Topology Reconstruction Approach for Distributed Blind Equalization Over Sensor Network
Sulin Chi, Tetsuya Shimamura · 2024
Distributed blind equalization in wireless sensor networks (WSNs) attracts much attention in the field of in-network processing since the data signal can be recovered without the training signal and the desired output at the receiver. Therefore, it has a wide range of applications in various fields such as science, medicine, and agriculture. However, the performance of the distributed blind equalization is susceptible to the transmission channel conditions, especially for the ill-channel conditions. To overcome this problem, we propose a topology reconstruction approach to discard the affect of the ill-channel condition based on the analysis of the received signal for all sensors in each local sensor network (LSN). In addition, the weights of each sensor node are re-designed in order to further improve the prediction accuracy of blind equalization. In the computer simulations, average mean square error (a-MSE) and average symbol error rate (a-SER) are used to evaluate the performance of the conventional and proposed methods. Compared with the conventional (state-of-the-art and non-cooperative) methods for distributed blind equalization, the proposed method achieves better a-MSE and a-SER.