Compact Algorithm Research Articles

Efficient Algorithms for Stream Compaction on GPUs

Stream compaction, also known as stream filtering or selection, produces a smaller output array which contains the indices of the only wanted elements from the input array for further processing. With the tremendous amount of data elements to be filtered, the performance of selection is of great concern. Recently, modern Graphics Processing Units (GPUs) have been increasingly used to accelerate the execution of massively large, data parallel applications. In this paper, we designed and implemented two new algorithms for stream compaction on GPU. The first algorithm, which can preserve the relative order of the input elements, uses a multi-level prefix-sum approach. The second algorithm, which is non-order-preserving, is based the hybrid use of the prefix-sum and the atomics approaches. We compared their performance with other parallel selection algorithms on the current generation of NVIDIA GPUs. The experimental results show that both algorithms run faster than Thrust, an open-source parallel algorithms library. Furthermore, the hybrid method performs the best among all existing selection algorithms on GPU and can be two orders of magnitude faster than the sequential selection on CPU, especially when the data size is large.

Compact spatio-spectral algorithm for single image super-resolution in hyperspectral imaging

Hyperspectral imaging (HSI) is used in a wide range of applications such as remote sensing, space imagery, mineral detection, and exploration. Unfortunately, it is difficult to acquire hyperspectral images with high spatial and spectral resolution due to instrument limitations. The super-resolution techniques are used to reconstruct low-resolution hyperspectral images. However, traditional superresolution (SR) approaches do not allow direct use of both spatial and spectral information, which is a decisive for an optimal reconstruction. This paper proposes a single image SR algorithm for HSI. The algorithm uses the fact that the spatial and spectral information can be integrated to make an accurate estimate of the high-resolution HSI. To achieve this, two types of spatio- pectral downsampling, and a three-dimensional interpolation are proposed in order to increase coherence between the spatial and spectral information. The resulting reconstructions using the proposed method are up to 2 dB better than traditional SR approaches.

A Fast-Multipole Unified Technique For The Analysis Of Potential Problems With The Boundary Element Methods

The proposed developments are based on a consistent implementation of the conventional, collocation boundary element method (BEM). A scheme is used to expand a generic (not problem-dependent) fundamental solution about hierarchical levels of source and field poles, which is particularly advantageous to make the technique seamlessly applicable to 2D and 3D problems of elasticity or potential, in terms of different types of curved elements for generally complicated geometry and topology. The proposed compact algorithm is more straightforward to lay out and seems to be more efficient than the ones available in the technical literature – particularly because the outermost loop refers to field nodes and geometry, in what may be called a reverse implementation. Some numerical results are shown for the conventional BEM, with validation and assessment for a few simple, but very large-scale, 2D potential problems with complicated geometry and topology for constant, linear and quadratic elements. Since iterative solvers are not required in this first step of numerical simulations, an isolated assessment of accuracy, computational effort and storage allocation of the proposed fast multipole technique becomes possible.

BFT: a placement algorithm for non‐rectangle task model in reconfigurable computing system

Task scheduling and placement problem is one of the most significant and time-consuming parts in reconfigurable computing (RC) system. Many investigators have explored on the subject, and most of the traditional studies are concentrated on the rectangle task model, which is inconsistent with objective task shape placed in a field programmable gate array (FPGA) but simplifies the system complexity. Rectangle task model produces inner fragments which reduces utilisation of reconfigurable resources in an FPGA. In this study, a task model transformation strategy and an innovative best-fit transformation (BFT) placement algorithm are proposed for a non-rectangle task model to improve the performance of an RC system in rejection rate and total execution time. According to simulation experiments, BFT algorithm reduced the rejection rate by 15% and 7% compared with that of the first-fit algorithm and the best-fit algorithm, respectively. Multi-shape placement algorithm and 3D compaction algorithm are also cited to compare with the BFT algorithm. The result shows that the BFT algorithm has less total execution time in short laxity period and lower rejection rate in large laxity period. Compared with 3D compaction algorithm, the proposed algorithm reduced the total execution time up to 10.79%.

