By Moritz Luca Schmid | June 22, 2018
I expanded the 2×2 V-BLAST zero-forcing scheme to a general MxM scheme. For a general MxM matrix, the inverse cannot be precalculated by hand as I did for the 1×1 and 2×2 scheme. The inversion of a possibly large channel matrix can quickly get a very complex problem which needs a lot of computation power, especially if it is recalculated frequently, for example each micro second when assuming a coherence time of 1μs and therefore a CSI update rate of 1MHz.