Dr. Joseph Lewis, in cooperation with Lockheed martin, developed a generic framework for building representations of an environment using a wireless sensor network. The distributed, emergent, layered, perceptual architecture, or DELPA, creates a large number of competing hypotheses regarding the environment being modeled. The use case examined was a sensor network spread across outdoor landscape and used to detect wild fire. This paper discusses a method of weighting the hypotheses, called the synthetic particle filter, in order to determine which hypotheses are most likely to accurately represent the environment. In the case of detecting wild fires, the synthetic particle filter examines the spatial characteristics of the hypotheses not considered by the rest of the architecture. The particle filter was tested using in a variety of simulated environments and hypotheses. The particle filter performed well when sorting hypotheses with good spatial characteristics from those that had poor spatial characteristics. It performed less well when tasked with sorting hypotheses with similar spatial characteristics
