Here are three Keysight launches at MWC that are worth a second look
On the ground at MWC, where our team was all week, there was one unifying theme across keynotes, demonstrations, and hallway conversations: AI.
AI workloads are taking over the network and telcos, data centers, and enterprises are looking to tap into AI-native solutions to accommodate the shift.
Naturally, the announcements that followed were one too many. I decided to take a divide and conquer approach as I sift through the hype and find the best MWC test and measurement announcements to bring to you.
Starting with Keysight, which, like several other T&M suppliers in the space, made a flurry of announcements at the Congress that, besides reflecting the broader AI network push, also showed the company’s effort to deepen its role in AI-centric network deployments.
An unified AI-RAN testing workflow with Samsung, powered by Nvidia
With Nvidia spearheading the AI-RAN (Radio Access Network) push through the AI-RAN Alliance, field trials and tests are already underway, paving the path for early commercialization. According to Nvidia’s State of AI in Telecommunications report, 66% respondents are investing in or considering deploying AI for RAN operations.
While only 37% said deploying it is an investment priority for them, 53% confirmed using AI for boosting spectral efficiency in the RAN, and 50% in multi-tenant infrastructure with the goal to simultaneously run AI and RAN workloads.
Workgroups at the AI-RAN Alliance are working towards establishing standardized test and validation frameworks that’d allow operators and vendors to accelerate validation and reduce deployment risks. To that end, Keysight and Samsung Electronics jointly developed an AI-RAN testing platform to allow engineers to validate AI-RAN modules with a single workflow.
Crucially, the workflow aims to eliminate the fragmented approach around data collection, model training, and benchmarking that exists today. With data coming in from distributed environments, comparing results or validating PHY-level functions are especially challenging.
Keysight’s solution uses the Keysight’s AI RAN Simulation Toolset to provide an end-to-end workflow that replaces the fragmented process with automatic data generation, AI/ML model training, and repeatable benchmarking capabilities. The result is, engineers can confidently validate RAN behavior using clear insights before field testing starts.
The solution, demoed at MWC, is powered by the Nvidia Aerial Testbed for over-the-air AI-RAN research, the Nvidia Aerial Omniverse Digital Twin — and AI/ML models jointly developed by the three parties.
Chief technologist at Keysight, Balaji Raghothaman, said in a statement, “AI in the RAN only delivers value when it can be validated with confidence. Working with Nvidia and Samsung, we’re demonstrating a streamlined, automated workflow that unifies data generation, AI/ML training, and benchmarking, helping operators and vendors accelerate deployment, reduce risk, and make more informed decisions as they introduce AI-driven RAN capabilities.”
A prototype for AI-driven uplink performance, with MediaTek
Building on the RAN work, Keysight collaborated with MediaTek to create a prototype for advancing AI-driven uplink optimization and model life cycle management for RAN.
Maintaining consistent uplink performance has become a thorny challenge for operators as coverage conditions continue to vary across regions. While AI models are designed to help with this, a model trained to work in one environment often fails to perform in the discrete conditions of another environment. These issues weigh heavily on network operations.
The solution, demoed at MWC, seeks to address this and fast-track the transition to AI-native networks. The prototype is designed to boost uplink performance with “RAN-assisted AI decision-making” while also evolving AI models as required through retraining and over-the-air updates, Keysight said.
VP and GM of Wireless Test Group, Peng Cao said, “Together with MediaTek, we have demonstrated a novel AI-enabling-AI technique which helps address increasing demand on mobile networks’ uplink performance, capacity and consistent end users’ experiences driven by AI and AR applications. This involves training and testing the AI-enhanced transmitter model in the UE with high-fidelity real-world training datasets generated from Keysight’s RF digital twin channel emulation solutions with 3D ray tracing capabilities.”
Pre-6G interoperability validation with Ericsson
Although not specifically an AI release, this one is significant as AI-nativity is a big piece of 6G.
With 6G standardization work underway, Keysight demoed a pre-6G troubleshooting and interoperability validation tool with Ericsson that seeks to provide a full-stack solution for “pre-6G” systems. Operators can validate early 6G implementations, with capabilities like interoperability validation between prototype base stations and root cause analysis.
The solution taps Keysight’s WaveJudge Wireless Analyzer Solution that analyzes signals, surfacing threshold deviations and other KPI insights, allowing for faster triage and resolutions.
“Early validation is critical to advancing next-generation wireless. By working closely with Ericsson, we are enabling early exploration of new ideas alongside the standards development process. Keysight’s WaveJudge solutions help identify issues quickly and accelerate our customers’ path from research to real-world deployment,” said Lucas Hansen of Keysight.
