Theoretical Analysis of Interference by Non Colocated eNodeB
Here I write about a simulation analysis of the system performance deterioration caused by the following four types of interference: eNodeB->UE, UE->eNodeB, eNodeB->eNodeB, and UE->UE when two systems are not co-located.
From simulation result with no filtering, when the eNodeBs are not co-located, the most serious interference is from eNodeB->eNodeB. The network coverage and capacity losses caused by the other three types of interference are smaller than 2% even if the guard band is not provided. Hence, our focus of analysis is the interference between eNodeBs.
Interference Simulation Analysis with Different ACIRs
When the interfering eNodeB is close to the victim eNodeB, the eNodeB->eNodeB interference can causes serious performance deterioration. This is especially the case when the eNodeBs are from different systems and operating under different bandwidth. However, as the eNodeBs’ separation and Adjacent Channel Interference Ratio increase, the network coverage loss and capacity reduction gradually decrease.
Assuming the offending eNodeB is in transmitting state and the victim eNodeB is in receiving state. Both uplink coverage and capacity loss will occur.
Coupling loss between the eNodeBs is calculated based on the free space propagation model by taking into consideration of eNodeB antenna gains caused by the direction angle and tilt. Figure below shows the simulation result of coverage change due to distance variations between the interfering eNodeB and interfered eNodeB.