Collaborative Edge Computing to Bring Microservices in Smart Rural Areas

Several reasons have been given to justify the inaccessibility or poor connectivity of the internet and the limits of technological resources in rural areas, such as the high cost of physical equipment, facilities and their deployments, low income, low population density, the distance linked to the geographical distribution of certain agglomerations, and so on. Due to these multiple difficulties, in this paper, we focus on the concept of smart rural area, proposing a three-phase model based on the emerging Collaborative Edge Computing (CEC) paradigm. Our approach aims at looking for the best Edge nodes where to deploy microservices. The first phase, i.e., discovery is responsible to find out federated Edge nodes, the second phase, i.e., authentication, is responsible to verify the identity of discovered devices, in the end, the third phase, i.e., the selection is responsible to choose the best Edge node in which to deploy the microservice. In particular, discovery and authentication agents were developed considering Extensible Messaging and Presence Protocol (XMPP) and Security Assertion Making Language (SAML) respectively, whereas the selector agent was developed using a Gossip consensus algorithm.