Reducing False Detections in Radar Systems: A Two-Stage Approach with Costas Sequences

Radar systems are one of the fundamental building blocks of autonomous navigation, surveillance, and collision avoidance applications. But, angular resolution limitation yields merging of targets which is the limitation in resolving close targets at long ranges. The paper design advanced signal processing technique with the data set of radar by using MATLAB engine. Combining modified Costas sequences with a two-stage positioning, the approach notably enhances the resolution and separation of targets.Furthermore, the achieved results present a significant decrease in sidelobe levels with a PSLR (Peak Sidelobe Level Ratio) improvement from − 13 dB to − 22 dB, which causes less false alerts. Moreover, Doppler processing further improves velocity estimation, which can be used to efficiently identify dynamic objects like pedestrians, bicycles, and vehicles. A heatmap visualization also accentuates target density in intricate settings, emphasizing the advantages of the proposed method. The results offer a more robust and precise radar sensing solution, improving target tracking for autonomous systems.