Real-time System Applications Research Articles

Iterative Algorithm Induced Deep-Unfolding Neural Networks: Precoding Design for Multiuser MIMO Systems

Optimization theory assisted algorithms have received great attention for precoding design in multiuser multiple-input multiple-output (MU-MIMO) systems. Although the resultant optimization algorithms are able to provide excellent performance, they generally require considerable computational complexity, which gets in the way of their practical application in real-time systems. In this work, in order to address this issue, we first propose a framework for deep-unfolding, where a general form of iterative algorithm induced deep-unfolding neural network (IAIDNN) is developed in matrix form to better solve the problems in communication systems. Then, we implement the proposed deep-unfolding framework to solve the sum-rate maximization problem for precoding design in MU-MIMO systems. An efficient IAIDNN based on the structure of the classic weighted minimum mean-square error (WMMSE) iterative algorithm is developed. Specifically, the iterative WMMSE algorithm is unfolded into a layer-wise structure, where a number of trainable parameters are introduced to replace the high-complexity operations in the forward propagation. To train the network, a generalized chain rule of the IAIDNN is proposed to depict the recurrence relation of gradients between two adjacent layers in the back propagation. Moreover, we discuss the computational complexity and generalization ability of the proposed scheme. Simulation results show that the proposed IAIDNN efficiently achieves the performance of the iterative WMMSE algorithm with reduced computational complexity.

Wavelet Scattering Network-Based Machine Learning for Ground Penetrating Radar Imaging: Application in Pipeline Identification

Automatic and efficient ground penetrating radar (GPR) data analysis remains a bottleneck, especially restricting applications in real-time monitoring systems. Deep learning approaches have good practice in automatic object identification, but their intensive data requirement has reduced their applicability. This paper developed a machine learning framework based on wavelet scattering networks to analyze GPR data for subsurface pipeline identification. Wavelet scattering network is functionally equivalent to convolutional neural networks, and its null-parameter property is intended for non-intensive datasets. A double-channel framework is designed with wavelet scattering networks followed by support vector machines to determine the existence of pipelines on vertical and horizontal traces separately. Classification accuracy rates arrive around 98% and 95% for datasets without and with noises, respectively, as well as 97% for considering surface roughness. Pipeline locations and diameters are convenient to determine from the reconstructed profiles of both simulated and practical GPR signals. However, the results of 5 cm pipelines are sensitive to noises. Nonetheless, the developed machine learning approach presents promising applicability in subsurface pipeline identification.

Encapsulation of Co3 O4 Nanocone Arrays via Ultrathin NiO for Superior Performance Asymmetric Supercapacitors.

Designing of multicomponent transition metal oxide system through the employment of advanced atomic layer deposition (ALD) technique over nanostructures obtained from wet chemical process is a novel approach to construct rational supercapacitor electrodes. Following the strategy, core-shell type NiO/Co3 O4 nanocone array structures are architectured over Ni-foam (NF) substrate. The high-aspect-ratio Co3 O4 nanocones are hydrothermally grown over NF following the precision controlled deposition of shell NiO considering Co3 O4 nanocone as host. NiO thickness of 5 nm exhibits the highest specific capacity of 1242 C g-1 (2760 F g-1 ) at current density 2 A g-1 , which is greater than pristine Co3 O4 @NF (1045.8 C g-1 or 2324 F g-1 ). The rate capability with 5 nm NiO/Co3 O4 @NF nanocone structures is about 77% whereas Co3 O4 @NF retains 46 % of capability at 10 A g-1 . The ultrathin ALD 5 nm NiO accelerates both rate capability and 95.5% cyclic stability after 12 000 charge-discharge cycles. An asymmetric device fabricated between 5 nm NiO/Co3 O4 @NF (positive) || activated carbon (negative) achieves an energy density of 81.45 Wh kg-1 (4268 W kg-1 ) with good cycling device stability. Additionally, LEDs can be energized by two ASC device in series. This work opens the path in both advanced electrode material and surface modification of earth-abundant systems for efficient and real-time supercapacitor applications.

Determination of parameters of models of signal deviation from neighboring samples by convex functions

The paper considers the Nyquist-Shannon-Kotelnikov theorem, the Khlistunov criterion of the greatest deviation, the corresponding deterministic formulas for finding a constant sampling interval of a signal in time and their drawbacks in traditional applications in real-time control systems. Mathematical models of the behavior of a signal or a controlled process between two adjacent samples from it along the trajectory of the greatest deviation in the form of convex functions (half-wave of a sinusoid, a power function of the modulus of time), which are free from these shortcomings, are considered. Constraints are found on the number and magnitude of the time parameters of these models, such as maximum speed, acceleration and sharpness, which are determined by analyzing the signal spectrum and the Nyquist interval. Formulas are obtained, examples of calculations and graphs are given, which illustrate the dependence of the time sampling interval on the accuracy of the linear approximation of the signal and allow increasing the number of signals processed in the system without increasing requirements for its performance.

