No abstract

No abstract

This paper presents the performance analysis of a decode-and-forward (DF) cooperative relaying (CR) scheme using quadrature amplitude modulation (QAM) in the presence of Rayleigh fading and Bernoulli-Gaussian impulsive noise. The exact symbol error probability (SEP) expression for direct transmission (DT) and SEP lower bound for DF-CR are first derived and then used to establish the optimum power allocation (OPA) for the source and the relay by exhaustive search. Analytical and simulation results for various scenarios with DT and DF-CR under the same bandwidth efficiency and power consumption are in good agreement and indicate that the lower bound SEP is very tight for the optimal Bayes receiver and CR in an impulsive noise environment can be beneficial at a certain degree depending on impulse power and impulse rate. Furthermore, OPA brings a negligible performance improvement as compared to equal power allocation under investigated conditions.Keywords-decode-and-forward cooperative relaying scheme; impulsive noise; Bernoulli-Gaussian process; lower bound; Bayes receiver I.

Energy savings in orthogonal frequency division multiplexing (OFDM) systems is an active research area. In order to achieve a solution, we propose a new cooperative relaying scheme operated on a per subcarrier basis. This scheme improves the bit error rate (BER) performance of the conventional signal‐to‐noise ratio (SNR)‐based selection relaying scheme by substituting SNR with symbol error probability (SEP) to evaluate the received signal quality at the relay more reliably. Since the cooperative relaying provides spatial diversity gain for each subcarrier, thus statistically enhancing the reliability of subcarriers at the destination, the total number of lost subcarriers due to deep fading is reduced. In other words, cooperative relaying can alleviate error symbols in a codeword so that the error correction capability of forward error correction codes can be fully exploited to improve the BER performance (or save transmission energy at a target BER). Monte‐Carlo simulations validate the proposed approach.

No abstract