Methods and systems for performing compressed time domain joint channel estimation in a multi-user MIMO LTE wireless network include receiving training signals from a plurality of users, estimating a maximum delay spread for the received data according to a coherence bandwidth of the received data, limiting the received data in the time domain to the estimated maximum delay spread, selecting and estimating an active tap from the limited data set, and subtracting a contribution of the selected active tap from the reduced data set. These steps can be repeated until the residual signal falls below a specified minimum. The network can be a C-RAN network. The training data can be SRS or DMRS data. Limiting the received data ensures that only a few significant taps are analyzed, so that the system is not under determined and can be analyzed for accurate channel estimation using any of several existing algorithms.