This is a theoritical question . Iam unable to find how exactly decoding of users at the base station occurs in LTE uplink MU-MIMO. Since it is known that a single RB can be accessed by many users at the same time . So. at receiver how receiver can distinguish these co-scheduled users who are allotted same RB . How rates can be assigned such that these co-scheduled users can be decoded . I meant everything i posted is in particular to Up-link where we have to limit control overhead . Please some one help me in this it would be of great help in my project... . If there is some source which explaind decoding procedure please mention so that i can go through it and learn .
I had the same question in mind when I was reading about MU-MIMO. I did some research, but could not find any detailed answer. However, I did read somewhere that eNB differentiates the users in MU-MIMO mode with the help of uplink reference signals; not much details were present.
I am not sure, but I think eNB does not use UL DMRS, but use UL SRS (sounding reference signals) instead to differentiate among users, as SRS use Zadoff-Chu sequences.
eNB creates the pair of the users for MU-MIMO on the basis of the correlation of their channel response. If the correlation is below a specified threshold, the MU-MIMO pair is made. At the eNB receiver side, when two UEs (paired for MU-MIMO) transmit simultaneously on the same frequency and time resources (same RB), the pilot or RS of each UE will show a separate channel response on equalization. The UE with higher SNR will be decoded after equalization and SIC (successive interference cancellation) is used to estimate the second UE's signal. The combining technique at the receiver should be IRC instead of MRC for MU-MIMO to mitigate the interference in this scenario.