Reuse Mechanism Research Articles

DLformer: A Dynamic Length Transformer-Based Network for Efficient Feature Representation in Remaining Useful Life Prediction.

Representation learning-based remaining useful life (RUL) prediction plays a crucial role in improving the security and reducing the maintenance cost of complex systems. Despite the superior performance, the high computational cost of deep networks hinders deploying the models on low-compute platforms. A significant reason for the high cost is the computation of representing long sequences. In contrast to most RUL prediction methods that learn features of the same sequence length, we consider that each time series has its characteristics and the sequence length should be adjusted adaptively. Our motivation is that an "easy" sample with representative characteristics can be correctly predicted even when short feature representation is provided, while "hard" samples need complete feature representation. Therefore, we focus on sequence length and propose a dynamic length transformer (DLformer) that can adaptively learn sequence representation of different lengths. Then, a feature reuse mechanism is developed to utilize previously learned features to reduce redundant computation. Finally, in order to achieve dynamic feature representation, a particular confidence strategy is designed to calculate the confidence level for the prediction results. Regarding interpretability, the dynamic architecture can help human understand which part of the model is activated. Experiments on multiple datasets show that DLformer can increase up to 90% inference speed, with less than 5% degradation in model accuracy.

MDFL: Multi-Domain Diffusion-Driven Feature Learning

High-dimensional images, known for their rich semantic information, are widely applied in remote sensing and other fields. The spatial information in these images reflects the object's texture features, while the spectral information reveals the potential spectral representations across different bands. Currently, the understanding of high-dimensional images remains limited to a single-domain perspective with performance degradation. Motivated by the masking texture effect observed in the human visual system, we present a multi-domain diffusion-driven feature learning network (MDFL) , a scheme to redefine the effective information domain that the model really focuses on. This method employs diffusion-based posterior sampling to explicitly consider joint information interactions between the high-dimensional manifold structures in the spectral, spatial, and frequency domains, thereby eliminating the influence of masking texture effects in visual models. Additionally, we introduce a feature reuse mechanism to gather deep and raw features of high-dimensional data. We demonstrate that MDFL significantly improves the feature extraction performance of high-dimensional data, thereby providing a powerful aid for revealing the intrinsic patterns and structures of such data. The experimental results on three multi-modal remote sensing datasets show that MDFL reaches an average overall accuracy of 98.25%, outperforming various state-of-the-art baseline schemes. Code available at https://github.com/LDXDU/MDFL-AAAI-24.

A column generation approach for the lexicographic optimization of intra-hospital transports

AbstractOver the last few years, the efficient design of processes in hospitals and medical facilities has received more and more attention, particularly when the improvement of the processes is aimed at relieving the workload of medical staff. To this end, we have developed a method to determine optimal allocations of intra-hospital transports to hospital transport employees. When optimizing transport plans in hospitals, there are various optimization goals to strive for. Therefore, we used a lexicographic approach to solve this multi-criteria optimization problem. To calculate optimal transport plans in a sufficiently short computation time, we have decomposed the problem at hand with the Dantzig–Wolfe reformulation and solved the resulting pricing subproblem with an enumerative column generation approach. To improve the efficiency of the column generation process, we have investigated and implemented different pruning methods, dominance rules and a column reuse mechanism for the online setting of the application at hand. In an extensive computational study, we first evaluated the efficiency of the different pruning methods before we compared our solution approach with the standard branch-and-bound column search approach, classical column generation methods and the solution of an integrated MIP model solved by a commercial solver. Finally, we present performance indicators of the transport optimization tool, which was developed from our method and is now productively used in a German hospital.