Offline routing and spectrum allocation algorithms for elastic optical networks

Elastic optical networks (EONs) have emerged as a promising solution for future high speed networks because of their ability to efficiently manage network resources and provide better spectrum utilization to cope with the recent rapid change in traffic behavior and the tremendous growth in bandwidth demand. The routing and spectrum allocation (RSA) problem is one of the key challenges for the effective design and control of EONs. Recently, the offline RSA problem has been mapped to a multiprocessor scheduling problem and solved using a well-known list scheduling heuristic. In this paper, we address an RSA multiprocessor scheduling formulation with the objective of minimizing the total amount of spectrum needed to serve the traffic demand in chain and mesh networks. The quality of the solution using list-scheduling algorithms is very sensitive to the ordering of the tasks in the list. We propose four list-ordering algorithms (the longest then widest compact algorithm, area compact algorithm, longest then left edge compact algorithm, and area then left edge compact algorithm) to enhance the performance of an existing list-scheduling algorithm called the compact scheduling algorithm. Our proposed algorithms are based on the left edge algorithm and the combination of the problem dimensions (i.e., the bandwidth and links). We evaluate the performance and the efficiency of our proposed algorithms across a range of demand distributions for two different network topologies (i.e., chain and National Science Foundation networks).Experimental results show that our algorithms outperform existing algorithms and provide close to optimal solutions, which are within 1–2% of the lower bound.

Classifying Data to Reduce Long-Term Data Movement in Shingled Write Disks

Shingled magnetic recording (SMR) is a means of increasing the density of hard drives that brings a new set of challenges. Due to the nature of SMR disks, updating in place is not an option. Holes left by invalidated data can only be filled if the entire band is reclaimed, and a poor band compaction algorithm could result in spending a lot of time moving blocks over the lifetime of the device. We propose using write frequency to separate blocks to reduce data movement and develop a band compaction algorithm that implements this heuristic. We demonstrate how our algorithm results in improved data management, resulting in an up to 45% reduction in required data movements when compared to naive approaches to band management.

Hybrid Firefly Algorithm Harmony Search for Feature Selection with BCNF for Multiple Subtables and EM-GMM for Top Down Initial Partitioning

Several anonymization techniques, such as generalization and bucketization, have been designed for privacy preserving microdata publishing. Recently it has become major importance in research areas and numerous approaches have been proposed to preserve privacy in literature. In order to solve this issue, recent work study the problem of privacy preservation hazard for mined Conditional Functional Dependency (CFD) against (d,l) inference model using Compact frequent pattern growth branch sort algorithm (CFPGBS) methods. Boyce–Codd normal Form (BCNF) method is presented for publishing multiple subtables and it is anonymized through (d, l)- inference model. But the major problem of the work is that initial partitioning is done without considering the importance of the feature, so it contains all quasi identifier features. It becomes more complex to perform initial partitioning if the data samples becomes large,thus may also reduce the privacy results. To overcome this issue in this paper presents a hybrid feature selection method which removes unimportant quasi identifier in the final Boyce–Codd normal Form (BCNF) table. The proposed hybrid feature selection method follows the procedure of hybrid firefly algorithm with harmony search (HFAHS) algorithm to select optimal feature in the table for initial partition of top down approach. For selected features expectation Maximization (EM) based Gaussian mixture distribution (EM-GMM) based top down initial partition is done to group the partition data from the result d,l -inference model along with changed publishing data from BCNF. Experimental results show that the proposed BCNF- d,l -inference model can adeptly anonymize the microdata with less information loss when compared to existing methods against CFD. The effectiveness and privacy results of proposed methods are also compared with existing conventional methods. Our workload experiments confirm that proposed preserves better utility, privacy than d,l -inference model against CFD attack and is more effective in workloads with selection of optimal quasi identifiers.

A single chip FPGA-based solution for controlling of multi-unit PMSM motor with time-division multiplexing scheme

The use of multiple unit controllers for parallel processing of multi-unit motor drive systems can significantly reduce the execution time of the control algorithm. However, this approach does not only increase the system cost but also incurs in additional cost of hardware and software interconnections. This paper presents a fully integrated single chip field programmable gate array (FPGA) based solution for controlling of an independent multi-unit permanent magnet synchronous motor (PMSM) drive system with space vector pulse width modulation (SVPWM) based vector control. For multi-unit motor systems, the complexity of control algorithms often exceeds the resource availability of low-cost FPGA devices. Thus, a system-level time-division multiplexing scheme applicable for multi-motor control systems is proposed. Using the proposed method, large identical complex control algorithms can be simplified into a single compact algorithm, which is fitted and configured into a low-cost FPGA. Simulation modeling and experimental results are shown, confirming the effectiveness of a multi-unit PMSM motor drive system using an inexpensive controller based on system-level time-division multiplexing scheme, which can operate simultaneously with robustness under different operating conditions.

