Search

Energy consumption of embedded applications has rapidly increased with the advancement of technology and computing. There is a little improvement in energy consumption as compared to computing and storage capacity. Although computing performance has been continuously increasing, power/energy consumption is more critical in the design of real-time embedded systems. Real-time embedded applications need a power management technique to judicially balance the energy consumption and computing performance. It should be done in such a way that the system performance improves along with an increase in the lifespan of the system. The proposed methodology presented in this paper deals with the minimization of energy for time-critical embedded applications. Simulation studies, along with theoretical analysis, have been carried out to show the effectiveness of the proposed three-phase reliable energy-aware scheduling method. It is observed that the proposed approach provides better tolerance (approximately four times) and consumes less energy (35% to 45%) for a wide range of applications.

With the rise of Big Data, there has been a significant effort in increasing compute power through GPUs, TPUs, and heterogeneous architectures. As a result, many applications are memory bound, i.e., they are bottlenecked by the movement of data from main memory to compute units. One way to address this issue is through data prefetching, which relies on accurate prediction of memory accesses. While recent deep learning models have performed well on sequence prediction problems, they are far too heavy in terms of model size and inference latency to be practical for data prefetching. Here, we propose clustering-driven compact LSTM models that can predict the next memory access with high accuracy. We introduce a novel clustering approach called Delegated model that can reliably cluster the applications. For each cluster, we train a compact meta-LSTM model that can quickly adapt to any application in the cluster. Prior LSTM-based work on access prediction has used orders of magnitude more parameters and developed one model for each application (trace). While one (specialized) model per application can result in more accuracy, it is not a scalable approach. In contrast, our models can predict for a class of applications by trading off specialization at the cost of few retraining steps at runtime, for a more generalizable compact meta-model. Our experiments on 13 benchmark applications demonstrate that clustering-driven ensemble compact meta-models can obtain accuracy close to specialized models using few batches of retraining for majority of the applications.

Graph Neural Networks (GNNs) are powerful deep learning models to generate node embeddings on graphs. When applying deep GNNs on large graphs, it is still challenging to perform training in an efficient and scalable way. We propose a novel parallel training framework. Through sampling small subgraphs as minibatches, we reduce training workload by orders of magnitude compared with state-of-the-art minibatch methods. We then parallelize the key computation steps on tightly-coupled shared memory systems. For graph sampling, we exploit parallelism within and across sampler instances, and propose an efficient data structure supporting concurrent accesses from samplers. The parallel sampler theoretically achieves near-linear speedup with respect to number of processing units. For feature propagation within subgraphs, we improve cache utilization and reduce DRAM traffic by data partitioning. Our partitioning is a 2-approximation strategy for minimizing the communication cost compared to the optimal. We further develop a runtime scheduler to reorder the training operations and adjust the minibatch subgraphs to improve parallel performance. Finally, we generalize the above parallelization strategies to support multiple types of GNN models and graph samplers. The proposed training outperforms the state-of-the-art in scalability, efficiency and accuracy simultaneously. On a 40-core Xeon platform, we achieve 60x speedup (with AVX) in the sampling step and 20x speedup in the feature propagation step, compared to the serial implementation. Our algorithm enables fast training of deeper GNNs, as demonstrated by orders of magnitude speedup compared to the Tensorflow implementation. We open-source our code at https://github.com/GraphSAINT/GraphSAINT., 43 pages, 8 figures. arXiv admin note: text overlap with arXiv:1810.11899

Collective behavior appearing abundantly in nature at various scales, such as swarming of insects, flocking of birds or schooling of fish, has motivated an interdisciplinary research thrust towards artificial swarms thanks to its numerous advantages. Although a lot of work has been done on coming up with controllers for these multi-agent systems towards optimizing coordination and planning, the inverse problem of identifying behaviors and models from observing a swarm is still not explored enough. Efficient tools for the analysis of the complex spatial-temporal dataset an observed swarm generates, are still missing. In this paper, we propose a methodology to solve the problem of identifying the leader in a swarm governed by leader-follower control framework using only the macroscopic view of the entire swarm. A methodology that combines clustering and probability methods is used to analyze the spatial-temporal data of a swarm of robots to narrow down on a subset of the robots that includes the leader. The clustering parameters are optimized over multiple simulated behaviors. The results show that this automated system can narrow down on a subset of agents (cluster) that includes the leader of the swarm with high accuracy. Applications of the proposed methodology include automatic identification of leader and swarm dynamics in both artificial and biological swarms, which can lead to a better understanding of collective behaviors and predictions of future behaviors.

In this pandemic era, businesses are highly affected due to low availability of manpower. In this situation, IoT is making a significant contribution to managing and controlling human activities in an industrial perspective and daily life. In the era of machine to machine communication, IoT is contributing to the interaction of machine and humans in a more innovative way. This chapter focuses on providing the entrepreneurial approach to manage the IoT devices used in the different sectors like health, agriculture, home appliances, research and communication as an intelligent machine. It will also focus on the impact and support of cloud computing in the implementation of IoT in the above-mentioned areas. This chapter elaborates the key feature of cloud computing services used to overcome the limitation of IoT resources in different implementation domains. This chapter also focuses on the major contribution of cloud computing on handling the IoT and making it more intelligent to encourage innovation and research in the IoT context.