A Knowledge Guided Transfer Strategy for Evolutionary Dynamic Multiobjective Optimization

The key task in dynamic multiobjective optimization problems (DMOPs) is to find Pareto-optima closer to the true one as soon as possible once a new environment occurs. Previous dynamic multiobjective evolutionary algorithms (DMOEAs) normally focus on DMOPs with regular environmental changes, but neglect widespread random one, limiting their applications in real-world fields. To address this issue, a knowledge guided transfer strategy based DMOEA is proposed in the paper. First, knowledge described as a two-tuple is extracted under each historical environment and preserved to a knowledge pool. Redundant knowledge is recognized and adaptively removed so as to guarantee the diversity of the pool. Second, a knowledge matching strategy is developed to re-evaluate the representative of each stored knowledge under a new environment, with the purpose of finding the most valuable one to promote positive knowledge transfer. Third, an improved knowledge transfer mechanism based on subspace alignment is introduced. By integrating it with knowledge reuse mechanism, a hybrid transfer strategy is constructed to adaptively select the most suitable one in terms of the similarity degree of selected knowledge on the current environment, and then generate a new initial population. Experiments on 20 benchmark problems demonstrate that the knowledge guided transfer strategy outperforms five state-of-the-art algorithms, achieving good versatility in solving DMOPs with both regular and random changes.

An optimal container update method for edge‐cloud collaboration

AbstractEmerging computing paradigms provide field‐level service responses for users, for example, edge computing, fog computing, and MEC. Edge virtualization technologies represented by Docker can provide a platform‐independent, low‐resource‐consumption operating environment for edge service. The image‐pulling time of Docker is a crucial factor affecting the start‐up speed of edge services. The layer reuse mechanism of native Docker cannot fully utilize the duplicate data of node local images. In this paper, we propose a chunk reuse mechanism (CRM), which effectively targets node‐local duplicate data during container updates and reduces the volume of data transmission required for image building. We orchestrate the CRM process for cloud and remote‐cloud nodes to ensure that the resource overhead from container update data preparation and image reconstruction is within an acceptable range. The experimental results show that the CRM proposed in this paper can effectively utilize the node local duplicate data in the synchronous update of containers in multiple nodes, reduce the volume of data transmission, and significantly improve container update efficiency.

Universal wavelength reuse mechanism for optical networks-on-chip based on a cooperative game

Optical networks-on-chip (ONoCs) based on wavelength division multiplexing (WDM) technology have lower end-to-end (ETE) delay, larger bandwidth, and higher throughput than NoCs. However, the utilization and parallelism of individual wavelengths are essential since WDM technology would require more optical devices to be integrated with ONoCs, leading to higher on-chip overhead. In this paper, we propose an optimal model of communication node cooperation based on game theory, in which the loss of the optical signal on the transmission path for a specific optical receiver sensitivity establishes the payoff function of each participant. In this cooperative game model, the communication nodes are regarded as the participants of the cooperative game. By analyzing the principle of optical router port occupation, the coordinate-based coalition assignment algorithm is proposed to implement fast construction of coalitions and reach the Nash equilibrium of the cooperative game. In Nash equilibrium, the utilization and parallelism of individual wavelengths are higher, and the network’s performance is optimized and improved. Cooperative-game-based ONoCs have more significant throughput and lower ETE delay. The optical signal loss and minimum output power of the optical transmitter can be further reduced at a given optical receiver sensitivity. By applying the cooperative game to ONoCs, the network’s optimization effect becomes more evident as the network scale increases and the number of communication nodes grows.

Multi-Scale MLP-Mixer for image classification

MLP-Mixer is a vision architecture that solely relies on multilayer perceptrons (MLPs), which despite their simple architecture, they achieve a slightly inferior accuracy to the state-of-the-art models on ImageNet. Given that the MLP-Mixer segments each input image into a fixed number of patches, small-scale MLP-Mixers are preferred due to attaining better accuracy because the image is segmented into more patches. However, this strategy significantly increases the computational burden. Nevertheless, this paper argues that even in the same dataset, each image has a different recognition difficulty due to its characteristics. Therefore, in the ideal case, choosing an independently scaled MLP-Mixer for each image is the most economical computational approach. Hence, this paper experimentally verifies the objective existence of this phenomenon, which inspires us to propose the Multi-Scale MLP-Mixer (MSMLP) that utilizes a suitably scaled MLP-Mixer for each input image. MSMLP comprises several MLP-Mixers of different scales. During testing, these MLP-Mixers are activated in order of scale from large to small (increasing number of patches and decreasing patch size). In addition, to reduce redundant computations, a feature reuse mechanism is designed between neighboring MLP-Mixers so that the small-scale MLP-Mixer downstream can reuse the features learned by the larger-scale MLP-Mixer upstream. Finally, extensive experiments on the public dataset CIFAR10/100 reveal that our method’s theoretically estimated computational cost and actual inference speed are significantly higher than those of MLP-Mixer.

