The purpose of this topic is to develop novel highly parallelizable exact and heuristic solutions to the sensor fusion, path planning, and target assignment problems for multiple munitions. In the early part of cooperative control, most of the research was done through a centralized decision process where all the raw data from the unmanned aerial vehicles were collected and processed in a single data base. When it became apparent that the communication architecture could not support the high bandwidth requirements of moving all relevant information to a central processor, the research community began to look at decentralized heuristics for combining measurement data, coordinating flight plans, and other cooperative endeavors. A huge disadvantage of decentralized solutions is that it is suboptimal. They are never better than the solution derived using all available information and their solution quality is difficult to evaluate. Another bottleneck is the limited computational resources on munition. With these earlier difficulties in mind, we are starting to look at using the wirelessly connected munitions as a computer grid that is capable of solving the cooperative control problems in a centralized and optimal manner. The idea of a computer grid on a wireless network brings unique challenges since data links are noisy and continually reconfiguring itself, some processor may leave the grid temporarily or permanently. We are looking to develop and analyze centralized algorithms for collaborative control of munition systems that can robustly adjust to a constantly changing computing environment.

 

key words

Optimization; Sensor fusion; Path planning; Target assignment; Computer grid; Wireless;

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral and Senior applicants