Performance Improvement in Multiprocessors using Two Row Boundary Allocation Method and Online Dynamic Compaction Algorithm

this paper, two row boundary (TRB) allocation algorithm and limited top-down compaction (LT-DC) migration method are proposed. The first scheme, attempts to allocate the free nodes in the center of the mesh and decrease the problem of external fragmentation. The next mechanism use task migration to improve the performance of existing sub-mesh allocation strategies. It should be noted that in this process three key metrics are considered. They are average execution time, average response time, and average wait time. In fact, we perform rigorous simulation experiments based on practical workloads as reported in the literature to quantify all our proposed schemes and compare them against standard schemes existing in the literature. Based on the results, we make clear recommendations on the choice of the strategies.

Automated Functional Test Generation for Digital Systems Through a Compact Binary Differential Evolution Algorithm

At present, the functional verification of a device represents the highest cost during manufacturing. To reduce that cost, several methods have been suggested. In this work we propose a method which produces a set of binary test sequences by means of a Compact binary Differential Evolution algorithm (Compact-BinDE). The strategy employed is based on the use of coverage models and cost functions in the verification process, which are built with relevant conditions or coverage points representing the full device behavior. The main problem is to cover all difficult situations since the relationships between the test points and the input data in the design are not trivial. The test generation method is included with a proposed verification platform based on a simulation representing a hybrid method. The main contribution of this work is that the method obtains test vector sequences that maximize the coverage percentage on run-time device simulation with an efficient search in the binary domain. Also, different to the previous works that used meta-heuristics, the proposed method by means of the Compact-BinDE algorithm can reduce the simulation time used to obtain test sequences that exercise the coverage points. The results show that the proposed method represents a good alternative to generate test sequences to cover the coverage points during the functional verification.

Image-based field monitoring of Cercospora leaf spot in sugar beet by robust template matching and pattern recognition

This paper presents a novel image algorithm using template matching and pattern recognition frameworks for monitoring Cercospora leaf spot (CLS) development on sugar beets on a single leaf scale under real field conditions. Due to the variety and complexity of the open field, it is a great challenge to achieve continuous and robust foliar disease observation in real field conditions. We propose a novel and compact algorithm, composed of two frameworks and a post-processing. The algorithm has continuous and highly discriminative capabilities for observing the process of disease in a single leaf from plant-level time sequence images. The first framework is based on robust template matching by orientation code matching (OCM), which implements successive tracking of a single leaf from a beet plant against severe illumination changes and non-rigid plant movements. The second framework uses a pattern recognition method of support vector machine (SVM) for achieving further disease classification from clutter field background. Prior to SVM, we propose a three feature combination of L∗, a∗, Entropy×Density, which has strong discrimination power to classify CLS disease from the clutter scene containing sandy soil, leaves, leaf stalks, and specular reflection. Additionally, post-processing is introduced to filter false positive noise to enhance the precision of the classification. Field experiment results demonstrate the feasibility and applicability of the proposed algorithm for disease monitoring under real field conditions. Meanwhile, comparative results with other conventional matching methods and feature combinations show the effectiveness of our proposed algorithm in both foliage tracking and disease classification.

Numerical Algorithms with High Spatial Accuracy for the Fourth-Order Fractional Sub-Diffusion Equations with the First Dirichlet Boundary Conditions

In this paper, a compact algorithm for the fourth-order fractional sub-diffusion equations with first Dirichlet boundary conditions, which depict wave propagation in intense laser beams, is investigated. Combining the average operator for the spatial fourth-order derivative, the L1 formula is applied to approximate the temporal Caputo fractional derivative. A novel technique is introduced to deal with the first Dirichlet boundary conditions. Using mathematical induction method, we prove that the presented difference scheme is unconditionally stable and convergent by the energy method. The convergence order is $$O(\tau ^{2-\alpha }+h^4)$$O(?2-?+h4) in $$L_2$$L2-norm. The outline for the two-dimensional problem is also considered. Finally, some numerical examples are provided to confirm the theoretical results.

A Clustering Algorithm based on Feature Weighting Fuzzy Compactness and Separation

Aiming at improving the well-known fuzzy compactness and separation algorithm (FCS), this paper proposes a new clustering algorithm based on feature weighting fuzzy compactness and separation (WFCS). In view of the contribution of features to clustering, the proposed algorithm introduces the feature weighting into the objective function. We first formulate the membership and feature weighting, and analyze the membership of data points falling on the crisp boundary, then give the adjustment strategy. The proposed WFCS is validated both on simulated dataset and real dataset. The experimental results demonstrate that the proposed WFCS has the characteristics of hard clustering and fuzzy clustering, and outperforms many existing clustering algorithms with respect to three metrics: Rand Index, Xie-Beni Index and Within-Between(WB) Index.

