Current Issue : January - March Volume : 2018 Issue Number : 1 Articles : 5 Articles
Cyber-Physical system devices nowadays constitute a mixture of Information Technology\n(IT) and Operational Technology (OT) systems that are meant to operate harmonically under a security\ncritical framework. As security IT countermeasures are gradually been installed in many embedded\nsystem nodes, thus securing them from many well-know cyber attacks there is a lurking danger that\nis still overlooked. Apart from the software vulnerabilities that typical malicious programs use, there\nare some very interesting hardware vulnerabilities that can be exploited in order to mount devastating\nsoftware or hardware attacks (typically undetected by software countermeasures) capable of fully\ncompromising any embedded system device. Real-time microarchitecture attacks such as the cache\nside-channel attacks are such case but also the newly discovered Rowhammer fault injection attack\nthat can be mounted even remotely to gain full access to a device DRAM (Dynamic Random Access\nMemory). Under the light of the above dangers that are focused on the device hardware structure,\nin this paper, an overview of this attack field is provided including attacks, threat directives and\ncountermeasures. The goal of this paper is not to exhaustively overview attacks and countermeasures\nbut rather to survey the various, possible, existing attack directions and highlight the security\nrisks that they can pose to security critical embedded systems as well as indicate their strength on\ncompromising the Quality of Service (QoS) such systems are designed to provide....
Fiber Bragg Grating (FBG) sensors have been increasingly used in the field of Structural Health Monitoring (SHM) in recent\nyears. In this paper, we proposed an impact localization algorithm based on the Empirical Mode Decomposition (EMD) and\nParticle Swarm Optimization-Support VectorMachine (PSO-SVM) to achieve better localization accuracy for the FBG-embedded\nplate. In our method, EMD is used to extract the features of FBG signals, and PSO-SVM is then applied to automatically train a\nclassification model for the impact localization. Meanwhile, an impact monitoring system for the FBG-embedded composites has\nbeen established to actually validate our algorithm. Moreover, the relationship between the localization accuracy and the distance\nfrom impact to the nearest sensor has also been studied. Results suggest that the localization accuracy keeps increasing and is\nsatisfactory, ranging from 93.89% to 97.14%, on our experimental conditions with the decrease of the distance. This article reports\nan effective and easy-implementing method for FBG signal processing on SHM systems of the composites....
Convolution is the most computationally intensive task of the Convolutional Neural\nNetwork (CNN). It requires a lot of memory storage and computational power. There are different\napproaches to compute the solution of convolution and reduce its computational complexity. In this\npaper, a matrix multiplication-based convolution (ConvMM) approach is fully parallelized using\nconcurrent resources of GPU (Graphics Processing Unit) and optimized, considerably improving the\nperformance of the image classifiers and making them applicable to real-time embedded applications.\nThe flow of this CUDA (Compute Unified Device Architecture)-based scheme is optimized using\nunified memory and hardware-dependent acceleration of matrix multiplication. Proposed flow is\nevaluated on two different embedded platforms: first on an Nvidia Jetson TX1 embedded board\nand then on a Tegra K1 GPU of an Nvidia Shield K1 Tablet. The performance of this optimized\nand accelerated convolutional layer is compared with its sequential and heterogeneous versions.\nResults show that the proposed scheme significantly improves the overall results including energy\nefficiency, storage requirement and inference performance. In particular, the proposed scheme on\nembedded GPUs is hundreds of times faster than the sequential version and delivers tens of times\nhigher performance than the heterogeneous approach....
The Chip Multiprocessors (CMPs) architecture moves from multi-core to many-core architecture to\nprovide higher computing performance, and more reliable systems. Moreover, the CMPs trend also\nmove from 2D CMPs to 3D CMPs architecture in order to obtain higher performance, more reliability,\nreduced cache access latency, and increased cache bandwidth when compared with 2D CMPs.\nTherefore, in this work we present a 3D many-core CMP architecture which executes heavy loaded\ntasks in order to improve the system performance. However, executing heavy loaded tasks demands\nincreasing in system power consumption which results in increasing the on-chip thermal hotspots. The\nthermal hotspots in the 3D many-core CMPs cause performance degradation, reducing reliability,\ndecreasing the chip life spam. Therefore, Runtime Thermal Management (RTM) in the 3D many-core\nCMPs has become crucial to control the thermal hotspots without any performance degradation. In\nthis paper, a new runtime task migration technique is proposed to control hotspots in the 3D manycore\nCMPs. The proposed technique migrates the hottest tile with the optimal coldest tile in the core\nlayer. The optimal coldest tileis selected by considering theDynamic Random Access Memory\n(DRAM) banks' access distribution level in the cache layer. The simulation results indicate up to 33\n\n(on average 13)\nreduction in the cores' temperature of the target 3D many-core CMP. Moreover, the\nproposed technique efficiency is clarified in the simulation results that the maximum temperature of\ncores in the core and cache layers are both less than the maximum temperature limit, 80....
Operational analytics is all about answering business questions while doing\nbusiness and supporting business users across the organization, from shop\nfloor users to management and executives. Therefore, business transactions\nand analytics must co-exist together in a single platform to empower business\nusers to drive insights, make decisions, and complete business processes in a\nsingle application and using a single source of facts without toggling between\nmultiple applications. Traditionally transactional systems and analytics were\nmaintained separately to improve throughput of the transactional system and\nthat certainly introduced latency in decision making. However, with innovation\nin the SAP HANA platform, SAP S/4HANA embedded analytics enables business\nusers, business analysts, and management to perform real-time analytics on\nlive transactional data. This paper reviews technical architecture and key components\nof SAP S/4HANA embedded analytics. This paper reviews technical\narchitecture and key components of SAP S/4HANA embedded analytics....
Loading....