Compression and Combining Based on Channel Shortening and Reduced-Rank Techniques for Cooperative Wireless Sensor Networks

This paper investigates and compares the performance of wireless sensor networks where sensors operate on the principles of cooperative communications. We consider a scenario where the source transmits signals to the destination with the help of L sensors. As the destination has the capacity of processing only U out of these L signals, the strongest U signals are selected, while the remaining (L - U) signals are suppressed. A preprocessing block similar to channel shortening (CS) is proposed in this paper. However, this preprocessing block employs a rank-reduction technique instead of CS. By employing this preprocessing, we are able to decrease the computational complexity of the system without affecting the bit-error-rate (BER) performance. From our simulations, it can be shown that these schemes outperform the CS schemes in terms of computational complexity. In addition, the proposed schemes have a superior BER performance as compared with CS schemes when sensors employ fixed-gain amplification. However, for sensors that employ variable-gain amplification, a tradeoff exists in terms of BER performance between the CS scheme and these schemes. These schemes outperform the CS scheme for a lower signal-to-noise ratio.