Implementation of a New Lightweight Encryption Design for Embedded Security

Lightweight cryptography is an interesting field that strikes the perfect balance in providing security, higher throughput, low-power consumption, and compactness. In recent years, many compact algorithms like PRESENT, CLEFIA, SEA, TEA, LED, ZORRO, Hummingbird, and KANTAN have made the mark to be used as lightweight crypto engines. In this paper, we present the design of a new lightweight compact encryption system based on bit permutation instruction group operation (GRP), which is widely studied and extensively researched. Using the S-box of PRESENT, we have added the confusion property for GRP, because all the existing algorithms using bit permutation instructions do not have this confusion property. By comparing the existing S-boxes of compact algorithms and its cryptanalysis, a new hybrid system is proposed in this paper that provides more compact results in terms of both memory space and gate equivalents. A hybrid cryptosystem, which consists of GRP and S-box of PRESENT, is designed and implemented on a 32-bit processor. This fusion has resulted in a lightweight cipher that is the most compact implementation, till now, in terms of memory requirement. We have tested and verified this on an LPC2129 processor. Various S-boxes of recently used lightweight algorithms, such as PRESENT and CLEFIA, are designed and analyzed to create a perfect fusion that should be resistant to attacks. Using the S-box of PRESENT, it helps in further reducing the gate complexity. This hybrid model results in 2125 gate equivalents, which is better than other light variant models like DESXL, CLEFIA, and AES. Moreover, GRP properties are very helpful not only to attain the desired avalanche effect, but also as it results in a compact implementation in hardware. This paper proposes a novel approach that will have a positive impact in the field of lightweight encryption protocols.

On the Use of Compact Approaches in Evolution Strategies

Compact evolutionary algorithms have proven to be an efficient alternative for solving optimization problems in computing environments with low processing power. In this kind of solution, a probability distribution simulates the behavior of a population, thus looking for memory savings. Several compact algorithms have been proposed, including the compact genetic algorithm and compact differential evolution. This work aims to investigate the use of compact approaches in other important evolutionary algorithms: evolution strategies. This paper proposes two different approaches for compact versions of evolution strategies. Experiments were performed and the results analyzed. The results showed that, depending on the nature of problem, the use of the compact version of Evolution Strategies can be rewarding.

Enhanced compact artificial bee colony

Challenges in many real-world optimization problems arise from limited hardware availability, particularly when the optimization must be performed on a device whose hardware is highly restricted due to cost or space. This paper proposes a new algorithm, namely Enhanced compact Artificial Bee Colony (EcABC) to address this class of optimization problems. The algorithm benefits from the search logic of the Artificial Bee Colony (ABC) algorithm, and similar to other compact algorithms, it does not store the actual population of tentative solutions. Instead, EcABC employs a novel probabilistic representation of the population that is introduced in this paper. The proposed algorithm has been tested on a set of benchmark functions from the CEC2013 benchmark suite, and compared against a number of algorithms including modern compact algorithms, recent population-based ABC variants and some advanced meta-heuristics. Numerical results demonstrate that EcABC significantly outperforms other state of the art compact algorithms. In addition, simulations also indicate that the proposed algorithm shows a comparative performance when compared against its population-based versions.

A COMPARATIVE STUDY FOR RHODOPSIN PROTEIN FOLDING PROBLEM

Proteins are very important components in any living cells. A number of diseases such as Retinitis pigmentosa, Stargadt-like macular degeneration and Doyne Honeycomb Retinal Dystrophy (DHRD) diseases are shown to result from misfunctioning of proteins. Protein folding problem is a way to predict the best and optimal 3D molecular structure (tertiary structure) of a protein which is then considered to be a sign for the protein’s proper functionality. This comparative study’s purpose is to calculate the protein’s energy using the Empirical Conformational Energy Program for Peptides (ECEPP) package and experiments were performed on the Rhodopsin proteinusing three different evolutionary algorithms in order to find the best energy in parallel with the best structure for the protein and a comparison for the results obtained from the three algorithms was performed. It was found that the best result was -11.8 obtained from the Extended Compact Genetic Algorithm (ECGA). ECGA has proved from the obtained results to be the best algorithm from the chosen algorithms in the comparative study in obtaining the Rhodpsin protein’s energy and its equivalent structure.

