Error Exponents for the Correlated Nakagami Fading Channel with Maximal -Ratio Combining

In this paper, we derive the random coding error exponent, or reliability function, for flat Nakagami fading channels with receiver antenna diversity. We consider maximal ratio combining (MRC) with an arbitrary number of diversity branches in correlated Nakagami fading channels, where the instantaneous signal-to-noise ratios (SNR's) of the diversity branches are not necessarily independent of or identically distributed. Specifically: 1) these SNR's can be arbitrarily correlated; 2) the SNR distributions can be from different Nakagami families, i.e., fading parameters (m's) are not necessarily equal; and 3) the average SNR's (averaged over the fading) of the branches are not necessarily equal. The proposed problem is mad analytically tractable by transforming the physical diversity branches into the "virtual branch" domain which makes it feasible to derive the error exponent in correlated Nakagami fading. We consider channel inputs which belong to continuous alphabets.