Search

Diabetes has become one of the most severe deceases due to an increasing number of diabetes patients globally. A large amount of digital data on diabetes has been collected through various channels. How to utilize these data sets to help doctors to make a decision on diagnosis, treatment and prediction of diabetic patients poses many challenges to the research community. The thesis investigates mathematical models with a focus on neural networks for large scale diabetes data modelling and analysis by utilizing modern computing technologies such as grid computing and cloud computing. These computing technologies provide users with an inexpensive way to have access to extensive computing resources over the Internet for solving data and computationally intensive problems. This thesis evaluates the performance of seven representative machine learning techniques in classification of diabetes data and the results show that neural network produces the best accuracy in classification but incurs high overhead in data training. As a result, the thesis develops MRNN, a parallel neural network model based on the MapReduce programming model which has become an enabling technology in support of data intensive applications in the clouds. By partitioning the diabetic data set into a number of equally sized data blocks, the workload in training is distributed among a number of computing nodes for speedup in data training. MRNN is first evaluated in small scale experimental environments using 12 mappers and subsequently is evaluated in large scale simulated environments using up to 1000 mappers. Both the experimental and simulations results have shown the effectiveness of MRNN in classification, and its high scalability in data training. MapReduce does not have a sophisticated job scheduling scheme for heterogonous computing environments in which the computing nodes may have varied computing capabilities. For this purpose, this thesis develops a load balancing scheme based on genetic algorithms with an aim to balance the training workload among heterogeneous computing nodes. The nodes with more computing capacities will receive more MapReduce jobs for execution. Divisible load theory is employed to guide the evolutionary process of the genetic algorithm with an aim to achieve fast convergence. The proposed load balancing scheme is evaluated in large scale simulated MapReduce environments with varied levels of heterogeneity using different sizes of data sets. All the results show that the genetic algorithm based load balancing scheme significantly reduce the makespan in job execution in comparison with the time consumed without load balancing.

Newly developed communication systems put strict requirements on the performance of RF power amplifiers. A key issue for the development of RF PA is the inherent nonlinearity of power amplifiers hindering its integration with the well established small-signal development infrastructure which forms a closely interlinked chain of measurement systems, small-signal models and CAD based simulation and design software. The linkage between these components is provided by common small-signal data import and export file formats ensuring a bidirectional data exchange without any loss of the small-signal information. However, no equivalent infrastructure exists for a large-signal design process inhibiting the development of RF power amplifiers and other nonlinear components. This work demonstrates a coherent methodology aiming to provide equivalent infrastructure for large signal design process as already exists in small signal design process. As first part of the methodology, a new approach is proposed to directly import measured current and voltage waveforms, obtained from a typical Large Signal Network Analyser (LSNA) system recently developed in Cardiff University, into nonlinear CAD simulator for power amplifier design. This approach offers an efficient solution for using large signal characteristic data in CAD-based simulation and PA design as the simulation/design accuracy is guaranteed by measurement and the no lengthy data processing is required. The approach is implemented in Agilent ADS simulator and its validity is comprehensively verified on different devices and device technologies. Moreover, the potential of it in predicting device large signal performance when interpolation or extrapolation is needed is explored. As second part of the methodology, a new large signal nonlinear behavioural approach is proposed from behavioural modelling perspective as a complementation to the direct waveform utilization approach. The proposed modelling approach features in impressive simulation speed while maintaining excellent simulation accuracy. The modelling approach is developed on the basis of polynomial approximation and theoretical analysis shows that the approach can be considered as reasonable extension of S parameter design tool in large signal environment. It's demonstrated in this work that the model is extracted from large signal waveform data with specially designed parameter extraction procedure. The extracted model is verified on several devices and repeatable accuracy can be obtained even on high power devices such as 100w LDMOS. It's illustrated in this work that the above two distinctive approaches can be combined and nicely considered as parts of an integrated nonlinear measurement data utilization strategy. Such a strategy provides a fast and time efficient path to accurate CAD-based nonlinear design even at power levels relevant for base station applications.