Context Data Fusion Model Enlightened Multi-Scale Capsule Network for Fruit Diseases Identification

Pests and diseases of tomato fruits are the foremost factors leading to the reduction in produce quantity. Effective and prompt differentiation of these lesion symptoms is essential for agricultural production. Due to the lack of a global receptive field, commonly used convolution neural network models restrict the capability on tomato images under complicated backgrounds. In this paper, we present an efficient and robust method for field diagnosis of tomato fruits based on the context data fusion and capsule network. First, in order to make better the point of inadequate images, the Cycle-GAN algorithm is used to realize the augmentation of the diseases dataset. To improve the quality of the generated images, the feature reconstruction loss function is proposed for the Cycle-GAN method, named Fruit-GAN. The proposed Fruit-GAN learns features of healthy and diseased tomatoes, respectively, and forms lesions based on healthy tomato fruit images. Adding lesion tomato images produced on Fruit-GAN to training sets improves the diagnostic capability compared to classical data augmentation techniques. Subsequently, context fusion studies were used to construct intermediate feature maps from different layers into the fruit diseases recognition compact representation to efficiently leverage the feature reuse mechanism. Finally, the fusion model’s enlightened multi-scale capsule network is integrated to form deep features and determine the disease diagnosis. After analyzing the hyper-parameters, the integral technique provides 96.16% accuracy. This research of tomato diseases diagnosis demonstrates the context capsule network model is efficient and establishes a foundation for tomato fruits determination in natural growing environment.

Method of increasing the reliability of knowledge-oriented systems software through code reuse mechanisms

The issues of possibility of application of the code reuse mechanism in the process of designing and developing software of knowledge-oriented systems have been considered. The necessity of an individual research on the development of software for analytical and information resources that provide support to all participants in the process of knowledge-oriented systems software development has been proved. The necessity of using the mechanisms of reuse of previously developed and verified program code as an element of the information resource has been shown. In frame of the developed method of improving the reliability of knowledge-oriented systems software through the of code reuse mechanisms, a new visual form of function libraries representation in the shape of a single software shell has been proposed. As a basic framework for solving the problem of information resource development, usage of Unified Modeling Language (UML), which is based on the paradigm of object-oriented programming has been proposed. Mentioned choice is justified by the fact that UML is an integral part of a unified software development process and is essentially an open standard that uses graphical notation to create an abstract model of the system. The obtained theoretical provisions are imaged into a running example, which shows one of the possible options for organizing libraries of functions as an element of the information resource. The use case diagrams, interaction diagrams, sequence diagrams, class diagrams have been developed and justified. Based on the obtained practical results, a block diagram of the method, which, in contrast to existing solutions, includes a procedure for generating of proofreads for basic UML-diagrams in accordance with requirements of the programming environments has been proposed.

