In this paper, radar detection of Swerling target of type 3 in GO-distributed clutter via dynamic programming based track-before-detect (DP-TBD) is considered. Conventional DP-TBD suffers from significant performance loss due to the high frequency of target-like outliers. To enhance the detection performance of DP-TBD, the log-likelihood ratio (LLR) of the measurements is derived and used in the integration process of DP-TBD taking the place of amplitude that used in the conventional DP-TBD strategy. Various simulations are used to examine the detection performances of different DP-TBD strategies. Simulation results show that significant performance improvement can be achieved, especially for vary heavy-tailed clutter.
展开▼