Packet reservations in real-time multipath transmissions

Subject of Research. The paper presents research of mechanisms of multipath redundant transmissions in computer networks. Method. The developed application level protocol gives the possibility to increase the probability of timely and faultless data delivery in computer networks which work within the conditions of delivery time limits using multipath redundant transmissions of packet copies via different channels. Main Results. The protocol with redundant multipath transmissions opportunity based on UDP transport protocol is developed. The efficiency analysis of the presented solutions is carried out on the basis of simulation modeling in OMNeT++ environment. Models of the proposed protocol and computer network with redundant transmissions opportunity are developed. Experiments with these models are carried out and efficiency area of the developed protocol application is determined. The multiplicative criterion that shows time reserve of timely and faultless transmitted packets is used as an efficiency indicator. The presented protocol shows efficiency in various working cases with the intensity and redundancy coefficient change of packet arrival. Practical Relevance. The developed algorithms and mechanisms can be used for program implementation of protocol with redundant multipath transmissions aimed at its application in cyberphysic real-time systems.

Channel and Feature Selection for a Motor Imagery-Based BCI System Using Multilevel Particle Swarm Optimization.

Brain-computer interface (BCI) is a communication and control system linking the human brain and computers or other electronic devices. However, irrelevant channels and misleading features unrelated to tasks limit classification performance. To address these problems, we propose an efficient signal processing framework based on particle swarm optimization (PSO) for channel and feature selection, channel selection, and feature selection. Modified Stockwell transforms were used for a feature extraction, and multilevel hybrid PSO-Bayesian linear discriminant analysis was applied to optimization and classification. The BCI Competition III dataset I was used here to confirm the superiority of the proposed scheme. Compared to a method without optimization (89% accuracy), the best classification accuracy of the PSO-based scheme was 99% when less than 10.5% of the original features were used, the test time was reduced by more than 90%, and it achieved Kappa values and F-score of 0.98 and 98.99%, respectively, and better signal-to-noise ratio, thereby outperforming existing algorithms. The results show that the channel and feature selection scheme can accelerate the speed of convergence to the global optimum and reduce the training time. As the proposed framework can significantly improve classification performance, effectively reduce the number of features, and greatly shorten the test time, it can serve as a reference for related real-time BCI application system research.

Performance comparison of partitioned and global approaches for weakly hard real-time systems

One way to minimizing resource requirements is through the careful management and allocation, in example, scheduling. Research on weakly hard real-time scheduling on multiprocessor has been extremely limited; most prior research on real-time scheduling on weakly hard real-time has been confined to uniprocessors. The need for multiprocessor is due to issues that impose limits on the performance of a single processor. As real-time application systems increasingly come to be implemented upon multiprocessor environments, thus, this study applies multiprocessor scheduling approach for verification of weakly hard real-time tasks and to guaranteeing the timing requirements of the tasks. In fact, within the multiprocessor, the task allocation and migration problems seem even harder than in uniprocessor case; thus, in order to cater that problem, the sufficient and efficient scheduling algorithm supported by accurate schedulability analysis technique is present to provide weakly hard real-time guarantees. The proposed approach involves the two existing multiprocessor real-time scheduling algorithms combining with the hyperperiod analysis and deadline models; weakly hard constraints and μ-pattern. In this paper, the Matlab simulation tool is used in order to validate the result of analysis. From the experimental and performance evaluation results, it proved that the proposed approach is satisfied the tasks deadlines with less number of misses.

