Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231243
Muhammad Arbab Arshad, Sakib Shahriar, A. Sagahyroon
Convolutional Neural Network (CNN) is a subclass of deep neural network that has gained popularity in recent years. CNN has revolutionized the execution of tasks such as natural language processing, image classification, and voice recognition. However, the performance of CNNs is often limited by the hardware available for training large sets of data. Graphical Processing Units (GPUs) have been shown to achieve good performance with CNN-based applications, however, GPU is expensive and is not suitable for all applications. In recent years, and for various reasons researchers have shifted their focus to Field Programmable Gate Arrays (FPGAs) and even other edge devices like microcontrollers to execute CNN models. This paper provides a survey of a number of applications where FPGAs are used in the implementation of various CNN-based models. The survey provides the reader with a compact and informative insight into recent efforts in this domain.
{"title":"On the Use of FPGAs to Implement CNNs: A Brief Review","authors":"Muhammad Arbab Arshad, Sakib Shahriar, A. Sagahyroon","doi":"10.1109/iCCECE49321.2020.9231243","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231243","url":null,"abstract":"Convolutional Neural Network (CNN) is a subclass of deep neural network that has gained popularity in recent years. CNN has revolutionized the execution of tasks such as natural language processing, image classification, and voice recognition. However, the performance of CNNs is often limited by the hardware available for training large sets of data. Graphical Processing Units (GPUs) have been shown to achieve good performance with CNN-based applications, however, GPU is expensive and is not suitable for all applications. In recent years, and for various reasons researchers have shifted their focus to Field Programmable Gate Arrays (FPGAs) and even other edge devices like microcontrollers to execute CNN models. This paper provides a survey of a number of applications where FPGAs are used in the implementation of various CNN-based models. The survey provides the reader with a compact and informative insight into recent efforts in this domain.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123513829","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231172
G. Almeida, Zhaochu Yang, P. Mendes, T. Dong
This work presents a current starved voltage-controlled oscillator based on standard 0.13μm CMOS process. Integrated within a power management circuit, the proposed oscillator provides an average periodic signal with a frequency of 84.81kHz. Additionally, to assure a stable periodic signal from the environmental instability a voltage reference is designed. The proposed architecture makes full use of subthreshold and deep triode MOSFETs to avoid the employment of any passive component. A voltage reference exhibits an output reference of 258.34mV in response to a 1.0 - 3.2V voltage supply. It shows a line sensitivity of 0.49%/V at 27°C. The output of the architecture leads to a robust time-control local oscillator, which can be employed in power management circuit for energy harvesting systems and be used on wireless sensor networks and implantable medical devices.
{"title":"A CMOS Current Starved VCO for Energy Harvesting applications","authors":"G. Almeida, Zhaochu Yang, P. Mendes, T. Dong","doi":"10.1109/iCCECE49321.2020.9231172","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231172","url":null,"abstract":"This work presents a current starved voltage-controlled oscillator based on standard 0.13μm CMOS process. Integrated within a power management circuit, the proposed oscillator provides an average periodic signal with a frequency of 84.81kHz. Additionally, to assure a stable periodic signal from the environmental instability a voltage reference is designed. The proposed architecture makes full use of subthreshold and deep triode MOSFETs to avoid the employment of any passive component. A voltage reference exhibits an output reference of 258.34mV in response to a 1.0 - 3.2V voltage supply. It shows a line sensitivity of 0.49%/V at 27°C. The output of the architecture leads to a robust time-control local oscillator, which can be employed in power management circuit for energy harvesting systems and be used on wireless sensor networks and implantable medical devices.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115799336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231071
A. Uka, X. Polisi, J. Barthès, A. Halili, Florenc Skuka, N. Vrana
Medical field depends heavily on understanding and analyzing microscopy images of cells to better diagnose diseases, to evaluate the effectiveness of various medical treatments and to determine their health under stress. The amount of data that needs to be analyzed has increased and computer assisted analysis has become crucial as it would be very labor intensive for the medical practitioners otherwise. Many of the images are acquired using brightfield microscopy with no staining in order to avoid all the side effects. The unstained images have some associating challenges as they suffer from random nonuniform illumination, low contrast, relatively high transparency of the cytoplasm. The initial challenge of the large amount of data calls for the use of deep learning algorithms, whereas the other structural challenges call for the need to carefully train the convolutional neural networks in order to have a reliable system of evaluation. We have prepared a dataset of 20.000 images and we have tested the trained models on datasets with different number of images (N=300-8000). Here is this work we present classification of the cell health using convolutional neural networks and monitor the effect of the preprocessing steps on the overall accuracy.
