Efficient computafion of MRC diversity performance in Nakagami fading channel with arbitrary parameters

Abstract

Based on the assumptions of zero-mean white noise and independent sources, we can find the estimate SZ of Q from the minima of(l/N)C,N=, E,Z(0, w) (with respect to 0) which lie below a chosen threshold. Sources whose powers cause the corresponding minima to lie above the threshold (i.e. weak sources) will not be detected. Simulation results: Simulations have been run on the computer for L = 24, N = 100 (i.e. a ULA of 24 sensors using 100 snapshots) for both a single source and two independent equipower sources, corrupted by additive white noise of mean zero when dih = 0.5. The results are shown in Tables 1 and 2. The accuracy of the estimates and the resolution is not as high as those of MUSIC and associated subspace methods [I, 21. However, the method has two advantages over subspace methods: (i) the processing is simpler because on-line eigendecomposition of the data correlation matrix is not needed, and (ii) the number of sources present need not be known a priori.

Adaptive generalized selection combining (A-GSC) receivers

“…In particular, the receiver first estimates the channel gains and phases for all the diversity paths. Then, starting from the best path, the receiver tries to increase the combined SNR γ c above the threshold γ T by adding more diversity branches unless the SNR of the next branch is below the value μγ (1) . More specifically, if γ (1) > γ T , then only one branch will be used.…”

“…Otherwise, the receiver will test the SNR of the next strongest branch (i.e. γ (2) ) against the value μγ (1) . If γ (2) < μγ (1) then the selection process terminates and the second branch is rejected, i.e.…”

“…an L-branch SC combiner is used. If γ (2) > μγ (1) the receiver adds the SNR of those two branches, i.e. γ c = γ (1) + γ (2) and tests it against the output threshold γ T .…”

“…γ c = γ (1) + γ (2) and tests it against the output threshold γ T . This procedure is continued until either the combined SNR is above γ T , or the SNR of a single branch is below μγ (1) , or all L branches are activated. In other words, the selection process terminates, if the combined SNR after the activation of the first l branches is above the threshold, or the ratio of the SNR of the (l + 1)th branch to that of the first branch is below the required threshold μ ∈ [0, 1].…”

“…The optimal diversity combining scheme is the well-known maximal ratio combining (MRC) [1]- [3], which attains the highest output signal-to-noise ratio (SNR) of any combining scheme, independently of the distribution of the branch signals. The performance gain of MRC comes at the cost of increased hardware complexity and power consumption since all the available signal replicas are combined.…”