Mapping Parallel Iterative Algorithms onto Workstation Networks

For communication-intensive parallel applications, the maximum degree of concurrency achievable is limited by the communication throughput made available by the network. In previous work [HPS94], we showed experimentally that the performance of certain parallel applications running on a workstation network can be improved significantly if a congestion control protocol is used to enhance network performance. In this paper, we characterize and analyze the communication requirements of a large class of supercomputing applications that fall under the category of fixed-point problems, amenable to solution by parallel iterative methods. This results in a set of interface and architectural features sufficient for the efficient implementation of the applications over a large-scale distributed system. In particular, we propose a direct link between the application and network layer, supporting congestion control actions at both ends. This in turn enhances the system's responsiveness to network congestion, improving performance. Measurements are given showing the efficacy of our scheme to support large-scale parallel computations.

A novel MAP-MRF approach for multispectral image contextual classification using combination of suboptimal iterative algorithms

In this paper we present a novel approach for multispectral image contextual classification by combining iterative combinatorial optimization algorithms. The pixel-wise decision rule is defined using a Bayesian approach to combine two MRF models: a Gaussian Markov Random Field (GMRF) for the observations (likelihood) and a Potts model for the a priori knowledge, to regularize the solution in the presence of noisy data. Hence, the classification problem is stated according to a Maximum a Posteriori (MAP) framework. In order to approximate the MAP solution we apply several combinatorial optimization methods using multiple simultaneous initializations, making the solution less sensitive to the initial conditions and reducing both computational cost and time in comparison to Simulated Annealing, often unfeasible in many real image processing applications. Markov Random Field model parameters are estimated by Maximum Pseudo-Likelihood (MPL) approach, avoiding manual adjustments in the choice of the regularization parameters. Asymptotic evaluations assess the accuracy of the proposed parameter estimation procedure. To test and evaluate the proposed classification method, we adopt metrics for quantitative performance assessment (Cohen`s Kappa coefficient), allowing a robust and accurate statistical analysis. The obtained results clearly show that combining sub-optimal contextual algorithms significantly improves the classification performance, indicating the effectiveness of the proposed methodology. (C) 2010 Elsevier B.V. All rights reserved.

ACCELERATING ITERATIVE ALGORITHMS FOR SYMMETRICAL LINEAR COMPLEMENTARITY-PROBLEMS

We propose a method for accelerating iterative algorithms for solving symmetric linear complementarity problems. The method consists in performing a one-dimensional optimization in the direction generated by a splitting method even for non-descent directions. We give strong convergence proofs and present numerical experiments that justify using this acceleration.

Relaxation of alternating iterative algorithms for the Cauchy problem associated with the modified Helmholtz equation

We propose two algorithms involving the relaxation of either the given Dirichlet data or the prescribed Neumann data on the over-specified boundary, in the case of the alternating iterative algorithm of ` 12 ` 12 `$12 `&12 `#12 `^12 `_12 `%12 `~12 *Kozlov91 applied to Cauchy problems for the modified Helmholtz equation. A convergence proof of these relaxation methods is given, along with a stopping criterion. The numerical results obtained using these procedures, in conjunction with the boundary element method (BEM), show the numerical stability, convergence, consistency and computational efficiency of the proposed methods.

Efficient algorithms for learning kernels from multiple similarity matrices with general convex loss functions

In this paper we consider the problem of learning an n × n kernel matrix from m(1) similarity matrices under general convex loss. Past research have extensively studied the m = 1 case and have derived several algorithms which require sophisticated techniques like ACCP, SOCP, etc. The existing algorithms do not apply if one uses arbitrary losses and often can not handle m > 1 case. We present several provably convergent iterative algorithms, where each iteration requires either an SVM or a Multiple Kernel Learning (MKL) solver for m > 1 case. One of the major contributions of the paper is to extend the well knownMirror Descent(MD) framework to handle Cartesian product of psd matrices. This novel extension leads to an algorithm, called EMKL, which solves the problem in O(m2 log n 2) iterations; in each iteration one solves an MKL involving m kernels and m eigen-decomposition of n × n matrices. By suitably defining a restriction on the objective function, a faster version of EMKL is proposed, called REKL,which avoids the eigen-decomposition. An alternative to both EMKL and REKL is also suggested which requires only an SVMsolver. Experimental results on real world protein data set involving several similarity matrices illustrate the efficacy of the proposed algorithms.

