Exploiting the Rules of the TF-MUSIC and Spatial Smoothing to Enhance the DOA Estimation for Coherent and Non-stationary Sources

Abstract

The paper introduces a combination of several techniques and methods to tackle the problem of Direction-of Arrival (DOA) estimation for coherent and non-stationary signals under severe noise conditions. Until now, these properties of the signals were studied by many researchers, yet the solutions were developed separately. At the first stage of the algorithm, the Spatial Time-Frequency Distribution (STFD) matrix is derived, which accounts for non-stationarity of the signals and allows denoising. The forward-backward spatial smoothing is applied to the STFD matrix to solve the problem of signals coherency. The main principle of the Root MUltiple SIgnal Classification (Root-MUSIC), namely, solving for the roots of the polynomial to obtain the signals DOA is exploited. The algorithm was simulated with coherent chirp signals with different Signal-to Noise Ratio (SNR) values to observe its efficiency compared to the conventional Root-MUSIC and Time-Frequency MUSIC TF-MUSIC) methods.

More by Ruslan Zhagypar

• LyRA-FL: A Scheduler for Federated Fine-Tuning of Foundation Models in Wireless NetworksIEEE Transactions on Mobile Computing · 2026
• Clustering and Aggregation for Wireless Federated Learning Under Label-Flipping AttacksIEEE Wireless Communications Letters · 2026
• UAV-Assisted Unbiased Hierarchical Federated Learning: Performance and Convergence AnalysisIEEE Transactions on Machine Learning in Communications and Networking · 2025 · Citations: 22
• Communication-Aware Optimizers for FedMeta Learning in Aerial Wireless NetworksIEEE Wireless Communications Letters · 2024