Effective social robots should leverage humor’s unique ability to improve relationship connections and dispel stress, but current robots possess limited (if any) humorous abilities. In this paper, we aim to supplement one aspect of autonomous robots by giving robotic systems the ability to "read the room" to assess how their humorous statements are received by nearby people in real time. Using a dataset of the audio of crowd responses to a robotic comedian over multiple performances (first presented in past work), we establish human-labeled joke success ground truths and compare individual human rater accuracy against the outputs of lightweight Machine Learning (ML) approaches that are easy to deploy in real-time joke assessment. Our results indicate that all three ML approaches (naive Bayes, support vector machines, and single-hidden-layer feedforward neural networks) performed significantly better than the baseline approach used in our past work. In particular, support vector machines and neural network approaches are comparable to a human rater in the task of assessing if a joke failed or not in certain cases. The products of this work will inform self-assessment techniques for robots and help social robotics researchers test their own assessment methods on realistic data from human crowds.

Energy storage systems (ESS) are effective solutions to reduce the cost of smart grid operations due to their ability to store and supply electricity on demand. Traditionally, ESS have been used to implement a plethora of cost reduction techniques for smart grid operations such as temporal supply-demand shifting, frequency regulation, voltage regulation etc. However, the stationary nature of ESS limits their flexibility, potentially leading to low utilization and risk being a stranded asset. In this work, we explore the use of Mobile ESS in implementing such cost reduction techniques in a grid consisting of multiple microgrids. We develop an algorithmic framework to assign multiple Mobile ESS to various microgrids of the smart grid across several days. Our framework attempts to maximize the cost reduction achieved due to Mobile ESS assignment minus the routing cost. We show that our algorithm is a polynomial time $1/e-$1/e-approximation for the NP-Hard problem of optimal assignment of MESS.

Short-term probabilistic forecasts of the trajectory of the COVID-19 pandemic in the United States have served as a visible and important communication channel between the scientific modeling community and both the general public and decision-makers. Forecasting models provide specific, quantitative, and evaluable predictions that inform short-term decisions such as healthcare staffing needs, school closures, and allocation of medical supplies. Starting in April 2020, the US COVID-19 Forecast Hub (https://covid19forecasthub.org/) collected, disseminated, and synthesized tens of millions of specific predictions from more than 90 different academic, industry, and independent research groups. A multi-model ensemble forecast that combined predictions from dozens of different research groups every week provided the most consistently accurate probabilistic forecasts of incident deaths due to COVID-19 at the state and national level from April 2020 through October 2021. The performance of 27 individual models that submitted complete forecasts of COVID-19 deaths consistently throughout this year showed high variability in forecast skill across time, geospatial units, and forecast horizons. Two-thirds of the models evaluated showed better accuracy than a naïve baseline model. Forecast accuracy degraded as models made predictions further into the future, with probabilistic error at a 20-week horizon 3-5 times larger than when predicting at a 1-week horizon. This project underscores the role that collaboration and active coordination between governmental public health agencies, academic modeling teams, and industry partners can play in developing modern modeling capabilities to support local, state, and federal response to outbreaks. Significance Statement This paper compares the probabilistic accuracy of short-term forecasts of reported deaths due to COVID-19 during the first year and a half of the pandemic in the US. Results show high variation in accuracy between and within stand-alone models, and more consistent accuracy from an ensemble model that combined forecasts from all eligible models. This demonstrates that an ensemble model provided a reliable and comparatively accurate means of forecasting deaths during the COVID-19 pandemic that exceeded the performance of all of the models that contributed to it. This work strengthens the evidence base for synthesizing multiple models to support public health action.

In today’s world, as the population increases so do diseases and more and more people are becoming dependent on medicines for their living. With medicine being of such great importance, it is necessary for one to take his doses on time. But older people and virus-infected people find it difficult to remember their medicine schedule. To overcome this problem, this proposed work “Smart Medicine Dispenser” stores their daily medicine schedule and notifies them on the basis of the schedule about when and which medicine to take. Using Wi-Fi, it can also notify the caretaker of the patient about whether the patient has taken his medicine or not. It comprises a box containing several small sections in which we can store medicines according to their daily dose. It uses a real-time clock (RTC) module and an alarm system to make sure that the patient takes his dose on time. The experimental results show that the proposed model gives better results in comparison with existing work.

The Internet of Things (IoT) is a system of interrelated computing devices, mechanical and digital machines provided with unique identifiers and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction. Nowadays, IoT plays a very important role in the field of the corporate sector, to make smart devices like smart watches, smart houses, and smart city and security purposes. One of the application of smart city which have smart light, smart parking system, smart garbage management system the author will propose in this research paper. Here, IoT works with technology like big data, cloud, and wireless sensors. The combination of these techniques leads to making a smart city. In this proposed model, the smart light will be used for energy consumption, smart parking to monitor the parking traffic, smart garbage management where we reduce the contaminated substances. The author proposes a real-time model that will enhance efficiency.

This paper explores the concept of web animations with the motive to generate exportable video files using a browser. We explain the process underlying animations such as tweening and controlling individual states of an object's animation cycle. We can composite many animations to follow a single timeline and capture individual frames at periodic intervals. Such frames, when stitched together, form a video. We will also explain the integration of an audio layer than runs in synchronization with the animations for better perception.

As HPC progresses toward exascale, writing applications that are highly efficient, portable, and support programmer productivity is becoming more challenging than ever. The growing scale, diversity, and heterogeneity in compute platforms increases the burden on software to efficiently use available distributed parallel resources. This burden has fallen on developers who, increasingly, are experts in application domains rather than traditional computer scientists and engineers. We propose CASPER-Compiler Abstractions Supporting high Performance on Extreme-scale Resources-a novel domain-specific compiler and runtime framework to enable domain scientists to achieve high performance and scalability on complex HPC systems. CASPER extends domain-specific languages with machine learning to map software tasks to distributed, heterogeneous resources, and provides a runtime framework to support a variety of adaptive optimizations in dynamic environments. This paper presents an initial design and analysis of CASPER for synthetic aperture radar and computational fluid dynamics domains.

