Using information-theoretic constructions, it is possible to characterise the security of a communication system. This is called physical layer security.The intrinsic randomness of the wireless channel allows for provable security guarantees in the presence of an eavesdropper. As telecommunications requirements and technologies evolve, questions about point to point systems are re-framed in ways which have not yet been explored. In this thesis we analyse the robustness of particular future wireless technologies against eavesdropping at the physical layer.In the first of the original research chapters the secrecy capacity of a Gaussian multiple antenna system is considered. Despite the importance of the secrecy capacity metric, the general solution remains an open problem.This thesis resolves the secrecy capacity to be concave in a particular region in the single antenna eavesdropper regime. This allows for efficient computation of the secrecy capacity and gives communication rates which are secure.In the second research chapter, we analyse a multiple antenna, multiple access scheme. We show that the system is inherently secure, since the eaves-droppers signal-to-noise ratio decreases with the number of users, amongst other results.The third research chapter introduces a novel channel coding scheme,combining constant weight arithmetic coding with an existing combinatorial scheme. The codewords are designed to be low-power and robust against time dispersion. This has the advantage that several users may broadcast messages simultaneously. The codebook design uses characteristics of the legitimate channel, which the eavesdropper does not have access to. Simulation results show that the eavesdropper has a low probability of success.We conclude with a discussion of future work
Physical Layer Security for Next Generation Wireless Systems
Chakravarty, J. J. (Author). 29 Sept 2020
Student thesis: Doctoral Thesis › Doctor of Philosophy (PhD)