Signal Enhancement Method of Adaptive Filtering Algorithm in Microelectronic Sensor Array

This study optimizes signal processing in microelectronic sensor arrays through adaptive filtering algorithm, improves signal quality and signal-to-noise ratio, and suppresses noise interference. The study aims to solve the problem of effective enhancement of multi-channel sensor signals and provide technical support for high-precision data acquisition and analysis. This paper introduces a sub-band-based DFRCT (Discrete Fractional Cosine Transform)-LMS (Least Mean Squares) algorithm, which uses DFRCT to decompose the collected multi-channel signal in the frequency domain, divides the signal into multiple sub-bands, and performs independent noise suppression and enhancement processing on each sub-band signal. Then, the filter coefficients are dynamically adjusted through the LMS algorithm to minimize the error between the output signal and the expected signal, thereby improving the signal quality. Experimental results show that the signal processed by the proposed algorithm has significant improvements in SNR (Signal-to-Noise Ratio) and MSE (Mean Squared Error) compared with traditional Kalman filtering and Wiener filtering algorithms. When the signal sequence length is 10s, the SNR of the DFRCT-LMS algorithm is 30.5 dB and the MSE is 0.08, which is significantly better than Kalman filtering (SNR is 24.8 dB, MSE is 0.23) and Wiener filtering (SNR is 27.1 dB, MSE is 0.15).