The rapid development of Big Data coupled with slowing down of Moore’s law has made the memory performance a bottleneck in the von Neumann architecture. Machine learning has the potential to provide opportunities to address the memory performance issues, specifically through data access prediction. While recent works focusing on the prediction of memory accesses have used recurrent neural networks (RNN), there is a lack of a framework utilizing such prediction in a prefetcher. This paper introduces the RNN Augmented Offset Prefetcher (RAOP) framework, which consists of two parts: an RNN-based predictor and an offset prefetching module. By leveraging the RNN predicted access as a temporal reference, RAOP improves prefetching performance by executing offset prefetching for both the current address and the RNN predicted address. We implement the RNN in the predictor with a compressed long short-term memory (LSTM) model and demonstrate the effect of augmenting an RNN predictor to a simple next-line prefetcher in the prefetching module results in 3.22x, 4.2x, and 15.6% improvement in prefetch accuracy, coverage, and speedup. We further implement a best-offset prefetcher (BOP) in RAOP and compare it to several state-of-the-art prefetchers. Results show that RAOP achieves a mean 4.05% speedup by prefetching in last level cache, outperforming state-of-the-art prefetchers. By augmenting an RNN predictor to BOP, RAOP results in 6.5x, 9.2x, and 12.8% improvement in prefetch accuracy, coverage, and speedup.

While heterogeneous systems provide considerable opportunities for accelerating big data applications, the variation in processing capacities and communication latency of different resources makes it challenging to effectively map the applications on the platform. To generate an optimized mapping of the input application on a variety of heterogeneous platforms, we design a flexible annotated task interaction graph based framework which 1) allows modeling of mixed CPU and GPU architectures, and 2) identifies an efficient task-hardware mapping of the input application, given the dependencies and communication costs between tasks that constitute the applications. The annotated task interaction graph (ATIG) representation captures all the information that is necessary to execute the application and the meta-data, such as performance models for estimating runtime on a target resource and communication latencies. Our framework supports solving the problem of mapping tasks in the ATIG onto available resources by including variations of greedy algorithm and LP relaxations with rounding. We show that our framework can achieve high speedup, allowing domain experts to efficiently compile a broad set of programs to parallel and heterogeneous hardware.

Writing high-performance code requires significant expertise in the programming language, compiler optimizations, and hardware knowledge. This often leads to poor productivity and portability and is inconvenient for a non-programmer domain-specialist such as a Physicist. More desirable is a high-level language where the domain-specialist simply specifies the workload in terms of high-level operations (e.g., matrix-multiply(A, B)), and the compiler identifies the best implementation fully utilizing the heterogeneous platform. For creating a compiler that supports productivity, portability, and performance simultaneously, it is crucial to predict the performance of various available implementations (variants) of the dominant operations (kernels) contained in the workload on various hardware to decide (a) which variant should be chosen for each kernel in the workload, and (b) on which hardware resource the variant should run. To enable the performance prediction, we propose lightweight augmented neural networks for arbitrary combinations of kernel-variant-hardware. A key innovation is utilizing the mathematical complexity of the kernels as a feature to achieve higher accuracy. These models are compact to reduce training time and allow fast inference during compile-time and run-time. Using models with less than 75 parameters, and only 250 training data instances, we are able to obtain accurate performance predictions, significantly outperforming traditional feed-forward neural networks on 48 kernel-variant-hardware combinations. We further demonstrate that our variant-selection approach can be used in Halide implementations to obtain up to 1.7x speedup over Halide auto-scheduler.

Spectral-domain CNNs have been shown to be more efficient than traditional spatial CNNs in terms of reducing computation complexity. However they come with a 'kernel explosion' problem that, even after compression (pruning), imposes a high memory burden and off-chip bandwidth requirement for kernel access. This creates a performance gap between the potential acceleration offered by compression and actual FPGA implementation performance, especially for low-latency CNN inference. In this paper, we develop a principled approach to overcoming this performance gap and designing a low-latency, low-bandwidth, spectral sparse CNN accelerator on FPGAs. First, we analyze the bandwidth-storage tradeoff of sparse convolutional layers and locate communication bottlenecks. We then develop a dataflow for flexibly optimizing data reuse in different layers to minimize off-chip communication. Finally, we propose a novel scheduling algorithm to optimally schedule the on-chip memory access of multiple sparse kernels and minimize read conflicts. On a state-of-the-art FPGA platform, our design reduces data transfers by 42% with DSP utilization up to 90% and achieves inference latency of 9 ms for VGG16, compared to the baseline state-of-the-art latency of 68 ms.

Q-Table based Reinforcement Learning (QRL) is a class of widely used algorithms in AI that work by successively improving the estimates of Q-values - quality of state-action pairs, stored in a table. They significantly outperform Neural Network based techniques when the state space is tractable. Fast learning for AI applications in several domains (such as robotics), with tractable `mid-sized' Q-tables, still necessitates performing a large number of rapid updates. State-of-the-art FPGA implementations of QRL do not scale well with the increasing Q-Table state space. Thus, they are not efficient for such applications. In this work, we develop a novel FPGA based design of QRL and SARSA (State Action Reward State Action), scalable to large state spaces and thereby facilitating a large class of AI applications. Our architecture provides higher throughput while using significantly fewer on-chip resources. It is capable of supporting a variety of action selection policies that covers Q-Learning and variations of bandit algorithms and can be easily extended for multi-agent Q learning. Our pipelined implementation fully handles the dependencies between consecutive updates allowing it to process one sample every clock cycle. We evaluate our architecture for Q-Learning and SARSA algorithms and show that our designs achieve a high throughput of up to 180 million samples per second.

