Complexity Reduction Techniques Research Articles

DdRAD sequencing-based identification of inter-genepool SNPs and association analysis in Brassica juncea

BackgroundNarrow genetic base, complex allo-tetraploid genome and presence of repetitive elements have led the discovery of single nucleotide polymorphisms (SNPs) in Brassica juncea (AABB; 2n = 4x = 36) at a slower pace. Double digest RAD (ddRAD) - a genome complexity reduction technique followed by NGS was used to generate a total of 23 million paired-end reads from three genotypes each of Indian (Pusa Tarak, RSPR-01 and Urvashi) and Exotic (Donskaja IV, Zem 1 and EC287711) genepools.ResultsSequence data analysis led to the identification of 10,399 SNPs in six genotypes at a read depth of 10x coverage among the genotypes of two genepools. A total of 44 hyper-variable regions (nucleotide variation hotspots) were also found in the genome, of which 93% were found to be a part of coding genes/regions. The functionality of the identified SNPs was estimated by genotyping a subset of SNPs on MassARRAY® platform among a diverse set of B. juncea genotypes. SNP genotyping-based genetic diversity and population studies placed the genotypes into two distinct clusters based mostly on the place of origin. The genotypes were also characterized for six morphological traits, analysis of which revealed a significant difference in the mean values between Indian and Exotic genepools for six traits. The association analysis for six traits identified a total of 45 significant marker-trait associations on 11 chromosomes of A- and B- group of progenitor genomes.ConclusionsDespite narrow diversity, the ddRAD sequencing was able to identify large number of nucleotide polymorphisms between the two genepools. Association analysis led to the identification of common SNPs/genomic regions associated between flowering and maturity traits, thereby underscoring the possible role of common chromosomal regions-harboring genes controlling flowering and maturity in Brassica juncea.

Spike Counts Based Low Complexity SNN Architecture With Binary Synapse.

In this paper, we present an energy and area efficient spike neural network (SNN) processor based on novel spike counts based methods. For the low cost SNN design, we propose hardware-friendly complexity reduction techniques for both of learning and inferencing modes of operations. First, for the unsupervised learning process, we propose a spike counts based learning method. The novel learning approach utilizes pre- and post-synaptic spike counts to reduce the bit-width of synaptic weights as well as the number of weight updates. For the energy efficient inferencing operations, we propose an accumulation based computing scheme, where the number of input spikes for each input axon is accumulated without instant membrane updates until the pre-defined number of spikes are reached. In addition, the computation skip schemes identify meaningless computations and skip them to improve energy efficiency. Based on the proposed low complexity design techniques, we design and implement the SNN processor using 65 nm CMOS process. According to the implementation results, the SNN processor achieves 87.4% of recognition accuracy in MNIST dataset using only 1-bit 230 k synaptic weights with 400 excitatory neurons. The energy consumptions are 0.26 pJ/SOP and 0.31 μJ/inference in inferencing mode, and 1.42pJ/SOP and 2.63 μJ/learning in learning mode of operations.

Methodologies of Compressing a Stable Performance Convolutional Neural Networks in Image Classification

Deep learning has made a real revolution in the embedded computing environment. Convolutional neural network (CNN) revealed itself as a reliable fit to many emerging problems. The next step, is to enhance the CNN role in the embedded devices including both implementation details and performance. Resources needs of storage and computational ability are limited and constrained, resulting in key issues we have to consider in embedded devices. Compressing (i.e., quantizing) the CNN network is a valuable solution. In this paper, Our main goals are: memory compression and complexity reduction (both operations and cycles reduction) of CNNs, using methods (including quantization and pruning) that don’t require retraining (i.e., allowing us to exploit them in mobile system, or robots). Also, exploring further quantization techniques for further complexity reduction. To achieve these goals, we compress a CNN model layers (i.e., parameters and outputs) into suitable precision formats using several quantization methodologies. The methodologies are: First, we describe a pruning approach, which allows us to reduce the required storage and computation cycles in embedded devices. Such enhancement can drastically reduce the consumed power and the required resources. Second, a hybrid quantization approach with automatic tuning for the network compression. Third, a K-means quantization approach. With a minor degradation relative to the floating-point performance, the presented pruning and quantization methods are able to produce a stable performance fixed-point reduced networks. A precise fixed-point calculations for coefficients, input/output signals and accumulators are considered in the quantization process.

