Pilot Sequence-based Channel Estimation in Massive MIMO wireless communication networks under strong Pilot Contamination

This work provides a straightforward channel estimator to overcome an unrealistic property provided by Minimum Mean Square Error Estimator (MMSEE) for Multi-Cell (MC) Massive Multiple-Input Multiple-Output (M-MIMO) systems operating under Time-Division Duplex (TDD) protocol. Besides, this work is in purpose to study and analyze the current ideal Least-Squares Estimator (LSE), the current ideal MMSEE, and the Maximum Likelihood Estimator (MLE) under various circumstances and considering under Pilot Contamination (PC) problems. This work compared and evaluate the performance of the studied estimators using the metric Mean Square Error (MSE). The traditional LSE provides the worst performance under a high interference level since it is considerably affected by PC. In spite of the greater accuracy achieved by MMSEE in many studies in the literature. However, the MMSEE is relying on an unrealistic assumption, which can be explained by the complete knowledge of among cell large-scale fading (LSF) coefficients as an unrealistic hypothesis in practical use. The suggested estimator (i.e., the MLE) is introduced to overcome the unusable property on which the MMSEE is based. Besides, the MLE is introduced to provides higher performance than LSE. Furthermore, we investigate a scenario of LSF coefficient (i.e., a LSF depends on the distance at which the user is located from its serving Base Station (BS)), wherewith we assert our analysis. An analytical, simulated, and approximated, results are provided for MLE to affirm our study, whereas analytical and simulated results are given for both LSE and MMSEE to assert the presented theoretical expressions.