A novel approach based on a distributed dynamic graph modeling set up over a subdivision process to deal with distributed optimized real time carpooling requests

Abstract

Years ago, researchers started spending efforts in carpooling domain leading to many systems. Those systems are either operating or not and even if some of them succeeded, they still need improvement. As optimization and dynamicity are the most lacking aspects, we are interested in setting up an optimized dynamic carpool service. This problem's high complexity constitutes a big handicap through the way to perform efficient process. Consequently, it is here addressed based on a decentralized parallel requests' process and a distributed architecture is adopted. In this context, we propose a subdivision principle in order to decompose the served network and establish a Distributed Dynamic Graph. The latter models the available information about passengers and drivers and helps to establish a distributed parallel process. In addition, the multi-agent concept has been adopted to set up an automated framework where several communicating entities evolve and perform requests' management.