A Generic Foreground Calibration Algorithm For ADCs With Nonlinear Impairments

This paper presents a generic foreground calibration algorithm that estimates and corrects memoryless nonlinear impairments in both single channel and time-interleaved analog-to-digital converters (TIADCs), and which is capable of correcting for amplifier nonlinearity, comparator offsets, and capacitance mismatch for each channel. It operates by generating, and then using, a look-up table which maps raw ADC output decision vectors to linearized output. For TIADCs, the algorithm also uses information gained during the calibration phase to estimate timing and gain mismatches among the sub-ADCs. The problem of selecting an appropriate timing reference so as to relax the requirements on the time-skew correction circuitry is statistically analyzed, as is the corresponding impact on manufacturing yield. Accordingly, a new method is proposed having superior performance; for example, in the case of an eight sub-ADC TIADC system, the proposed scheme reduces the time skew correction requirement by 44% compared with conventional methods. The architecture is instrumented with some additional circuitry to facilitate built-in self-test, allowing manufacturing test time and cost reductions. Implementation aspects are discussed, and several complexity reduction techniques are presented along with synthesis results from a Verilog implementation of the calibration engine.

Comparison of Independent Component Analysis, Principal Component Analysis, and Minimum Noise Fraction Transformation for Tree Species Classification Using APEX Hyperspectral Imagery

Hyperspectral imagery provides detailed spectral information that can be used for tree species discrimination. The aim of this study is to assess spectral–spatial complexity reduction techniques for tree species classification using an airborne prism experiment (APEX) hyperspectral image. The methodology comprised the following main steps: (1) preprocessing (removing noisy bands) and masking out non-forested areas; (2) applying dimensionality reduction techniques, namely, independent component analysis (ICA), principal component analysis (PCA), and minimum noise fraction transformation (MNF), and stacking the selected dimensionality-reduced (DR) components to create new data cubes; (3) super-pixel segmentation on the original image and on each of the dimensionality-reduced data cubes; (4) tree species classification using a random forest (RF) classifier; and (5) accuracy assessment. The results revealed that tree species classification using the APEX hyperspectral imagery and DR data cubes yielded good results (with an overall accuracy of 80% for the APEX imagery and an overall accuracy of more than 90% for the DR data cubes). Among the classification results of the DR data cubes, the ICA-transformed components performed best, followed by the MNF-transformed components and the PCA-transformed components. The best class performance (according to producer’s and user’s accuracy) belonged to Picea abies and Salix alba. The other classes (Populus x (hybrid), Alnus incana, Fraxinus excelsior, and Quercus robur) performed differently depending on the different DR data cubes used as the input to the RF classifier.

Complexity reduction effects on transient, atomic plasmas

Plasma simulations involving collisional-radiative processes present many computational challenges because of the wide range in scales spanned by the governing set of plasma physics equations. These numerical challenges are primarily caused by limitations in resources such as memory and computational time. Collisional-radiative models deliver a part of the challenge because of the large number of levels and transitions that are associated with atomic plasmas. Multiple complexity reduction techniques have been introduced in the past to overcome this obstacle, including the quasi-steady-state solution, uniform grouping, and the recently-devised Boltzmann grouping. These reduction schemes were tested against full numerical simulations of a simplified, time-dependent argon model, with and without an external Planckian radiation field. Results will show the advantages and disadvantages of using complexity reduction techniques when employed in various plasma regimes.

A modified sequence capture approach allowing standard and methylation analyses of the same enriched genomic DNA sample

BackgroundBread wheat has a large complex genome that makes whole genome resequencing costly. Therefore, genome complexity reduction techniques such as sequence capture make re-sequencing cost effective. With a high-quality draft wheat genome now available it is possible to design capture probe sets and to use them to accurately genotype and anchor SNPs to the genome. Furthermore, in addition to genetic variation, epigenetic variation provides a source of natural variation contributing to changes in gene expression and phenotype that can be profiled at the base pair level using sequence capture coupled with bisulphite treatment. Here, we present a new 12 Mbp wheat capture probe set, that allows both the profiling of genotype and methylation from the same DNA sample. Furthermore, we present a method, based on Agilent SureSelect Methyl-Seq, that will use a single capture assay as a starting point to allow both DNA sequencing and methyl-seq.ResultsOur method uses a single capture assay that is sequentially split and used for both DNA sequencing and methyl-seq. The resultant genotype and epi-type data is highly comparable in terms of coverage and SNP/methylation site identification to that generated from separate captures for DNA sequencing and methyl-seq. Furthermore, by defining SNP frequencies in a diverse landrace from the Watkins collection we highlight the importance of having genotype data to prevent false positive methylation calls. Finally, we present the design of a new 12 Mbp wheat capture and demonstrate its successful application to re-sequence wheat.ConclusionsWe present a cost-effective method for performing both DNA sequencing and methyl-seq from a single capture reaction thus reducing reagent costs, sample preparation time and DNA requirements for these complementary analyses.

