Community

Samsung Open Source Blog

Result :33

[Upcoming event] SOSCON Global India Virtual Open Source Conference (5G, IoT and Edge Computing) :Date: 3rd July 2021Time 8.30 am to 2 pm (IST), 12:00 pm to 5:30 pm(KST)Link to Register : http://bit.ly/ComSoc_July3 Fast emerging technologies such as 5G, IoT, and Edge Computing are changing the landscape of the technology industry and bringing significant value to end consumers. At the same time, open-source platforms are rapidly evolving and playing an important role in meeting the agile demands of the industry. Join us to learn from the industry-leading experts on the value, innovation, and benefits of deploying emerging technologies using open source platforms.You will also learn how to develop and deploy Kubernetes Operator, LF Home Edge and O-RAN SC XApp through developer sessions. June 17, 2021
Open Source Kubernetes for the next level of network virtualization We as Samsung Research have recently presented a research achievement as a global team work with Samsung R&D Institute Poland and Samsung Research America for improving the Kubernetes open source project into the next level of competency in IEEE INFOCOM 2021.What is a background of this research?Container technology has revolutionized the way software is being packaged and run. The telecommunications industry, now challenged with the 5G transformation, views containers as the best way to achieve agile infrastructure that offers high throughput and low latency for 5G edge applications. These challenges make optimal scheduling of performance-sensitive workflows a matter of emerging importance.How to solve this? Do not worry, we have Kubernetes open source project!Kubernetes is a de-facto standard open source project for container orchestration in IT industries and has been deployed to various cloud solutions including AWS, Google, Microsoft, IBM, VMWare, and so on. It is now in Cloud Native Computing Foundation (CNCF) of Linux Foundation, and is a very dynamic project since over 400 commits are being contributed to it every month. Samsung is also contributing various ideas into this project with a collaboration with community members.Okay, we have an open source which is free, but is that enough? It is not so simple.It is only logical that there are also many new capabilities in the scheduling area. Grabowski and Tyczynski described Kubernetes scheduler and its features available as of Kubernetes v1.16 [1]. It is able to execute the same decision tree policy as Google’s Borg does with the proper configuration [2], but it is not the default choice in Kubernetes, so it requires additional operator’s intervention to activate it. Several improvements have been proposed to Kubernetes scheduler to make it more suitable for public cloud, but they are not considering any network metrics, or examining any inter-application relations and node communication delays, or requiring information about the service function chains that is provided manually, which are all crucial in most of telecommunication use cases in terms of high throughput, low latency orchestration operated in a fully automated manner.In addition, the most critical drawback of the existing scheduler enhancement proposals is their incompatibility to the default scheduler of Kubernetes. No matter how excellent the proposed scheduler guarantees, it requires a custom configuration of selectively activating that scheduler for a specific use case – this is a huge blocker in the telecommunication industry in terms of wide deployment of the scheduler solution. When you consider of realizing the better world of technologies with your great idea using open source like Kubernetes, the most important criteria is its generality of deployment in various use cases that in the end should also guarantee compatibility to the existing open source mainstream.So what have we been researching and developing for general deployment of the industries?We design and implement NetMARKS – Kubernetes scheduler extender that uses information collected by Service Mesh to schedule pods based on current network metrics. Our solution does not conflict with current Kubernetes scheduler plugins or other extenders, and it can be used simultaneously with them to ensure backward compatibility. NetMARKS can automatically discover inter-application relations as well as schedule the application in a way to save the inter-node network bandwidth and to reduce the application response delay by extending the default Kubernetes scheduler including network metrics awareness, to ensure an efficient placement of Service Function Chains (SFCs). In this context, we propose a novel approach to collect network metrics for the scheduler using Istio Service Mesh.Kubernetes for a baseline scheduler, but what is Istio and