LMMSE Estimation of Equivalent Noise Variance in Amplify-and-Forward Relay Communication Systems

[[abstract]]For amplify-and forward (AF) relay communication systems with coherent modulation, the destination node requires the channel state information and equivalent noise variances of all diversity links to perform maximal ratio combining. In this paper, we design a linear minimum mean squared error (LMMSE) estimator to estimate the equivalent noise variance of the source-relay-destination (SRD) link in AF relay networks with either a variable or fixed amplification gain. Pilot symbol assisted modulation is employed by the transmitter to facilitate estimation tasks at both the relay and destination nodes. In the case of variable amplification gain, our analyses and simulation results show the mean squared error (MSE) of the estimated equivalent noise variance of the SRD link does not necessarily decrease as the signal-to-noise (SNR) increase. Moreover, the estimation error of the equivalent noise variance of the SRD link results in SNR mismatch which degrades the bit-error-rate (BER) performance of AF relay networks. On the other hand, the AF relay system with a fixed amplification gain has a smaller MSE of the estimated equivalent noise and its BER performance is more robust for the mismatched SNR.[[notice]]補正完畢[[journaltype]]國內[[incitationindex]]EI[[ispeerreviewed]]Y[[booktype]]紙本[[countrycodes]]TW