AI-Native Wireless Access

AI-Native Wireless Access is a Sub-Project of the OCP AI Computing Continuum Project. Its vision is to make wireless access an inherently AI-native, autonomous, and distributed platform that embeds accelerated computing into the network edge, and its mission is to accelerate adoption of AI-native wireless access by defining open, common, and sustainable specifications, interfaces, and best practices under OCP. The sub-project connects the wireless last mile into the continuum through three interdependent activities: the Autonomous AI-Native Radio (AuRU), the Radio Abstraction Interface (RAI) with its containerization framework, and the Security Control Module (RU-SCM), each delivered as an open specification or reference design.

AI-Native Wireless Access

Get involved

Steering Committee Representative

Sean Varley

Project Leads

Subscribe to mailing list

About This Sub-Project

Artificial Intelligence and Machine Learning for Manufacturing, Operational, and Value-Driven Autonomation in xG or WiFi Wireless Access

AI-Native Wireless Access is a Sub-Project of the OCP AI Computing Continuum Project.

Vision

To lead the global transformation of wireless access, making it an inherently AI-native, autonomous, and distributed platform that seamlessly embeds accelerated computing into the network edge, thereby unlocking unprecedented levels of performance, efficiency, and innovation across all industries.

Mission

To accelerate the adoption and efficient integration of AI-native wireless access technologies within the AI Computing Continuum by defining open, common, and sustainable infrastructure specifications, interfaces, and best practices under the Open Compute Project (OCP) framework.

Goals

The shift from deploying cloud data centers to AI data centers requires augmenting traditional computational infrastructure (CPU) with accelerated compute capacity (XPU) that can scale up and down on standard platforms to execute AI models. This transformation within the data center is well underway, with new inference architectures still emerging to drive performance. To deliver on the range of training and inference use cases across industries, accelerated compute infrastructure will be geographically distributed at many locations, from centralized hyperscale data centers out to where data is created and consumed by users.

The work is split into three activities with the goal of connecting the wireless access last mile seamlessly into the AI Computing Continuum. Each activity depends on the others, so they are handled jointly:

  • Autonomous Radio (AuRU) — necessary architectural and design implementations, including augmentation supporting manufacturing processes
  • Radio Abstraction Interface (RAI) and the containerization of apps that can run in the radio unit environment
  • Security Control Module (RU-SCM) and the application of security best practices to a radio unit of aggregated parts

Work Products

The sub-project will develop open specifications and reference designs for the core components of the AI-Native Wireless Access solution. Key work products include:

  • Autonomous AI-Native Radio (AuRU) Reference Specification — detailed design and implementation guidelines for AuRU, including augmentation of manufacturing processes
  • Radio Abstraction Interface (RAI) Specification — a standardized interface definition and containerization framework to enable third-party application deployment on the radio unit
  • Security Control Module (RU-SCM) Design — best practices and architectural guidelines for applying security to a radio unit composed of aggregated parts

Scope

The sub-project focuses on the design, development, and implementation of an AI-native Wireless Access (AWA) solution, leveraging AI/ML for automation (operational, industrial, network-level, and so on) in xG or WiFi wireless access environments. The scope encompasses three core technical activities defining the hardware platform and firmware stack:

  • Autonomous AI-Native Radio (AuRU) : design of the AI-accelerated autonomous radio unit
  • Radio Abstraction Interface (RAI) : specification of the interface and its associated containerization framework
  • Security Control Module (RU-SCM) : development of the security control module for aggregated radio units

The ultimate goal is to create specifications that support the distributed accelerated compute required for AI models at the network edge, with a focus on incorporating Radio Units (RUs) within the scope of the AI Computing Continuum.

Resources

This Project is open to the public and we want to welcome all those who would like to be involved.

Disclaimer: Please do not submit any confidential information to the Project Community. All presentation materials, proposals, meeting minutes and/or supporting documents are published by OCP and are open to the public in accordance to OCP’s Bylaws and IP Policy. This can be found on the OCP OCP Policies page. If you have any questions please contact OCP.

This new subproject was refactored from the initiative formerly known as Evenstar.

Vision & Mission

To lead the global transformation of wireless access, making it an inherently AI-native, autonomous, and distributed platform that seamlessly embeds accelerated computing into the network edge, thereby unlocking unprecedented levels of performance, efficiency, and innovation across all industries. We will accelerate the adoption and efficient integration of AI-native wireless access technologies within the AI Computing Continuum by defining open, common, and sustainable infrastructure specifications, interfaces, and best practices under the Open Compute Project (OCP) framework.