how it works in NetMARKS?Istio has a sidecar-based service mesh design as shown in Fig. 1. This architecture assumes that there is an adjacent container, as a sidecar for every application that acts as a proxy and intercepts all the network traffic. Istio uses a powerful proxy named Envoy [3]. It provides multiple functionalities including mutual TLS, L3/L4 traffic filtering, HTTP L7 filtering and routing, telemetry, and advanced load balancing with circuit breaking and automatic retries. To benefit from those features, Envoy Proxy has to be configured properly and has to be provided with certificates. This is done by istiod, which is an essential part of Service Mesh controlplane that governs all Envoy Proxies. Istio is configured using Kubernetes CRD which allows a user to interact with it using Kubernetes objects.Fig. 1. An architecture of Istio Service Mesh.As Envoy participates in communication between services, it adds some delay to it. Based on [4], Istio generally adds around 3-ms latency for a mesh with 1,000 requests per second across 16 connections. This result can be improved [5] using a Container Network Interface (CNI) plugin, which supports SOCKMAP from the Linux kernel, such as Calico [6] and Cilium [7].As the communication is always intercepted by at least one Envoy Proxy, the traffic can be easily traced and monitored. Every Envoy instance can be configured to collect metrics on the traffic that it is passing. These metrics are further collected and stored by Prometheus [8], which can be deployed as a part of Istio control plane. Users can use the data in multiple ways depending on their needs, e.g. for observability and diagnostics using Kiali [9], which visualizes the structure of mesh and presents flows between components.NetMARKS is intended to be fully backward compatible with kube-scheduler. The most effective way to achieve this is to utilize Kubernetes Scheduler Extension mechanism. Extenders, including NetMARKS, are implemented as separate HTTP services that are contacted by the scheduler whenever a new pod needs to be scheduled. It is assumed that Istio is enabled for all of cluster namespaces, and Prometheus service is deployed as a part of Istio control plane.NetMARKS uses data collected by the Prometheus to make the scheduling decision. It retrieves two metrics from there: istio_request_bytes_sum and istio_response_bytes_sum, which contain the number of bytes that is transferred in requests and responses respectively. The application entity does not exist in Kubernetes, so it is defined by Istio as a set of pods residing in the same namespace and having the same app label.Both requests and responses metrics are summed in (1) to calculate F_(t_1,t_2)^(A,B), which is an average data flow between applications A and B in time period [t_1,t_2 ] in bytes per second. 〖req〗_t^(A,B) is the number of bytes in requests from application A to application B until t in bytes, and 〖resp〗_t^(B,A)- same for responses from B to A, with respect to 〖req〗_t^(A,B).(1)Metrics are collected periodically while keeping the average flow, since the start of data collection using (2).(2)Let F(t_i) be the current flow at a given time t_i, after expanding F_(t_(i-1), t_i)^(A,B) in (2) with (1). Then, its final numerical model can be derived as follows.(3)This data flow metric is used to assess each feasible node that has not been filtered out. The algorithm has two input parameters: a Pod to be scheduled and a Node. Current state of pods assignment to nodes is required to determine where to bind the new Pod, so NetMARKS retrieves this information via Kubernetes API.Operation "Get all traffic neighbors of Pod" selects data flow metrics, where namespace and application label of the Pod match either source or destination of the traffic.The result of the algorithm can be understood as a sum of the traffic bandwidths between the Pod communication peers assigned to the Node. To calculate the Score, the algorithm iterates over all the pods that intersect between a set of pods assigned to the Node and a set of Pod communication partners. Then, it adds the flow data to the final result if namespaces are matched.Scheduling example: Consider 4 applications:- F: farm, producing grain- W: well, producing water- P: pig farm, producing pigs, requiring grain and water- S: slaughter, producing meat, requiring pigsFig. 