Iterative inversion of experimental data in weighted spaces

We present iterative algorithms for solving linear inverse problems with discrete data and compare their performances with the method of singular function expansion, in view of applications in optical imaging and particle sizing.

Synthesis of FPGA-based accelerators implementing recursive algorithms

O desenvolvimento de sistemas computacionais é um processo complexo, com múltiplas etapas, que requer uma análise profunda do problema, levando em consideração as limitações e os requisitos aplicáveis. Tal tarefa envolve a exploração de técnicas alternativas e de algoritmos computacionais para optimizar o sistema e satisfazer os requisitos estabelecidos. Neste contexto, uma das mais importantes etapas é a análise e implementação de algoritmos computacionais. Enormes avanços tecnológicos no âmbito das FPGAs (Field-Programmable Gate Arrays) tornaram possível o desenvolvimento de sistemas de engenharia extremamente complexos. Contudo, o número de transístores disponíveis por chip está a crescer mais rapidamente do que a capacidade que temos para desenvolver sistemas que tirem proveito desse crescimento. Esta limitação já bem conhecida, antes de se revelar com FPGAs, já se verificava com ASICs (Application-Specific Integrated Circuits) e tem vindo a aumentar continuamente. O desenvolvimento de sistemas com base em FPGAs de alta capacidade envolve uma grande variedade de ferramentas, incluindo métodos para a implementação eficiente de algoritmos computacionais. Esta tese pretende proporcionar uma contribuição nesta área, tirando partido da reutilização, do aumento do nível de abstracção e de especificações algorítmicas mais automatizadas e claras. Mais especificamente, é apresentado um estudo que foi levado a cabo no sentido de obter critérios relativos à implementação em hardware de algoritmos recursivos versus iterativos. Depois de serem apresentadas algumas das estratégias para implementar recursividade em hardware mais significativas, descreve-se, em pormenor, um conjunto de algoritmos para resolver problemas de pesquisa combinatória (considerados enquanto exemplos de aplicação). Versões recursivas e iterativas destes algoritmos foram implementados e testados em FPGA. Com base nos resultados obtidos, é feita uma cuidada análise comparativa. Novas ferramentas e técnicas de investigação que foram desenvolvidas no âmbito desta tese são também discutidas e demonstradas.

Relaxation procedures for an iterative MFS algorithm for the stable reconstruction of elastic fields from Cauchy data in two-dimensional isotropic linear elasticity

We investigate two numerical procedures for the Cauchy problem in linear elasticity, involving the relaxation of either the given boundary displacements (Dirichlet data) or the prescribed boundary tractions (Neumann data) on the over-specified boundary, in the alternating iterative algorithm of Kozlov et al. (1991). The two mixed direct (well-posed) problems associated with each iteration are solved using the method of fundamental solutions (MFS), in conjunction with the Tikhonov regularization method, while the optimal value of the regularization parameter is chosen via the generalized cross-validation (GCV) criterion. An efficient regularizing stopping criterion which ceases the iterative procedure at the point where the accumulation of noise becomes dominant and the errors in predicting the exact solutions increase, is also presented. The MFS-based iterative algorithms with relaxation are tested for Cauchy problems for isotropic linear elastic materials in various geometries to confirm the numerical convergence, stability, accuracy and computational efficiency of the proposed method.

Robust THP Transceiver Designs for Multiuser MIMO Downlink with Imperfect CSIT

We present robust joint nonlinear transceiver designs for multiuser multiple-input multiple-output (MIMO) downlink in the presence of imperfections in the channel state information at the transmitter (CSIT). The base station (BS) is equipped with multiple transmit antennas, and each user terminal is equipped with one or more receive antennas. The BS employs Tomlinson-Harashima precoding (THP) for interuser interference precancellation at the transmitter. We consider robust transceiver designs that jointly optimize the transmit THP filters and receive filter for two models of CSIT errors. The first model is a stochastic error (SE) model, where the CSIT error is Gaussian-distributed. This model is applicable when the CSIT error is dominated by channel estimation error. In this case, the proposed robust transceiver design seeks to minimize a stochastic function of the sum mean square error (SMSE) under a constraint on the total BS transmit power. We propose an iterative algorithm to solve this problem. The other model we consider is a norm-bounded error (NBE) model, where the CSIT error can be specified by an uncertainty set. This model is applicable when the CSIT error is dominated by quantization errors. In this case, we consider a worst-case design. For this model, we consider robust (i) minimum SMSE, (ii) MSE-constrained, and (iii) MSE-balancing transceiver designs. We propose iterative algorithms to solve these problems, wherein each iteration involves a pair of semidefinite programs (SDPs). Further, we consider an extension of the proposed algorithm to the case with per-antenna power constraints. We evaluate the robustness of the proposed algorithms to imperfections in CSIT through simulation, and show that the proposed robust designs outperform nonrobust designs as well as robust linear transceiver designs reported in the recent literature.

