Fusion Orchestration Guidelines (FOG) for Collaborative Computing and Network Data Fusion

In recent years, there has been the opportunity to execute data fusion methods onboard a multi-sensor system. Typically, the motivation was to process the sensor data at the collection location, but the processing power was limited. Hence, for many decades, the limitation of edge-based (or far edge) was only available to systems with small-data message processing. Currently, processing larger-size data at the device edge is possible and has been coordinated with text, imagery, seismic, acoustic, and radar sensing. Hence, future edge-based heterogeneous data fusion methods would be facilitated from various architectures developed from big data, cyber-physical sensing, machine learning, and software advancements. This paper compares cloud-fog-edge developments to support data fusion multi-domain architecture considerations. A feasibility study shows the benefits of fog computing using the analytical hierarchy processing (AHP) metrics of timeliness, accuracy, confidence/credibility, throughput, and security.