I wish you well and apologise for the inconvenience I caused. This communication is regarding the Vienna simulator and the system-level implementation of MMSE. I am aware that the purpose of the Vienna system level simulator is to reduce complexity. Still, I wish to test the theory by implementing MMSE rather than Zero Forcing, the default receiver in the system level simulator.
The Zero forcing receiver is implemented in the simulator file with the name calculate effective channel and receiver, and I am aware that the conversion to MMSE uses this equation (H*HT + noise), which plus the noise power the inner matrix of (H and H transpose). My issue is the type of noise I should use and whether this value is constant without considering the influence of inter-cluster interference. If constant, am I correct in assuming that the Noise power is simply KTB? If not, which power should be added?
Additionally, how can I trace which bandwidth to use in the equation of KTB as I am aware the implementation is done following Mutual Information Effective SNR Mapping (MIESM).
in the 5G system-level simulator, we consider thermal noise and a receiver noise figure. The noise power is calculated for the bandwidth of one resource block, which is the bandwidth considered by the calculations in the link quality model.
However, it seems like you are using the LTE system-level simulator, for which we do not provide support anymore. In the LTE simulator, two post-equalization SINR values per resource block are calculated, and the noise power is handed over to the SINR calculation as noise_W_half_RB.
When implementing the MMSE receiver, do not forget to also adapt the signal power fraction zeta and the inter-layer interference xi, which are not one and zero anymore, as for the Zero Forcing receiver.
I appreciate your answer clearly, for which I am glad. I now fully comprehend the implementation of the LTE version.
Concerning the implementation of MMSE at the 5G system level, are signal power fraction zeta and inter-layer interference xi considered the same as in the LTE version?
Regards,
MMSE is implemented neither in the 5G nor the LTE system-level simulator.
For the Zero Forcing receiver, the signal power fraction zeta is considered to be 1 since the Zero Forcing equalizer inverts the effects of the channel. Thus, the received signal power is identical to the transmitted signal power (excluding macroscopic fading effects - these are considered in the power term in the SINR calculation).
You can try to follow the steps in Section 3.1.1.1 in System Level Modeling, but replace the Zero Forcing receiver (denoted as G and later as H+) with an MMSE receiver for your implementation.