Smart education is the constituent of smart cities. Smart education is the use of computers in the classroom. However, there are many other factors to improve a child’s quality of education. One part is the amount of time the child spends on a bus from traveling bus to home. Although, with currently available technologies as the Internet of Things (IOT) and Android, with these advanced technologies we will be able to track or visibility in a child’s life. This paper presents the design and implementation of an IOT based system that permits parents, schools and regulatory bodies to track the conditions of comfort and safety inside a bus in a real-time manner. A variety of reports for various activities for bus comfort and attendance are then generated from this real-time providing complete visibility of buses.

nowadays development of IoT applications with robotics is an ongoing revaluation. This paper mainly focuses on the security, remote surveillance, and monitoring of our homes done by the surveillance robots. Remote surveillance has become the most important research topic over the past decade. Through this paper we put forward a surveillance robot that can be used in domestic areas and many other places. Robots are becoming important in our day to day life activities as they reduce the human labor and probability of error. We can control robots manually or they can be automatic based on the need of people. This paper focuses on design and implementation of mobile robot for obstacle detection and avoidance in a real-time basis.

In this paper, we present an IoT-based solution that can reduce the complexity of crowd estimation. About the human crowd estimation many techniques are in existence but nowadays more work is going on in this field because this is era of IoT and most of the organization is shifted toward IoT-based system. So in our proposed system we are using the Raspberry Pi-3 which are having quad-core processor that can be very useful and gives better result and accurate number even when the humans are very close to each other. This IoT-based model can easily be implemented in crowded areas and monitor the same. The camera module in this model also helps to differentiate between human and other bodies. As this is a mobile model, it can be easily fixed on the walls of street light and in the time of darkness or in night the camera captures clear images for process in the presence of street light. So that this model gives better result almost 70% better result in compare to exiting approaches.

The least squares based cost functions are sensitive to outliers in the measured data. The presence of outliers is considered as impulsive noise. In practical scenarios, the co channel interference, saturation effects, non linearity of the measuring instruments, atmospheric conditions and malfunction of sensors will result in outliers or impulsive noise. The robust function obtained by considering the error as the linear combination of sign preserving basis functions is found to be robust against outliers in the desired data. In many practical applications, the parameter to be estimated can be sparse in nature, i.e. only a few elements are large values and the rest are insignificantly small. In such sparse systems, if the prior information about the sparsity is known, then the known information can be incorporated in the cost function as a regularization function. A robust sparse diffusion algorithm is proposed in this work, which is robust against outliers in the desired data and performs better than the existing algorithms in sparsity underlying systems. Simulations performed for different cases of outliers conditions and sparsity conditions validate that the proposed method outperforms the state of the art methods.

Hash table is a fundamental data structure that provides efficient data store and access. It is a key component in AI applications which rely on building a model of the environment using observations and performing lookups on the model for newer observations. In this work, we develop FASTHash, a “truly” high throughput parallel hash table implementation using FPGA on-chip SRAM. Contrary to state-of-the-art hash table implementations on CPU, GPU, and FPGA, the parallelism in our design is data independent, allowing us to support p parallel queries (\(p>1\)) per clock cycle via p processing engines (PEs) in the worst case. Our novel data organization and query flow techniques allow full utilization of abundant low latency on-chip SRAM and enable conflict free concurrent insertions. Our hash table ensures relaxed eventual consistency - inserts from a PE are visible to all PEs with some latency. We provide theoretical worst case bound on the number of erroneous queries (true negative search, duplicate inserts) due to relaxed eventual consistency. We customize our design to implement both static and dynamic hash tables on state-of-the-art FPGA devices. Our implementations are scalable to 16 PEs and support throughput as high as 5360 million operations per second with PEs running at 335 MHz for static hashing and 4480 million operations per second with PEs running at 280 MHz for dynamic hashing. They outperform state-of-the-art implementations by 5.7x and 8.7x respectively.

With the rise of Big Data, there has been a significant effort in increasing compute power through GPUs, TPUs, and heterogeneous architectures. As a result, many applications are memory bound, i.e., they are bottlenecked by the movement of data from main memory to compute units. One way to address this issue is through data prefetching, which relies on accurate prediction of memory accesses. While recent deep learning models have performed well on sequence prediction problems, they are far too heavy in terms of model size and inference latency to be practical for data prefetching. Here, we propose extremely compact LSTM models that can predict the next memory access with high accuracy. Prior LSTM based work on access prediction has used orders of magnitude more parameters and developed one model for each application (trace). While one (specialized) model per application can result in more accuracy, it is not a scalable approach. In contrast, our models can predict for a class of applications by trading off specialization at the cost of few retraining steps at runtime, for a more generalizable compact meta-model. Our experiments on 13 benchmark applications demonstrate that three compact meta-models can obtain accuracy close to specialized models using few batches of retraining for majority of the applications.

Dengue virus is an important prevalent mosquito-borne, single positive-stranded RNA virus of the family Flaviviridae (Flavivirus), which is transmitted by infected Aedes female mosquitoes and causes dengue fever. The challenge to control the transmission of these diseases and prompt sight of this virus is an international health problem at present. As a result of the scientific significance of its recognition, recently, the use of biosensors to sense and measure the disease with quick detection is one of the best methods, which is comparatively simple and practical, with high sensibility, accuracy, and selectivity rather than the traditional techniques such as immunology and molecular-based approaches. The electrochemical biosensor-based recognition method from biomolecules using transducers is a simple method, with the possibility of commercialization, inexpensiveness, simplicity, and convenience. Hence, several types of biosensing methods using different biosensors with transducers such as voltammetric, potentiometric, amperometric, and impedimetric are used for the detection of dengue NS1 protein. In this study, we aimed to find out the best biosensing method for the detection and diagnosis of dengue virus. We expect that dengue NS1-based electrochemical diagnostic techniques will open the horizons for growth in the coming times, with thrilling new opportunities for the advancement of commercial design of electrochemical biosensors.