Model predictive control of ITER plasma current and shape using singular-value decomposition

A model predictive control (MPC) scheme for ITER plasma current and shape controller (PCSC) is presented. The controller is able to control a large number of geometrical plasma shape descriptors using output-space reduction based on singular-value decomposition (SVD). The online optimisation problems imposed by MPC are solved using the dual fast gradient method (dFGM) solver, which is shown to be computationally feasible when MPC complexity-reduction techniques are applied. A performance evaluation in simulation of the flat-top phase of ITER Scenario 1 is presented, showing a moderate general improvement of control, compared to a reference control scheme based on multivariable PID control with SVD. Moreover, the proposed MPC PCSC is capable of avoiding superconductive current saturations, and in some cases shows better performance regarding voltage saturations.

Complexity reduction techniques for quantified diagnosability of stochastic systems

In a discrete event stochastic system, the natural notion of diagnosability, called A-diagnosability, requires that each fault event is eventually detected with probability one. Several definitions of diagnosability degree have been derived from this notion. They examine the detection probability after a fault occurs. To check diagnosability and compute diagnosability degrees, one usually attaches to the original stochastic system the information of a so-called diagnoser, which is in general exponentially larger than the original system. In this paper, we show that the full complexity of such diagnosers is not necessary, and that one can rely on simpler systems, with up to an exponential gain in complexity.

Space diversity‐based low‐complexity sensing using phase inversion and its simple analyses

A computational complexity reduction technique for space diversity-based spectrum sensing is presented. First, the received signals are combined, and a statistic for signal detection is computed from the combined signal. Phase inversion is introduced to avoid attenuation of the combined signal. As the computation of only one statistic is required regardless of the number of receive antenna, the complexity can be reduced. Numerical examples and simple analyses verify the effectiveness of the presented technique.

Reducing Memory Footprints in Explicit Model Predictive Control using Universal Numbers

Explicit Model Predictive Control (MPC) is an effective alternative to reduce the on-line computational demand of traditional MPC. The idea of explicit MPC is to pre-compute the optimal MPC feedback law off-line and store it in a form of look-up table which is to be used in on-line phase. One of the main bottlenecks in an implementation of explicit MPC is memory required to store optimal solutions. This limit its applicability to systems with few states, small number of constraints, and short prediction horizons. In this paper, we present a novel way of reducing the memory footprint of explicit MPC solutions. The procedure is based on encoding all data (i.e., the critical regions and the feedback laws) as universal numbers (unums), which can be viewed as a memory-efficient extension of IEEE floating point standard. By doing so, we illustrate that the total memory footprint can be reduced by 80% without losing control accuracy. An additional advantage of proposed approach is, it can be applied on top of existing complexity reduction techniques.

Complexity Reduction of Multiscale UTLM Cell Clusters

Time-domain electromagnetic simulations employing unstructured tetrahedral meshes offer smooth boundary approximations and graded meshes for multiscale problems. However, multiscale effects may arise not only as a consequence of fine geometry but also from CAD and mesh generation artifacts, and it is critical that the simulation algorithms can be employed in their presence without unduly compromising their computational performance. The ability of the unstructured transmission line modeling (UTLM) algorithm to coalesce small computational cells into larger entities is a key enabler for the approach. This paper demonstrates the use of complexity reduction techniques to both notably reduce the preprocessing time required for this and, as a consequence, substantially extend its capability.

Effects of methylation-sensitive enzymes on the enrichment of genic SNPs and the degree of genome complexity reduction in a two-enzyme genotyping-by-sequencing (GBS) approach: a case study in oil palm (Elaeis guineensis).