{"title":"Effect of Preprocessing on Performance of Neural Networks for Microscopy Image Classification","authors":"A. Uka, X. Polisi, J. Barthès, A. Halili, Florenc Skuka, N. Vrana","doi":"10.1109/iCCECE49321.2020.9231071","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231071","url":null,"abstract":"Medical field depends heavily on understanding and analyzing microscopy images of cells to better diagnose diseases, to evaluate the effectiveness of various medical treatments and to determine their health under stress. The amount of data that needs to be analyzed has increased and computer assisted analysis has become crucial as it would be very labor intensive for the medical practitioners otherwise. Many of the images are acquired using brightfield microscopy with no staining in order to avoid all the side effects. The unstained images have some associating challenges as they suffer from random nonuniform illumination, low contrast, relatively high transparency of the cytoplasm. The initial challenge of the large amount of data calls for the use of deep learning algorithms, whereas the other structural challenges call for the need to carefully train the convolutional neural networks in order to have a reliable system of evaluation. We have prepared a dataset of 20.000 images and we have tested the trained models on datasets with different number of images (N=300-8000). Here is this work we present classification of the cell health using convolutional neural networks and monitor the effect of the preprocessing steps on the overall accuracy.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117052868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231197
Jing Wang
Image restoration in poor weather conditions can assist military combatants to efficiently and accurately perform object detection, object recognition and object tracking. Moreover, in security systems, traffic navigation, etc. it also has high application value. Aiming at the problem of image distortion caused by different poor weather conditions like dust, rain, snow, fog, haze, etc. this paper proposes a new deep neural network based image restoration technology, a residual aggregation module is constructed for extracting the detailed features. Furthermore, dense connection is applied to combine low-dimensional features and generate high-dimensional features. The experimental results show that the network achieves superior results in image de-raining(IDR) compared with Deep Detail Network(DDN) and Dual Convolutional Neural Network(DualCNN) while obtaining favorable performances in image de-noising, image de-hazing, image de-blurring, image de-raindrops and other tasks.
{"title":"Image Restoration on Residual Aggregation Network in Poor Weather Condition","authors":"Jing Wang","doi":"10.1109/iCCECE49321.2020.9231197","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231197","url":null,"abstract":"Image restoration in poor weather conditions can assist military combatants to efficiently and accurately perform object detection, object recognition and object tracking. Moreover, in security systems, traffic navigation, etc. it also has high application value. Aiming at the problem of image distortion caused by different poor weather conditions like dust, rain, snow, fog, haze, etc. this paper proposes a new deep neural network based image restoration technology, a residual aggregation module is constructed for extracting the detailed features. Furthermore, dense connection is applied to combine low-dimensional features and generate high-dimensional features. The experimental results show that the network achieves superior results in image de-raining(IDR) compared with Deep Detail Network(DDN) and Dual Convolutional Neural Network(DualCNN) while obtaining favorable performances in image de-noising, image de-hazing, image de-blurring, image de-raindrops and other tasks.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131040583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231119
U. Hudomalj, C. Mandla, M. Plattner
This paper presents FPGA implementations of two standard image preprocessing algorithms, namely image filtering and image averaging. The implementations allow processing of high-frame-rate and high-resolution image streams in real-time. The developed implementations are evaluated in terms of resource usage, power consumption, and achievable frame rates. The performance of the developed implementation of image filtering algorithm is compared with implementation provided by MATLAB’s Vision HDL Toolbox. The algorithms are evaluated on Microsemi’s Smartfusion2 Advanced Development Kit. The development board includes a SmartFusion2 M2S150 SoC FPGA. For verification, the board is connected to an industrial camera which uses the Camera Link interface. Limitations of processing image streams with FPGA platforms are discussed.
