Design and Analysis of Stable Distributed Dynamic Scan Scheduling Policies∗

The problem of how to dynamically schedule a receiver to detect multiple emitters is modeled as a sensor resource allocation problem. It is shown how to specify a variety of receiver-based dynamic scan scheduling strategies that dynamically focus a sensor in different frequency bands so as to develop and maintain good information about an emitter environment. We contrast dynamic and “fixed” scheduling policies, show how to incorporate emitter priorities, and we use simulation analyses to provide insights into the design and behavior of the policies. We explain how a “response surface methodology” can be used to design optimal dynamic scan schedulers. Moreover, we introduce the idea of using the “simultaneous perturbation stochastic approximation” method for optimal dynamic scan scheduler design when there are many emitters. Next, it is shown how to characterize and analyze stability properties (e.g., boundedness) of a variety of dynamic scan scheduling methods. We discuss a “universal stabilizing mechanism” for dynamic scan schedulers and propose two dynamic scan scheduling methods that exploit emitter environment information in order to try to optimize scheduler performance. Moreover, we provide some stability and simulation results on “distributed dynamic scan scheduling” where there are multiple receivers that try to coordinate their actions to maintain good information about an emitter environment. In this case the simulation results clearly show that there are advantages to “collaborative” distributed dynamic scan scheduling. The paper closes with a brief summary of the key ideas and a list of potential future directions. ∗Version: 7/03/02. The section on distributed dynamic scan schedulers was completed by application of the results developed in collaboration with Mr. Alvaro Gil on distributed resource allocation for another application domain; the author gratefully acknowledges the contributions of his work via this collaboration. †The author may be contacted at (614)-292-5716, fax: (614)-292-7596, [email protected], http://eewww.eng.ohio-state.edu/~passino