Search

In this paper we analyze the energy efficiency of single-hop, multi-hop, cooperative selective decode-and-forward, cooperative incremental decode-and-forward, and even the combination of cooperative and non-cooperative schemes, in wireless sensor networks composed of several nodes. We assume that, as the sensor nodes can experience either non line-of-sight or some line-of-sight conditions, the Nakagami-m fading distribution is used to model the wireless environment. The energy efficiency analysis is constrained by a target outage probability and an end-to-end throughput. Our results show that in most scenarios cooperative incremental schemes are more energy efficient than the other methods. [ABSTRACT FROM AUTHOR]

We consider the use of spatial diversity to improve the performance of analog joint source-channel coding in wireless fading channels. The communication system analyzed in this paper consists of discrete-time all-analog-processing joint source-channel coding where Maximum Likelihood (ML) and Minimum Mean Square Error (MMSE) detection are employed. By assuming a fast-fading Rayleigh channel, we show that MMSE performs much better than ML at high Channel Signal-to-Noise Ratios (CSNR) in single-antenna wireless systems. However, such performance gap can be significantly reduced by using multiple receive antennas, thus making low complexity ML decoding very attractive in the case of receive diversity. Moreover, we show that the analog scheme can be considerably robust to imperfect channel estimation. In addition, as an alternative to multiple antennas, we also consider spatial diversity through cooperative communications, and show that the application of the Amplify-and-Forward (AF) protocol with single antenna nodes leads to similar results than when two antennas are available at the receiver and Maximal Ratio Combining (MRC) is applied. Finally, we show that the MMSE implementation of the analog scheme performs very close to the unconstrained capacity of digital schemes using scalar quantization, while its complexity is much lower than that of capacity-approaching digital systems. [ABSTRACT FROM PUBLISHER]