Electrical and Electronic Engineering - Theses
Permanent URI for this collection
Search Results
1 - 2 of 2
-
ItemAn evaluation of system penalties from polarisation mode dispersion( 2007)Polarisation mode dispersion (PMD) arises from the polarisation properties of an optical signal interacting with the non-degenerate polarisation modes of an optical fibre. When an optical signal is launched into a fibre, the power is split between the polarisation modes. The amount of power contained in each mode depends on the relationship between the launch state of polarisation and the axes of the fibre's polarisation modes. Assuming that some optical power is contained in each polarisation mode, the difference in propagation constants between the modes causes pulse broadening to data bits contained within an optical signal. The inter-symbol interference that results from PMD-induced pulse broadening is the first-order manifestation of PMD. However, in deployed optical systems, the coordinate axes of the fibre's polarisation modes vary with length, and are sensitive to mechanical stresses and temperature variations. The resulting all-order PMD causes more complicated signal distortions, which are stochastic in nature. This thesis investigates the impact of first- and all-order PMD in optical systems using the PMD-induced system penalty. To achieve this, tools are first developed to aid the investigation. Firstly, a novel channel characteristic is introduced, for monitoring PMD-induced penalties. Secondly, the channel characteristic is used to develop a new theoretical framework, to interrelate different methods for monitoring PMD to the PMD-induced penalty. The techniques investigated include degree of polarisation, RF spectral power monitoring and eye closure. The framework is confirmed experimentally, and then used to demonstrate that the novel channel characteristic is the most practical method for monitoring PMD-induced penalties. Finally, the impact of the penalty measurement is recognised as a factor that influences the PMD-induced penalty. As a result, a new theoretical basis is derived to quantify the differences between the various methods of measuring system penalties, including eye closure penalties, Q-penalties, power penalties and optical signal-to-noise ratio penalties. This basis is also experimentally verified. In the second half of the thesis the tools that have been developed are used to experimentally investigate the system impact of first- and all-order PMD on two different systems. The first system does not employ PMD compensation, while the second system is compensated for PMD, using a maximum likelihood sequence estimation-based electronic equaliser. In both cases, the work extends previous investigations of PMD-induced penalties, by investigating the impact of first- and all-order PMD over many launch states of polarisation. The results from the study of PMD-induced penalties in the PMD uncompensated system are used to derive a new PMD-induced penalty model, which is then critically compared to existing models. The new model is shown to better represent both first- and all-order PMD-induced system penalties. The impact of all-order PMD in the two optical systems is reported in detail, and it is shown that in most circumstances first-order PMD is the dominant cause of system penalty. In addition, the results provide experimental evidence for previous simulation based studies of all-order PMD, and some new findings are reported.
-
ItemBroadband millimetre-wave fibre-radio access networks( 2000)Millimetre-wave (mm-wave) fibre-radio networks are emerging as a promising technology for providing customer access to future interactive broadband multimedia services. At mm-wave frequencies, propagation effects through air limit radio cell sizes so an implementation of a mm-wave fibre-radio network would require large numbers of simple, low-cost antenna base stations (BSs). To achieve this goal, it is advantageous to share some of the electronic functionality and processing at a central office (CO) rather than at each BS, with the connection between the CO and BSs provided by an optical fibre network The integration of the optical and radio technologies poses a new set of technical challenges. These include mm-wave generation schemes using optics, effects of fibre distribution on radio signals, implementation of low cost BSs, network architectures that provide lower cost links with improved system performance, and system modelling techniques. This thesis advances each of these areas. The optical chirp associated with modulation is shown to be able to improve fibre link performances that are usually limited by chromatic dispersion. Furthermore, a novel modulation scheme generating an optical single sideband- with-carrier (OSSB-WC) spectrum and a novel frequency tripling self-heterodyning scheme are shown to also significantly reduce the effects of fibre dispersion. These optical sources are also inc1uded in demonstrations of a number of novel bi-directional links. A descriptive analysis is used to determine the expected system performance using each technique. This same analysis is extended to predict component specifications and system algorithms that are necessary to achieve certain levels of system performance in a variety of mm-wave fibre-radio link scenarios. The thesis also examines the optical-radio interface and determines suitable techniques that can be used to transport different radio modulation formats, inc1uding multilevel QAM and spread spectrum signals. The physical optoelectronic interface is also considered and a novel system function and hybrid integrated circuit package design is described. A major contribution of this thesis is a detailed investigation into wavelength-division-multiplexing (WDM) based mm-wave fibre-radio networks. Simple uni-directional ring/bus and star-tree architectures are proposed and implemented and the deterioration effects of linear and homodyne crosstalk are also considered so that the limitations of each architecture can be determined. The final major contribution of this thesis is to examine the modelling of mm-wave fibre-radio systems, which is especially important for dimensioning networks. One critical parameter that does require consideration in the models is the deterioration of signal quality from intermodulation distortion in multi-carrier systems. This thesis extends conventional fixed load models that have been developed for low frequency fibre-radio networks and applied them to mm-wave systems. This fixed load model was the first attempt to link physical layer limitations to system performance. While the model derived is simple and fast to simulate, it does not model the behaviour of users in the system. In this thesis, the first model to link physical layer limitations and user behaviour has been implemented and the significant changes in the expected system performance are shown. The model can also be used to predict the required component specifications and for examining system algorithms that are necessary to achieve certain levels of system performance in a variety of mm-wave fibre-radio link scenarios.