Documents

– Charter TBA

Workstreams

TBA

Regular Project Calls

This project meets TBA. This call is open to the public (please check the call calendar).

– Call Calendar

Recordings from Past Calls

To be added as the new project begins

Past Evenstar Events

Specs and Designs

Link to the Specs and Designs page (none yet)

Documents

Evenstar Open RU (Radio Unit) Charter

Introduction: The purpose of this charter is to establish the framework for an Open Compute Strategic Initiative focused on the collaborative design and development of a modular Open RU (Radio Unit) that is supportive of relevant xG technologies, and tailored for specific market segments . This project aims to deliver a multi vendor sup[ply chain for Open 5g radios fostering innovation, interoperability, and cost-effectiveness in the rapidly evolving field of 5G radio technologies.

Radio Design Principles: Set clear targets for high-performance specifications, ensuring that the Open RU meets or exceeds industry standards for data throughput, low latency, and reliability. Implement cost-effective design strategies without compromising the overall performance of the Open RU, encouraging efficient use of resources and materials. Explore economies of scale and collaborative procurement strategies to make the Open RU accessible and affordable for a broad range of users, including smaller operators in select market segments. Collaborate and partner with Open Air Interface (OAI), 3GPP, TIP, and Open RAN (O-RAN).

The following principles are essential but will be leveraged in this order of priority: Modularity: Design an Open RU with a modular architecture to allow for easy customization and upgradability including a variety of software, firmware, FPGA and hardware options. Establish and specify standardized interfaces and communication protocols to ensure interoperability. Ensure that key components, such as radio transceivers, antennas, and processing units, are interchangeable and can be upgraded independently.

Power Efficiency: Prioritize the development of power-efficient design principles for the Open RU, considering the environmental impact and energy consumption. Explore innovative technologies and techniques to optimize power usage without compromising performance, addressing the unique challenges of 5G networks.

Technology Agnosticism: Develop the Open RU to be agnostic to G technologies, supporting seamless integration with various generations of mobile networks (2G, 3G, 4G, and 5G). Encourage compatibility with both standalone and non-standalone 5G deployments.

Community Engagement: Encourage a vibrant and diverse community of engineers, product developers, and 5G enthusiasts to collaborate on the project. Foster an open and inclusive environment for sharing ideas, code, and designs.

Documentation and Education: Develop comprehensive documentation to facilitate easy adoption and implementation of the Open RU. Provide educational resources and workshops to empower developers and engineers interested in contributing to or utilizing the Open RU.

Security and Reliability: Prioritize developing secure and reliable features, considering the unique challenges and requirements of 5G networks. Implement best practices for cybersecurity and resilience against potential threats.

Ecosystem Support: Collaborate with ecosystem partners, including telecom operators, equipment manufacturers, and software developers, to ensure broad industry support for the Open RU. Promote the development of third-party applications and services that enhance the functionality of the Open RU. Establish and maintain an open baseline test suite and test benches for manufacturing and prototype creation. Specifically for FPGA use this is critical.

Governance: Project Leadership: Establish a leadership committee responsible for overseeing the project’s direction and ensuring alignment with the goals outlined in this charter. Encourage representation from diverse backgrounds to ensure a holistic approach to decision-making.

Contributor Guidelines: Define clear guidelines for contributors, following OCP code of conduct, OCP code review processes, and OCP contribution requirements. Implement a meritocratic system to recognize and reward valuable contributions to the project.

Transparent Decision-Making: Conduct open and transparent discussions for major decisions, involving the community in decision-making processes. Utilize public forums, mailing lists, and regular meetings to keep the community informed and engaged.

Conclusion: Open RUs serve as a foundation for the collaborative development of a multi-vendor supply chain delivering a modular open radio unit , supportive of relevant xG technologies for diverse market segments. By fostering an open and inclusive environment, we aim to accelerate innovation, drive interoperability, and empower a global community of contributors. Together, we will shape the future of 5G radio technologies and products.

Regular Project Calls

Every other Friday at 7-8am PT

Call Calendar

These meeting are recorded via audio and video. By participating you consent that these recordings may be made publicly available. Any presentation materials, proposals and meeting minutes are published on the respective project’s wiki page and are open to the public in accordance to OCP’s Bylaws and IP Policy. This can be found at http://www.opencompute.org/about/ocp-policies/. If you have any questions please contact OCP.

AI-Native Wireless Access calendar