A unified approach for representing wireless channels using EM-based finite mixture of gamma distributions

We present a unified framework to evaluate the error rate performance of wireless networks over generalized fading channels. In particular, we propose a new approach to represent different fading distributions by mixture of Gamma distributions. The new approach relies on the expectation-maximization (EM) algorithm in conjunction with the so-called Newton-Raphson maximization algorithm. We show that our model provides similar performance to other existing state-of-art models in both accuracy and simplicity, where accuracy is analyzed by means of mean square error (MSE). In addition, we demonstrate that this algorithm may potentially approximate any fading channel, and thus we utilize it to model both composite and non-composite fading models. We derive novel closed form expression of the raw moments of a dual-hop fixed-gain cooperative network. We also study the effective capacity of the end-to-end SNR in such networks. Numerical simulation results are provided to corroborate the analytical findings.