A Decentralized and Scalable Model for Resource Discovery in IoT Network

The growing number of communicating devices in the Internet of Things (IoT) network requires an efficient resource discovery scheme without relying on centralized entity that may turn into a bottleneck affecting the system efficiency. In this paper we propose a distributed model for resource discovery in IoT. The model is based on structured peer-to-peer (p2p) scheme and follows the general system trend of fog computing. It supports multi-attribute queries and utilizes a Distributed Hash Table (DHT) as an overlay to organize the discovery process in a distributed manner. A specific method for identifier generation has been introduced to ensure the privacy of objects. Additionally, an address propagation model in the system reduces the overhead in the network by allowing the local lookup instead of global lookup for pre-connected objects. Our preliminary evaluation shows that the proposed model has a lower latency comparing to the cloud based resource discovery.