Problem Of Compression Research Articles

Distributed Learning Algorithms for Optimal Data Routing in IoT Networks

We consider the problem of joint lossy data compression and data routing in distributed Internet of Things (IoT). Heterogeneous sources compress their data using a source-specific lossy compression scheme, where heterogeneity is meant in terms of signal type and/or transmission rates. The compressed data is thus disseminated in a multi-hop fashion until it reaches a data collector (the IoT gateway). The problem we address is to compute a suitable rate-distortion working point for the compression scheme at the source nodes, while jointly assessing the most energy efficient routing paths for the data they transmit, under channel access, distortion and capacity constraints. This is formulated as a multi-objective optimization problem that is solved through distributed learning algorithms , where source coding and routing configurations emerge as the result of local interactions among the network devices. Our final algorithm is based on the alternating direction method of multipliers (ADMM), which is accelerated using the most recent findings from the literature. As a result, it has faster convergence (up to three times) to the global optimum than standard ADMM. Numerical results are discussed for selected network scenarios, emphasizing the interrelations that exist between signal reconstruction quality at the IoT gateway and total transport energy in the network.

FORMATION OF WAVELET - TRANSFORM SIGNALS TAKING INTO ACCOUNT THRESHOLD FUNCTIONS IN THE DIGITAL STREAM COMPRESSION PROBLEM

In order to reduce the digital stream, the problem of improving the quality of digital filtering based on Discrete Wavelet Transformation (DWT) was analyzed using different options for setting the threshold function and the choice of wavelet basis. Given the considerable interest in this problem and the numerous studies concerning the search for ways to optimize the suppression of interference present in signals and images, the topic continues to be relevant and important. For such tasks as image compression, operations in processing and synthesis of different signals, in the analysis of images of different nature, in reducing (compressing) large amounts of information, and to protect information is often used wavelet transform, implemented on the basis of mirror filters. For the formation of wavelet-transform signals taking into account the threshold functions in the problem of digital stream compression, a standard approach to solving the problem of signal purification from interference and random distortions using Daubechies wavelet and adjusting the signal decomposition coefficients based on wavelet functions using soft and rigid threshold task options. It is shown that the use of complex bases provides an advantage both in terms of threshold filtering error and in terms of reducing the risk of accidental distortion in the reconstruction of the useful signal by wavelet coefficients. Appropriate steps were taken for the test signal and experimental data. When using the threshold function, the large modulus (most significant) wavelet coefficients remain unchanged, and the small ones are reset to zero. The change in the amplitude of the recovered signal leads in the latter case to a decrease in the absolute values of all wavelet coefficients, including large modulo. For those applications where it is important to keep the amplitude characteristics constant, this approach is not suitable, but there are problems where it is more important to maintain the regularity of the signal than to accurately reproduce its amplitude. Rhis is the filtering of images from various obstacles, where the method of "soft" task of the threshold function is a widely used approach. When analyzing signals, the constant amplitude is also not always a mandatory requirement. An audio signal can be amplified after filtering, and pre-cleaning it from interference is more important than changing the amplitude characteristics. The use of complex wavelet transform allowed to obtain error reduction and lower threshold level when adjusting wavelet coefficients, and can be recommended for analysis of complex wavelet transform methods as an effective tool for cleaning signals and images of different nature for further research.

Finite-element simulation of cyclic compression of a cylinder with account for energy dissipation

The article considers the solution of the associated thermoelastic problem of dynamic compression of a cylindrical sample from hyperelastic material in an axisymmetric formulation using the Abaqus Simulia finite element modeling software package. The authors present the results of a numerical experiment, carry out its analysis, make a conclusion if it is necessary to take into account energy dissipation in the study of materials operating under cyclic loading.

Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing

The past decade has witnessed a prosperity of sparsity-inspired face hallucination methods that use sparse prior and instances to generate High-Resolution (HR) faces. However, they need numerous Low-Resolution (LR) and HR instance pairs and adopt approximate sparse coding, which will bring bias to the recovery and suffer from high computational burden. In this paper we advance a Single Face Image Hallucination (SFIH) method from a new perspective of Non-linear Learning Compressive Sensing (NLCS), which can recover HR faces from a surprisingly small number of HR faces. The nonlinear sparse coding of facial images is explored, and a Deep AutoEncoder (DAE) network is constructed for learning a kernel function from a single HR instance set. SFIH is then reduced to an analytic compressive recovery problem by reformulating linear sparse coding as a nonlinear DAE model. By exploring the nonlinear sparsity in the feature space, NLCS can accurately and rapidly recover HR facial images with large magnification factor and exhibit robustness to LR-HR instance pairs mapping. Some experiments are taken on realizing 3X, 6X, 9X amplification of face images, and the results prove its efficiency and superiority to its counterparts.

