Recent advances in track fusion techniques

Abstract

Multiple sensor data fusion has become an essential part of target tracking in recent years. The use of multiple sensors on single or multiple platforms requires fusion of data to provide a true synergistic use of the sensors. Applications include airborne target tracking, ground based sensor netting, fusion of JTIDS data, integrated navigation, and multi-spectral seekers in missiles. A comparison of the relative advantages of track fusion (sensor-level fusion) and measurement fusion (central-level fusion) suggests that although measurement fusion is technically superior, in many scenarios track fusion is more appropriate (e.g. the target tracking algorithms may be optimised to the sensor). This paper considers recent advances in track fusion algorithm design from experience in the development of generic track fusion algorithms for sensors with very different characteristics. (4 pages)