ETHOS: ML-enabled RAN Testing & Validation Framework

Wed, 17 Jun 2026 00:00:00 +0000

Overview

ETHOS (Evaluation and Testing for Heterogeneous Open-Source RAN) is an NTIA (National Telecommunications and Information Administration) public wireless supply chain innovation project developing advanced testing and validation frameworks for ML-enabled radio access networks (RAN). The project creates standardized methodologies for evaluating real-world performance of 5G/6G software-defined RAN deployments.

Funding & Timeline

Funder: NTIA Public Wireless Supply Chain Innovation Fund
Award Amount: Multi-million dollar investment
Project Duration: Through 2028
Focus: RAN software validation and testing
Status: Active development

Project Objectives

1. Testing Framework Development

Standardized test methodologies for RAN software
Interoperability testing protocols
Performance benchmarking standards
Repeatability and reproducibility framework

2. Multi-dimensional Evaluation

Software Stability: Robustness and reliability assessment
Interoperability: Cross-vendor compatibility validation
Energy Efficiency: Power consumption optimization
Communication Performance: Throughput, latency, quality metrics

3. ML-enabled RAN Testing

AI/ML algorithm validation
Model robustness evaluation
Adversarial scenario testing
Online learning safety

Key Research Areas

RAN Testing Methodology

Controlled test environments
Reproducible test scenarios
Baseline performance metrics
Regression testing protocols

Open RAN Validation

O-RAN compliant testing
Vendor interoperability
Software modularity validation
Integration testing

ML Model Evaluation

Training data quality assessment
Model generalization testing
Failure mode analysis
Performance prediction accuracy

Field Performance Validation

Real-world deployment testing
Network measurement campaigns
User experience evaluation
Operational insights

Technology Stack

Testing Infrastructure

Programmable testbeds (RENEW, Houdini integration)
Controlled propagation environments
Real-time monitoring systems
Data logging and analysis

Analysis Tools

Machine learning evaluation frameworks
Performance analytics pipeline
Visualization and reporting
Anomaly detection systems

Rice Wireless Connection

Faculty Lead: Rahman Doost-Mohammady (Research Assistant Professor)

Key contributions from Rice Wireless:

RAN software testing expertise
ML algorithm evaluation methods
Testbed integration and operation
Open RAN protocol validation

Impact & Significance

ETHOS addresses critical challenges in 5G/6G deployment:

Supply Chain Security: Validated software provenance
Quality Assurance: Standardized testing methodologies
Innovation Acceleration: Faster vendor integration
Interoperability: True O-RAN realization

Industry Collaboration

Stakeholders

Network equipment vendors
Software developers
Wireless operators
Open RAN alliance members

Adoption Pathways

Testing as industry standard
Certification program potential
Technology transfer to vendors
Community feedback integration

Learn More: |

When Does AI Truly Help in Networking? Reflections from Maestro and Wixor

Fri, 15 May 2026 12:00:00 -0500

Speaker

Dr. Puneet Sharma — HPE Fellow, Vice President, and Director of the Networking and Distributed Systems Lab at HPE Labs.

Dr. Sharma leads research in Edge-to-Cloud-to-Exascale infrastructure, multi-cloud resource orchestration, AI for infrastructure, 5G/Wi-Fi, and security. Since joining HP Labs in 1998 after earning his PhD from the University of Southern California, he has driven innovations in software-defined networking (SDN), GPU virtualization, container orchestration, edge computing, Private 5G, and AI/ML systems, enabling major technology transfers across HPE business units. A recognised global thought leader, Puneet has authored 150+ papers, holds 100+ patents, co-authored IETF RFCs, and delivered keynote talks at IEEE and industry events. He is an IEEE Fellow, ACM Distinguished Scientist, and a two-time Tsinghua AI 2000 Most Influential Scholar. He earned his B.Tech from IIT Delhi.

Abstract

Artificial Intelligence has quickly become a powerful tool for addressing complex systems problems, and networking has been no exception. Yet, the journey from promising idea to practical impact often reveals unexpected challenges and deeper insights. This talk reflects lessons learned from applying AI to networking problems, drawing on experiences from our Maestro and Wixor projects. These efforts tackled issues such as Wi-Fi QoE orchestration and cellular (5G) resource scheduling, showing both the promise and the limits of AI in networked systems.

Several themes emerged across these projects: the critical role of formulating the right learning problem, the tradeoffs between model complexity and system usability, and the value of embedding domain knowledge into AI-driven solutions. Along the way, we also discovered that applying AI to networking is not just about better algorithms — it can reshape how we think about networks themselves.

This talk will take a step back to distill these experiences into broader reflections: when does AI truly help in networking, what pitfalls should we anticipate, and how can we design systems that are both intelligent and trustworthy?

Details

When: Friday, May 15, 2026, 12:00 PM – 1:00 PM CST
Where: O’Connor Engineering and Sciences Building, Room 406

5G | Rice Wireless