Aether Reaches Maturity and is Now Ready for Broad Developer Engagement
Mobile World Congress Demo Highlights Aether’s Unique Ability to Support Industry 4.0 Apps
Furthering its mission to seed the industry with innovative open source platforms to advance 5G and software-defined open networking, the Open Networking Foundation (ONF) announced that its Aether Private 5G + Edge Cloud platform, and related component projects SD-Core™, SD-RAN™ and SD-Fabric™ have now all been released under the permissive Apache 2.0 open source license.
Aether is the first open source 5G Connected Edge platform for enabling enterprise digital transformation. Aether provides 5G mobile connectivity and edge cloud services for distributed enterprise networks. Aether represents a complete, open 5G solution, addressing RAN through Core, democratizing availability of a robust and complete software-defined 5G platform for developers.
Recommended AI News: Mint Group and AvePoint Partner to Deliver Smart Compliance for Information Management
In just 2 years, Aether has achieved significant milestones and demonstrated numerous industry firsts:
- Aether was selected for the $30M DARPA Pronto project for building secure 5G
- Aether was deployed in an ongoing field trial with Deutsche Telekom in Berlin
- Aether has been deployed in over 15 locations, operating 7×24 and delivering production-grade uptime
- Aether is the only private 5G solution leveraging the benefits of open RAN for private enterprise use cases
In the process of achieving these remarkable milestones, Aether has matured to the point where it is ready to be released to the community for broad consumption.
Aether is built upon a number of world-class component projects that are each in their own right best-in-class. Today, all the component projects are also being open sourced, including:
SD-Core 4G/5G dual-mode cloud native mobile core
SD-Fabric SDN P4 Programmable Networking Fabric
SD-RAN Open RAN implementation with RIC and xApps
Demonstrating What is Possible
The ONF stand #1F66 at Mobile World Congress (MWC) will showcase an Aether deployment, demonstrated as a cloud managed offering optimized for enterprise private networks. In the demo, devices (UEs), such as mobile phones, cameras, sensors and IoT devices, can be aggregated into device-groups, and each device-group assigned a 5G slice and connected to specific edge applications thereby extending the slicing concept to individual applications and services. Each slice is attached to specific Industry 4.0 application(s), thereby creating distinct slices for different use cases and guaranteeing each slice secure isolation for security along with bandwidth, latency, quality of service (QoS) and resource assurances.
Two Industry 4.0 applications are demonstrated running over the Aether Private 5G using Intel technologies enabling ONF SD-RAN and SD-Core ranging from Intel Xeon Scalable processors, Intel vRAN accelerator ACC100, Intel Tofino Intelligent Fabric Processors, to software offerings such as Intel’s FlexRAN reference architecture, Intel Smart Edge Open, and the Intel Distribution of OpenVINO. The first is a security application built on Intel’s Distribution of OpenVINO toolkit, an intelligent AI/ML edge platform running on Aether 5G and leveraging Aether’s Industry 4.0 APIs to dynamically change the network slice to suit the application’s real time requirements for connectivity. The demo first creates an application slice for video surveillance, grouping together a set of streaming cameras. Next, whenever a human is detected in the field of vision for a camera, the solution automatically increases the 5G bandwidth for the impacted camera and instructs the camera to increase its resolution so a high-def recording can be made. With this approach, 5G wireless capacity is reserved for mission critical applications, and bandwidth is dynamically allocated precisely when and where needed. All of this is performed in real time without human intervention.
Anomaly detection is featured in the second Industry 4.0 application. Based on Intel’s Anomalib, an Aether application slice carries a mission critical video feed of a manufacturing / packaging line which is channeled into an anomaly detection edge-app. The edge app is trained using defect free data and uses probabilistic AI to detect anomalies like spoiled fruit (bananas). Given the time critical nature of detecting faults early, the app is built to work in a real time fashion over Aether to deliver results at line rate for typical industrial and packaging production lines.
Recommended AI News: 3D Systems & Saremco Dental AG Enter Strategic Partnership to Advance Digital Dentistry
Getting Started with Aether
It is easy to get started using Aether Private 5G. Aether can be deployed by developers on a laptop using Aether-in-a-Box, a simple pre-packaged end-to-end development environment including RAN through mobile core. By making such a complete solution available in a footprint that can run on a laptop, developers can get started with minimal friction. Developers can then organically grow the test deployment at their own pace into a fully disaggregated production-grade deployment, including production-grade RAN radios, disaggregated UPF edge processing and cloud-native mobile core. This makes it easy to get started with Aether, while assuring developers that Aether can scale to meet the needs of even the most demanding commercial applications.
Thanking the Community
Aether has been an amazing collaboration between ONF engineering resources and an active community that includes: Aarna Networks, AirHop, AT&T, Binghamton University, China Mobile, China Unicom, Ciena, Cohere Technologies, Cornell University, Deutsche Telekom, Edgecore Networks, Facebook Connectivity, Foxconn, Google, GSLab, HCL, Intel, NTT Group, Microsoft, Princeton University, Radisys, Sercomm, Stanford University and Tech Mahindra.
Recommended AI News: Nokia Launches Intelligent RAN Operations to Manage the Power of 5G with Machine Learning
[To share your insights with us, please write to sghosh@martechseries.com]
Comments are closed.