Sphere detection algorithm is an efficient for MIMO system. As we know, the complexity of the algorithm increased when the dimension increased. Therefore, How it is working in large-scale MIMO systems?
One of the nice things with massive MIMO is that the interference caused between users is compressed by the transceiver linear processing (e.g., zero-forcing). By treating the remaining interference as additive noise, the decoding can be performed separately for each other. Any algorithm can then be used for decoding, including sphere decoding.
As Emil mentioned zero forcing can nullify interference. remaining decoding can be performed by and decoder, even with sphere decoding. The complexity of SD would be the same as of conventional mimo.
can we relate impact of pilot contamination on capacity in massive MIMO system
The exact capacity is unknown, but there are many lower bounds on the capacity in the massive MIMO literature. These lower bounds are impacted by pilot contamination. Additional interference is caused between users that use the same pilot.
Does anyone have this sphere detection algorithm for Massive MIMO system matlab code? thank you
We can exploit the spatial structure of mmWave channels to formulate the precoding/combining problem as a sparse reconstruction problem. And then apply compressed sensing to solve the underdetermined linear system of equation. Generally, CS has two key advantages, it minimizes mean square error (MSE) and enables the reconstruction of sparse or compressible signals with far fewer samples than those required by Nyquist theorem. The essential ingredients of CS to be linked together are:(1) the signal to be reconstructed needs to have a sparse representation, and (2) the reconstruction process. One such algorithm is the orthogonal matching pursuit (OMP)