Before The Weak Target Detection Radar Tracking Technology Research

The development of stealth technology makes stealth targets Radar Cross Section(RCS) increasingly reduced. And it will make radar’s power scope and detectionperformance greatly weakened. It is possible by increasing the transmit power, antennaaperture and gain to compensate for the reduction of the radar sensitivity due to thedecrease of the target RCS, but these will be restricted by the real problem. How toimprove the capability of radar weak target detection and tracking becomes anextremely important and serious mission in the modern radar system.Comparatively speaking, it is flexible and low cost means if use signal processingmethods for small target detection and tracking. In the strong clutter, strong interference,low signal to noise ratio, track-before-detect algorithm uses multiple scans accumulatedto improve signal to noise ratio, thus improving the capability for small target detectionand tracking. The main work and innovation are summarized as follows:1. It introduces the track-after-detect and track-before-detect algorithms and sumsup tow advantages by compared the algorithms.2. For the issue of track-before-detect algorithm of one-target based on dynamicalprogramming, a new track-before-detect algorithm based on improved dynamicalprogramming algorithm is proposed to detect weak target. The improvement in the newalgorithm is that the latter successive states are estimated when deciding the transformstate, besides processed two times of CFAR and limited the area of search.3. It can cover coherent integration for non-coherent integration for the uniformmotion target which far away from the radar. The improvement in the new algorithm isthat for take full advantage of echo signal’s phase information cover coherentintegration for non-coherent integration and a novel motion compensation method basedon keystone transform is adopted. The simulation of real data confirms that thisalgorithm improves the detection performance of radar target.4. For the issue of the classical PF-TBD algorithms a new track-before-detectalgorithm based on improved particle filter algorithm is proposed. For take fulladvantage of the information of measure an updating strategy is proposed and the updating strategy is that newborn particles are uniformly distributed within the set withhigh-intensity bins. Besides the bootstrap particle filtering has been applied and it canescape successfully from the sample impoverishment problem.