Artificial Locust Swarm Routing Algorithm: Decision Making in Path Search Problem

The good known problems and solutions nowadays require to implement some new approaches due to changing circumstances. The primary factor of these changes is a drastic growth of data volume and transition from simple data processing models and algorithms to big data ones. According to a limited computing resources (memory, CPU load, time boundaries etc.) there is a need arises to follow required limits but still solve problems involving some larger amount of data. This paper considers the problem of a path search on a graph (also known as a problem of routing on a network) with respect to a large amount of vertices and edges. Difficulty of the problem lies in both many thousands of data entries describing graph topology and necessity to satisfy reasonable time boundaries to perform the whole search on-the-fly. Such complexity requires developing a new and more efficient approaches in decision making during path search to avoid combinatorial explosion (burst) within known approaches to path search. The artificial locust swarm routing algorithm with special assessment procedures was offered as a solution that enables less time and memory consuming approach. Case study analysis was performed with navigation data of airways network.