{"title":"FPGA Implementations for Real-Time Processing of High-Frame-Rate and High-Resolution Image Streams","authors":"U. Hudomalj, C. Mandla, M. Plattner","doi":"10.1109/iCCECE49321.2020.9231119","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231119","url":null,"abstract":"This paper presents FPGA implementations of two standard image preprocessing algorithms, namely image filtering and image averaging. The implementations allow processing of high-frame-rate and high-resolution image streams in real-time. The developed implementations are evaluated in terms of resource usage, power consumption, and achievable frame rates. The performance of the developed implementation of image filtering algorithm is compared with implementation provided by MATLAB’s Vision HDL Toolbox. The algorithms are evaluated on Microsemi’s Smartfusion2 Advanced Development Kit. The development board includes a SmartFusion2 M2S150 SoC FPGA. For verification, the board is connected to an industrial camera which uses the Camera Link interface. Limitations of processing image streams with FPGA platforms are discussed.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130008491","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231177
Minami Yoda, Shuji Sakuraba, Y. Sei, Yasuyuki Tahara, Akihiko Ohsuga
Internet of Things (IoT) for smart homes enhances the convenience of our life; however, it also introduces the risk of leakage of privacy data in the house. A user wants to protect their privacy data from leakage. However, the analysis of IoT devices requires technical knowledge; therefore, it is challenging for the users to detect any vulnerability by themselves. In this study, we propose a lightweight method to detect the hardcoded username and password in IoT devices using static analysis. This method can detect the 1st vulnerability from 2018 OWASP TOP 10 for the IoT device. The hardcoded login information can be obtained by comparing the user input with strcmp or strncmp. Thus, previous studies analyzed the symbols of strcmp or strncmp to detect the hardcoded login information. However, these studies require time because of the usage of complicated algorithms such as symbolic execution. To develop a lightweight algorithm, we focus on a network function, such as the socket symbol in firmware, because the IoT device is compromised when it is invaded by someone via the Internet. We propose two methods to detect the hardcoded login information, i.e., string search and socket search. In string searching, it finds a function that uses strcmp or strncmp symbol. In socket searching, it finds a function that is referenced by socket symbol. In the experiment, we measured the ability of our method by searching six firmware in the real world that has a backdoor. we ran three methods: string search, socket search, and whole search to compare two methods. As a result, all methods found login information from four of six firmware. Our method reduces an analysis time that when the whole search takes 38mins to complete, our methods finish 4-6min.
智能家居的物联网(IoT)增强了我们生活的便利性;然而，它也引入了家庭隐私数据泄露的风险。用户希望保护他们的隐私数据不被泄露。然而，物联网设备的分析需要技术知识;因此，用户自己检测漏洞是一项挑战。在这项研究中，我们提出了一种轻量级的方法，通过静态分析来检测物联网设备中的硬编码用户名和密码。此方法可以检测到2018年OWASP TOP 10中IoT设备的第一个漏洞。通过将用户输入与strcmp或strncmp进行比较，可以获得硬编码的登录信息。因此，以往的研究通过分析strcmp或strncmp的符号来检测硬编码的登录信息。然而，这些研究需要时间，因为使用复杂的算法，如符号执行。为了开发轻量级算法，我们将重点放在网络功能上，例如固件中的套接字符号，因为当物联网设备被某人通过互联网入侵时，它会受到损害。我们提出了两种检测硬编码登录信息的方法，即字符串搜索和套接字搜索。在字符串搜索中，它查找使用strcmp或strncmp符号的函数。在套接字搜索中，它查找由套接字符号引用的函数。在实验中，我们通过在现实世界中搜索六个具有后门的固件来测量我们的方法的能力。我们运行了三种方法:字符串搜索、套接字搜索和整体搜索来比较两种方法。结果，所有方法都从六个固件中的四个找到了登录信息。我们的方法减少了分析时间，当整个搜索需要38分钟完成时，我们的方法完成4-6分钟。
