Abstract:

To address issues in traditional subband detection algorithms in the field of dynamic digital channelized reception field, such as the need for prior information about signals and noise. An eigenvalue fusion based dynamic channel subband detection algorithm is proposed. Firstly, based on random matrix theory (RMT), the maximum, minimum, and average feature values in the sampled covariance matrix are employed, incorporating a fusion parameter α to construct a fusion detection statistical value. Subsequently, an efficient detection threshold is derived through the limit distribution of the minimum feature value. Based on this, a subband blind detection algorithm based on feature-value fusion is designed, which is named as the α-maximum-average-minimum eigenvalue (α-MAME) algorithm. In the experiment, the algorithm's performance is simulated and verified under various receiving conditions of dynamic digital channels. The experimental results indicate that compared to existing algorithms, the proposed subband detection algorithm demonstrates superior detection performance under low signal-to-noise ratios and low-dimensional conditions.

Key words: dynamic digital channelization, subband detection, eigenvalue detection, random matrix