2 shows relation and traffic flows between applications. Assuming that F, W, and S are already running on nodes, the algorithm calculates the score in order to place pod with P as follows:- Node1: State={F},Score=30+50=80- Node2: State={W},Score=40+70=110- Node3: State={S},Score=60+100=160- Node4: State={},Score=0The node3 receives the highest score, because placing pig farm together with slaughter allows to use intra-node communication, the expected volume of which is bigger than others.Has NetMARKS been validated from rigorous experiments? Yes!!We validated our solution using different workloads and processing layouts. Based on our analysis, NetMARKS can reduce application response time up to 37 percent and save up to 50 percent of inter-node bandwidth in a fully automated manner. This significant improvement is crucial to Kubernetes adoption in 5G use cases, especially for multi-access edge computing and machine-to-machine communication.Publications: https://infocom.info/day/3 ↓(Day3: Containers and Data Centers session “NetMARKS: Network Metrics-AwaRe Kubernetes Scheduler Powered by Service Mesh”)What is the role of Samsung Open Source Group?We have been contributing our idea into various open source projects in technical domains such as telecommunication, IoT, cloud, robotics, artificial intelligence, and so on with various developers across Samsung Electronics and others from various companies and R&D institutes, in order to make the better technical innovation and the bigger ecosystem in an open collaboration. If you are interested in our recent activities, please visit our official website https://opensource.samsung.com/.Reference[1] W. T. Marek Grabowski, “Kubernetes scheduling features or how can i make the system do what i want,” in KubeCon + CloudNativeCon Europe, 2017. ↓[Online]. Available: https://www.youtube.com/watch?v=bbPcb2JuJPw[2] A. Verma, L. Pedrosa, M. R. Korupolu, D. Oppenheimer, E. Tune, and J. Wilkes, “Large-scale cluster management at Google with Borg,” in Proceedings of the European Conference on Computer Systems (EuroSys), Bordeaux, France, 2015.[3] “The envoy proxy.” [Online]. Available: https://www.envoyproxy.io/[4] “Istio performance best practices.” [Online]. ↓Available:https://istio.io/latest/blog/2019/performance-best-practices/[5] N. Kapocius, “Performance studies of kubernetes network solutions,” in 2020 IEEE Open Conference of Electrical, Electronic and Information Sciences (eStream), 2020, pp. 1–6.[6] “Project calico.” [Online]. Available: https://www.projectcalico.org/[7] “Project cilium.” [Online]. Available: https://cilium.io/[8] “The prometheus project.” [Online]. Available: https://prometheus.io/[9] “Kiali - service mesh management for istio.” [Online]. Available: https://kiali.io/* Original Contents : https://research.samsung.com/blog/Open-Source-Kubernetes-for-the-next-level-of-network-virtualization June 17, 2021
The ONAP Open Source Software community releases its 8th version, with significant contribution from Samsung The Open Networking Automation Platform (ONAP) is a comprehensive platform for real-time, policy-driven service orchestration and automation for networks and edge services. It is being developed as an open source software project with neutral governance under the auspices of The Linux Foundation. Started back in early 2017, it has recently made available its 8th release, codenamed ‘Honolulu’.This release brings cloud native and modular capabilities with deeper 5G and O-RAN integration for Communication Service Providers and Enterprise. The integration with O-RAN (another open initiative, aimed at bringing flexibility and more vendor choice to Radio Access networks) will enable RAN operators to optimize their radio network and reduce their cost of operation. Newly introduced 5G network slicing features in ONAP will open the door for new services to be offered by mobile network operators. The end-to-end 5G functionality in this release spans across the Radio, Transport and Core network domains which is crucial for providing the long anticipated 5G services of immersive AR/VR, Factory automation and connected vehicles.Over 200 developers from 25 organizations collaborated to produce Honolulu, the eighth release of the Linux Foundation ONAP project. Samsung played a significant role in contributions for this release, that amounted to it being one of the top-5 contributing companies:As a leader in wireless networks market, our engineers were able to use their expertise to improve ONAP and make it suitable for managing and orchestrating modern networks. Some of the main areas of contribution for Samsung in this release were Modularity, O-RAN integration, and Cloud Native Network Functions (CNFs).Modularity continues to be an important aspect of ONAP. It allows users to pick and choose the components they need for their specific use case and Honolulu continues to advance modularity. The ONAP Operation Manager (OOM) project is a key enabler of this modularity. It provides an easy method for users to indicate which of the many ONAP modules they would like to deploy in their environment. Samsung engineers contributed numerous improvements to this project during the Honolulu release. One of our