LOF Clustering-Based Adaptive CFAR Detection for Multi-Target ISAC

In multi-target integrated sensing and communication (ISAC) scenarios, conventional constant false alarm rate (CFAR) detectors often encounter performance degradation due to their reliance on fixed parameter settings, which lack adaptability to dynamic environments. By leveraging the local outlier factor (LOF) clustering algorithm, an adaptive detector named LOF-CFAR is proposed to deal with this challenge. Considering the sea clutter environment, LOF-CFAR identifies the interfering targets and sea spikes as undesirable outliers within the reference window, effectively mitigating the influence of target masking and improving the accuracy of clutter background estimation. Through simulation comparisons with conventional detectors that struggle to adapt to varying conditions, the proposed LOF-CFAR method demonstrates superior detection performance in multi-target scenarios, even without prior knowledge about the distribution and quantity of interfering targets.