This chapter provides the functional understanding of a novel gene silencing approach in bacteria by CRISPR interference commonly known as CRISPRi. The chapter begins by briefly introducing the CRISPR-Cas system with an emphasis on type II system which is mainly employed in CRISPRi. It then introduces the concept of CRISPRi with detailed description of various critical components and mechanism of gene silencing by this approach. In the subsequent section, major concerns are discussed that need to be considered while implementing this approach in bacteria. The chapter ends with potential application and limitations of the CRISPRi technique in bacteria.

To improve the reliability of content shared on social media, effective strategies for mitigating the diffusion of fake news are increasingly necessary. Traditionally, to counter false belief a competing cascade approach is used. This approach assumes that the opposite belief is already known, and thus, not applicable to newly spreading fake news. Another approach is to block nodes and links of the network to impede the flow of fake news (rumor/influence blocking). However, a more active way to battle the dissemination of fake news is to propagate the corresponding real news, since people who receive the real news in tandem with the fake news are less likely to believe in fake news. Such a setting is especially useful on a messaging platform such as WhatsApp, where the news item flows as a private message and the correction of fake news and its propagation must be performed by the users within the network as they receive it. To achieve this goal, we propose network-based recruitment for active fake news correction (RecANt) to find a set of individuals of a pre-defined size to be incentivized for actively fact-checking and passing on the real news so as to reach the maximum number of nodes in the network. These individuals should be such that they are likely to receive the fake news so that they can test its credibility, and when they propagate the corresponding real news, it reaches a large number of individuals. We prove that RecANt is NP-Hard with a monotone and submodular objective, leading to a polynomial time greedy algorithm (AFC) which provides a (1 – 1/e – ϵ)-approximation. We further optimize the runtime of AFC by developing a fast graph-pruning heuristic (RAFC) that performs as well as AFC in checking the spread of fake news while reducing the runtime significantly. Simulations on several networks demonstrate that our approach outperforms popular social network centrality measures and state-of-the-art information diffusion algorithm.

Accurate prediction of crime incidents can assist the police in better planning of prevention strategies and scheduling deployment. The problem is often studied as a spatio-temporal regression problem approached by dividing the area of interest into a grid of uniform cells, and performing regression on timeseries of each cell. We propose that changing the method of division of the area can significantly improve crime prediction. We demonstrate this using a heterogeneous division of the area obtained by our partitioning algorithm that takes into account the density of crime. We further show that existing measures do not provide a fair comparison of two methods that partition the area in two different ways. To address this severe drawback in crime prediction evaluation, we propose a novel measure which is based on optimal allocation of resources relying on the prediction and then checking the actual number of crimes that would have been avoided by the allocation. Essentially, our measure answers the question of which model would have assisted in preventing most number of actual crimes if allocation were to be done using the predicted crimes. We also prove that a greedy algorithm results in the optimal allocation resources, thus making our evaluation computationally lightweight. Experiments on real-world datasets demonstrate that heterogeneous division of the area results in improved crime prediction while drastically decreasing the number of models to be trained compared to uniform grid division.

Process industrial loads are increasing due to expansion of existing manufacturing units. Increasing of loads will increase the power requirement which is reflected at the distribution level. Feeding the additional loads existing switchgear to be modified and for meeting load requirement additional source of distribution transformer to be added, which it may not be an economical solution. To meet the additional load requirement of industries, two transformers need to be operated in parallel in continuous mode. Bus fault current at LV switchgear is the main constrain with parallel operation of the distribution transformer. Due to increase in fault current, existing switchgears bus bar design capacities may not be adequate, generally in India fault current at LV side switchgear bus bars will be designed for 50kA withstanding capacity. Approach is to reduce the fault current below design fault current with parallel operation of transformers. This approach is mainly safeguard the switchgear equipment during parallel operation of transformers and also avoiding adding additional transformers in distribution network. This paper proposes the method of utilizing series FACTS device Interline Power Flow controller (IPFC), connected in LV side to control the power flow and reduce the fault current on switchgear bus bar during transformer parallel operation. The controller used as reference design fault current method and injected the required voltage in a distribution side during fault condition to achieve desired fault current. The distribution network consider for analysis is one of the Indian 800MW Thermal power plant Water system LV switchgear and analyzed short circuit current on switchgear bus bar with and without IPFC by using ETAP and MATLAB/SIMULINK model and compared the bus short circuit current for different ratings of transformers connected in parallel to LV switchgear.

To accelerate inference of Convolutional Neural Networks (CNNs), various techniques have been proposed to reduce computation redundancy. Converting convolutional layers into frequency domain significantly reduces the computation complexity of the sliding window operations in space domain. On the other hand, weight pruning techniques address the redundancy in model parameters by converting dense convolutional kernels into sparse ones. To obtain high-throughput FPGA implementation, we propose SPEC2 -- the first work to prune and accelerate spectral CNNs. First, we propose a systematic pruning algorithm based on Alternative Direction Method of Multipliers (ADMM). The offline pruning iteratively sets the majority of spectral weights to zero, without using any handcrafted heuristics. Then, we design an optimized pipeline architecture on FPGA that has efficient random access into the sparse kernels and exploits various dimensions of parallelism in convolutional layers. Overall, SPEC2 achieves high inference throughput with extremely low computation complexity and negligible accuracy degradation. We demonstrate SPEC2 by pruning and implementing LeNet and VGG16 on the Xilinx Virtex platform. After pruning 75% of the spectral weights, SPEC2 achieves 0% accuracy loss for LeNet, and, This is a 10-page conference paper in 26TH IEEE International Conference On High Performance Computing, Data, and Analytics (HiPC)

Now a day, most of the electronic devices are battery operated and the performance of these devices is depend upon battery power. For the concern of feasible scheduling in real time operated system, need to manage battery power. The energy consumption in mobile devices is depends upon power management. In this article we proposed an optimal real time scheduling algorithm with constraint of harvesting energy, where every task set having flexible speed to manage the energy consumption. The experimental result shows that this optimal algorithm allows the task to execute on assigned speed that will reduce the overall energy consumption. Proposed algorithm give better management of harvesting energy and also gives 10- 15% better result in term of available energy after execution of task in compare to existing DVFS algorithm.