Request, Coalesce, Serve, and Forget: Miss-Optimized Memory Systems for Bandwidth-Bound Cache-Unfriendly Applications on FPGAs

Applications such as large-scale sparse linear algebra and graph analytics are challenging to accelerate on FPGAs due to the short irregular memory accesses, resulting in low cache hit rates. Nonblocking caches reduce the bandwidth required by misses by requesting each cache line only once, even when there are multiple misses corresponding to it. However, such reuse mechanism is traditionally implemented using an associative lookup. This limits the number of misses that are considered for reuse to a few tens, at most. In this article, we present an efficient pipeline that can process and store thousands of outstanding misses in cuckoo hash tables in on-chip SRAM with minimal stalls. This brings the same bandwidth advantage as a larger cache for a fraction of the area budget, because outstanding misses do not need a data array, which can significantly speed up irregular memory-bound latency-insensitive applications. In addition, we extend nonblocking caches to generate variable-length bursts to memory, which increases the bandwidth delivered by DRAMs and their controllers. The resulting miss-optimized memory system provides up to 25% speedup with 24× area reduction on 15 large sparse matrix-vector multiplication benchmarks evaluated on an embedded and a datacenter FPGA system.

AxP: A HW-SW Co-Design Pipeline for Energy-Efficient Approximated ConvNets via Associative Matching

The reduction in energy consumption is key for deep neural networks (DNNs) to ensure usability and reliability, whether they are deployed on low-power end-nodes with limited resources or high-performance platforms that serve large pools of users. Leveraging the over-parametrization shown by many DNN models, convolutional neural networks (ConvNets) in particular, energy efficiency can be improved substantially preserving the model accuracy. The solution proposed in this work exploits the intrinsic redundancy of ConvNets to maximize the reuse of partial arithmetic results during the inference stages. Specifically, the weight-set of a given ConvNet is discretized through a clustering procedure such that the largest possible number of inner multiplications fall into predefined bins; this allows an off-line computation of the most frequent results, which in turn can be stored locally and retrieved when needed during the forward pass. Such a reuse mechanism leads to remarkable energy savings with the aid of a custom processing element (PE) that integrates an associative memory with a standard floating-point unit (FPU). Moreover, the adoption of an approximate associative rule based on a partial bit-match increases the hit rate over the pre-computed results, maximizing the energy reduction even further. Results collected on a set of ConvNets trained for computer vision and speech processing tasks reveal that the proposed associative-based hw-sw co-design achieves up to 77% in energy savings with less than 1% in accuracy loss.

Hyperspectral image classification with dual attention dense residual network

ABSTRACT Recent years, convolutional neural networks (CNNs) have attracted broad attention in hyperspectral image (HSI) classification. Most CNN-based HSI classification methods use image patches as the inputs of network. As the size increases, input patches may contain inhomogeneous pixels, which will cause damage to the classification results. In general, it needs to develop hierarchical architectures for CNNs to learn high-level feature representation. However, excessively deep architectures will cause overfitting and gradient vanishing, thus degrades the generalization performance of CNN. To address the above issues, a dual attention dense residual network (DADRN) is proposed for HSI classification. First, a dual attention module (DAM) is designed to reduce the damage of inhomogeneous pixels. It consists of a spectral attention unit and a spatial attention unit, and can achieve adaptive feature refinement. Second, a dense residual subnetwork (DRN) composed of dense convolutional blocks (DCBs) is proposed to extract more discriminative features from the output feature maps of the DAM. Specifically, the DCB achieves feature reuse mechanism through dense connection and channel concatenation operation, and the DRN adopts dense residual connections to alleviate overfitting and gradient vanishing. Experimental results on three benchmark HSI datasets demonstrate that the competitive advantage of proposed method over several state-of-the-art classification methods.

Compressive strength of concrete with partial replacement of Fine Aggregate with crushed Electronic waste plastic

Electronic waste (E-waste) of about 53.6 MT is generated globally with 7.3 kg per capita, wherein India generates about 1.72 MT. Out of the total generated e-waste, only 17.4% is recycled properly. The e-waste constitutes both organic and inorganic compounds while handling. Plastic waste is one of the inorganic compounds of e-waste, which has majorly become a massive problem owing to its non-degradable property. Moreover, a very less percentage of this is being reused/recycled every year. This also has many negative impacts on the environment so there is a necessity to study its uses alternately. Therefore, it is preliminarily desired to attempt in replacing the fine aggregate to certain percentage viz 0%, 10%, 15%, 20% by using available plastic components of e-waste available from circuit boards, body cabinets of a computer after crushing. In this paper, an attempt has been made to observe the effects of partially replacing fine aggregate with e-waste (plastic). The impact of alternative material to fine aggregate in material compositions is determined by testing the physical (Density) & strength (Compressive and Flexural) properties of concrete. Testing for these impacts helps in understanding the extent up to which these wastes can be used in concrete. Thereby, the usage of natural materials in construction practices can be reduced. Finally, it is inferred that the usage of e-waste plastic particles as a part of the reuse mechanism helps in improving sustainability due to less usage of sand.