engineers, Krzysztof Opasiak from Samsung Research R&D Institute-Poland (SRPOL), holds a ‘committer’ position in this project. During this release he demonstrated his expertise and dedication to ONAP quality by reviewing numerous incoming patches. His work was crucial for driving the Honolulu release to its successful completion.ONAP continues to be Service Management and Orchestration (SMO) of choice in O-RAN use cases. A key enhancement in the Honolulu release was broader support for the non-RT RIC functionality and the O-RAN A1 standard, both of which are implemented in the CCSDK and SDN-C projects. The O-RAN A1 interface provides a flexible way for RAN operators to manage wide area RAN network optimization reducing capex investment needs. Both the enhanced A1 interface controller and A1 policy capabilities are now usable in ONAP with a Near-Real-Time Radio Intelligent Controller (nRTRIC) thanks to the contributions of Samsung engineers. This functionality is also used downstream in O-RAN-Software Community (OSC) non-RT RIC project, strengthening alignment between ONAP & OSC.The Honolulu release has important updates to support Cloud Native Network Functions (CNF). The functionality includes configuration of Helm based CNFs and seamless day 1, 2 operations. The Configuration API allows a user to create, modify and delete Kubernetes. resource templates and their base parameters and the Profile API allows for sophisticated day 0 configuration. Samsung engineers leveraged their expertise in understanding the architecture of modern CNFs and contributed to the implementation of the interfaces between ONAP and Kubernetes.Peer code review is one of the pillars of successful open source software projects like ONAP. Samsung engineers contributing to ONAP are well versed in this methodology and practice it daily for both their internal and external software development work. During the Honolulu release our Samsung colleagues were among the leaders in code reviews, putting the company in the top-4 companies in terms of ONAP code reviews. Two Samsung engineers, Krzysztof Opasiak and Bartek Grzybowski were among the top-10 reviewers.One of the areas that is often overlooked in projects like ONAP is system integration and testing. While it rarely gets a mention during the announcement of a software release, the quality of the integration and testing infrastructure of a project could make the difference between break and make. We are proud to say that the dedication of Samsung engineers that worked on the ONAP Integration project made a significant impact and was an important factor in the success of this release. During the work on the Honolulu release, the ONAP community reflected on the work done during the previous release (‘Guilin’). Our efforts have been recognized by the community and a few of our colleagues won several awards.Pawel Wieczorek, Senior Software Engineer from the Open Source Group in Samsung R&D Institute-Poland has been awarded in the Automation Testing & Test Coverage category for supporting the Integration Project Technical Lead.Another award, the ‘ Scalability, Security, Stability and Performance (S3P) Improvements’ category has been granted to the whole Integration Team in which our company has a few active members:- Bartek Grzybowski, Senior DevOps Engineer from the Network Cloud Team in SRPOL- Lasse Kaihlavirta, Senior Software Architect from the Network Cloud Team in SRPOL- Pawel Wieczorek, Senior Software Engineer from the Open Source Group in SRPOLWe are honored and humbled to receive these awards from the community and are very proud of our colleagues achievements.Honolulu is the first release to follow the new release cadence that will allow for more frequent releases, up to three times a year. The process behind this release cadence is yet another area where Samsung team members led the way for the ONAP community. The next ONAP release planned for 2H of 2021 and will be named Istanbul. It will explore additional intent-based networking and AI/ML integrations.* Original Contents : https://research.samsung.com/blog/The-ONAP-Open-Source-Software-community-releases-its-8th-version-with-significant-contribution-from-Samsung June 17, 2021
