Asymptotically Tight Performance Bounds of Diversity Receptions Over $\alpha$-$\mu$ Fading Channels With Arbitrary Correlation

Abstract

Exact performance analyses over correlated α-μ fading channels result in intractable multifold integrals or infinite series. In this paper, upper and lower bounds are derived for outage probabilities and error rates of maximum-ratio combining (MRC), selection combining (SC), and equal-gain combining (EGC) over the α-μ fading channels with arbitrary correlation. The derived bounds are in closed form and asymptotically tight. Our applied correlation structure uniquely determines the joint probability density function of the α-μ random variables, thus it is a generalization of the other correlation models based on branches' hyperpower or power correlation. Based on these closed-form bounds, one can easily estimate the performance of MRC, SC, and EGC in large signal-to-noise ratio region, where performing Monte Carlo simulation is time consuming. These new bounds also reveal a fact that SC can in fact outperform EGC under some certain channel conditions. More importantly, these bounds reveal which factors determine the asymptotic performance of diversity reception systems, and provide criteria for system design.