Efficient spectrum use in cognitive radio networks using dynamic spectrum management

Abstract

Radiofrequency spectrum is a finite resource that consists of the frequencies in the range 3 kHz to 300 GHz. It is used for wireless communication and supports several applications and services. Whether it is at the personal, community or society level, and whether it is for applications in consumer electronics, building management, smart utility networks, intelligent driving systems, the Internet of Things, industrial automation and so on, the demand for wireless communication is increasing continuously. Together with this increase in demand, there is an increase in the quality of service requirements in terms of throughput, and the reliability and availability of wireless services. Industrial wireless sensor networks, for example, operate in environments that are usually harsh and time varying. The frequency spectrum that is utilised by industrial wireless protocols such as WirelessHART and ISA 100.11a, is also used by many other wireless technologies, and with wireless applications growing rapidly, it is possible that multiple heterogeneous wireless systems will need to operate in overlapping spatiotemporal regions in the future. Increased radiofrequency interference affects connectivity and reduces communication link quality. This affects reliability and latency negatively, both of which are core quality service requirements.

Getting multiple heterogeneous radio systems to co-exist harmoniously in shared spectrum is challenging. Traditionally, this has been achieved by granting network operators exclusive rights that allow them to access parts of the spectrum assigned to them and hence the problems of co-existence and limited spectrum could be ignored. Design time multi-access techniques have also been used. At present, however, it has become necessary to use spectrum more efficiently, to facilitate the further growth of wireless communication. This can be achieved in a number of ways. Firstly, the policy that governs the regulation of radiofrequency spectrum must be updated to accommodate flexible, dynamic spectrum access. Secondly, new techniques for multiple-access and spectrum sharing should be devised. A revolutionary new communication paradigm is required, and one such paradigm has recently emerged in the form of Cognitive Radio technology. Traditional methods to sharing spectrum assume that radios in a wireless network work together in an unchanging environment. Cognitive radios, on the other hand, can sense, learn and adapt. In cognitive radio networks, the interactions between users are taken into account, in order for adjustments to be made to suit the prevailing radio environment.

In this thesis, the problem of spectrum scarcity and coexistence is addressed using cognitive radio techniques, to ensure more efficient use of radio-frequency spectrum. An introduction to cognitive radio networks is given, covering cognitive radio fundamentals, spectrum sensing, dynamic spectrum management, game theoretic approaches to spectrum sharing and security in cognitive radio networks. A focus is placed on wireless industrial networks as a challenging test case for cognitive radio. A study on spectrum management policy is conducted, together with an investigation into the current state of radio-frequency spectrum utilisation, to uncover real and artificial cases of spectrum scarcity. A novel cognitive radio protocol is developed together with an open source test bed for it. Finally, a game theoretic dynamic spectrum access algorithm is developed that can provide scalable, fast convergence spectrum sharing in cognitive radio networks. This work is a humble contribution to the advancement of wireless communication.