On simulating propagation for OFDM/MIMO systems with the MR-FDPF model

Radio propagation tools are needed to help operators to find the best setting of their network, including but not limited to access points positioning, radiated power optimization, and channel selection. Hence, different approaches proposed in the literature deal with this issue. Empirical models suffer from an unacceptable lack of accuracy while deterministic models have to cope with exponential computational complexity. Geometric models like ray tracing [1], [2] have been extensively developed as they offer a good trade-off between computational load and accuracy but they fail in simulating properly severe environments where multiple diffractions and numerous reflections hold. Second, more accurate methods based on the resolution of the Maxwell´s equations have been implemented [3], [4] but they suffer from a high computational load. The MR-FDPF (MultiResolution Frequency Domain Partial Flows) approach [5] was proposed to fill in the gap by developing a multi-resolution preprocessing for the frequency domain TLM approach. However, the current challenge for radio propagation tools not relies on providing signal power levels but beyond on providing a realistic prediction of the system performance, e.g. bit error rate (BER) or throughput estimation. From a snapshot of the current main standards, it can be highlighted that most of them exploit OFDM and MIMO principles. Any system level simulator should rely with OFDM and MIMO prediction requirements. This paper investigates the appropriateness of the MR-FDPF approach for such task.