Towards Efficient Neuromorphic Hardware: Unsupervised Adaptive Neuron Pruning

To solve real-time challenges, neuromorphic systems generally require deep and complex network structures. Thus, it is crucial to search for effective solutions that can reduce network complexity, improve energy efficiency, and maintain high accuracy. To this end, we propose unsupervised pruning strategies that are focused on pruning neurons while training in spiking neural networks (SNNs) by utilizing network dynamics. The importance of neurons is determined by the fact that neurons that fire more spikes contribute more to network performance. Based on these criteria, we demonstrate that pruning with an adaptive spike count threshold provides a simple and effective approach that can reduce network size significantly and maintain high classification accuracy. The online adaptive pruning shows potential for developing energy-efficient training techniques due to less memory access and less weight-update computation. Furthermore, a parallel digital implementation scheme is proposed to implement spiking neural networks (SNNs) on field programmable gate array (FPGA). Notably, our proposed pruning strategies preserve the dense format of weight matrices, so the implementation architecture remains the same after network compression. The adaptive pruning strategy enables 2.3× reduction in memory size and 2.8× improvement on energy efficiency when 400 neurons are pruned from an 800-neuron network, while the loss of classification accuracy is 1.69%. And the best choice of pruning percentage depends on the trade-off among accuracy, memory, and energy. Therefore, this work offers a promising solution for effective network compression and energy-efficient hardware implementation of neuromorphic systems in real-time applications.

Neuro-genetic programming for multigenre classification of music content

A machine learning approach based on hybridization of genetic programming and neural networks is used to derive mathematical models for music genre classification. We design three multi-label classifiers with different trade-offs between complexity and accuracy, which are able to identify the degree of belonging of music content to ten different music genres. Our approach is innovative as it entirely relies on simple analytical functions and a reduced number of features. Resulting classifiers have an extremely low computational complexity and are suitable to be easily integrated in low-cost embedded systems for real-time applications. The GTZAN dataset is used for model training and to evaluate the accuracy of the proposed classifiers. Despite of the reduced number of features used in our approach, the accuracy of our models is found to be similar to that of more complex music genre classification tools previously published in the literature.

A Quantitative Approach to Analyze Modifiability in Software Architectural Design of Agile Application Systems

The present-day software application systems require a high degree of agility during development and operational phases due to the advancements in software technologies and also because of the need to support the variation points in software architecture. A single architectural style will not be adequate for the architectural design of such application systems. This paper proposes a composite architectural style involving three different architectural styles, namely model view controller, pipes and filters and reflection architectural pattern. A metric to evaluate the extent to which the proposed architectural design is modifiable is defined and formulated. The number of direct connections between the components and their modes of operation are the various factors that will determine the extent to which the architectural design is modifiable. The model has been tested successfully for a prototype document processing application system. The composite architectural design is quite generic and it can be used for any real-time application system where the three modes of operation - data stream mode, user interaction and dynamic invocation mode exist together.

Implementation of deep neural networks on FPGA-CPU platform using Xilinx SDSOC

Deep Convolutional Neural Networks (CNNs) are the state-of-the-art systems for image classification due to their high accuracy but on the other hand their high computational complexity is very costly. The acceleration is the target in this field nowadays for using these systems in real time applications. The Graphics Processing Units is the solution but its high-power consumption prevents its utilization in daily-used equipment moreover the Field Programmable Gate Array (FPGA) has low power consumption and flexible architecture which fits more for CNN implementations. This work discusses this problem and provides a solution that compromises between the speed of the CNN and the power consumption of the FPGA. This solution depends on two main techniques for speeding up: parallelism of layers resources and pipelining inside some layers. On the other hand, we added a new methodology to compromise the area requirements with the speed and design time by implementing CNN using Xilinx SDSOC tool (including processor and FPGA on the same board). Implementing design using HW/SW partitioning will enhance time design based on high level language(C or C++) in Vivado HLS (High Level Synthesis). It also fits for more large designs than using FPGA only and faster in design time.

Segmentation of Malayalam Handwritten Characters into Pattern Primitives and Recognition using SVM

This paper describes a lexical analysis (segmentation) approach in Pattern Recognition for Online Handwritten Character Recognition (OHCR) in Malayalam. The subunits (Pattern Primitives) in the single stroke vowel characters in Malayalam are identified and marked with pattern primitives to obtain a reference set of characters. Segmentation of the handwritten character samples into pattern primitives is made using a Combined Approach of Ramer Douglas Peucker algorithm and Eight Direction Freeman Code as per reference set. Features that are unique in the primitives of a character are extracted. The discriminating features identified are the direction of first primitive, segment count, cusp in second primitive, crossing in third primitive, and cusp in seventh primitive. The experiments were conducted on 100 samples per character that showed exact segmentation as per the reference set. With a five dimension feature set, the study achieved a recognition rate of 95.77% for five-fold cross-validation using Support Vector Machine with RBF kernel. The study shows that the segmentation of characters into pattern primitives is an effective method to realize accurate Malayalam OHCR systems for real-time applications.

