Samsung Research (SR) is thrilled to announce that the research paper titled “Network GDT: GenAI based Digital Twin for Automated Network Performance Evaluation” has won the Best Paper Award at the Institute of Electrical and Electronics Engineers International Conference on Communications (IEEE ICC) 2025, one of the most prestigious flagship conferences in the field of telecommunications. This year’s conference was held in Montreal, Canada, from June 8 to June 12, bringing together global thought leaders, researchers, and industry pioneers to showcase cutting-edge advancements in communications technology.

Authored by a cross-institutional team of researchers from Samsung R&D Institute India-Bangalore (SRI-B), SR, Indian Institute of Technology Madras, and George Mason University in the United States, the award-winning paper presents a pioneering approach to automating the performance evaluation of AI/ML features in Beyond 5G (B5G) networks using Generative AI-based Digital Twins.

Best Paper award received by authors in IEEE ICC 2025 held in Montreal, Canada

In modern telecommunications, integrating AI-based features into networks has become essential but evaluating their impact under real-world conditions is often complex, time-consuming, and resource-intensive. The Network GDT framework solves this problem by leveraging conditional Generative Adversarial Networks (cGANs) to generate realistic synthetic network data and simulate performance scenarios.

A key innovation introduced in the paper is the Digital Twin Augmenting Condition (DTAC) framework. DTAC encodes new AI/ML features and environmental conditions, enabling the digital twin to adapt to various real-world scenarios. This dual-phase framework allows AI-based solutions to be proactively tested in a virtual, yet highly accurate, environment before field deployment, significantly reducing risk, cost, and deployment time.

The platform was rigorously evaluated using a congestion prediction use case. It achieved over 97% prediction accuracy in scenarios involving critical KPIs such as Downlink Physical Resource Block (PRB) utilization, Radio Resource Control (RRC) user count, and throughput, demonstrating the robustness and fidelity of the generated digital twin. Such results underscore the system’s potential to become a cornerstone in the AI life cycle of telecom networks. Beyond its technical novelty, Network GDT addresses a major operational challenge in the telecom industry: how to scale, evaluate, and deploy AI-driven enhancements quickly and confidently in dynamic network environments such as v-RAN, C-RAN, and O-RAN.

“Congratulations to the authors for winning the Best Paper Award at IEEE ICC 2025, an outstanding achievement that brings global recognition to the cutting-edge research coming out of SRI-B,” said Mohan Rao, Corporate Vice President and Head of SRI-B. “At SRI-B, we foster a collaborative and innovation-driven environment that empowers teams to tackle complex industry challenges with bold ideas and scientific rigor. This success highlights our continued focus on AI-native networks and impactful academic partnerships. We are proud of our researchers for setting a benchmark in AI-led telecom innovation on the world stage.”

From left to right: Dr. Madhan, Ashish Jain, CVP Mohan Rao, Dr. Sukhdeep Singh, Swaraj Kumar

“Network GDT is a technical leap that enables real-time, risk-free validation of AI features using Generative AI before they ever touch a live network,” said Dr. Sukhdeep Singh of SRI-B. “This award is a proud moment for our global Samsung team and our academic partners who pushed the boundaries of possibility in telecom AI.”

“We engineered the Digital Twin with cGAN and DTAC not just to be accurate but also to be adaptable to evolving network scenarios,” added Dr. Peter Moonki Hong of SR. “Winning this award is a recognition of our vision to democratize network intelligence through scalable AI platforms.”

Receiving the Best Paper Award at IEEE ICC 2025, recognized globally as a leading platform for communications research, marks a significant achievement for the authors and their institutions. This commendation is a testament to Samsung’s vision of empowering networks with advanced AI/ML technologies, highlighting the organization’s commitment to leading innovation from research to real-world impact.

Swaraj Kumar presenting the paper in the conference