A framework for the optimization of self-organized distributed autonomous agents on a dynamic network of heterogeneous intelligent sensors and actuators.

摘要

Conventional methods for network discovery, process allocation, and processing do not apply to large heterogeneous sensory networks with complex topologies that have dynamic parameters such as topology, bandwidth, processing capability, and energy availability. It is not feasible to manually design and allocate distributed processes on a network of intelligent sensors with continuously changing parameters. To scale, the system must be self-aware and autonomous and the decisions for process allocation and network topology must be localized. While reacting to changes in network topology, energy, and processing constraints, the sensors must also be reacting to external stimuli and performing their tasks. This research presents a new agent behavior-based stimulus response framework for network discovery, process deployment, and execution of tasks for distributed intelligent sensor systems.