Blind SINR Estimation Algorithm Based on Joint Eigenvalue-Steering Vector Decoupling in Non-Cooperative Interference

To address the limitations of traditional Signal-to-Interference-plus-Noise Ratio (SINR) estimation methods under non-cooperative malicious interference, such as their confinement to specific modulation signals and the narrow effective estimation range of Signal-to-Noise Ratio (SNR), this paper proposes an adaptive robust SINR estimation algorithm based on joint decoupling of eigenvalues and steering vectors. The algorithm first realizes the joint estimation of the number of sources and noise power under finite snapshots via Marchenko-Pastur (MP) law-based iteration, then completes the coarse estimation of desired signal power through steering vector-eigenvector matching and performs strategy switching according to the coarsely estimated Signal-to-Interference Ratio (SIR). In scenarios where the desired signal power is non-dominant, the robust Capon method with adaptive diagonal loading is adopted for fine estimation, and, finally, the interference power is decoupled using the matrix trace identity. The algorithm is independent of modulation characteristics, applicable to both constant-modulus and non-constant-modulus signals, and only requires prior information of the Direction of Arrival (DOA) of the desired signal. Simulation results demonstrate that under complex non-ideal conditions, including 30 dB Interference-to-Noise Ratio (INR), steering vector mismatch, and finite snapshots, the average Root Mean Square Error (RMSE) of the proposed algorithm across the entire SNR range is only 0.83 dB, which significantly outperforms mainstream blind estimation algorithms such as RMT-BSE, M2M4, and SSME. Its robustness and universality are more in line with the requirements of non-cooperative anti-jamming communications.