In Modern days, real time system plays an important role in our modern and digital society. The success of any real time application is totally depending upon the selection of optimal scheduling algorithm. In real time application every task should have the nature of deadlines and time when they arrived, on the basis of these parameters we observe the response time of different scheduling algorithm then we select the optimal algorithm for a particular application. So in this paper our aim is to reduce the complexity of real time system researcher for selection of scheduling algorithm for a particular application. This artificial intelligence based approach is an extent the state of any real time system in the area of scheduling. This approach works in any uni-processor system. Introduction

The world of control is an exciting field that has exploded with new technologies where the Internet of Things vision becomes a reality. Today, IoT with AI is used to make applications such as intelligent transportation, smart security, and smart home. The wireless communication technique is beneficial in IoT nowadays. In this paper, the author proposed a model for android control robot car using wireless technology. Obstacle detection is advantageous in terms of avoiding accidents; the author has been introduced in this paper. This robot car also detects the AQI index of the place. The data is shared on the BLYNK cloud app as well for further decisions.

Runtime-reconfigurable software coupled with reconfigurable hardware is highly desirable as a means towards maximizing runtime efficiency without compromising programmability. Compilers for such software systems are extremely difficult to design as they must leverage different types of hardware at runtime. To address the need for static and dynamic compiler optimization of workflows matched to dynamically reconfigurable hardware, we propose a novel design of the central component of a dynamic software compiler for software defined hardware. Our comprehensive design focuses not just on static knowledge but also on semi-supervised extraction of knowledge from program executions and developing their performance models. Specifically, our novel dynamic and extensible knowledge base 1) continuously gathers knowledge during execution of workflows 2) identifies optimal implementations of workflows on optimal (available) hardware configurations. It plays a hub role in storing information from, and providing information to other components of the compiler, as well as the human analyst. Through a rich tripartite graph representation, the knowledge base captures and learns extensive information on decomposition and mapping of code steps to kernels and mapping of kernels to available hardware configurations. The knowledge base is implemented using the $C++$ Boost Library and is capable of quickly processing offline and online queries and updates. We show that our knowledge base can answer queries in $1 ms$ regardless of the number of workflows it stores. To the best of our knowledge, this is the first design of a dynamic and extensible knowledge base to support compilation of high-level languages to leverage arbitrary reconfigurable platforms.

Graph sampling is an important tool to obtain small and manageable subgraphs from large real-world graphs. Prior research has shown that Induced Edge Sampling (IES) outperforms other sampling methods in terms of the quality of subgraph obtained. Even though fast sampling is crucial for several workflows, there has been little work on parallel sampling algorithms in the past. In this paper, we present parIES - a framework for parallel Induced Edge Sampling on shared-memory parallel machines. parIES, equipped with optimized load balancing and synchronization avoiding strategies, can sample both static and streaming dynamic graphs, while achieving high scalability and parallel efficiency. We develop a lightweight concurrent hash table coupled with a space-efficient dynamic graph data structure to overcome the challenges and memory constraints of sampling streaming dynamic graphs. We evaluate parIES on a 16-core (32 threads) Intel server using 7 large synthetic and real-world networks. From a static graph, parIES can sample a subgraph with > 1.4B edges in

Flow routing over inter-datacenter networks is a well-known problem where the network assigns a path to a newly arriving flow potentially according to the network conditions and the properties of the new flow. An essential system-wide performance metric for a routing algorithm is the flow completion times, which affect the performance of applications running across multiple datacenters. Current static and dynamic routing approaches do not take advantage of flow size information in routing, which is practical in a controlled environment such as inter-datacenter networks that are managed by the datacenter operators. In this paper, we discuss Best Worst-case Routing (BWR), which aims at optimizing the tail completion times of long-running flows over inter-datacenter networks with non-uniform link capacities. Since finding the path with the best worst-case completion time for a new flow is NP-Hard, we investigate two heuristics, BWRH and BWRHF, which use two different upper bounds on the worst-case completion times for routing. We evaluate BWRH and BWRHF against several real WAN topologies and multiple traffic patterns. Although BWRH better models the BWR problem, BWRH and BWRHF show negligible difference across various system-wide performance metrics, while BWRHF being significantly faster. Furthermore, we show that compared to other popular routing heuristics, BWRHF can reduce the mean and tail flow completion times by over $1.5\times$ and $2\times$, respectively., Comment: Accepted for publication in the 2019 57th Annual Allerton Conference on Communication, Control, and Computing (Allerton)

In this digital era, here the sum of data is generate with store has prolonged inside a less period of time. The data in this era generated high speed leads to many challenges. The size is primarily and periodically, just the measurement that bounces in the big data position. In this survey, we have attempt to give a broad explanation of big data that captures its other unique and important features. We have discussed 4V’s model. Also, the latest technologies uses in big data, like Hadoop, HDFS, MapReduce and different type of methods used by big data. Finally various benefits of using big data analysis and include some feature of cloud computing.

