As the emergence of new radio convergence technologies is accelerating and the trend replacement cycle is getting faster, concerns about strengthening the capabilities of personnel in the radio field are deepening. As the Radio Broadcasting and Communications Education Center is taking the lead in radio convergence education, the changes in electromagnetic wave design and analysis methods brought about by generative AI technology have attracted attention.

On the 14th, the 4th Radio Future Leaders Seminar was held at DDP in Seoul, hosted by the Radio Promotion Association (RAPA) Radio Broadcasting and Communications Education Center.

This seminar was prepared to provide the latest status of the radio convergence industry and information on the capabilities required for employment to job seekers and employees in the radio industry, and was presented by SK Telecom, Autonomous A2G, Samsung Electronics, and Telit.

This event, which was held both online and offline, recorded approximately 140 registered participants, and brought together job seekers and employees from various fields of radio convergence to share insights on industry prospects and capabilities in each field.

In his opening remarks, Kim Gu-nyeon, the director of RAPA Radio Broadcasting and Communication Education Center, said, “This seminar is a place prepared for those working in the radio industry who need to continue to develop, and we will strive to create a sustainable educational environment.” He added, “I hope this will promote the development of radio utilization and become a place for harmony among those working in the radio convergence field.”

■ Park Hak-byeong, Senior, “ChatGPT electromagnetic wave design and analysis book will be established”
While the emergence of generative AI is expected to create convergence synergy in other industrial fields, there is not much known about how it can be used in real-world applications. Especially for developers and engineers in the radio field, it was difficult to imagine the connection between radio and ChatGPT.

Meanwhile, Samsung Electronics Senior Vice President Park Hak-byeong conducted a case study on electromagnetic wave analysis and design using generative AI since the beginning of this year and presented it at this seminar.

In the past, EMI was not a major issue in securing performance during the PCB design and manufacturing process, but today, EMI/EMC design complexity is increasing due to trends such as higher frequencies and faster signal speeds.

In the past, electromagnetic design and analysis were time-consuming and costly due to redesign and use of resistive components. Currently, simulation and design rule checks have been used to reduce time and cost in product development, but this requires capable and experienced experts, and guidance and education are necessary, which inevitably requires a lot of training time.

Senior Park predicted that in the future, when utilizing generative AI, it will be possible to dramatically reduce time and cost compared to current methods by utilizing LLM services for design guides and simulations and by utilizing in-house models, and that “while experts who are familiar with Matlab, Python, and APIs were able to do this in the past, now even general developers will be able to use generative AI to perform simulations and modeling in natural language, and results will be available little by little over the next year or so,” raising expectations for a phase change in generative AI in electromagnetic wave design and analysis in the future.

Even for EMI experts who do not know AI or coding, Senior Park explained the process of deriving analysis values and modeling through the process of △ defining tasks △ preparing learning data △ fine tuning using ChatGPT and GPT Builder during the seminar.

Senior Park predicted that “generator AI will take hold in electromagnetic design and analysis in the future,” and that “the future will be shaped by the changes and transformations that this technology will bring.” “We need to think about the capabilities of engineers,” he advised.

■ “In the AI era, the material of opportunity is ‘big data’”
SK Telecom Vice President Ha Do-hoon introduced the company’s Geovision Puzzle and emphasized, “Big data is the raw material for opportunities in the era of generative AI.”

The amount and value of information coming from mobile communication devices such as cell phones is limitless, and big data generated through base station data is expected to have infinite usability and potential. Radio-based big data not only provides real-time, time-based, and period-based information on floating population and population density, but also allows for the interpretation of various geographic-based data such as gender, consumption patterns, living areas, and lifestyles.

As the possibility of utilizing this in business opens up, Vice President Ha said, “With Geovision Puzzle, we can understand population residence, movement, and stay, and based on information on congestion and floating population, kickboard sharing platform companies can use it for kickboard relocation, etc.” He also explained an example of business utilization, saying, “We can understand population movement information and characteristics by subway exit, and target outdoor advertising and commercial district analysis.”

Currently, it is reported that geographic information big data is showing 70% of the total demand in the public sector, such as in the establishment of local government policies, and is being used for adjusting bus routes within the region, providing low-floor buses as a priority, and identifying tourist routes by country.

“Telecommunication-based big data generates 28 petabytes (28,000 terabytes) a day, so it is necessary to store and utilize this,” said Vice President Ha. “Anyone can become a startup CEO if they come up with utilization ideas one step ahead of others.”