Tomas Andersson

Thermal cycling of indented ceramic materials

Abstract A new technique, based on the limited propagation of small precracks, has been used to investigate the thermal shock and thermal cycling resistance of several ceramic materials quenched into boiling water. The ΔTc’s of the materials are defined from the crack propagation and extension after a single quench. Statistically significant data are obtained by studying a large number of cracks in each specimen. Few specimens are thus required. The materials are ranked according to the individual values of ΔTc. In order to estimate the effect of different quenching rates on the values of ΔTc the results are compared with data from quenches into water at room temperature using the same technique. The experimentally determined values of ΔTc are analysed in terms of R and R′ parameters, specimen thickness and cooling rates.

Design of n-Channel Multiple Description Vector Quantizers

A new design approach for multiple description vector quantizers over more than two channels is presented. The design is inspired by the concept of channel optimized vector quantization. While most previous works have split the decoder into several independent entities, identifying the appropriate channel model makes it straightforward to implement the multiple description design problem using only one decoder. Our simulation results compare systems with 2, 4 and 8 channels. We demonstrate significant gains over previous designs, as well as over a benchmark scheme based on separate quantization and forward erasure-correcting error control

Frequency estimation utilizing the Hadamard transform

Fast analog to digital conversion with only one bit per sample not only makes high sampling rates possible but also reduces the required hardware complexity. For short data buffers or block lengths, it has been shown that tone frequency estimators can be implemented by a simple table look-up. We present an analysis of such tables using the Hadamard transform. As an outcome of the analysis, we propose a class of nonlinear estimators of low complexity. Their performance is evaluated using numerical simulations. Comparisons are made with the proper Cramer-Rao bound and with the table look-up approach.

Robustness of wave-fitting with respect to uncertain parameter values

A simple criterion for model order selection has been derived. This criterion holds when the model is linear in the unknown parameters. The correction term in the MSSE holds exactly for linear models. A generalization of the result to include nonlinear parameters have been studied in the special case with one nonlinear parameter. Numerical illustrations of the theoretical results have shown good agreement with the theoretical analysis. From the two examples above we can conclude that estimation of the DC-level is superfluous if |C| <radic sigma2/N in sinewave-fitting, as well as in linear models with a CRB associated to the DC-level given by sigma2/N. The assumption under which the results have been derived includes a Gaussian requirement on the white noise additive model imperfections, and that the model Mscr2 describes the signal correctly. Thus, the result may not be applicable for example in ADC-testing where the considered noise model is questionable (Kollar and Blair, 2004)

A COVQ-based image coder for channels with bit errors and erasures

We illustrate how channel optimized vector quantization (COVQ) can be used for channels with both bit-errors and bit-erasures. First, a memoryless channel model is presented, and the performance of COVQs trained for this channel is evaluated for an i.i.d. Gaussian source. Then, the new method is applied in implementing an error-robust sub-band image coder, and we present image results that illustrate the resulting performance. Our experiments show that the new approach is able to outperform a traditional scheme based on separate source and channel coding.

Multiple Description Image Coding Based on Multi-Stage Vector Quantization

A new design approach for multiple description coding, based on multi-stage vector quantizers, is presented. The design is not limited to systems with two descriptions, but is also well suited for the n-descriptions case. Inspired by the concept of channel optimized vector quantization, the design can easily be tailored to suit different erasure channels, e.g. packet erasure channels with memory (burst-losses). The optimization procedure used in the design takes a sample-iterative approach. All stage codebooks are updated simultaneously for each vector in the training database. The resulting algorithm has the behaviour of a simulated annealing algorithm, with several good properties, e.g. it usually provides codebooks with good index assignments. Image results are presented for systems with 2 and 4 channels. The image coder is based on a subband transform followed by 64-dimensional vector quantization, to illustrate the capacity of the design to handle large problem sizes.