In general it was observed that, in a macro-femto deployment with CSG cells, ICIC techniques do not improve the performance of macro UEs considerably, when compared to the scenario without ICIC schemes in the system (even for the case when there is high UE clustering in the CSG LPNs - 60% UE clustering). However, when TDRP and FDRP ICIC schemes are contrasted with each other, it was observed that the FDRP scheme slightly outperforms the TDRP scheme by providing 12% gain in macro UE T.P in DL, primarily due to the fact that no CRS are transmitted in the part of the bandwidth not used by CSG LPNs.