Distributed Data Fusion Overarching Design Concerns and Some New Approaches
Nicholson D., Reece S., Rogers A., Roberts S., Jennings N.
This chapter explores some of the design concerns associated with distributed data fusion (DDF) systems and describes how they could be overcome. It introduces several critical design concerns that must be resolved if a multi-agent DDF system is to succeed in practice. The chapter describes the resolution of information recycling concerns a technique known as bounded covariance inflation (BCI). It examines the needs for coordinated actions within a DDF system and describes how this can be achieved with the max-sum algorithm. The chapter deals the chapter with some perspectives on the new design challenges that will be raised by future DDF systems that achieve effect by tightly interleaving human and software agent endeavors. The sensors themselves must be equipped with capability to self-organize and coordinate sometime after deployment once the local environment in which they find themselves has been determined. Three sensor management strategies were implemented: local, centralized, and decentralized.