Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Fundamentals Of Statistical Signal Processing

Monte Carlo methods are revolutionizing the on-line analysis of data in fields as diverse as financial modeling, target tracking and computer vision. These methods, appearing under the names of bootstrap filters, condensation, optimal Monte Carlo filters, particle filters and survival of the fittest, have made it possible to solve numerically many complex, non-standard problems that were previously intractable. This book presents the first comprehensive treatment of these techniques, including convergence results and applications to tracking, guidance, automated target recognition, aircraft navigation, robot navigation, econometrics, financial modeling, neural networks, optimal control, optimal filtering, communications, reinforcement learning, signal enhancement, model averaging and selection, computer vision, semiconductor design, population biology, dynamic Bayesian networks, and time series analysis. This will be of great value to students, researchers and practitioners, who have some basic knowledge of probability. Arnaud Doucet received the Ph. D. degree from the University of Paris-XI Orsay in 1997. From 1998 to 2000, he conducted research at the Signal Processing Group of Cambridge University, UK. He is currently an assistant professor at the Department of Electrical Engineering of Melbourne University, Australia. His research interests include Bayesian statistics, dynamic models and Monte Carlo methods. Nando de Freitas obtained a Ph.D. degree in information engineering from Cambridge University in 1999. He is presently a research associate with the artificial intelligence group of the University of California at Berkeley. His main research interests are in Bayesian statistics and the application of on-line and batch Monte Carlo methods to machine learning. Neil Gordon obtained a Ph.D. in Statistics from Imperial College, University of London in 1993. He is with the Pattern and Information Processing group at the Defence Evaluation and Research Agency in the United Kingdom. His research interests are in time series, statistical data analysis, and pattern recognition with a particular emphasis on target tracking and missile guidance.

/pdf/sequential-monte-carlo-methods-in-practice-3kfxeebn8e.pdf

Sequential Monte Carlo methods in practice

物件導向軟體之架構(Object-Oriented Software Construction)探討

This paper presents an overview of progress in the area of multiple input multiple output (MIMO) space-time coded wireless systems. After some background on the research leading to the discovery of the enormous potential of MIMO wireless links, we highlight the different classes of techniques and algorithms proposed which attempt to realize the various benefits of MIMO including spatial multiplexing and space-time coding schemes. These algorithms are often derived and analyzed under ideal independent fading conditions. We present the state of the art in channel modeling and measurements, leading to a better understanding of actual MIMO gains. Finally, the paper addresses current questions regarding the integration of MIMO links in practical wireless systems and standards.

/pdf/from-theory-to-practice-an-overview-of-mimo-space-time-coded-431vesj02q.pdf

From theory to practice: an overview of MIMO space-time coded wireless systems

Digital VLSI implementation of the timing and carrier synchronization of digital receivers at very high data rates is addressed. The implementation of the synchronization algorithms for 100 Mb/s digital receiver for coded 8-PSK modulation is described. The digital receiver operates on a signal which is down-converted to baseband with a fixed-frequency oscillator and which is sampled at a fixed rate (no VCOs). The timing is recovered by estimating the timing offset and shifting the received samples by means of a digital interpolation. The carrier synchronization is carried out after matched filtering by estimating the carrier phase offset and rotating the samples accordingly. >

VLSI implementation of synchronization algorithms in a 100 Mbit/s digital receiver

A reduced-complexity equalizer with soft output which combines the principle of breadth-first trellis search algorithms and maximum a posteriori equalization is derived. The algorithm is suited for long channel impulse responses. Its performance is studied on two standardized mobile radio channels. >

https://ieeexplore.ieee.org/iel2/657/6293/00245327.pdf

Soft output M-algorithm equalizer and trellis-coded modulation for mobile radio communication

In digital TDMA-based packet radio channel access schemes operating over slowly time-variant frequency-selective multipath radio channels, it is necessary to achieve fast initial acquisition of receiver synchronization. The authors are concerned with improving the frame sync performance of the ML-algorithm [Fechtel and Meyr, 1993] for feedforward joint frame sync and frequency offset estimation. Emphasis is placed upon improved "non-data-dependent" (NDD) reduced-complexity variants of the ML-estimator. This is achieved by abandoning the test for signal periodicity in favor of a different preamble structure that yields a more pronounced peak of the autocorrelation at the expense of a larger preamble length. The average frame error rate and frequency estimation performance of the new frame/frequency synchronizers is assessed via simulation. For typical examples, the new algorithms are shown to yield frame sync performance gains of several decibels so that asynchronous packet transmission in the lower SNR regions becomes feasible.

Improved frame synchronization for spontaneous packet transmission over frequency-selective radio channels

We consider the problem of estimating the propagation delays of a synchronous direct-sequence code division multiple access (DS-CDMA) system operating over a multipath fading channel. In a mobile receiver, the task of delay estimation can be divided into an acquisition of all multipath delays and subsequent tracking of the individual delays e.g. with a RAKE structure. Both tasks are especially challenging in indoor scenarios which commonly exhibit low delay-spread and thus require a high resolution of the estimation algorithms. A novel delay acquisition algorithm is presented, which is able to resolve multipaths whose delay difference is below one chip duration with high probability of acquisition and low computational complexity. It is based on a decomposition of the time-averaged correlation matrix of the output of a sliding correlator into signal and noise subspaces, with subsequent MUSIC spectrum computation and maximum search. The performance of the algorithm is assessed by means of computer simulations.

/pdf/low-complexity-high-resolution-subspace-based-delay-3uw24wwl63.pdf

Low complexity high resolution subspace-based delay estimation for DS-CDMA

In digital transmission systems employing PLL repeaters, timing jitter is produced mostly by intersymbol interference (ISI). Despite the fact that IS1 represents a narrow-band disturbance to the PLL for a long chain of N repeaters, it is responsible for cycle slips of the PLL repeater. It is demonstrated that the cycle slip rate is the key parameter determining the error performance of the system. It is shown that the accumulated jitter for a large chain becomes a narrow-band process, which can be replaced by an equivalent low-pass jitter-noise process, which lends itself to analytical treatment. Numerical results of the cycle slip rate are given, based on the solution of a time-dependent Fokker-Planck equation. The results are useful for a wide range of loop transfer functions and phase-detector characteristics. The theoretical findings are confirmed by experimental results.

Heinrich Meyr

Papers

VLSI implementation of synchronization algorithms in a 100 Mbit/s digital receiver

Soft output M-algorithm equalizer and trellis-coded modulation for mobile radio communication

Improved frame synchronization for spontaneous packet transmission over frequency-selective radio channels

Low complexity high resolution subspace-based delay estimation for DS-CDMA

Synchronization Failures in a Chain of PLL Synchronizers