{"title":"Detection of the Hardcoded Login Information from Socket Symbols","authors":"Minami Yoda, Shuji Sakuraba, Y. Sei, Yasuyuki Tahara, Akihiko Ohsuga","doi":"10.1109/iCCECE49321.2020.9231177","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231177","url":null,"abstract":"Internet of Things (IoT) for smart homes enhances the convenience of our life; however, it also introduces the risk of leakage of privacy data in the house. A user wants to protect their privacy data from leakage. However, the analysis of IoT devices requires technical knowledge; therefore, it is challenging for the users to detect any vulnerability by themselves. In this study, we propose a lightweight method to detect the hardcoded username and password in IoT devices using static analysis. This method can detect the 1st vulnerability from 2018 OWASP TOP 10 for the IoT device. The hardcoded login information can be obtained by comparing the user input with strcmp or strncmp. Thus, previous studies analyzed the symbols of strcmp or strncmp to detect the hardcoded login information. However, these studies require time because of the usage of complicated algorithms such as symbolic execution. To develop a lightweight algorithm, we focus on a network function, such as the socket symbol in firmware, because the IoT device is compromised when it is invaded by someone via the Internet. We propose two methods to detect the hardcoded login information, i.e., string search and socket search. In string searching, it finds a function that uses strcmp or strncmp symbol. In socket searching, it finds a function that is referenced by socket symbol. In the experiment, we measured the ability of our method by searching six firmware in the real world that has a backdoor. we ran three methods: string search, socket search, and whole search to compare two methods. As a result, all methods found login information from four of six firmware. Our method reduces an analysis time that when the whole search takes 38mins to complete, our methods finish 4-6min.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134531020","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231090
Alexander Marinšek, L. De Strycker
Europe’s massive shift towards sustainable energy production has triggered a variety of new large scale projects, and wind energy is a crucial part of the effort to achieve a carbon-free future. However, because of their low financial impact and relatively high measurement campaign costs, small scale projects are often deemed impractical beforehand. To help small communities gain insight on the wind energy conditions in their surroundings, the present work briefly introduces a measuring station (MEST) concept based on affordable electronic components and proposes a solution to alleviating the effects of inevitable measurement data inconsistency on the energy yield analysis. By leveraging open source machine learning models and establishing a link with the publicly available ERA5-Land climate database, missing wind speed measurement data is reconstructed at an accuracy of up to 0.11 $frac{m}{s}$. The impact of data reconstruction on the estimated energy production of a wind turbine (WT) erected at the measuring location is then evaluated using the measurement data acquired by a MEST prototype and the ERA5-Land data recorded during October and November 2019. The results indicate that at a location experiencing moderate wind speeds, the estimated energy output of the WT is increased by up to 2 % in comparison with other data analysis procedures. Although the minute underestimation is not of great importance to the success of the analysis, the inaccuracies at higher wind speeds have a far more profound effect on the WT’s estimated energy output, and they can stop a potentially successful wind energy project from gaining further attention.
