I tried to track the precoding matrix, and I found it generated by “1/sqrt(2) * eye(2)” Lines 105 and 106 in generic scenario file. This method seems to assign one antenna at base station to one antenna at user to handle one stream.
My question, how could I change the precoding matrix such that I can generate a MU-MISO / MU-MIMO scenario that uses ZF or the MMSE precoder with more than 2 antennas at base station?
the part of code from the scenario file you described is used when there is no PMI feedback activated. In this case, the precoding amtrix from the scenario file is employed in each transmission frame. Further, in the current version 1.1, spatial multiplexing is only supported in the single user sense, meaning each user is allocated distinct time-frequency resources.
We are currently working on the implementation of MU-MIMO for future releases. However, this transmission mode is merely an issue of generation of the precoding matrix, but rather a problem of scheduling several users the same time-frequency resources.