Full Program »
Automatic service deployment in resource constrained ad-hoc networks with configurable QoS
Future digital services are envisioned to be supported by non-traditional infrastructures. New physical infrastructures are resource constrained, lack of power supply or fixed communication links and tend to be pervasive. In this context, traditional strategies for service deployment, based on dashboards from where new services are distributed among all components making up the communication networks, are not feasible. New communication networks deployed in smart agriculture, pervasive sensing, smart environments and other similar scenarios are, in fact, based on ad hoc networks which are only related to the global networks and the Internet through data mules. These data mules may be any existing vehicle, but drones are probably the most interesting case, although it is also the most complicated one (as the introduce an additional spatial variable: the height). In these new networks, services are typically embedded into physical components and they are specifically programmed to only perform certain activities, so no new services could be considered without removing the existing hardware nodes. Therefore, in this paper it is proposed a new service deployment mechanism to be integrated into ad hoc networks with constrained resources. Services will be deployed through software agents and a new specific protocol. Besides, this protocol allows to deploy services with different categories, so ad hoc networks could support services with different Quality-of-Service. The proposed solution enables the service deployment from drones. Finally, in order to evaluate the proposed mechanism, it is performed an experimental validation based on simulation scenarios.