A fast stream transaction system for real-time IoT applications

Due to the recent prevalence of IoT (Internet of Things) technologies, various IoT devices connect to the Internet and continuously send their generated data to remote processing computers such as video data or sensor data. The transaction rate is one of the main factors to improve the performance of some IoT applications. For instance, in surveillance systems, the probability to catch a thief increases as the processing computer analyzes the video with a higher transaction rate. To improve the transaction rate, some methods reduce the transaction time between a processing computer and stream data sources under a static transaction interval. However, the transaction rate can be further improved by changing the transaction interval dynamically depending on the transaction time. In this paper, we propose a method to improve the transaction rate by changing the transaction interval dynamically. In our proposed method, a processing computer sometimes changes the transaction interval to be the same length as the average transaction time. Moreover, our proposed method adopts a progressive quality improvement (PQI) approach to reduce the transaction time. We measured the transaction rate of our proposed method by both a simulator and an implemented system. We confirmed that our proposed method can improve the transaction rate by 4.4 times and the transaction time by 21% at least compared with the conventional method. Moreover, we confirmed that the average frame rate increases 22% compared with a simple method in a real situation.

Adaptive Artificial Potential Fields with Orientation Control Applied to Robotic Manipulators

This paper proposes the integration of an Adaptive Artificial Potential Fields algorithm with a new end effector orientation control technique for real-time robot path planning. The development of autonomous robotic systems has undergone several advances in path planning algorithms. These systems generate object collision-free paths in the robot’s workspace. In this context, the Artificial Potential Fields technique has been the focus of improvements in recent years due to its simplicity of application and efficiency in real-time systems, since it does not require a global mapping of the robot’s workspace. In spite of its efficiency, this technique is susceptible to local minimum problems of different natures, such as Goals Non-Reachable with Obstacles Nearby (GNRON). To solve this problem, we suggest the use of an improvement called Adaptive Artificial Potential Fields used in conjunction with the proposed end effector orientation control technique, which allows reaching a desired orientation of the end effector. The resulting force, generated from the Adaptive Artificial Potential Field, guides the robot end effector to the goal. The Robot Operating System (ROS) framework and a collaborative robot manipulator UR5 are used to validate the proposed method on an approaching task for an object on a 3D printer tray.

A Tag Based Random Order Vector Reduction Circuit

Vector reduction is a very common operation to reduce a vector into a single scalar value in many scientific and engineering application scenarios. Therefore a fast and efficient vector reduction circuit has great significance to the real-time system applications. Usually the pipeline structure is widely adopted to increase the throughput of the vector reduction circuit and achieve maximum efficiency. In this paper, to deal with multiple vectors of variable length in random input sequence, a novel tag based fully pipelined vector reduction circuit is firstly proposed, in which a cache state module is used to queer and update the cache state of each vector. However, when the quantity of the input vector becomes large, a larger cache state module is required, which consumes more combinational logic and lower the operating frequency. To solve this problem, a high speed circuit is proposed in which the input vectors will be divided into several groups and sent into the dedicated cache state circuits, which can improve the operating frequency. Compared with other existing work, the prototype circuit and the improved circuit based on the prototype circuit can achieve the smallest Slices × us (<; 80% of the state-of-the-art work) for different input vector lengths. Moreover, both circuits can provide simple and efficient interface whose access timing is similar to that of a RAM. Therefore the circuits can be applied in a greater range.

Fast Sparse Aperture ISAR Autofocusing and imaging via ADMM based Sparse Bayesian Learning.

Sparse aperture ISAR autofocusing and imaging is generally achieved by methods of compressive sensing (CS), or, sparse signal recovery, because non-uniform sampling of sparse aperture disables fast Fourier transform (FFT)-the core of traditional ISAR imaging algorithms. Note that the CS based ISAR autofocusing methods are often computationally heavy to execute, which limits their applications in real-time ISAR systems. The improvement of computational efficiency of sparse aperture ISAR autofocusing is either necessary or at least highly desirable to promote their practical usage. This paper proposes an efficient sparse aperture ISAR autofocusing algorithm. To eliminate the effect of sparse aperture, the ISAR image is reconstructed by sparse Bayesian learning (SBL), and the phase error is estimated by minimum entropy during the reconstruction of ISAR image. However, the computation of expectation in SBL involves a matrix inversion with an intolerable computational complexity of at least O(L3). Here, in the Bayesian inference of SBL, we transform the time-consuming matrix inversion into an element-wise matrix division by the alternating direction method of multipliers (ADMM). An auxiliary variable is introduced to divide the computation of posterior into three simpler subproblems, bringing computational efficiency improvement. Experimental results based on both simulated and measured data validate the effectiveness as well as high efficiency of the proposed algorithm. It is 20-30 times faster than the SBL based sparse aperture ISAR autofocusing approach.

