A base station MMSE combining performance evaluation using Γ distribution function

This paper presents an alternative approach to approximately estimate the performance of the MMSE (minimum mean square error) combining at the base station. The performance of the MMSE combining is evaluated by the CDF (cumulative distribution function) of the output SINR (signal to noise and interference ratio). To evaluate the performance of the MMSE combining, the definition of the equivalent parameters is proposed. Based on the Γ distribution approximation (the moment matching), the equivalent parameters are found by 2 ways: (1) direct computation from simulation SINR, and (2) the signal subspace approach. As a result, the CDF approximated by using the equivalent parameters is compared with the CDF directly derived by the Monte-Carlo simulation. The moment matching is proved to be an effective approach to find the equivalent parameters. However, the accuracy of the approximated CDF by the moment matching dramatically degraded if we consider the situation where the desired signal can not be efficiently combined. In the other words, the equivalent parameters do not give the accurate results when the degree of freedom for desired signal decreases. Nevertheless, they still offer the ability to roughly evaluate the performance of the MMSE combining