Newest ROS2 Distribution "Galactic Geochelone" Released! May 2021 comes with many new and exciting changes: easing of COVID-19 restrictions in parts of the world, a warm and eventful Kentucky Derby, and the newest release of ROS2, Galactic Geochelone! Galactic is the 7th release of the new Robot Operating System (ROS) 2.0, built on DDS to be the production grade middleware for robotics with all of the amazing tools you came to love in ROS.ROS2 was redesigned from the ground up to handle non-ideal networks found in real-world production environments, communicate between multiple robots in a fleet securely, operate on low-power embedded systems, and leverage the real-time capacity of several different DDS vendors. Galactic is another step forward in its journey; extending existing capabilities and expanding new ones!Released on May 23rd, Galactic provides many improvements that have been in the works for the last year. The expected end of life of Galactic will be November 2022, at which time the next LTS distribution of ROS2 will have been released (codenamed “H-Turtle”). Most notably to Galactic, Cyclone DDS is now the default ROS MiddleWare (RMW) vendor. Eclipse Cyclone DDS [1] is a DDS implementation donated to the Eclipse Foundation by ADLINK, building off of their success with Vortex DDS. It has rapidly developed into a community favorite having a great out-of-the-box experience for users and true zero-copy IPC leveraging Eclipse’s Iceoryx project. Other DDS vendors that have Tier 1 support include eProsima’s Fast-DDS [2] and RTI’s Connext DDS [3]. Both of these DDS implementations are still available as binaries for easy install, unit testing support by Open Robotics, and only require setting a single environmental variable to utilize in your system.The ROS2 client library and its recursive dependencies have received a Quality Level 1 rating as set out in REP 2004[4], the highest level of quality possible. This translates to high unit and integration test coverage, strict policies regarding the quality of code introduced, process control, and stability. This quality level rating applies not only to rclcpp, but all of its dependencies (e..g rcl, rclc, rmw, etc), marking quality and stability up and down the entire middleware stack. Additionally, ROS2 pushes the default C++ version from C++14 to C++17.Two particularly great feature updates occurred in Galactic. It is now possible to externally configure Quality of Service (QoS) settings in ROS2 configuration files and rosbag2 has become substantially more reliable. Rather than statically defining QoS settings in your ROS2 code, you can now opt-in to allow for dynamically overriding these settings at run-time with your parameter file to try different profiles or reconfigure for unique use-cases [5]. rosbag2 was notably “buggy” in Foxy and older distributions. Even in high-stress networking environments, Galactic’s rosbag2 was able to successfully record 100% of messages (closer to 30% for Foxy). Many new features were also added including compression, playback control settings, commandline utilities, and regular expression support for topics to record and playback [6].These contributions, and many more, were brought to fruition due to the establishment of the ROS2 Technical Steering Committee (TSC) [7]. The TSC is a group of research and industry leaders who are contributing to the success of ROS2 and the related ecosystem, including Samsung, LG, Microsoft, Amazon, and many others. Samsung Electronics is a member of this TSC, leading the development of cutting-edge localization and mobile robotics technology.TSC Member Companies [8]The driving reason so many users choose ROS isn’t just the middleware, timing, logging, and launch systems-- it is the rich ecosystem of robotics capabilities built on ROS. The vast majority of your favorite projects have already been ported to ROS2, many of which have also received significant upgrades. Even now, major projects like Navigation2 (aka Nav2), MoveIt2, and ROS2 Control are receiving significant attention from industry and the community to create next-generation capabilities.Nav2 has had a substantial number of updates from Foxy to Galactic, including 3 new algorithms for planning and control, 7 new behavior tree nodes, and overhauls of external APIs [9]. New Hybrid-A*, A*, and Regulated Pure Pursuit controllers enable Ackermann and legged vehicles to have first-class support in Nav2. A new Theta* planner allows for planning to prioritize straight-line any-angle plans quickly through constrained spaces.The default behavior tree of Nav2 was redesigned to be more responsive to preemption calls and execute a minimum set of recovery behaviors to resolve an issue rather than iterate through all possible behaviors. New behavior tree nodes now make it possible to dynamically change planner, controller, or the new ‘goal checker’ plugins from an external application or in unique behavior tree contexts.Further, with the introduction of the new Costmap Filters API, we have support for Keepout Zones and Speed Restricted Zones out-of-the-box with tutorials for setup and use [10].The new Nav2 Waypoint Follower enables basic robot autonomy by following specific waypoints, stopping at each, and executing the new ‘Task Executor’ plugins to do anything from pausing, to taking a photo, to picking up a box with an attached robot