Samsung Using AI and ML in the Design of Next-Generation Wireless Networking Equipment

Summary

In this talk, Dr. Sung Young Choi from Samsung Research discusses the advancements in 6G technology, driven by AI and machine learning. With 5G still in its early stages, Samsung is already exploring 6G, which promises to connect over 500 billion devices by 2030. This will require innovations in signal processing, wireless communication, and network optimization. Dr. Choi highlights the use of AI to enhance wireless signal quality and the role of NI's software-defined radios in testing these innovations. The future of mobile networks will enable immersive experiences like extended reality and holograms, significantly boosting data rates and reducing latency.

Takeaways

• 😀 AI and ML are crucial in the development of sophisticated products like rockets and wireless networking equipment, as they enable new levels of automation.
• 😀 The future of wireless technology is focused on 6G, which aims to significantly improve mobile communications with hyper-connected experiences.
• 😀 By 2030, over 500 billion machines are expected to be connected to the internet, creating new technical requirements for mobile networks beyond those that connect people.
• 😀 6G networks will leverage AI and ML for tasks such as network planning and radio frequency (RF) optimization.
• 😀 Applications for 6G, such as immersive extended reality and mobile holograms, will demand innovations like one terabit per second peak data rates and reduced latency.
• 😀 One of the key challenges for wireless communications is the degradation of signal strength over distance, which can be mitigated with AI-driven solutions.
• 😀 AI-based non-linearity compensation (AINC) is an innovative technology that helps correct distortion in wireless signals, allowing for higher data rates and better spectrum utilization.
• 😀 AI models in AINC adaptively learn distortion patterns for each mobile device in real-time, improving the signal quality and performance of wireless communication systems.
• 😀 The use of software-defined radios (SDRs), such as NI's USRP products, allows Samsung to test their AI solutions in real network environments and refine their technologies.
• 😀 With the help of SDR technology, Samsung has achieved a 1.9 times improvement in transmission distance, showcasing the potential of these innovations to revolutionize wireless communications.

Q & A

What role does artificial intelligence (AI) and machine learning (ML) play in product development and testing?

AI and ML are increasingly used to define product features, automating processes and leading to faster market delivery of sophisticated products, like rockets. These technologies create new challenges for design and testing engineers, making accurate models of systems and their environments crucial.

What is Samsung's approach towards 6G, and what are the major trends they foresee for its development?

Samsung is already working on 6G technologies, aiming for deployments by 2030. They highlight two key trends: the increasing number of machines (500 billion by 2030) connected to the internet, and the rapid advancement of AI and ML technologies, which will enable new applications for hyper-connected experiences like immersive extended reality and high-fidelity mobile holograms.

How does 6G differ from 5G in terms of technological requirements?

6G will need to meet new technical requirements, such as expanding peak data rates up to one terabit per second and reducing air latency to 100 microseconds. It will also require advancements in areas like antenna technologies and higher-frequency bands to address signal propagation issues.

What are some of the innovations in AI and ML that Samsung is exploring for 6G?

One exciting innovation is AI-based non-linearity compensation (AINC), which helps correct signal distortion in mobile devices. This AI algorithm allows base stations to learn distortion patterns of mobile transmitters, improving signal quality, sustaining higher data rates over long distances, and optimizing spectrum usage.

How does AI-based non-linearity compensation (AINC) work?

AINC uses an AI algorithm to reverse-engineer distortion caused by mobile device amplifiers. It corrects the received signal at the base station by training the AI model during runtime, adaptively adjusting for each mobile device, thus enhancing data transmission quality and power consumption.

What challenges do wireless signal strength and distortion present in mobile networks?

Wireless signal strength decreases with distance, and higher power signals can lead to distortion, which makes decoding difficult. To overcome this, systems either reduce power to minimize distortion or use expensive distortion-free amplifiers, which are not always practical.

How does Samsung’s use of AI help in overcoming signal distortion challenges?

Samsung's AI technology compensates for distortion at the base station by analyzing and correcting the distorted signals received from mobile devices. This allows for cleaner signals, even over long distances, and ensures efficient spectrum use and better power consumption.

How does NI's software-defined radio (SDR) assist in Samsung’s 6G research?

NI's SDR platform allows Samsung to test their AI-driven ideas in real-world networks using a test bed. It emulates both mobile devices and base stations, enabling the testing of various configurations across different antennas and waveforms to improve transmission distance and signal quality.

What improvements have Samsung's AI-based techniques demonstrated in terms of transmission distance?

Samsung's AI techniques have demonstrated a 1.9 times improvement in transmission distance, showcasing the potential of their AI-powered solutions to enhance wireless communication capabilities for 6G.

How does the future of mobile networks look with the incorporation of AI and 6G technologies?

The future of mobile networks, with AI and 6G technologies, will lead to hyper-connected experiences such as immersive extended reality and mobile holograms. These advancements will require innovations in bandwidth, latency, signal propagation, and AI-driven optimizations to meet the growing demand for faster, more reliable communication.