Spectrum coordination for intelligent wireless Internet of Things networks

Abstract

Diversity of radio access technologies, such as ZigBee, Bluetooth, LTE and Wi-Fi, together with growing requirements for their simultaneous use, significantly increase complexity of Internet of Things (IoT) wireless networks. A number of open challenges affect practical deployments, such as simultaneous use of multiple technologies, intelligent coordination of a subset of nodes, coexistence of different technologies using the same spectrum, efficient management of (simultaneously used) heterogeneous radio links, etc. This paper will consider Semantic Technology (ST), as a promising approach to coordination in such complex wireless infrastructures, especially in cases where interference models are not well understood. A Neural Network (NN) will be used for the network state estimation and ST for reasoning about required actions. It is based on semantic data sets mining such that coordination decisions may be driven by predictions instead of using physical spectrum sensing devices. ST facilitates reasoning about coordination, application priority, frequency selection and dynamic spectrum access. Because of capability of the NN to solve regression and classification problems, potentially problematic network states could be proactively avoided instead of reactively corrected particularly in priority critical applications.