Performance of linear receivers based on superimposed training

In this paper, we derive a performance comparison between two training-based schemes for MIMO semi-blind channel estimation. The two schemes are the conventional superimposed training scheme and the more recently proposed data-dependent superimposed pilot scheme. For both schemes, a closed-form for the outage probability and a lower bound of the bit error rate are given. We also determine for the data-dependant superimposed training scheme the optimal allocation of power between pilot and data. Once the optimal data and training power are set, we prove analytically that the optimum data-dependent superimposed scheme always outer- performs the conventional superimposed training scheme.