arm. Complementary, a new action API was added enabling Navigate Through Poses behavior. Rather than point-to-point navigation or waypoint following, the Navigate Through Poses action will allow you to specify an arbitrary number of intermediate hard-pose constraints for your robot to navigate through and execute the path without stopping at each via-point. When combined with the Smac Hybrid-A* planner, this creates a kinematically feasible and smooth path between any number of intermediate points.A full list of new features in Nav2 can be found in [9].Looking into the future, Spring 2022 we can expect the release of the next LTS release of ROS2, HTurtle. By then, a whole new host of exciting features, capabilities, and algorithms will be available to build amazing robot systems using ROS2.Reference[1] https://github.com/eclipse-cyclonedds/cyclonedds[2] https://github.com/eProsima/Fast-DDS[3] https://www.rti.com/products[4] https://ros.org/reps/rep-2004.html#quality-level-1[5] http://design.ros2.org/articles/qos_configurability.html[6] https://github.com/ros2/rosbag2[7] https://docs.ros.org/en/foxy/Governance.html[8] All trademarks are the property of their respective owners.[9] https://navigation.ros.org/migration/Foxy.html[10] https://navigation.ros.org/tutorials/index.html* Original Contents : https://research.samsung.com/blog/Newest-ROS2-Distribution-Galactic-Geochelone-Released June 17, 2021
Embedded IoT 2021 - Intro of the research insight from the presentation based on LF Edge HomeEdge IntroductionEvery home is flooded with consumer electronic devices which have capability to connect to internet. These devices can be accessed from anywhere in the world and thus making the home smart. As per a survey the size of smart home market would be $6 Billion by 2022. As per Gartner report 80% of all the IoT projects to include AI as a major component.In conventional cloud based setup following two things need to be considered with utmost importance:Latency – As per study there is latency of 0.82ms for every 100 miles travelled by dataData Privacy – Average total cost of data breach is $3.92 MillionWith devices of heterogeneous hardware capabilities the processing which were traditionally performed at cloud are being moved to devices if possible as shown below.These lead us to work towards developing an edge platform for smart home system. As an initial step the Edge Orchestration project was developed under the umbrella of LF Edge.Before getting into the details of Edge orchestration, let us understand few terms commonly used in this.Edge computing - a distributed computing paradigm that brings computation and data storage closer to the location where it is needed, to improve response times and save bandwidthDevices : Our phones, tablets, smart TVs and so that are now a part of our homesInferencing - act or process of reaching a conclusion from known facts.ML perspective - act of taking decisions / giving predictions / aiding the user based on a pre trained model.Home Edge – TargetHome Edge project defines uses cases, technical requirements for smart home platform. The project aims to develop and maintain features and APIs catering to the above needs in a manner of open source collaboration. The minimum viable features of the platform are below:Dynamic device/service discovery at “Home Edge”Service OffloadingQuality of Service guarantee in various dynamic conditionsDistributed Machine LearningMulti-vendor InteroperabilityUser privacyEdge Orchestration targets to achieve easy connection and efficient resource utilization among Edge devices. Currently REST based device discovery and service offloading is supported. Score is calculated based on resource information (CPU/Memory/Network/Context) by all the edge devices. The score is used for selecting the device for service offloading when more than one device offer same service. Home Edge also supports Multi Edge Communication for multi NAT device discovery.Problem StatementAs mentioned above the service offloading is based on score of all the devices. Score is in-turn is calculated using CPU/Memory/Network/Context. These values are populated in the db initially when home edge is installed in the device. And hence these are more static values.Static Score is a function ofNumber of CPU cores in a machineBandwidth of each CPU coreNetwork bandwidthRound trip time in communication between devicesFor example say at time t1, device B has available memory ‘X’ and hence this value is stored in db. Say at time t2 device A requests device B for score and at t2 the available memory in device B is ‘X-Y’. But the score is calculated using ‘X’ which has been stored initially.Say in another scenario device A requests device B for service offloading. Device B starts executing the service. But say in the