{"title":"Towards efficient wind energy monitoring: Learning more from open source data","authors":"Alexander Marinšek, L. De Strycker","doi":"10.1109/iCCECE49321.2020.9231090","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231090","url":null,"abstract":"Europe’s massive shift towards sustainable energy production has triggered a variety of new large scale projects, and wind energy is a crucial part of the effort to achieve a carbon-free future. However, because of their low financial impact and relatively high measurement campaign costs, small scale projects are often deemed impractical beforehand. To help small communities gain insight on the wind energy conditions in their surroundings, the present work briefly introduces a measuring station (MEST) concept based on affordable electronic components and proposes a solution to alleviating the effects of inevitable measurement data inconsistency on the energy yield analysis. By leveraging open source machine learning models and establishing a link with the publicly available ERA5-Land climate database, missing wind speed measurement data is reconstructed at an accuracy of up to 0.11 $frac{m}{s}$. The impact of data reconstruction on the estimated energy production of a wind turbine (WT) erected at the measuring location is then evaluated using the measurement data acquired by a MEST prototype and the ERA5-Land data recorded during October and November 2019. The results indicate that at a location experiencing moderate wind speeds, the estimated energy output of the WT is increased by up to 2 % in comparison with other data analysis procedures. Although the minute underestimation is not of great importance to the success of the analysis, the inaccuracies at higher wind speeds have a far more profound effect on the WT’s estimated energy output, and they can stop a potentially successful wind energy project from gaining further attention.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124740783","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231160
F. Ahmed, F. Khelifi, Ashref Lawgaly, A. Bouridane
Estimating the acquisition date of digital photographs is crucial in image forensics. The task of dating images by processing their contents should be reasonably accurate as this can be used in court to resolve high profile cases. The goal of temporal forensics analysis is to find out the links in time between two or more pieces of evidence. In this paper, the problem of picture dating is addressed from a machine learning perspective, precisely, by adopting a deep learning approach for the first time in temporal image forensics. In this work, the acquisition time of digital images is estimated in such a way that the analyst can identify the timeline of unknown digital photographs given a set of pictures from the same source whose time ordering is known. By applying Convolutional Neural Networks (CNN), namely the AlexNet and GoogLeNet architectures in both feature extraction and transfer learning modes, results have shown that the networks can successfully learn the temporal changes in the content of the digital pictures that are acquired from the same source. Interestingly, although images are divided into non-overlapping blocks in order to increase the number of training samples and feed CNNs, the obtained estimation accuracy has been from 80% to 88%. This suggests that the temporal changes in image contents, modelled by CNNs, are not dependent on block location. This has been demonstrated on a new database called ‘Northumbria Temporal Image Forensics’ (NTIF) database which has been made publicly available for researchers in image forensics. NTIF is the first public database that captures a large number of images at different timeslots on a regular basis using 10 different digital cameras. This will serve the research community as a solid ground for research on picture dating and other image forensics applications.
{"title":"Temporal Image Forensic Analysis for Picture Dating with Deep Learning","authors":"F. Ahmed, F. Khelifi, Ashref Lawgaly, A. Bouridane","doi":"10.1109/iCCECE49321.2020.9231160","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231160","url":null,"abstract":"Estimating the acquisition date of digital photographs is crucial in image forensics. The task of dating images by processing their contents should be reasonably accurate as this can be used in court to resolve high profile cases. The goal of temporal forensics analysis is to find out the links in time between two or more pieces of evidence. In this paper, the problem of picture dating is addressed from a machine learning perspective, precisely, by adopting a deep learning approach for the first time in temporal image forensics. In this work, the acquisition time of digital images is estimated in such a way that the analyst can identify the timeline of unknown digital photographs given a set of pictures from the same source whose time ordering is known. By applying Convolutional Neural Networks (CNN), namely the AlexNet and GoogLeNet architectures in both feature extraction and transfer learning modes, results have shown that the networks can successfully learn the temporal changes in the content of the digital pictures that are acquired from the same source. Interestingly, although images are divided into non-overlapping blocks in order to increase the number of training samples and feed CNNs, the obtained estimation accuracy has been from 80% to 88%. This suggests that the temporal changes in image contents, modelled by CNNs, are not dependent on block location. This has been demonstrated on a new database called ‘Northumbria Temporal Image Forensics’ (NTIF) database which has been made publicly available for researchers in image forensics. NTIF is the first public database that captures a large number of images at different timeslots on a regular basis using 10 different digital cameras. This will serve the research community as a solid ground for research on picture dating and other image forensics applications.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132364247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231052
Maiass Zaher, Aymen Hasan Alawadi, S. Molnár
The emerging technologies leveraging Data Center Networks (DCN) and their consequent traffic patterns impose more necessity for improving Quality of Service (QoS). In this paper, we propose Sieve, a new distributed SDN framework that efficiently schedules flows based on the available bandwidth to improve Flow Completion Time (FCT) of mice flows. In addition, we propose a lightweight sampling mechanism to sample a portion of flows. In particular, Sieve schedules the sampled flows, and it reschedules only elephant flows upon threshold hits. Furthermore, our framework allocates a portion of the flows to ECMP, so that the associated overhead can be mitigated in the control plane and ECMP-related packet collisions are fewer as well. Mininet has been used to evaluate the proposed solution, and Sieve provides better FCT up to 50% in comparison to the existing solutions like ECMP and Hedera.