Learning from distribution-changing data streams via decision tree model reuse

In many real-world applications, data are collected in the form of streams. As a result of the evolving nature of dynamic environments, the distribution of data streams generally changes over time. Such distribution changes hinder the application of conventional machine learning approaches because the fundamental assumption of independent and identical distribution does not hold in these scenarios. This paper proposes an algorithm based on the decision tree model reuse mechanism for learning from distribution-changing data streams. The proposed algorithm is essentially an online ensemble method that maintains a model pool and updates it by performing decision tree model reuse. The main idea is to exploit the useful knowledge in historical data to help resist the negative effects of distribution changes. We validate the effectiveness of the proposed approach through experiments on synthetic and real-world datasets.

Semantics of multiway dataflow constraint systems

Multiway dataflow constraint systems (MDCS) is a programming model where statements are not executed in a predetermined order. Rather, individual methods are selected from specific method sets and then executed to achieve a desired global state. The selection is done by a planner, which typically bases the choice of methods on the history of updates to the global state. MDCS is well suited for describing user interface logic where choosing what code to execute depends in complicated ways on the history of user interactions and on data availability. User interfaces are the domain of examples in this paper.Much of the research into MDCS has been on planning algorithms and their efficiency. Here we investigate a semantic setting for MDCS, introducing dataflow constraints as modules with explicit goals and related method sets. MDCS is defined in a similar manner, with an explicit goal and a set of supporting dataflow constraints. This enables verification and testing of methods and dataflow constraints against the goals. The exposition is based on abstract syntax for an idealised programming language with global variables. On top of this we define a modular reuse mechanism for dataflow constraints based on Goguen-Burstall institution theory. We show how this setup enables reuse in user interfaces; traditionally code that defines user interface logic is almost invariably non-reusable.

Image super-resolution based on dense convolutional auto-encoder blocks

Deep convolutional neural networks (DCNNs) have recently boosted the performance of image super-resolution (SR) by learning deep non-linear mappings from low-resolution images to their high-resolution counterparts. In general, these methods learn the mapping relationship in image space of a single scale. In this paper, we consider that features across different scales can provide various types of information for SR. Thus, we propose a novel network that extracts features of different spatial resolutions for image super-resolution. We successfully build a model to learn non-linear mappings across feature spaces of various spatial resolutions. Specifically, we propose a dense convolutional auto-encoder block (DCAE), which includes several auto-encoder (AE) units and a squeeze unit, as the basic component of our model. The AE units exploit features of different resolutions through paired encoding and decoding layers. Further, we employ skip connections to combine features of the same spatial scale in one AE unit and dense connections across successive AE units in one DCAE block to establish a temporal feature reuse mechanism. The squeeze units combine features in a DCAE block and the previous DCAE block to achieve long-term temporal feature reuse. Furthermore, we extend our work by performing multi-scale supervised training to build a single framework for SR of all scale factors. Comprehensive experiments show that the proposed method outperforms state-of-the-art methods.