mid of execution moves away from the network. In such cases it would be unsuccessful offloading.We could see in the above situations that the main purpose of Home Edge is gone for a task.And these lead to the proposal of dynamic score calculation which is more effective and efficient.Properties of a home based Distributed Inferencing SystemWe propose an effective method for score calculation. The proposed method for score calculation is based on following points:1. Model Specific Device Selection2. Efficient Data Distribution3. Efficient Churn Prediction4. Non Rigid and Self ImprovingBased on the above four properties we defined a feedback based system that can help us achieve optimal data distribution for parallel inferencing in a heterogeneous system.High Level SystemFollowing diagram shows the high level architecture of the system. Let us take a use case and see the flow of tasks. For example when a user queries to Speaker “Let me know is any visitors today?”. The User query is indigested by the query manager to understand the query. The user query is to identify the number of visitors who visited today. The Data Source needs to be identified for query execution say the Doorbell Camera/Backyard CCTV. Then the model required for this purpose is identified and the devices which have the service (model). Based on the past performance, static score, trust of the device and number times previously the model has ran on the device, the devices are selected. Priority of the devices are calculated and based on the same the data is split among the devices. The result from the devices are aggregated to calculate the final output.Understanding ScoreTotal Score is a function of Static score and Memory usage of the deviceChanges in Score can help us predict estimated runtimes at the current timeUnderstanding Time EstimatesTime Estimates are calculated based on previous time records, average and current scoreData Points which are stored :1. Average Score2. Average time taken per inference3. Number of times a model has been deployed on a device4. Success probability of a model running on a device – TrustEstimating runtime on DevicesRuntime estimation is done post sufficient data points being collected in the systemDelta Factor - A liner relationship factor to approximate the change in runtimes when score changes is calculatedEstimated time - using the Delta factor, Average Score and Average Runtime and the Current score of the deviceUnderstanding Churn Probability EstimationLogical buckets for every 5 minute frames is created to know the availability of devicesIn case a bucket is polling a device for nth time and the was available k times before the new probability would be1. (K+1)/(n+1), if device is available2. K/(n+1), if device unavailableCalculating PriorityPost estimation of runtimes, The probability of devices moving out of the system are checked and devices whose probability falls low are not kept in the consideration listAll devices in the consideration list are sorted based on the product of the trust and inverse of runtimesTop K devices out of N are selected and data is distributed among themData DistributionData on the devices is distributed in the inverse ratio of the estimated runtimes to minimize the total time the system takes to make the complete inferenceFeedbackPost execution the actual runtimes and scores of the devices are used to adjust the average in the databasesIn case of failure next device is selected and the trust value of the device is adjusted in databaseTest Run on SimulatorAim – To demonstrate heuristic based data distribution among devicesTo Test – Note the devices that work best for a model, checking the selection irrespective of device scoresWeb based simulator was developed to demonstrate the proposed method. Following is a sample screenshot of the simulatorIn the due course of our work we also found few models execute better in specific devices. To check up on this we had run two use case – Camera Anomaly detection and Person identification. And we found that the models used in both the scenarios gave better execution results in different devices as shown below.* Original Contents at here : https://research.samsung.com/blog/Embedded-IoT-2021-Intro-of-the-research-insight-from-the-presentation-based-on-LF-Edge-HomeEdge June 17, 2021
Samsung Open Source Companions 2기 모집 ( ~5/9) Samsung Open Source Companions 2기를 모집하고 있습니다. 많은 관심 가져주세요!> 지원하기 April 20, 2021
K-호미로 배우는 오픈소스! "공유"와 "협력"의 이야기 공유경제는 아는데 오픈소스는 잘 모르겠다면?K호미로 공유텃밭을 가꾸며 오픈소스 전문가가 직접 알려주는 오픈소스!문과생도 가능?! 나이와 전공에 관계없는, 코딩을 몰라도 참여할 수 있는 공유와 협력의 오픈소스 생태계 이야기, 한번 들어볼까요?0:12 오프닝0:56 오픈소스가 뭐죠?1:57 오픈소스는 왜 하나요?3:44 퀴즈1) 오픈소스가 쓰이지 않는 분야는?4:22 퀴즈2) 오픈소스를 이용할 때 주의할 점은?5:15 오픈소스의 미래 모습6:26 문과도 참여 가능한가요??7:12 요약정리※ 본 촬영은 사회적 거리두기 2단계 방역수칙을 준수하였습니다. April 2, 2021