Non-Linear Complementary SO(3) Filters for Attitude Estimation and Navigation of Ground Vehicles

The use of Inertial Navigational System (INS) has been proven to be suitable for vehicular stability and control. The same system can be used for inertial based navigation in the absence of GPS. In this paper, the problem of obtaining good attitude estimates from low cost sensors used for car navigation in the absence of GPS data is discussed. The states to be estimated are using angular velocity and linear accleration signals obtained from the sets of gyros and accelerometers of the INS. The special orthogonal group, the SO(3)-based complementary filters, have been used as the estimators as they are most suited for embedded systems to generate highly efficient algorithms for navigation. The INS has also been integrated with a set of magnetometers to assist in achieving global navigation. This integration requires kinematics equations as well as the inclusion of the gyro and accelerometer calibration and filtering. By using the quatronion representation, not only highly compact algorithms for integration can be generated, but it can also estimate and remove the effects of other biases and misalignments caused by, for instance, inaccurate installations and inherent sensors problems. The results obtained through simulation indicate better performance then Kalman filter approach as well as iterative recursive least square approach even with low grade sensors. The results are comparable with attitude estimation using roll index but with much less computations and better performance.

An algorithm for the grain-level modelling of a dry sand particulate system

This paper is composed of two parts: the generation of the sand particulate system and insights into the grain-level response under static and dynamic loadings. First, the algorithms for the generation of sand particles are presented, considering the randomness in their shape and distribution. Improvements to the robustness of the algorithms are obtained using controlling parameters. Second, we employ the take-and-place algorithm, placing sand grains into the specimen and checking how they overlap to form the initial model. In order to improve the porosity of the specimen, we develop the compaction algorithm: self-compaction by gravity and artificial compaction by mechanical vibration and pressure. The steps for the generation of a finite element grid are also introduced. Third, the grain-level configurations of the dry sand particulate system (aspects such as porosity, friction and contact) are taken into account in modelling. Results show that the grain-level responses of grains, i.e. deformation, fracture and damage of sand grains, impose significant effects on the mechanical behavior of dry sand under static and dynamic loadings.

Evaluating the mechanical compaction of quartzarenites: The importance of sorting (Llanos foreland basin, Colombia)

Mechanical compaction is the main porosity-reducing process in sandstones, including high-reservoir-quality rigid-grain sandstones. For such sandstones, the extrapolation of theoretical or experimental compaction algorithms needs calibration with rocks having well constrained burial histories. Evaluating the compaction of these rocks is achieved by comparing current intergranular volume (IGV) with depositional IGV, which is strongly dependent on sorting. However, because sandstone sorting is difficult to measure accurately, its impact on depositional porosity and compaction state is largely underestimated. We use the quartzarenites of the Oligocene Carbonera Formation in the subsurface of the hydrostatically-pressured Llanos basin to illustrate the importance of sorting when evaluating the compaction of rigid-grain sandstones. IGV and sorting were measured in core samples using a combination of transmitted-light and cathodoluminescence images, resulting in improved accuracy over standard procedures. The compaction state of clean quartzarenites at given depths is best described using IGV-versus-sorting plots, which are used to derive compaction curves for specified sorting values. The IGV-versus-sorting trends are displaced to lower IGV values with increasing burial depth. The differences in IGV caused by differences in sorting exceed the differences in IGV resulting from 1000 m of burial, illustrating the high impact of sorting when evaluating compaction. Contrasting with published experimental results, the compaction of the Llanos basin ductile-grain-poor quartzarenites is independent of grain size, and grain rearrangement is the main compaction mechanism during the first ∼1.6 km of burial. Based on the Llanos data, we have generated IGV-versus-depth curves for clean pure quartzarenites of specific sorting, which can be used to predict their maximum primary porosity up to moderate burial depths. Differences with other published burial curves are probably related to unaccounted variations in sorting, ductile-grain content and framework-strengthening cements. However, the Llanos basin quartzarenites contain virtually no cements, explaining their high degree of compaction relative to other rigid-grain sandstones, and making them ideal to isolate the effects of compaction on the IGV of quartzarenites. The Llanos basin data suggest that, below ∼2.5 km of depth, clean well- to moderately well sorted quartzarenites continue reducing their IGV by mechanical compaction below the 26% limit, which should apply only to extremely well sorted, rigid grain, uncemented sandstones.