Advances in next generation sequencing have facilitated a large-scale single nucleotide polymorphism (SNP) discovery in many crop species. Genotyping-by-sequencing (GBS) approach couples next generation sequencing with genome complexity reduction techniques to simultaneously identify and genotype SNPs. Choice of enzymes used in GBS library preparation depends on several factors including the number of markers required, the desired level of multiplexing, and whether the enrichment of genic SNP is preferred. We evaluated various combinations of methylation-sensitive (AatII, PstI, MspI) and methylation-insensitive (SphI, MseI) enzymes for their effectiveness in genome complexity reduction and enrichment of genic SNPs. We discovered that the use of two methylation-sensitive enzymes effectively reduced genome complexity and did not require a size selection step. On the contrary, the genome coverage of libraries constructed with methylation-insensitive enzymes was quite high, and the additional size selection step may be required to increase the overall read depth. We also demonstrated the effectiveness of methylation-sensitive enzymes in enriching for SNPs located in genic regions. When two methylation-insensitive enzymes were used, only 16% of SNPs identified were located in genes and 18% in the vicinity (± 5 kb) of the genic regions, while most SNPs resided in the intergenic regions. In contrast, a remarkable degree of enrichment was observed when two methylation-sensitive enzymes were employed. Almost two thirds of the SNPs were located either inside (32–36%) or in the vicinity (28–31%) of the genic regions. These results provide useful information to help researchers choose appropriate GBS enzymes in oil palm and other crop species.Electronic supplementary materialThe online version of this article (doi:10.1007/s11032-016-0572-x) contains supplementary material, which is available to authorized users.

Reduced-complexity fingerprinting in WLAN-based indoor positioning

In this paper, a two-stage complexity reduction technique is implemented for a fingerprinting-based indoor positioning system. Several computation reduction techniques are applied during the offline and online phases of fingerprinting. Specifically, dimensionality reduction algorithms, clustering techniques and fast search strategies are integrated to achieve an ultimate reduction in the computational requirements of fingerprinting. The computational cost of fingerprinting is first reduced by restricting the location fingerprints to signal strength values received from informative access points (APs). Afterwards, clustering techniques are employed to speed up the online search for the target best match. Finally, selective matching between the target RSS and the pre-stored fingerprints is proposed to reduce the computational cost even further. In particular, this paper studies different dimensionality reduction methods and chooses the method that minimizes the positioning error. Moreover, a hybrid search solution of clustering and fast search strategies is proposed to minimize the search operations to find a user position in the radio map.

Computational complexity reduction techniques for real-time and high-resolution medical ultrasound imaging using the beam-space Capon method

The beam-space (BS) Capon method is an adaptive beamforming technique that reduces computational complexity. However, the complexity is not low enough for real-time imaging. Reducing the number of time-delay and transformation processes from element-space to BS signal processing is required. We propose a technique that replaces the time-delay processes by the multiplication of steering vectors and covariance matrices. In addition, we propose a compensation technique for estimating the intensity accurately. In an experimental study using a 2.0 MHz transmission frequency on a 15 × 10.4 mm2 region of interest, the first side-lobe level, the −6 dB beam width, the intensity’s estimation error, and the calculation time of the conventional method were −15 dB, 0.70 mm, 3.2 dB, and 656 ms. Those of the proposed method were −17 dB, 0.36 mm, 1.6 dB, and 81 ms, respectively. Using our method on three CPUs achieves imaging of 37 frames/s.

Dynamic estimation of the modeling error statistics in Diffuse Optical Tomography

Diffuse Optical Tomography (DOT) is a non-invasive imaging technique that suffers from a typical large-scale and ill-posed inverse problem with low spatial resolution. In DOT, the inverse problem is computationally intensive and decreasing the computation complexity and making it well-posed is the one of the most challenging research areas. More precisely, one of the well-known complexity reduction techniques is defined as applying modelling error originated from discretization of forward problem. Applying the discretization error in Bayesian inference has already been discussed; the method in which the likelihood is modified by an off-line prior density estimation. This paper implements a new method to enhance the modelling error approach using an iterative scheme to update statistical parameters of modelling discrepancy in DOT. The algorithm is very similar to Ensemble Kalman Filter. Moreover, the reconstruction process in the applied method is conducted by a small sample size rather than off-line method. Hence, the computation complexity is decreased and the algorithm converges in few iterations. The efficiency of the proposed method is illustrated by simulations.