On Recoverability of Randomly Compressed Tensors With Low CP Rank

Our interest lies in the recoverability properties of compressed tensors under the \textit{canonical polyadic decomposition} (CPD) model. The considered problem is well-motivated in many applications, e.g., hyperspectral image and video compression. Prior work studied this problem under somewhat special assumptions---e.g., the latent factors of the tensor are sparse or drawn from absolutely continuous distributions. We offer an alternative result: We show that if the tensor is compressed by a subgaussian linear mapping, then the tensor is recoverable if the number of measurements is on the same order of magnitude as that of the model parameters---without strong assumptions on the latent factors. Our proof is based on deriving a \textit{restricted isometry property} (R.I.P.) under the CPD model via set covering techniques, and thus exhibits a flavor of classic compressive sensing. The new recoverability result enriches the understanding to the compressed CP tensor recovery problem; it offers theoretical guarantees for recovering tensors whose elements are not necessarily continuous or sparse.

A Simple Modification of HLLEM Approximate Riemann Solver Applied to the Compressible Euler System at Low Mach Number

Due to its attractive properties in computing compressible flows, such as sharp capturing of discontinuities, satisfying entropy condition and positivity preservation, the HLLEM approximate Riemann solver has been widely applied for simulations of many compressible flow problems. However, when it comes to weakly compressible or incompressible flows, the HLLEM scheme cannot give physically correct solutions. In the current study, a simple low Mach number fix is applied to improve the accuracy and stability of HLLEM approximate Riemann solver in the low Mach limit. As a result, a modified HLLEM scheme called LM-HLLEM scheme is proposed. Numerical results demonstrate that the proposed LM-HLLEM scheme is able to compute various flow problems accurately and robustly ranging from compressible to low Mach incompressible flows.

Optimal Network Compression

This paper introduces a formulation of the optimal network compression problem for financial systems. This general formulation is presented for different levels of network compression or rerouting allowed from the initial inter-bank network. We prove that this problem is, generically, NP-hard. We focus on objective functions generated by systemic risk measures under systematic shocks to the financial network. We conclude by studying the optimal compression problem for specific networks; this permits us to study the so-called robust fragility of certain network topologies more generally as well as the potential benefits and costs of network compression.

Large Database Compression Based on Perceived Information

Lossy compression algorithms trade bits for quality, aiming at reducing as much as possible the bitrate needed to represent the original source (or set of sources), while preserving the source quality. In this letter, we propose a novel paradigm of compression algorithms, aimed at minimizing the information loss perceived by the final user instead of the actual source quality loss, under compression rate constraints. As main contributions, we first introduce the concept of perceived information (PI), which reflects the information perceived by a given user experiencing a data collection, and which is evaluated as the volume spanned by the sources features in a personalized latent space. We then formalize the rate-PI optimization problem and propose an algorithm to solve this compression problem. Finally, we validate our algorithm against benchmark solutions with simulation results, showing the gain in taking into account users’ preferences while also maximizing the perceived information in the feature domain.

An Image Compression Method for Video Surveillance System in Underground Mines Based on Residual Networks and Discrete Wavelet Transform

Video surveillance systems play an important role in underground mines. Providing clear surveillance images is the fundamental basis for safe mining and disaster alarming. It is of significance to investigate image compression methods since the underground wireless channels only allow low transmission bandwidth. In this paper, we propose a new image compression method based on residual networks and discrete wavelet transform (DWT) to solve the image compression problem. The residual networks are used to compose the codec network. Further, we propose a novel loss function named discrete wavelet similarity (DW-SSIM) loss to train the network. Because the information of edges in the image is exposed through DWT coefficients, the proposed network can learn to preserve the edges better. Experiments show that the proposed method has an edge over the methods being compared in regards to the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM), particularly at low compression ratios. Tests on noise-contaminated images also demonstrate the noise robustness of the proposed method. Our main contribution is that the proposed method is able to compress images at relatively low compression ratios while still preserving sharp edges, which suits the harsh wireless communication environment in underground mines.

Mixed finite volume element-upwind mixed volume element of compressible two-phase displacement and its numerical analysis

A fundamental topic in numerical simulation of two-phase displacement is discussed in this paper. The mathematical model for the compressible problem is defined mainly by two nonlinear partial differential equations: a parabolic equation for the pressure and a convection–diffusion equation for the saturation. The Darcy velocity is determined by the pressure, and affects the whole physical process. The system is solved by a composite numerical scheme. The conservative mixed volume element is used for the first equation. The computational accuracy is improved for the Darcy velocity. The second equation is solved by a conservative upwind mixed volume element, where the mixed volume element and upwind approximation treat the diffusion and convection, respectively. The upwind method preserves the high computational accuracy, and numerical dispersion and nonphysical oscillation are eliminated. The saturation and its adjoint vector function are obtained simultaneously. An important feature in numerical scheme, the conservation of mass, is proved. By the traditional theoretical work of numerical analysis such as a priori estimates of differential equations, the optimal order error estimate is obtained. Finally, numerical tests show the effectiveness and practicability, then the present method possibly solves the challenging problems as a powerful tool.

