Aiming at the problem that the local optimal detector (LOD) has complex detection structure and poor robustness for signal detection in Non-Gaussian noise background

an adaptive limiter detector (ALMD) is proposed based on the improvement of traditional limiter. Firstly

the detection performance of weak signal is systematically studied. Then the non-Gaussian background is modeled by the mixed Gaussian model. On this basis

the analytical expression between the clipping threshold and the detection performance is obtained. Finally

the optimal threshold is determined which leads to a significantly improvement of the detection performance. Simulation results shows that the ALMD has similar performance compared to LOD but provide relatively efficient and robustness.