Found a paper about Radar-to-ESM correlation written by Chinese scientists in 2000. I first saw this paper at the library of a British university.

http://www.wanfangdata.com.cn/qikan/periodical.articles/hkxb-e/hkxb2000/0002/000204.htm

However, this is not about radar data fusion at measurement/plot or track level.

—–

QUADRI-THRESHOLD ALGORITHM OF CO-LOCATED RADAR-TO-ESM CORRELATION

WANG Guo-hong , HE You
(Institute of Radar 402 , Naval Aeronautical Engineering Academy, Yantai 264001, P.O. Box 109, Shandong Province, China)
MAO Shi-yi
(Beijing University of Aeronautics and Astronautics, Beijing 100083, China)
WANG Yong-an
(4330 Factory, Qixia 264001, Shandong Province, China )

Abstract:　A quadri-threshold radar-to-ESM correlation algorithm is presented and the corresponding four thresholds (high threshold, middle threshold, low threshold, and threshold margin) are given. The proposed algorithm is especially suited for the scenarios where each radar track is specified by different numbers of measurements. Using a simple and effective discriminant function based on the fuzzy synthesis function, the proposed quadri-threshold radar-to-ESM correlation algorithm first selects the two most likely radar tracks and then makes the soft decision based on the selected two tracks. The introduction of the threshold margin in the presented algorithm can further reduce mis-classification errors. Simulation results show the feasibility of the algorithm.

Key words:　radar; ESM; correlation
CLC number:　TP202　　Document code:　A
Article ID: 1000-9361(2000)02-0086-05

In recent years, radar-to-ESM correlation has been widely studied［1～9]. Use of radar and ESM sensors increases the likelihood of target acquisition and reduces the vulnerability to jamming. Radar-to-ESM correlation was first studied by Coleman［1］. Trunk and Wilson［2］ considered the problem of associating a set of n ESM angle measurements with one of m possible radar tracks, where each radar track was specified by a different number of measurements and the cumulative probability of the random variable with a chi-squared density was used as the discriminant function. In order to alleviate the computational burden, the closeness score for correct association in Ref.［3］ and fuzzy synthetic function in Refs.［4,5］ were used as discriminant function, respectively.

Based on the fuzzy synthetic function, a tripe-threshold decision algorithm of radar-to-ESM correlation was presented in Ref.［5]. In order to reduce the mis-classification when radar tracks are close in azimuth, a quadri-threshold radar-to-ESM correlation algorithm is proposed in this paper based on the work of Ref.［5]. The proposed quadri-threshold radar-to-ESM correlation decision rule is based on the two most likely radar tracks and, therefore, can further reduce mis-classification errors.

—–

Cheers,
Sunho