Investigating thermal developing zone of compressible laminar flow inside hot tube

Purpose Increasing the temperature of gas flows passing through hot tubes is one of the industrial interests. Operations in the gas phase with high temperature variations involve engineers with the compressible fluids problems. The paper aims to discuss this issue. Design/methodology/approach In this study, a mathematical three-dimensional turbulent model is applied for investigating the heat transfer and laminar gas flow inside the thermal developing zone of a hot tube. The Favre Averaged Navier–Stokes and energy equations and also the Reynolds Stress Model are numerically solved to obtain the fluid velocity and temperature profiles inside this the tube. This model is validated using the experimental data and also well-known formulas in this science. Findings Finally, effects of inlet volumetric flow rate, heating conditions of the tube wall and tube angle on the temperature and velocity distributions of the gaseous phase inside this zone are investigated. Originality/value The compressible laminar gas flow and also heat transfer in the thermal developing zone of a hot tube is studied using a three-dimensional turbulent model.

Stability of out-of-phase solitons and laser pulse self-compression in active multicore fibers

The out-of-phase soliton distribution of the wave field was found for a multicore fiber (MCF) from an even number of cores located in a ring. Its stability is proved both with respect to small wave field perturbations, including azimuthal ones and to small deformations of the MCF structure. As an example of using this soliton distribution, the problem of laser pulse compression in an active MCF is studied. The optimal fiber parameters, the minimum duration of the output pulse, and the compression length have been found, which are in good agreement with the results of numerical simulation. In order to achieve high energies in the output laser pulse, the requirements for MCF deformations are determined.

Central upwind scheme based immersed boundary method for compressible flows around complex geometries

The present work combines easy to implement high resolution central upwind scheme (CUS) with a ghost cell based discrete forcing immersed boundary method (IBM) and proposes a novel CUS-IBM finite volume solver to simulate high-speed compressible inviscid/viscous flows around complex geometries on Cartesian grids. Proposed CUS-IBM solver has been tested for a variety of inviscid and viscous compressible flow problems, ranging from transonic to supersonic flow speeds (0.625 < M < 3), involving flow over different geometries, i.e., bump, airfoil and cylinder. Results from our computations clearly show that solutions computed with the proposed method are in good agreement with the numerical data available in literature for the various inviscid/viscous and external/internal compressible flow problems. The present study demonstrates that the proposed CUS-IBM exhibits nominal second-order accuracy in space and is an easy to implement, robust and promising alternative to the methods involving Riemann solver and characteristics decomposition.

Research on Ship Trajectory Compression Algorithm Based on Cumulative Heading Variation

In this paper, we proposed an improved Douglas-Peucker algorithm based on cumulative heading variation to improve the ship trajectory data compression algorithm. This method calculates the cumulative heading change of the post-point and the base point in turn based on the starting point of the route. If the cumulative heading change of the post-point is greater than the set threshold value, the point is stored in the feature point library. With this point as the new base point, the above calculation is repeated until the last point of the route. According to the obtained feature points, the route is divided into several sections, each section is separated by Douglas-Peucker algorithm. Compression is performed to obtain the final trajectory. This method can avoid the problems of trajectory compression caused by traditional Douglas-Peucker algorithm, such as crossing shoreline and dangerous area in inland river bend, and can better meet the needs of inland ship trajectory compression.

Deep neural networks compression learning based on multiobjective evolutionary algorithms

This work addresses the problem of deep neural network compression, which is a promising technique to reduce the number of network parameters and/or speed up the network evaluation process significantly. Most of the existing methods rely on domain experts’ experience for the selection of hyperparameters such as the rank and sparsity ratio of the weight matrix in order to get an appropriate compression result without serious performance downgrade. However, they usually suffer from heavy computational loads due to the large number of tests in revealing the best hyperparameters. In this work, we propose an efficient approach to network compression from the perspective of multiobjective evolution. The contributions in the paper are twofolds: (1) We build a multiobjective compression learning model that considers the model classification error rate and compression rate as two objectives in the optimization, which can provide a compromise of the tradeoffs between these two objectives. (2) A mechanism for approximate compressed model generation is devised in the framework of expensive multiobjective optimization, which is able to reduce the high model training costs involved in the optimization process. Experiments are carried out to confirm the superiority of the proposed algorithm.

