High-Fidelity Emulation Advances AI-Native Networks

At Mobile World Congress 2026 in Barcelona, Keysight Technologies will demonstrate validation and emulation solutions designed to accelerate AI-native wireless systems and next-generation network deployment. The demonstrations span 5G-Advanced evolution toward 6G, AI-driven radio access network (RAN) optimization, and non-terrestrial network (NTN) device validation under realistic, reproducible test conditions.

5G-Advanced to 6G: Digital Twins and AI-Native Air Interfaces
As research transitions from 5G-Advanced to early 6G exploration, reproducible lab validation is critical for evaluating new spectrum bands, waveforms and AI-driven radio techniques. Keysight will showcase high-precision radio frequency (RF) digital twins that replicate real-world propagation environments within controlled test setups. These digital twins enable realistic channel emulation and closed-loop testing, supporting development of AI-native air interfaces and repeatable lab workflows.

A separate demonstration will focus on machine learning-based channel state information (CSI) compression. CSI compression is essential in systems with high-order MIMO and wide bandwidths, where feedback overhead can limit spectral efficiency. ML-based approaches aim to reduce signaling load while maintaining beamforming accuracy.

Additional software demonstrations include pre-6G FR3 air-interface performance validation, Integrated Sensing and Communications (ISAC) simulation-to-emulation workflows, and end-to-end 5G-Advanced eRedCap device testing. Keysight will also highlight quantum photonics test capabilities and low-noise multichannel RF signal generation for advanced research environments.

AI Validation Across Networks and Infrastructure
AI integration into wireless systems introduces new validation challenges, particularly when models operate in closed control loops or under dynamic traffic conditions. Keysight will demonstrate validation of AI and machine learning functions across devices, infrastructure and network layers under emulated real-world conditions.

Use cases include agentic AI for autonomous wireless test execution, high-order MIMO performance evaluation, GPU inference benchmarking and AI-driven network threat detection. These workflows aim to quantify latency, throughput and inference performance while ensuring that AI-assisted decisions meet operational reliability targets.

By combining network emulation with AI workload benchmarking, engineers can assess system behavior before field deployment, reducing performance uncertainty and integration risk.

AI-RAN Scaling and Automated Network Operations
In AI-RAN architectures, artificial intelligence is embedded within radio access network components for optimization, scheduling and automation. Keysight will present validation workflows that integrate testing, data collection, AI/ML training and benchmarking into a unified process.

Demonstrations will show how agentic AI, telecom-specific large telecom models (LTMs) and digital twins can automate 5G and emerging 6G network operations. Software demonstrations will include end-to-end AI and RAN validation, AI workload scheduling within AI-RAN environments and real-time coordination across terrestrial and non-terrestrial networks.

Such validation is designed to support dynamic coverage deployment, adaptive capacity management and improved network resilience under variable load and propagation conditions.

End-to-End Validation for Non-Terrestrial Networks
Keysight will also present end-to-end validation of NR-NTN devices and modems. NTN deployments introduce additional complexity due to propagation delay, Doppler effects, multi-orbit handover and positioning requirements.

Through network emulation, device and modem manufacturers can evaluate compliance with commercial NTN service requirements before launch. Demonstrations will include NTN quality of service evaluation, live-to-lab positioning, navigation and timing (PNT) testing, multi-orbit handover scenarios and operationally representative NTN scenario emulation.

By enabling repeatable lab validation of satellite-based connectivity, these workflows aim to shorten development timelines and reduce deployment risk for next-generation space-based communications systems.

Through high-fidelity emulation, digital twins and AI-enabled validation workflows, Keysight’s demonstrations at Mobile World Congress 2026 illustrate how connected intelligence can be developed and verified across terrestrial and non-terrestrial 5G-Advanced and pre-6G environments.