Arm at the Edge: Telco and IoT Akraino Blueprints debut at ONS 2019
By Tina Tsou, co-chair, Akraino Edge Stack Technical Steering Committee & Enterprise Architect, Arm. A version of this post also appeared on the Arm Community blog
Last week at Open Networking Summit in San Jose, there was a lot of buzz about the Akraino Edge Stack Project. Launched in 2018, Akraino Edge Stack was developed to create an open source software stack that supports high-availability cloud services optimized for edge computing systems and applications.
Significant progress has been made in this community since the launch, and many member companies showcased their Akraino blueprints last week. Arm is very active in the Akraino community and is excited about the Akraino Edge Stack Blueprints.
Here are some additional details about the four blueprints that were shown:
1) SDN Enabled Broadband Access (SEBA) on Ampere-based servers
For users of Virtual broadband access (XGS-PON which is a higher bandwidth, symmetric version of GPON), Ampere demonstrated SDN Enabled Broadband Access (SEBA) validation on Arm. The SEBA blueprint in the Akraino Edge Stack Project, can run applications of Virtual broadband access – vOLT access and aggregation for 5000 edge locations. There are three servers per POD, with x86 and Arm (with 8-16 cores each). The power consumption is restricted to less than 1 kW and includes NEBS compliance and 48V DC. The Ampere eMAG-based server delivers competitive performance per watt with 32 Armv8 CPU cores at 3+ GHz with Turbo.
The foundation of the SEBA validation on Arm demo is built on Integration Edge Cloud (IEC) blueprint family. The Integrated Edge Cloud (IEC) enables new functionality and business models on the network edge with benefits such as lower latency for end users, less load on the network since more data an be processed locally, and full utilization of the computational power of the edge devices.
VMWare proposed multi-cloud xConnection to interconnect different kinds of clouds of IT and Telco. The IEC had several deployment models that each support different business case such as telco/enterprise edge cloud (ex. MEC or brand office data center) or telco/enterprise remote edge locations (ex. SD-WAN, IoT Gateways). The demonstration included Ubuntu, Kubernetes, and Calico on Arm.
2) ELIOT (Edge Lightweight and IoT) Blueprint on Huawei IoT Gateway
We were excited to partner with Huawei to demonstrate the ELIOT: Edge Lightweight and IoT Blueprint Family. The ELIOT blueprint was designed to service the need of many diverse business applications that require a converged IoT gateway, and Enterprise WAN edge use of SD-WAN solutions or universal CPU (uCPE). The IoT gateway can be deployed in smart cities, smart homes, connected farming, agriculture logistics industrial, and Industrial IoT. SD-WAN, WAN edge, uCPE are designed to be used for hybrid WAN, hybrid cloud deployment, and BYOD. ELIOT is very scalable, from 1 single unit to 10K, 100K, 1000K, or more. ELIOT also supports diverse types of edge applications in many industries and market segments, including but not limited to: telcos, operators, service/cloud providers, medicine, smart cities, industrial IoT, home, and enterprise. The cloud/network infrastructure for ELIOT includes containers, Kubernetes, and the Kubernetes ecosystem. At the same time, the blueprint is designed to use lightweight operating systems and container runtime environments.
“The IoT gateway and enterprise edge SD-WAN gateway are two great examples of computing or power resource constrained edge nodes. The ELIOT project provides end-to-end light-weight open source blueprints for deploying and managing these use cases, built on any processor architecture, to foster a vibrant ecosystem around edge gateways in both hardware and software.”
–– Bill Ren, Chief Open Source Liaison Officer, Huawei
3) Micro-MEC Blue Print from Nokia
Nokia, Arm, and other ecosystem partners within the Akraino/LF Edge community have formed an edge blueprint for a Smart Cities platform called Micro-Mec (uMEC) targeted for a range of use cases. Nokia is using an Arm-based Marvell CN83xx uMEC design to show a highway traffic monitoring application. The uMEC enables new functionalities and business models on the network edge. The benefits of running applications on the network edge include lower latencies for end users, less load on the network since more data can be processed locally, and better security/privacy since sensitive data need not be transferred to a centralized location.
All these new services support the business case for building new high-speed networks which in turn enable new things. The uMEC has several deployment models that each support different business cases including:
- Fixed installation as part of 5G NR base station enabling new services that require low latency such as AR/VR.
- As an extension of the previous, the “Smart City” deployments have additional functions such as weather stations, cameras, displays, or drone charging stations. The control software for these functions would run on the uMEC.
- In an Industry 4.0 use case set, the uMEC is deployed as part of a 5G network and would provide a platform for running services for the factory floor.
- In a train, the uMEC could collect and store surveillance camera data for later uploading.
“Nokia is very excited to demo the the new edge blueprint for operators and smart cities called Micro Multiaccess Edge (uMEC). It demonstrates how even a very small ARM- based system can implement a Smart City use case, and complements the industry standard Open Edge hardware that is used in the Radio Edge Cloud and the 5G Radio Access Network,” said Tapio Tallgren, Project Technical Leader for uMEC and Akraino Technical Community
4) Tencent Future Network Lab Connected Vehicle Blueprint
The Connected Vehicle Blueprint focuses on the MEC platform, which is the backbone for V2X (Vehicle to Everything) applications. The blueprint can be used in multiple use cases, including but not limited to:
- Accurate Locations: The blueprint is designed to deliver more than 10X finer granularity in location. GPS is 5-10 meters level location which can be improved to <1 meter.
- Smart Navigation: Real-time traffic information update reduces the latency from minutes to seconds to provide the more efficient path for drivers.
- Safer Driving: Insight into potential risks which can’t be seen by drivers’ eyes.
- Reduced Violation in Traffic Rules: Let the driver understand the traffic rules in some specific area. For example, change in lines for an upcoming and narrow street, avoiding opposite way drive in the one way road, and avoiding carpool lanes for a single driver.
The blueprint can be flexibly deployed in multiple environments, including bare metal, virtual machine, and container based on commodity hardware (Arm/x86 server). The major software component of this blueprint is Tars, a Linux Foundation microservice Framework project. Tars can be deployed in Arm and x86 servers. For more detail information for Tars, refer to the github link.
Learn more about connected vehicle blueprint on Google Drive or read the blog post here.
“Tencent continuously promotes network innovation from various application perspectives. We believe that application-based network innovation promotes a stronger ecosystem, which brings tremendous benefits to our customers and stakeholders.” –– Zhang Yun Fei, Director of Future Network Lab, Tencent
“Open source is an important technical strategy for Tencent. As both a platinum member and board member of the Linux Foundation, Tencent continuously makes contributions to the Linux Foundation and its projects. After the Tars project contributed to the LF in 2018, and recent Akraino blueprint, Tencent will continue to contribute several new open source projects focused on cache and configuration. We welcome additional participation from more Linux Foundation member companies!” — Xin Liu, Linux Foundation Board member, and Tencent General Manager
We are excited about these great Blueprint demonstrations and the others shown last week. I want to acknowledge the Akraino Edge Stack Project Technical Steering Committee, PTLs, committers, and contributors, for their support with our activities at the conference. It is truly a team effort! Special recognition goes to:
- Aaron Byrd, AT&T
- Matt Taylor, Ampere
- Trevor Tao, Arm
- Gabriel Yu, Huawei
- Tapio Tallgren, Nokia
- Robert Qiu, Tencent
- Xinhui Li, VMware
- Ken Yi, DiDi
- Wenhui Zhang, PSU
