2021 International Conference on Information and Communication Technology Convergence (ICTC)最新文献_第10页

Evaluating the Performance of Deep Learning Inference Service on Edge Platform 边缘平台上深度学习推理服务的性能评估

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620870

Hyun-Hwa Choi, Jae-Geun Cha, Seung Hyun Yun, Dae-Won Kim, Sumin Jang, S. Kim

{"title":"Evaluating the Performance of Deep Learning Inference Service on Edge Platform","authors":"Hyun-Hwa Choi, Jae-Geun Cha, Seung Hyun Yun, Dae-Won Kim, Sumin Jang, S. Kim","doi":"10.1109/ICTC52510.2021.9620870","DOIUrl":"https://doi.org/10.1109/ICTC52510.2021.9620870","url":null,"abstract":"Deep learning inference requires tremendous amount of computation and typically is offloaded the cloud for execution. Recently, edge computing, which processes and stores data at the edge of the Internet closest to the mobile devices or sensors, has been considered as new computing paradigm. We have studied the performance of the deep neural network (DNN) inference service based on different configurations of resources assigned to a container. In this work, we measured and analyzed a real-world edge service on containerization platform. An edge service is named A!Eye, an application with various DNN inferences. The edge service has both CPU-friendly and GPU-friendly tasks. CPU tasks account for more than half of the latency of the edge service. Our analyses reveal interesting findings about running the DNN inference service on the container-based execution platform; (a) The latency of DNN inference-based edge services is affected by CPU-based operation performance. (b) Pinning CPUs can reduce the latency of an edge service. (c) In order to improve the performance of an edge service, it is very important to avoid PCIe bottleneck shared by resources like CPUs, GPUs and NICs.","PeriodicalId":299175,"journal":{"name":"2021 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"20 80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125325696","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}

引用次数: 0

Conducted Electromagnetic Interference Energy Harvesting for Green Communication of IoT Sensors 物联网传感器绿色通信的传导电磁干扰能量收集

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9621086

Changhee Hyoung, Junghwan Hwang

引用次数: 0

Benefits, Challenges and Practical Concerns of IoT for Smart Manufacturing 物联网对智能制造的好处、挑战和现实关注

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620771

Vivian Ukamaka Ihekoronye, C. I. Nwakanma, Goodness Oluchi Anyanwu, Dong-Seong Kim, Jae-Min Lee

引用次数: 3

An Efficient Policy Improvement in Human Interactive Learning Using Entropy 基于熵的人机交互学习有效策略改进

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620856

Sungyun Park, Dae-Wook Kim, Sang-Kwang Lee, Seong-il Yang

{"title":"An Efficient Policy Improvement in Human Interactive Learning Using Entropy","authors":"Sungyun Park, Dae-Wook Kim, Sang-Kwang Lee, Seong-il Yang","doi":"10.1109/ICTC52510.2021.9620856","DOIUrl":"https://doi.org/10.1109/ICTC52510.2021.9620856","url":null,"abstract":"Human knowledge is used in reinforcement learning (RL), which reduces the amount of time taken by the learning agent to achieve its goal. The TAMER (Training an Agent Manually via Evaluative Reinforcements) algorithm allows a human to provide a reward to an autonomous agent through a manual interface while watching the agent performs the action. Because a policy, the agent have, is updated based on human rewards, it approximates how a human trainer gives rewards to the agent. For policy update, events that occurred during learning were selected. Furthermore, while selecting events, the temporal distance from the event to the human reward is considered. Thus, the events that only occurred in a certain time interval before the human trainer gives a reward are selected. However, this approach of considering only the time factor demands quite many human rewards for the policy. The policy update with high complexity make the human trainer exhausted during improvement of policy. Therefore, we propose a new method of selecting events, which considers the entropy value over the distribution of Q-values, in addition to the time factor. For the policy update in our proposed event selection method, we reuse the events despite of long temporal distance since human reward when their each human reward is negative and entropy value (over the distribution of Q-values) is low. To compare the effectiveness of the proposed method with the classic TAMER, we implement an experiment with the policy initialized to an incorrect weight. The results show that the TAMER algorithm, using our proposed selection of events, efficiently improves the policy.","PeriodicalId":299175,"journal":{"name":"2021 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132736355","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}

引用次数: 0

Smart-mDAG: An Intelligent Scheduling Method for Multi-DAG Jobs Smart-mDAG:多dag作业的智能调度方法