利用数据中心网络(DCN)的新兴技术及其随之而来的流量模式对提高服务质量(QoS)提出了更大的必要性。在本文中，我们提出了一种新的分布式SDN框架Sieve，该框架基于可用带宽有效地调度流量，以提高流量完成时间(Flow Completion Time, FCT)。此外，我们提出了一种轻量级的采样机制来对一部分流进行采样。特别地，Sieve调度采样流，它只在达到阈值时重新调度大象流。此外，我们的框架将一部分流分配给ECMP，这样可以减轻控制平面中的相关开销，并且与ECMP相关的数据包冲突也更少。Mininet已经被用来评估提议的解决方案，与现有的解决方案(如ECMP和Hedera)相比，Sieve提供了更好的FCT，高达50%。
{"title":"Class-based Flow Scheduling Framework in SDN-based Data Center Networks","authors":"Maiass Zaher, Aymen Hasan Alawadi, S. Molnár","doi":"10.1109/iCCECE49321.2020.9231052","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231052","url":null,"abstract":"The emerging technologies leveraging Data Center Networks (DCN) and their consequent traffic patterns impose more necessity for improving Quality of Service (QoS). In this paper, we propose Sieve, a new distributed SDN framework that efficiently schedules flows based on the available bandwidth to improve Flow Completion Time (FCT) of mice flows. In addition, we propose a lightweight sampling mechanism to sample a portion of flows. In particular, Sieve schedules the sampled flows, and it reschedules only elephant flows upon threshold hits. Furthermore, our framework allocates a portion of the flows to ECMP, so that the associated overhead can be mitigated in the control plane and ECMP-related packet collisions are fewer as well. Mininet has been used to evaluate the proposed solution, and Sieve provides better FCT up to 50% in comparison to the existing solutions like ECMP and Hedera.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121043475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-08-17DOI: 10.1109/iCCECE49321.2020.9231154
Fabian Scheidl
In the field of emerging software architectures, there has been a dramatic push towards flexible and sand-boxed software modules that allow systems to safely execute untrusted code in a guaranteed side-effect free manner. Latest developments have further given rise to portable and statically validatable representations of software in a bytecode format like WebAssembly. In order to ease the segue into the domain of embedded systems, this paper explores the feasibility of a novel and easily retargetable streaming quasi-singlepass on-target-compiler topology with concurrent bytecode validation. For this, a generalized compile-time virtual stack is employed which is logically partitioned into separately emittable blocks (named valent-blocks). This forms the foundation of a corresponding runtime for resource constrained systems that demonstrate the need for predictable, resource-efficient and fast sandboxed execution of hot-loaded software. This paper further benchmarks the resultant performance against current popular competing standalone WebAssembly runtimes.
{"title":"Valent-Blocks: Scalable High-Performance Compilation of WebAssembly Bytecode For Embedded Systems","authors":"Fabian Scheidl","doi":"10.1109/iCCECE49321.2020.9231154","DOIUrl":"https://doi.org/10.1109/iCCECE49321.2020.9231154","url":null,"abstract":"In the field of emerging software architectures, there has been a dramatic push towards flexible and sand-boxed software modules that allow systems to safely execute untrusted code in a guaranteed side-effect free manner. Latest developments have further given rise to portable and statically validatable representations of software in a bytecode format like WebAssembly. In order to ease the segue into the domain of embedded systems, this paper explores the feasibility of a novel and easily retargetable streaming quasi-singlepass on-target-compiler topology with concurrent bytecode validation. For this, a generalized compile-time virtual stack is employed which is logically partitioned into separately emittable blocks (named valent-blocks). This forms the foundation of a corresponding runtime for resource constrained systems that demonstrate the need for predictable, resource-efficient and fast sandboxed execution of hot-loaded software. This paper further benchmarks the resultant performance against current popular competing standalone WebAssembly runtimes.","PeriodicalId":413847,"journal":{"name":"2020 International Conference on Computing, Electronics & Communications Engineering (iCCECE)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126538146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：