Non-linear Transceiver Designs with Imperfect CSIT Using Convex Optimization

In this paper, we consider non-linear transceiver designs for multiuser multi-input multi-output (MIMO) down-link in the presence of imperfections in the channel state information at the transmitter (CSIT). The base station (BS) is equipped with multiple transmit antennas and each user terminal is equipped with multiple receive antennas. The BS employs Tomlinson-Harashima precoding (THP) for inter-user interference pre-cancellation at the transmitter. We investigate robust THP transceiver designs based on the minimization of BS transmit power with mean square error (MSE) constraints, and balancing of MSE among users with a constraint on the total BS transmit power. We show that these design problems can be solved by iterative algorithms, wherein each iteration involves a pair of convex optimization problems. The robustness of the proposed algorithms to imperfections in CSIT is illustrated through simulations.

Singular value decomposition based computationally efficient algorithm for rapid dynamic near-infrared diffuse optical tomography

Purpose: A computationally efficient algorithm (linear iterative type) based on singular value decomposition (SVD) of the Jacobian has been developed that can be used in rapid dynamic near-infrared (NIR) diffuse optical tomography. Methods: Numerical and experimental studies have been conducted to prove the computational efficacy of this SVD-based algorithm over conventional optical image reconstruction algorithms. Results: These studies indicate that the performance of linear iterative algorithms in terms of contrast recovery (quantitation of optical images) is better compared to nonlinear iterative (conventional) algorithms, provided the initial guess is close to the actual solution. The nonlinear algorithms can provide better quality images compared to the linear iterative type algorithms. Moreover, the analytical and numerical equivalence of the SVD-based algorithm to linear iterative algorithms was also established as a part of this work. It is also demonstrated that the SVD-based image reconstruction typically requires O(NN2) operations per iteration, as contrasted with linear and nonlinear iterative methods that, respectively, requir O(NN3) and O(NN6) operations, with ``NN'' being the number of unknown parameters in the optical image reconstruction procedure. Conclusions: This SVD-based computationally efficient algorithm can make the integration of image reconstruction procedure with the data acquisition feasible, in turn making the rapid dynamic NIR tomography viable in the clinic to continuously monitor hemodynamic changes in the tissue pathophysiology.

Noniterative inversion strategy for photoacoustic tomography: recovery of absorbed energy map from boundary pressure measurements

The inverse problem in photoacoustic tomography (PAT) seeks to obtain the absorbed energy map from the boundary pressure measurements for which computationally intensive iterative algorithms exist. The computational challenge is heightened when the reconstruction is done using boundary data split into its frequency spectrum to improve source localization and conditioning of the inverse problem. The key idea of this work is to modify the update equation wherein the Jacobian and the perturbation in data are summed over all wave numbers, k, and inverted only once to recover the absorbed energy map. This leads to a considerable reduction in the overall computation time. The results obtained using simulated data, demonstrates the efficiency of the proposed scheme without compromising the accuracy of reconstruction.

Fair Scheduling in Cellular Systems in the Presence of Noncooperative Mobiles

We consider the problem of ``fair'' scheduling the resources to one of the many mobile stations by a centrally controlled base station (BS). The BS is the only entity taking decisions in this framework based on truthful information from the mobiles on their radio channel. We study the well-known family of parametric alpha-fair scheduling problems from a game-theoretic perspective in which some of the mobiles may be noncooperative. We first show that if the BS is unaware of the noncooperative behavior from the mobiles, the noncooperative mobiles become successful in snatching the resources from the other cooperative mobiles, resulting in unfair allocations. If the BS is aware of the noncooperative mobiles, a new game arises with BS as an additional player. It can then do better by neglecting the signals from the noncooperative mobiles. The BS, however, becomes successful in eliciting the truthful signals from the mobiles only when it uses additional information (signal statistics). This new policy along with the truthful signals from mobiles forms a Nash equilibrium (NE) that we call a Truth Revealing Equilibrium. Finally, we propose new iterative algorithms to implement fair scheduling policies that robustify the otherwise nonrobust (in presence of noncooperation) alpha-fair scheduling algorithms.