Formal framework for distributed swarm computing: abstract model and properties

Swarm computing is an emerging computing paradigm suitable for solving difficult optimization problems by employing a nature-inspired search for solutions. A set of entities called swarm entities populates a digital space that emulates a real physical environment. The digital space acts as a container of swarm entities by providing the basic functionalities for the entities management including processing and migration. Swarm entities are modeled as computational objects that follow a specific set of behavioral laws of natural inspiration. The entities are organized as a swarm, i.e., they coordinate, collaborate and act by exchanging information either directly or indirectly via the environment, seeking to solve a difficult computational optimization problem. Our main research result reported in this paper is the proposal of a new formal computational model of a generic distributed framework for swarm computing. Our model captures the basic computational properties of the swarm using the formal language of Finite State Process algebra. The proposed model is simple, clear and abstract, i.e., technology independent. It can serve as a starting basis for further refinement and subsequent development of new concurrent and/or distributed implementations using the available distributed computing technologies.