Low-cost FPGA implementation of 2D digital pre-distorter for concurrent dual-band power amplifier

This paper proposes a low-cost hardware architecture based on concurrent dual-band digital pre-distorter (DPD). The architecture is implemented on field programmable gate array (FPGA) to compensate for the nonlinearity of the concurrent dual-band power amplifier (PA). This implementation introduces a novel model complexity reduction technique into system, namely, time-division multiplexing for out-of-band lookup tables (LUTs) sharing. Performances are evaluated with an experimental test setup using a wideband class-F PA. The dual-band signal center frequency separated by 80 MHz. Lower and upper center frequency are located at 2.61 GHz and 2.69 GHz, respectively. This novel DPD implementation maintains excellent performance, but uses hardware resources reduced by 29.17% compared with conventional approaches. The results show that the adjacent channel power ratio (ACPR) is less than 59 dBc and normalized mean square error (NMSE) is around 62dB for lower sideband (LSB) and 63dB for upper sideband (USB).

RAD Capture (Rapture): Flexible and Efficient Sequence-Based Genotyping.

Massively parallel sequencing has revolutionized many areas of biology, but sequencing large amounts of DNA in many individuals is cost-prohibitive and unnecessary for many studies. Genomic complexity reduction techniques such as sequence capture and restriction enzyme-based methods enable the analysis of many more individuals per unit cost. Despite their utility, current complexity reduction methods have limitations, especially when large numbers of individuals are analyzed. Here we develop a much improved restriction site-associated DNA (RAD) sequencing protocol and a new method called Rapture ( R: AD c APTURE: ). The new RAD protocol improves versatility by separating RAD tag isolation and sequencing library preparation into two distinct steps. This protocol also recovers more unique (nonclonal) RAD fragments, which improves both standard RAD and Rapture analysis. Rapture then uses an in-solution capture of chosen RAD tags to target sequencing reads to desired loci. Rapture combines the benefits of both RAD and sequence capture, i.e., very inexpensive and rapid library preparation for many individuals as well as high specificity in the number and location of genomic loci analyzed. Our results demonstrate that Rapture is a rapid and flexible technology capable of analyzing a very large number of individuals with minimal sequencing and library preparation cost. The methods presented here should improve the efficiency of genetic analysis for many aspects of agricultural, environmental, and biomedical science.

Recent advances in molecular marker techniques: Insight into QTL mapping, GWAS and genomic selection in plants

Recent advances in sequencing technology have brought several novel platforms for marker development and subsequent genotyping. The high-throughput and cost effective marker techniques have changed the entire scenario of marker applications. The huge genotypic data obtained with next generation sequencing (NGS) also demands analytical tools, statistical advances, and comprehensive understanding to cope with breeding applications. In the present review, we discussed different available marker techniques, their strengths, and limitations. Emphasis was given on software tools, analytical pipelines, workbenches, and online resources available for marker development. Comparison of SNP genotyping involving complexity reduction techniques like GBS, RRL, RAD, and array-based platforms were presented in a view to describe suitability for specific purposes. We found that genotyping by whole genome re-sequencing has great potential, and could be a routine application in the near future with continuously decreasing cost of sequencing. Microsatellites, still a valuable option for breeders, have also advanced with NGS. Here a catalogue of tools for microsatellite evaluation in short sequence reads was provided. The most common applications of molecular marker like QTL mapping, genome-wide association mapping (GWAS), and genomic selection were highlighted. The present review will be helpful for the effective utilization of available resources and for the planning of crop improvement programs employing molecular marker techniques.

A temporal compressive resource allocation technique for complexity reduction in PLC transceivers

AbstractThis work introduces a novel resource allocation technique for dealing with linear and periodically time‐varying power line channels when an orthogonal frequency division multiplexing scheme is applied. By exploiting the correlations within one cycle of the mains signal, among cycles of mains signal and a combination of these connections, the proposed technique can offer three distinct trade‐offs between computational complexity reduction and data‐rate loss. Numerical results, which are based on measured data, are used to analyse these trade‐offs, when these inter‐cycle and intra‐cycle relationship are taken into account. Also, we verify that the use of the normalised signal‐to‐noise ratio incurs very low performance degradation, and because of computational complexity rationale, its use is strongly recommended. Additionally, we show that those cases, in which the correlations among the cycles of the mains signal is relevant, offer the best trade‐off between computational complexity reduction and data‐rate loss. Finally, we show that the proposed technique can achieve the optimal data‐rate and offers substantial improvements in terms of computational complexity when compared with existing approaches, including the Institute of Electrical and Electronics Engineers 1901 standard. Copyright © 2015 John Wiley & Sons, Ltd.