Green fault detection scheme for IoT‐enabled wireless sensor networks

SummaryWireless sensor network (WSN) is one of the major components for the Internet of Things (IoT)‐based smart systems that accumulate meaningful information and send them to the end‐user. In IoT‐enabled WSNs, large numbers of sensor nodes are deployed in an on‐site environment to collect and transmit different physical parameters from the monitoring environments to a central repository. However, low‐cost sensing devices in IoT‐enabled WSNs are prone to fail due to energy depletion, software failures, and hardware failures. The existing fault detection approaches consume notable additional energy to detect and overcome the failures. As a result, the whole network is subject to premature death. This article proposes a green fault detection scheme for IoT‐enabled WSNs to detect faulty nodes and solve the premature death problem of the network. Furthermore, a faulty node reuse mechanism is proposed that significantly prolongs the network lifetime. We analyze the properties of the proposed algorithm mathematically and validate its performance through extensive simulation and real‐life experiments. The simulation and experimental results show improved performance of the proposed scheme in terms of fault detection accuracy, false alarm rate, and network lifetime.

Conditions required for multiple reuses of solvents in indirect carbonation

Abstract Reuse of solvents is essential to making indirect carbonation technology economical and environmentally friendly. Despite its importance, however, no studies have been conducted to confirm the feasibility of reuse of solvents and to determine the conditions for the reuse. This study investigated the optimal conditions for critical factors affecting the reuse of solvents in indirect carbonation and the causes that make it difficult to reuse solvents. For the study, three chelating agents (citrate, malonate, and adipate) and cement kiln dust (CKD) were adopted as solvents and raw material. The end-of-carbonation pH of 10.5 and the concentration of 0.1 M were found to be the optimal conditions necessary for the reuse of chelating agents. Under the optimal conditions, the repeated use of chelating agents as solvents could not only result in high CO2 storage and high yield and purity of CaCO3, but also maintain them the same. Under non-optimal conditions, however, the concentration of bicarbonate ions in the remaining solution after carbonation increased significantly, causing the reuse of solvents to become very difficult. We provided strong evidence to support the reuse mechanism of the chelating agent through FT-IR analysis. We believe that our research on the reuse of solvents will contribute much to the commercialization of indirect carbonation technology.

ICedge: When Edge Computing Meets Information-Centric Networking

In today's era of explosion of Internet of Things (IoT) and end-user devices and their data volume emanating at the network's edge, the network should be more in-tune with meeting the needs of these demanding edge computing applications. To this end, we design and prototype Information-Centric edge (ICedge), a general-purpose networking framework that streamlines service invocation and improves the reuse of redundant computation at the edge. ICedge runs on top of named-data networking, a realization of the information-centric networking vision, and handles the “low-level” network communication on behalf of applications. ICedge features a fully distributed design that: 1) enables users to get seamlessly on-boarded onto an edge network; 2) delivers application invoked tasks to edge nodes for execution in a timely manner; and 3) offers naming abstractions and network-based mechanisms to enable (partial or full) reuse of the results of already executed tasks among users, which we call “compute reuse,” resulting in lower task completion times and efficient use of edge computing resources. Our simulation and testbed deployment results demonstrate that ICedge can achieve up to $50\times $ lower task completion times leveraging its network-based compute reuse mechanism compared to cases, where reuse is not available.

Efficient Support of AXI4 Transaction Ordering Requirements in Many-Core Architecture

The Advanced Microcontroller Bus Architecture (AMBA) Advanced eXtensible Interface 4 (AXI4) protocol was initially a bus oriented interface designed for on-chip communication. To offer the possibility of utilizing the AXI4 based processors and peripherals in the on-chip network based system, we propose a whole system architecture solution to make the AXI4 protocol compatible with the Network-on-Chip (NoC) based communication interconnect in the many-core architecture. Due to the out-of-order transaction in the NoC interconnect, which conflicts with the ordering requirements specified by the AXI4 protocol, we especially focus on the design of the transaction ordering units, realizing a high-performance and low cost (area) solution to the ordering requirements by the sequence ID (seq_ID) reuse mechanism and a simple but smart seq_ID synchronization process. Besides, the micro-architectures and the functionalities of the transaction ordering units are described and explained in detail for ease of implementation. The experimental results in a C++ based system simulator show that, compared with the state-of-the-art works, our solution can maximally increase the system throughput by 66.0% and decrease the transaction queueing delay in the master-side ordering unit by 91.2%.