Theoretical and empirical study of various problems in system identification. Particular attention is paid to robust estimation of Multivariable and Nonlinear systems, and to error quantification.
This toolbox is a MATLAB-based software package for the estimation of dynamic systems.
A wide range of standard estimation approaches are supported. These include the use of non-parametric, subspace-based and prediction-error algorithms coupled (in the latter case) with either MIMO state space or MISO polynomial model structures.
Additionally, some new approaches are included. These include the support for bilinear and other Hammerstein-Wiener non-linear structures, and the use of the expectation-maximisation (EM) algorithm for time and frequency domain estimation of state space structures.
This is a hardware device designed to be used in the design and testing of wireless MIMO communications systems. It is connected to a PC via USB 2.0 or ethernet and uses an on-board FPGA to allow implementation of algorithms in logic, together with provision for multiple radio modules.
Markov Chain Monte-Carlo methods are used to calculate probability density functions for parameters in dynamic systems models. By virtue of computation of the true posterior density, these methods allow accurate quantification of estimation error, even for short data lengths.
The application of Metropolis-Hastings and Gibbs Sampling algorithms to CDMA Multi-User Detection. This approach offers near-maximum likelihood detection with soft-outputs. This project investigates the computational feasibility of this approach.
Orthgonal Frequency Division Multiplexing (OFDM)
is core to emerging and future wireless systems. Of note,
802.16, 802.20 and 3GPP LTE all depend upon OFDM.
The goal of this project is to generate core expertise in
this area, publish in leading conferences and journals
while securing valuable IP for the project participants.
Numerous ASIC prototypes will result.
This project offers a collection of software routines for solving quadratic programming problems that can be written in this form
x* = arg min 0.5x'Hx + f'x convex cost
s.t. Ax = b, linear equality constraints,
Lx <= k, general linear inequality constraints,
l <= x <= u, bound constraints.
The routines are written in C and callable from Matlab using the standard syntax. State-of-the-art solvers are available.
This project offers a suite of software routines that run under Matlab, which perform
various signal filtering and smoothing operations. This includes standard Kalman filtering
and Kalman smoothing routines.
This project develops quantifications for the frequency domain variance of prediction error system estimates. A key theme is to derive new approximations offering improved accuracy via the principles of reproducing kernel principles and orthonormal parametrizations.
Details of our attempt at the Wiener-Hammerstein Benchmark problem