Search

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]

We compare the energy efficiency of some transmission schemes in wireless sensor networks. By constraining the system to end-to-end throughput requirements, we assume that the nodes in multi-hop and cooperative schemes operate at a spectral efficiency which is twice that of the single-hop transmission. By taking into account the energy consumption of the RF circuitry, we show that cooperation may be considerably more energy efficient than non-cooperative schemes even in small transmission ranges, specially if a feedback channel is available. [ABSTRACT FROM AUTHOR]

Wireless Sensors Networks (WSNs) are an essential element of the Internet of Things (IoT), and are the main producers of big data. Collecting a huge amount of data produced by a resource-constrained network is a very difficult task, presenting several challenges. Big data gathering involves not only periodic data sensing, but also the forwarding of queries and commands to the network. Conventional network protocols present unfeasible strategies for large-scale networks and may not be directly applicable to IoT environments. Information-Centric Networking is a revolutionary paradigm that can overcome such big data gathering challenges. In this work, we propose a soft-state information-centric protocol, ICENET (Information Centric protocol for sEnsor NETworks), for big data gathering in large-scale WSNs. ICENET can efficiently propagate user queries in a wireless network by using a soft-state recovery mechanism for lossy links. The scalability of our solution is evaluated in different network scenarios. Results show that the proposed protocol presents approximately 84% less overhead and a higher data delivery rate than the CoAP (Constrained Application Protocol), which is a popular protocol for IoT environments. [ABSTRACT FROM AUTHOR]