Novel DSA algorithms and Virtual Machines for SDR

It is foreseen that in the near future the regulation of radio spectrum will undergo a change from a static and slow process to a more dynamic and faster process know as Dynamic Spectrum Access (DSA). In a DSA environment parts of the available spectrum will be made available for dynamic use of instantaneously unutilized frequencies. The main focus of the project Dynamic Telecommunications is research concerning methods for a cognitive radio (CR) to sense the environment, e.g. utilization of frequencies, and to use this information to adjust parameters of the radio. The sensing of the environment might be done autonomously by a radio node, or it might be a collaborative task of many radio nodes, helping each other by exchanging data. In either case, there are security issues, since the sensing could possibly be disturbed by a hostile opponent. Therefore, it is important to have security measures, intrinsic in the cognitive radio node, to mitigate such attacks. In this report, results are presented from the work on security measures and algorithms for dynamic frequency allocation. We propose an architecture based on Virtual Machines. The central security quality is the ability of a virtual machine monitor to separate and isolate partitions within the system. This can be used to separate different classes of information. Novel DSA algorithms have been developed in the project. These algorithms are shown to outperform traditional algorithms based on simple interference energy detection only. The algorithms will be implemented in a demonstrator to evaluate the performance over a real radio channel.