Hadamard matrices from Goethals — Seidel difference families with a repeated block

Purpose: To construct Hadamard matrices by using Goethals — Seidel difference families having a repeated block, generalizingthe so called propus construction. In particular we construct the first examples of symmetric Hadamard matrices of order 236.Methods: The main ingredient of the propus construction is a difference family in a finite abelian group of order v consisting offour blocks (X1, X2, X3, X4) where X1 is symmetric and X2 X3. The parameters (v; k1, k2, k3, k4; λ) of such family must satisfythe additional condition ki  λ  v. We modify this construction by imposing different symmetry conditions on some of theblocks and construct many examples of Hadamard matrices of this kind. In this paper we work with the cyclic group Zv of order v.For larger values of v we build the blocks Xi by using the orbits of a suitable small cyclic subgroup of the automorphism groupof Zv. Results: We continue the systematic search for symmetric Hadamard matrices of order 4v by using the propus construction.Such searches were carried out previously for odd v  51. We extend it to cover the case v53. Moreover we construct thefirst examples of symmetric Hadamard matrices of order 236. A wide collection of symmetric and skew-symmetric Hadamardmatrices was obtained and the corresponding difference families tabulated by using the symmetry properties of their blocks.Practical relevance: Hadamard matrices are used extensively in the problems of error-free coding, compression and masking ofvideo information. Programs for search of symmetric Hadamard matrices and a library of constructed matrices are used in themathematical network Internet together with executable on line algorithms.

Biological Assessment of Zn-Based Absorbable Metals for Ureteral Stent Applications.

The use of ureteral stents to relieve urinary tract obstruction is still challenged by the problems of infection, encrustation, and compression, leading to the need for early removal procedures. Biodegradable ureteral stents, commonly made of polymers, have been proposed to overcome these problems. Recently, absorbable metals have been considered as potential materials offering both biodegradation and strength. This work proposed zinc-based absorbable metals by firstly evaluating their cytocompatibility toward normal primary human urothelial cells using 2D and 3D assays. In the 2D assay, the cells were exposed to different concentrations of metal extracts (i.e., 10 mg/mL of Zn–1Mg and 8.75 mg/mL of Zn–0.5Al) for up to 3 days and found that their cytoskeletal networks were affected but were recovered at day 3, as observed by immunofluorescence. In the 3D ureteral wall tissue construct, the cells formed a multilayered urothelium, as found in native tissue, with the presence of tight junctions at the superficial layer and laminin at the basal layer, indicating a healthy tissue condition even with the presence of the metal samples for up to 7 days of exposure. The basal cells attached to the metal surface as seen in a natural spreading state with pseudopodia and fusiform morphologies, indicating that the metals were non-toxic.

Free element method and its application in CFD

Purpose The purpose of this paper is to propose a new strong-form numerical method, called the free element method, for solving general boundary value problems governed by partial differential equations. The main idea of the method is to use a locally formed element for each point to set up the system of equations. The proposed method is used to solve the fluid mechanics problems. Design/methodology/approach The proposed free element method adopts the isoparametric elements as used in the finite element method (FEM) to represent the variation of coordinates and physical variables and collocates equations node-by-node based on the newly derived element differential formulations by the authors. The distinct feature of the method is that only one independently formed individual element is used at each point. The final system of equations is directly formed by collocating the governing equations at internal points and the boundary conditions at boundary points. The method can effectively capture phenomena of sharply jumped variables and discontinuities (e.g. the shock waves). Findings a) A new numerical method called the FEM is proposed; b) the proposed method is used to solve the compressible fluid mechanics problems for the first time, in which the shock wave can be naturally captured; and c) the method can directly set up the system of equations from the governing equations. Originality/value This paper presents a completely new numerical method for solving compressible fluid mechanics problems, which has not been submitted anywhere else for publication.

New Artificial Neural Network Models for Bio Medical Image Compression

This article presents an image compression method using feed-forward back-propagation neural networks (NNs). Marked progress has been made in the area of image compression in the last decade. Image compression removing redundant information in image data is a solution for storage and data transmission problems for huge amounts of data. NNs offer the potential for providing a novel solution to the problem of image compression by its ability to generate an internal data representation. A comparison among various feed-forward back-propagation training algorithms was presented with different compression ratios and different block sizes. The learning methods, the Levenberg Marquardt (LM) algorithm and the Gradient Descent (GD) have been used to perform the training of the network architecture and finally, the performance is evaluated in terms of MSE and PSNR using medical images. The decompressed results obtained using these two algorithms are computed in terms of PSNR and MSE along with performance plots and regression plots from which it can be observed that the LM algorithm gives more accurate results than the GD algorithm.

Image Squeezing By The Wavelet Squeezing

the problem of effective compression and transmission of images remains one of the most important in the field of multimedia technologies and is very promising in recent years. The article attempts to give its own interpretation of this problem and to analyze the features of the existing methods of wavelet transform, as well as to present software designed for compression of images for their compact storage and transmission through communication channels at maximum speed.