An RF backscatterer system for real-time multi-sensing applications

In this paper, a power-efficient passive analog frequency modulation based RF system is proposed for real-time communication with multiple sensor elements for internet-of-things (IoT) based smart applications. The proposed RF tag with integrated sensors uses an onboard ultra low power oscillator to generate a single frequency modulation signal unique to each of the sensor elements. The generated single tone frequency is modulated to generate the backscattered RF signal, which is immune to the interference present in the propagating environment. The maximum power required for continuous operation of the sensing oscillator is 30 μW, which is harvested from the received RF signal. The designed multi-sensor RF tag has a sensitivity of −6 dBm and has the capacity to integrate up to six different sensors within the available single channel RFID bandwidth of 300 kHz. The wireless communication distance of 10 ft is achieved due to the low power consumption of the multi-sensor RF tag. A passive RF tag with two example sensing elements; temperature and luminescence are demonstrated. The temperature and luminescence change from 25 °C to 60 °C and from 0 to 20 footcandle (lumen/ft2) is represented by 75–100 kHz and 125–150 kHz frequency range, respectively.

A flexible piezoelectric nanogenerator using conducting polymer and silver nanowire hybrid electrodes for its application in real-time muscular monitoring system

In the present work, a flexible piezoelectric nanogenerators (PNGs) using poly(vinylidene fluoride) (PVDF) as an active layer with hybrid electrode is demonstrated. Here, a poly(3,4-ethylenedioxythiophene) (PEDOT) derivative, poly(2-hexyl-2,3-dihydrothieno[3,4-b][1,4]dioxine:dodecyl sulfate (PEDOT-C6:DS) was introduced. The polymer material has a good solubility and dispersion with organic solvents. Also, the polymer has a good electrical conductivity with hydrophobic surface and uniform conducting network on PVDF. The device performance of hybrid layer with a type of uniformly covered silver nanowire (AgNWs) network and synthesized PEDOT-C6:DS with different concentration as an electrode has been tested. This PNG exhibit superior energy harvesting performance with maximum piezoelectric output voltage and current 7.02 V and 1.11 μA, at 8 Hz operating frequency, respectively. The fabricated hybrid electrode PNG (1.5 cm × 2.5 cm) generates a maximum output power of 1.18 μW. The fabricated PNG device is capable to lit up white LEDs and charged a commercial capacitor of 1 μF with charged voltage 2.68 V in 80 s. Further, with the results obtained it is confirmed that the 1 wt% PEDOT-C6:DS is a promising material for hybrid electrodes in electronic applications.

Design and Implementation of Embedded Audio System based on Zynq-SOC

Nowadays smart world requires, Embedded audio system with optimum design metrics for smart applications in various fields like smart car, intelligent systems and Robotics etc. This paper describes the Design and implementation of embedded Audio system for real-time applications on SOC-FPGA with optimized design metrics (low power, low-cost, low development time, low area, high speed). An electret microphone / line in are used to feed the audio input. An audio codec from Analog devices named ADAU1761 which is integrated on the zynq-7020 board. In the proposed embedded audio system, the block design in Vivado2017.2 has been modeled with VHDL; application software developed using C language in SDK2017.2. This Audio system is the optimized solution for a wide range of smart applications.

Slurry bubble column measurements using advanced electrical capacitance volume tomography sensors

A novel approach to studying dynamic three phase systems using Electrical Capacitance Volume Tomography (ECVT) is proposed and verified against previously published pressure gauge techniques and hydrodynamics patterns. In this three-phase study, hold up of water, glass beads, and air are measured simultaneously in a slurry bubble column reactor. Application of ECVT to investigate three phase systems holds several advantages over pressure gauges including the ability to deliver non-invasive real time direct measurements without making assumptions about the hydrodynamics. Results support the conclusion that ECVT can be used in place of pressure gauges to obtain accurate holdup measurements of three-phase flow systems in real time applications. The holdup of each phase inside of a three-phase slurry bubble column were measured using pressure sensors and an ECVT sensor. The results suggest that ECVT can be used in place of pressure sensors to obtain accurate holdup measurements of three-phase flows in real time. • Pressure and ECVT phase holdup measurements closely match. • ECVT reconstructed solid holdup images used to analyze bed expansion and contraction. • Solids and gas holdup measurements reflect hydrodynamic principles. • ECVT presents a valid replacement for pressure measurements in three-phase flows.