Maximum-Likelihood Deconvolution: A Journey into Model-Based by Jerry M. Mendel

By Jerry M. Mendel

Convolution is crucial operation that describes the habit of a linear time-invariant dynamical method. Deconvolution is the unraveling of convolution. it's the inverse challenge of producing the system's enter from wisdom in regards to the system's output and dynamics. Deconvolution calls for a cautious balancing of bandwidth and signal-to-noise ratio results. Maximum-likelihood deconvolution (MLD) is a layout process that handles either results. It attracts upon rules from greatest chance, while unknown parameters are random. It ends up in linear and nonlinear sign processors that offer high-resolution estimates of a system's enter. All points of MLD are defined, from first ideas during this e-book. the aim of this quantity is to give an explanation for MLD as easily as attainable. to do that, the full conception of MLD is gifted when it comes to a convolutional sign producing version and a few particularly uncomplicated rules from optimization conception. previous methods to MLD, that are couched within the language of state-variable versions and estimation conception, are pointless to appreciate the essence of MLD. MLD is a model-based sign processing technique, since it is predicated on a sign version, specifically the convolutional version. The ebook specializes in 3 elements of MLD: (1) specification of a likelihood version for the system's measured output; (2) decision of a suitable probability functionality; and (3) maximization of that chance functionality. Many useful algorithms are bought. Computational elements of MLD are defined in nice element. large simulations are supplied, together with genuine info applications.

Show description

Read or Download Maximum-Likelihood Deconvolution: A Journey into Model-Based Signal Processing PDF

Best & telecommunications books

Trends in Enterprise Knowledge Management

Wisdom administration (KM) incorporates a wide variety of instruments and techniques which are on the middle of the knowledge and conversation society and supply strategies that depend as a lot on association as on know-how. This identify brings jointly contributions from authors from various nations who're well-known as prime figures during this box, either in an educational and a realistic experience.

Chaotic Signals in Digital Communications

Chaotic indications in electronic Communications combines basic heritage wisdom with cutting-edge equipment for utilizing chaotic signs and platforms in electronic communications. The e-book builds a bridge among theoretical works and useful implementation to assist researchers reach constant functionality in real looking environments.

Mobile Clouds: Exploiting Distributed Resources in Wireless, Mobile and Social Networks

This ebook discusses and explores the idea that of cellular cloud, developing an inspiring examine house for exploiting opportunistic source sharing, and masking from theoretical study methods to the advance of commercially ecocnomic rules. A cellular cloud is a cooperative association of dynamically attached conversation nodes sharing opportunistic assets.

Deep Space Communications

A set of a few of the Jet Propulsion Laboratory's area missions chosen to symbolize the planetary communications designs for a development of varied kinds of missions The textual content makes use of a case learn method of express the communications hyperlink functionality due to the planetary communications layout constructed via the Jet Propulsion Laboratory (JPL).

Extra resources for Maximum-Likelihood Deconvolution: A Journey into Model-Based Signal Processing

Example text

Al. (1989), Korrnylo (1979), Mahalanabis, et. al. (1982), and Goussard and Demoment (1987)]. 8 Update Wavelet Parameters Loglikelihood function ;e. , {a, b, s, q I z} are very complicated and depend in very nonlinear ways on wavelet parameters a and b. ( ) or JVt{ ). Newton-Raphson and Marquardt-Levenberg algorithms are examples of the latter. , Kormylo, 1979 and Mendel, 1983), we describe it here. A brief derivation of this algorithm is given in Chapter 7. We assume that our objective is to maximize JVt (a, b, s, q I z).

Then, we compute a time-varying threshold function t(k), which depends in a rather complicated way on the error-variance between uMV (k I N) and u(k). , ao, D(z; k) = [uMV (k I N)]2 - t(k). (4-6) The threshold detector decision strategy is: at each value of k, If [uMV (k I N)f - t(k) > 0 decide q(k) = 1; or If [uMV (k I N)]2 - t(k) < 0 decide q(k) = O. (4-7) If [uMV (k I N)]2 - t(k) = 0, we can decide q(k) = 1 or o. The resulting event sequence is denoted qTD. A derivation of this detector is given in Chapter 7.

These steps are then repeated recursively. Figure 4-4 depicts the steps of this method, which we shall refer to as a recursive block optimization method. Let Xj, Yi, denote the values of x and y at the i th iteration of this method, and Xj+ I, Yi+ 1 denote the values of x and y at the i+ 1 st iteration. For this method to be successful, we require Our description of this method applies equally as well to functions of more than two variables. ~;;;; I I I t I ~~----------~I~----------------------+~ X Keep x constant Starting point Figure 4-4.

Download PDF sample

Rated 4.77 of 5 – based on 46 votes