Measurement and Track Fusion with Disparate Sensors

It is an assumed fact that measurement fusion will always produce better results than track fusion at the system level. Decisions have been made given this assumed fact, and many trade-offs have been made due to communication limitations. For some systems, only the source tracks are communicated to the system level due to these constraints. It also depends to some degree on the expected characteristics of the flight path of the object that is being tracked. Therefore, the decision of whether to fuse measurements or source tracks at the system level is made given the communication limitations and the nature of the objects that are being tracked. This paper extends our previous work on this subject. For the first phase of this work, a three-sensor scenario was created with all measurements being sent to the system level where a composite tracker filtered in all of the measurements. In parallel for comparison, the source tracks were sent to the system level where a network tracker fused the source tracks together. The second phase of this work modified the scenario to increase the measurement rate of the local sensor, but only communicated the measurements to the system level at the same rate that the source tracks were transmitted. The rate of the local measurements was at 10 hertz, but measurements and source tracks were sent only at a one hertz rate. The results from this setup produced some interesting take-aways. Now in phase three, the scenario is enhanced to change to make the three sensors in the scenario disparate in their resolution capabilities. The setup, results, and conclusions for this updated scenario are presented in this paper. The overall goal of this effort is to explore the situations when measurement fusion is warranted given the extra "expense" of communications.