Analysing and reducing the limitations of continuous-variable quantum cryptography and quantum networks

Due to a recent influx of attention, the field of quantum information is rapidly progressing towards the point at which quantum technologies move from the laboratory to widespread community use. However, several difficulties must be overcome before this milestone can be achieved. Two such difficulties are addressed in this thesis. The first is the ever-growing security threat posed by quantum computers to existing cryptographic protocols and the second is the missing knowledge regarding the performance differences between quantum and classical communications over various existing network topologies. Continuous-variable (CV) quantum key distribution (QKD) poses a practical solution to the security risks implied by the advancement of quantum information theory, with the promise of provably secure communications. Unfortunately, the maximum range of many CV-QKD protocols is limited. Here, this limitation is addressed by the application of post-selection, firstly, to a scenario in which two parties communicate using terahertz frequency radiation in the atmosphere, and secondly, to measurement-device-independent QKD, in which two parties communicate through the medium of an untrusted relay. In both cases, the introduction of post-selection enables security over distances substantially exceeding those of equivalent existing protocols. The second difficulty is addressed by a comparison of the quantum and classical networking regimes of the butterfly network and a group of networks constructed with butterfly blocks. By computing the achievable classical rates and upper bounds for quantum communication, the performance difference between the two regimes is quantified, and a range of conditions is established under which classical networking outperforms its quantum counterpart. This allows for guidance to be provided on which network structures should be avoided when constructing a quantum internet.