Adaptive Mediation for Data Exchange in IoT Systems

Messaging and communication is a critical aspect of next generation Internet-of-Things (IoT) systems where interactions among devices, software systems/services and end-users is the expected mode of operation. Given the diverse and changing communication needs of entities, the data exchange interactions may assume different protocols (MQTT, CoAP, HTTP) and interaction paradigms (point to point, multicast, unicast). In this paper, we address the issue of supporting adaptive communications in IoT systems through a mediation-based architecture for data exchange. Here, components called mediators support protocol translation to bridge the heterogeneity gap. Aiming to provide a placement of mediators to nodes, we introduce an integer linear programming solution that takes as input: a set of Edge nodes, IoT devices, and networking semantics. Our proposed solution achieves adaptive placement resulting in timely interactions between IoT devices for larger topologies of IoT spaces.