To facilitate deep penetration of solar energy in smart grids, we need high observability of solar generation at the edges of the grid. Current advanced metering infrastructures (AMI) only monitor the aggregated measurements from net-metered households, but disaggregated consumption and solar generation components are required for grid optimizations. We propose an unsupervised disaggregation model for disaggregating solar generation from AMI measurements without the need of training data. The model requires only AMI measurements from consumers in a region and the solar irradiance as input, and models the consumption of consumers by neighboring households without rooftop photovoltaics (PV) to perform the disaggregation. We evaluate our results on a real life dataset from Austin, Texas. We show that our model is able to disaggregate consumption and solar generation measurements with 42.24% and 31.67% less mean squared error, respectively, in comparison to a baseline technique that uses supervised learning. This shows that our model is capable of disaggregating historical data even if the dataset has no training data and only contains minimal exogenous data.

Long flows contribute huge volumes of traffic over inter-datacenter WAN. The Flow Completion Time (FCT) is a vital network performance metric that affects the running time of distributed applications and the users' quality of experience. Flow routing techniques based on propagation or queuing latency or instantaneous link utilization are insufficient for minimization of the long flows' FCT. We propose a routing approach that uses the remaining sizes and paths of all ongoing flows to minimize the worst-case completion time of incoming flows assuming no knowledge of future flow arrivals. Our approach can be formulated as an NP-Hard graph optimization problem. We propose BWRH, a heuristic to quickly generate an approximate solution. We evaluate BWRH against several real WAN topologies and two different traffic patterns. We see that BWRH provides solutions with an average optimality gap of less than $0.25\%$. Furthermore, we show that compared to other popular routing heuristics, BWRH reduces the mean and tail FCT by up to $1.46\times$ and $1.53\times$, respectively., Comment: Accepted for publication in IEEE Communications Letters

The Graph Convolutional Network (GCN) model and its variants are powerful graph embedding tools for facilitating classification and clustering on graphs. However, a major challenge is to reduce the complexity of layered GCNs and make them parallelizable and scalable on very large graphs -- state-of the art techniques are unable to achieve scalability without losing accuracy and efficiency. In this paper, we propose novel parallelization techniques for graph sampling-based GCNs that achieve superior scalable performance on very large graphs without compromising accuracy. Specifically, our GCN guarantees work-efficient training and produces order of magnitude savings in computation and communication. To scale GCN training on tightly-coupled shared memory systems, we develop parallelization strategies for the key steps in training: For the graph sampling step, we exploit parallelism within and across multiple sampling instances, and devise an efficient data structure for concurrent accesses that provides theoretical guarantee of near-linear speedup with number of processing units. For the feature propagation step within the sampled graph, we improve cache utilization and reduce DRAM communication by data partitioning. We prove that our partitioning strategy is a 2-approximation for minimizing the communication time compared to the optimal strategy. We demonstrate that our parallel graph embedding outperforms state-of-the-art methods in scalability (with respect to number of processors, graph size and GCN model size), efficiency and accuracy on several large datasets. On a 40-core Xeon platform, our parallel training achieves $64\times$ speedup (with AVX) in the sampling step and $25\times$ speedup in the feature propagation step, compared to the serial implementation, resulting in a net speedup of $21\times$., Comment: 10 pages. 2019 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

Recently, architectures with scratchpad memory are gaining popularity. These architectures consist of low bandwidth, large capacity DRAM and high bandwidth, user addressable small capacity scratchpad. Existing algorithms must be redesigned to take advantage of the high bandwidth while overcoming the constraint on capacity of scratchpad. In this paper, we propose an optimized edge-centric graph processing algorithm for scratchpad based architectures. Our key contribution is significant reduction in (slower) DRAM accesses through intelligent reuse of scratchpad data. We trade off reduction in DRAM accesses for slightly higher scratchpad accesses. However, due to the much higher bandwidth of scratchpad, the total memory access cost (DRAM + scratchpad) is significantly reduced. We validate our analysis with experiments on real world graphs using a simulator which mimics the scratchpad based architecture using Single Source Shortest Path (SSSP) and Breadth First Search (BFS). Our experimental results demonstrate 1.7× to 2.7× reduction in DRAM accesses leading to an improvement of 1.4× to 2× in total memory (DRAM + scratchpad) accesses.

The sustainable wood produces −160 tonnes/GWh carbon dioxide emissions, as compared to coal with +964, petroleum oil with +726 and even renewable energy source, solar having +5 tonnes CO2 emissions/GWh. These figures are from European Union Report 2001, which increased exponentially with the passage of time. Nevertheless, research and development on reducing the CO2 emissions gives a substantial ray of hope. It is quite evident from the recent advancements of bioenergy that biomass being a potentially clean energy can prove to be the most energy efficient source, provided its conversion mechanism are little more improved. Being a promising renewable energy source, biomass has been able to grab much attention recently. For the efficient utilization of biomass, among the 3 biomass-to-energy conversion technologies named combustion, gasification and pyrolysis — the technology which has proved to be a convenient and useful one is the pyrolysis. In this review paper, the classification, mechanism, processes of biomass pyrolysis along with its associated latest technologies will be brought into limelight.