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9621176

Yifan Zhu, Bo Hu

{"title":"Smart-mDAG: An Intelligent Scheduling Method for Multi-DAG Jobs","authors":"Yifan Zhu, Bo Hu","doi":"10.1109/ICTC52510.2021.9621176","DOIUrl":"https://doi.org/10.1109/ICTC52510.2021.9621176","url":null,"abstract":"Job scheduling is a fundamental problem in cloud data center, which plays an essential role in the makespan, resource utilization and maintenance of scheduling security, it has received widespread attention. With the rapid increase of jobs' amount, higher requirements are put forward for scheduling efficiency and makespan. Meantime dependencies between tasks are closely related to makespan and throughput, and these dependent tasks form multiple DAG structures. Heuristic algorithms have limitation on adjusting the scheduling policy according to diverse dependencies, thus resulting the extension of makespan. In this paper, we propose an intelligent scheduling method for multi-DAG jobs using deep reinforcement learning, called Smart-mDAG. It is a job-specific scheduling method that adjusts the scheduling policy based on diverse dependencies to minimize the makespan. Firstly, we convert dependencies to numeric form through a feature extraction module to obtain the dependent information from the DAG. Secondly, we use cascaded neural networks to implement the fusion of scheduling information, so we can obtain the fitness between machines and tasks. With Alibaba Cluster Data V2018, we evaluate the performance of Smart-mDAG on a five-machines small cluster. The result shows that compared to control algorithms, Smart-mDAG can shorten the makespan for 70% jobs, and the optimal makespan of single job can be decreased to 65% of the past.","PeriodicalId":299175,"journal":{"name":"2021 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"53 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134480060","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}

引用次数: 1

3D hand motion and position estimation using ultrasonic receiver array and inertial sensors 基于超声接收阵列和惯性传感器的三维手部运动和位置估计

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9621004

Sang-Woo Seo, Seung-Bok Kwon

{"title":"3D hand motion and position estimation using ultrasonic receiver array and inertial sensors","authors":"Sang-Woo Seo, Seung-Bok Kwon","doi":"10.1109/ICTC52510.2021.9621004","DOIUrl":"https://doi.org/10.1109/ICTC52510.2021.9621004","url":null,"abstract":"In this study, we introduce a novel technique for hand pose and position estimation using ultrasonic and inertial sensors. The three-dimensional (3D) position of the hand is localized by a combination of ultrasonic-based position estimation and inertial measurement unit (IMU) sensor-based 3D acceleration and pose estimation. The proposed method roughly estimates the 3D position of a hand at a relatively low speed (10 Hz) using the delay-and-sum beamforming (DSBF) algorithm by placing two multi-channel ultrasonic sensors in a two-dimensional plane; it accurately estimates the real-time 3D position using a high-speed (187.5 Hz) inertial sensor mounted on the back of the hand. To maximize the real-time performance and cost-effectiveness of the proposed system, a cost-effective field-programmable gate array is used for the ultrasonic array-based 3D localization estimation module to optimize the DSBF algorithm through parallel operations. In addition, the motions of all joints of the human hand are estimated by the attitude and heading reference system algorithm using 12 IMU sensors. Because of the high accuracy and real-time performance of the hand pose estimation method, it is expected that the proposed system will be applied to interactive virtual reality tools.","PeriodicalId":299175,"journal":{"name":"2021 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133136073","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}

引用次数: 2

On the Formulation of Boundary Layer Functions 关于边界层函数的表述

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9621206

Assiya Zhumanazarova, Young-Im Cho

引用次数: 0

A HashChain-based Partial Matching Embedded Manifest for NDN 基于哈希链的NDN部分匹配嵌入式清单

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620987

Sae Hyong Park, Yong-Yoon Shin, Namseok Ko

引用次数: 0

Dynamic Spatial Area Design for Transportation Management at Smart Road Lighting Platform System in Korea 韩国智能道路照明平台系统交通管理的动态空间区域设计

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620787

Tae Wan Kim, Ingon Ryu, Hyunmi Lee, Jeong-Ah Jang

引用次数: 1

The Design of The Emerging 5G Using Hybrid GPON and XGS-PON Technology 基于混合GPON和XGS-PON技术的新兴5G设计

2021 International Conference on Information and Communication Technology Convergence (ICTC) Pub Date : 2021-10-20 DOI: 10.1109/ICTC52510.2021.9620906

Yohana Jayanti Aruan, P. T. Daely, Gino Sampedro, Jae-Min Lee, Dong-Seong Kim

{"title":"The Design of The Emerging 5G Using Hybrid GPON and XGS-PON Technology","authors":"Yohana Jayanti Aruan, P. T. Daely, Gino Sampedro, Jae-Min Lee, Dong-Seong Kim","doi":"10.1109/ICTC52510.2021.9620906","DOIUrl":"https://doi.org/10.1109/ICTC52510.2021.9620906","url":null,"abstract":"Emerging 5G systems will need to efficiently support new ultra-reliable low-latency communication (URLCC) and bandwidth requirements. To meet this challenging demand, an efficient and reliable communication system are required, which accelerated the development of high-capacity long-distance optical communication. High-capacity transport infrastructure is vital for an efficient telecommunication company's network operation. To fulfil the greater capacity, it is necessary to bring the optical fiber everywhere. They deliver 50% transport cost savings, perfectly matching the footprint of mobile cells, and have the high performance needed for the 5G world. In this paper, we considered high-speed and low latency passive optical network (PON) technology, due to its conciseness and cheapness and its reliable network topology structure, is widely acknowledged as the ultimate solution for fiber-to-the-home (FTTH). While the results show a promising improvement on the packet loss aspect, there is still a lot of room for improvement in other elements such as latency, jitter especially for some 5G use-cases, and for next-gen that comes after 5G.","PeriodicalId":299175,"journal":{"name":"2021 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"140 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123483622","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}

引用次数: 2