ISBN 9783843925457

978-3-8439-2545-7, Reihe Elektrotechnik

Steffen Rieß
Architecture Optimization and Implementation of a Radio Receiver with a Multistage Spectrum Sensing Technique as Part of a Low-Cost Spectrum Sensing Grid

294 Seiten, Dissertation Universität Erlangen-Nürnberg (2016), Softcover, A5

Zusammenfassung / Abstract

This work is focused on spectrum sensing based on the cognitive radio technology. Since the radio spectrum is a natural but limited resource, it has to be utilized as efficiently as possible. In this context, the UHF TV band from 470-790 MHz is considered because of its excellent radio wave propagation features. Recent changes in spectrum regulation issued new challenges especially to professional wireless microphone systems. These are due to a reduced availability of bandwidth while maintaining the high demands on robustness required for this application. The cognitive radio approach is a solution to ensure reliable operation in the future.

Therefore, this possibility was investigated during the German research project C-PMSE with the development and implementation of a cognitive radio system. A major property of such a system is the spectrum sensing capability. Depending on the observed application area, it is composed of a single sensor node, i.e. a radio receiver including antenna, or a large number of them referred to as spectrum sensing grid. The developed cognitive radio system is worldwide unique and is built at the fairground of Berlin, Germany.

An important component of the spectrum sensing grid is the radio receiver, which is able to capture the frequency band of interest. Three areas are identified that constitute the major contributions of this work beyond the state of the art. First, a dedicated radio receiver hardware is developed that supports the features of a software defined radio: compactness, producibility, upgradability, flexibility and reconfigurability. This is achieved by considering the complete signal processing chain, i.e. from the antenna input to the visualization of the measured result on the controller. Second, the occupation of the frequency spectrum is determined with a model-based design framework for a reconfigurable power spectral density calculation. And third, an advanced signal detection technique operating in SNR environments of less than -15 dB is intensively investigated by Monte Carlo computer simulations. For the evaluation of algorithms, a powerful simulation environment is developed with the ability to consider signal impairments induced by channel effects and synchronization errors. Together with the secondary user detection on the controller, the developed signal detection procedure is referred to as multistage spectrum sensing technique and is a novel contribution to spectrum sensing methods.