In online social systems, users with common affiliations or interests form social groups for discussing various topical issues. We study the relationships among these social groups, which manifest through users who are common members of multiple groups, and the evolution of these relationships as new users join the groups. Focusing on a certain number of the most popular groups, we model the group memberships of users as a subclass of bipartite networks, known as Alphabetic Bipartite Networks (α-BiNs), where one of the partitions contains a fixed number of nodes (the popular groups) while the other grows unboundedly with time (new users joining the groups). Specifically, we consider the evolution of the thresholded projection of the user-group bipartite network onto the set of groups, which accurately represents the inter-group relationships. We propose and solve a preferential attachment based growth model for evolution of α-BiNs, and analytically compute the degree distribution of the thresholded projection. We further investigate whether the predictions of this model can explain the projection degree distributions of user-group networks derived from several real social systems (Livejournal, Youtube and Flickr). The study also shows that the inter-group network is tightly knit, and there is an implicit semantic hierarchy within its structure, that is clearly identified by the method of thresholding. To the best of our knowledge, this is the first attempt to analytically model the dynamical relationships among groups in online social systems.

Minimization of energy consumption in the battery operated system is become a major issue. Most of the real-time systems consist of a battery operated microprocessor system with a limited battery life. So, energy consumption is becoming a critical issue in the design of embedded systems because of the popularity of portable devices such as mobile devices and personal digital assistants. In this paper we propose a approach to handle task synchronization for Real Time Systems with energy efficiency consideration. Our p r o p o s e d approach is a variation of the well known priority c e i l i n g protocol (PCP) which is to enforce mutually exclusive a c c e s s to shared resources. W e a r e using the concept of speed locking i n p r o p o s e d a p p r o a c h s o t h a t we can save the enrgy consumption.

Several applications including community detection in social networks and discovering correlated genes involve finding large subgraphs of high density. We propose the problem of finding the largest subgraph of a given density. The problem is a generalization of the Max-Clique problem which seeks the largest subgraph that has an edge density of 1. We define an objective function and prove that its optimization results in the largest graph of given density. We propose an algorithm that finds the subgraph by running multiple local search heuristics with random restarts. For massive graphs, where running the algorithm directly may be intractable, we use a sampling technique that reduces the graph to a smaller one which is likely to contain large dense subgraphs. We evaluate our algorithm on multiple real life and synthetic datasets. Our experiments show that our algorithm performs as well as the state-of-the-art for finding large subgraphs of high density, while providing density guarantees.

Online Social Networks (OSNs), which are among the most popular sites on the Web today, are experiencing an exponential rise in the number and activity of users in recent years, leading to problems of scalability and increasing spam. Most OSNs currently impose restrictions on the number of social links that a user can have, in order to reduce strain on the OSN infrastructure and to prevent spammers from linking to large number of users indiscriminately. However, such restrictions are increasingly being criticized for being too prohibitive for socially active and popular users. Hence there is a growing need for the OSN authorities to design restrictions that fulfill their objectives while minimizing user-dissatisfaction; this is evident from the innovative 'soft' cut-off imposed in Twitter instead of the traditional 'hard' cut-offs in other OSNs. In this scenario, the goal of the current work is to develop an analytical framework, taking the restriction in Twitter as a case-study, to aid the design of restrictions which balance the two conflicting objectives of controlling system-load and minimizing user-dissatisfaction. We consequently define a simple utility function for restrictions considering the two objectives, and find that Twitter's policy well balances both. However, we find that the imposed restriction has limited effectiveness in controlling the linking activity of spammers. From a Complex Network Theoretical perspective, this is the first analysis of 'soft' cut-offs in any sort of network, to the best of our knowledge.

Instance retrieval (IR) is the problem of retrieving specific instances of a particular object, like a monument, from a collection of images. Currently, the most popular methods for IR use Bag of words (BoW) features for retrieval. However, a prominent problem for IR remains the tendency of BoW based methods to retrieve near-identical images as most relevant results. In this paper, we define diversity in IR as variation of physical properties among most relevant retrieved results for a query image. To achieve this, we propose both an ITML algorithm that re-fashions the BoW feature space into one that appreciates diversity better, and a measure to evaluate diversity in retrieval results for IR applications. Additionally, we also generate 200 hand-labeled images from the Paris dataset, for use in further research in this area. Experiments on the popular Paris dataset show that our method outperforms the standard BoW model in many cases.

Sorting is a key kernel in numerous big data application including database operations, graphs and text analytics. Due to low control overhead, parallel bitonic sorting networks are usually employed for hardware implementations to accelerate sorting. Although a typical implementation of merge sort network can lead to low latency and small memory usage, it suffers from low throughput due to the lack of parallelism in the final stage. We analyze a pipelined merge sort network, showing its theoretical limits in terms of latency, memory and, throughput. To increase the throughput, we propose a merge sort based hybrid design where the final few stages in the merge sort network are replaced with “folded” bitonic merge networks. In these “folded” networks, all the interconnection patterns are realized by streaming permutation networks (SPN). We present a theoretical analysis to quantify latency, memory and throughput of our proposed design. Performance evaluations are performed by experiments on Xilinx Virtex-7 FPGA with post place-androute results. We demonstrate that our implementation achieves a throughput close to 10 GBps, outperforming state-of-the-art implementation of sorting on the same hardware by 1.2x, while preserving lower latency and higher memory efficiency.

The study of information dissemination on a social network has gained significant importance with the rise of social media. Since the true dynamics are hidden, various diffusion models have been exposed to explain the cascading behavior. Such models require extensive simulation for estimating the dissemination over time. In an earlier work, we proposed a unified model which provides an approximate analytical solution to the problem of predicting probability of infection of every node in the network over time. Our model generalizes a large class of diffusion process. We demonstrate through extensive empirical evaluation that the error of approximation is small. We build upon our unified model to develop an efficient method for influence maximization. Unlike most approaches, we assume that diffusion spreads not only via the edges of the underlying network, but also through temporal functions of external out-of-network processes. We empirically evaluate our approach and compare it against state-of-the-art approaches on real-world large-scale networks. The evaluation demonstrates that our method has significant performance gains over widely used seed-set selection algorithms.