While major global companies are fiercely competing to dominate the AI market, artificial intelligence (AI) technology is also serving as an opportunity for telecommunications companies to enter a new phase. Recently, domestic and foreign telecommunications companies are using AI technology in various ways along with non-telecommunications strategies to improve customer service, improve network operation efficiency, and discover new revenue-generating opportunities.

According to the 'Trends in AI in Telecommunications 2024' study released by NVIDIA on the 21st, telecommunications AI solutions are already improving both revenue and cost reduction. Since the commercialization of 5G in 2018, the number of 5G subscribers has steadily increased worldwide, but as the growth rate slows, telecommunications companies' de-telecommunication strategies are interpreted as a way to maximize corporate value in conjunction with AI strategies.

Of the 400 telecommunications professionals, 53% agreed that “AI adoption will be a source of competitive advantage.” Additionally, 31% of respondents said they have invested in at least six AI use cases by 2023 and will increase their investments this year. Additionally, 48% added that “achieving improved customer experience (CX) is a key goal.”

■ Telecommunications-specialized AI ' Telco LLM' discovers new businesses

With the recent formation of the enterprise-specific AI market, it is expected that telecommunications companies will be able to possess unrivaled strengths. Specialized AI is being pioneered in various service areas such as space, defense, education, logistics, finance, and e-commerce, and compared to general-purpose AI such as ChatGPT, it is designed to be applied to specific fields and exhibit high performance in those fields, making it more accurate and faster in processing.

Telecommunications companies are expanding new businesses based on telecommunications data that they have accumulated for over 20 years. It seems inevitable that telecommunications companies will jump into the AI industry based on their strong capital power, such as the telecommunications industry logic and knowledge required to learn and train AI.

In particular, the 'Telecom Large Language Model (TELCO LLM)', an AI large language model specialized in communications, is a language model learned based on massive data in the communications field. Compared to general LLMs, it is specialized in the communications field, so it is expected to be highly utilized in the communications business and service fields such as AICC (customer support contact center) and security. For example, since natural language processing is possible, it is possible to implement AI chatbot services, information extraction, document creation, and translation services.

SKT established a joint venture for joint development and business cooperation of the telecommunications industry 'Telco LLM' at MWC2024. The goal is to develop a multilingual LLM that supports languages around the world, starting with five languages: Korean, English, Japanese, German, and Arabic, through the 'Global Telco AI Alliance (GTTA)' with Deutsche Telekom, E&G Group, Singtel Group, and SoftBank.

SKT emphasized, “If Telco LLM is developed based on our ‘A.X,’ telecommunications companies around the world will be able to flexibly create generative AI services, such as AI chatbots, tailored to each country’s environment for a total of 1.3 billion subscribers.”

■ Improving network operation management efficiency with AI

AI technology in the communications field is expected to be applied not only to new services and marketing areas, but also to innovate network operation infrastructure.

Telecommunication companies are actively challenging the development of next-generation network infrastructure technologies such as AI-based network operation efficiency and 6G simulators and AI-applied open LAN optimization. AI analyzes massive communication data to detect and predict network abnormalities and assist in problem solving and optimization. It can learn patterns of network traffic surges in specific areas to predict overloads and take measures to expand capacity or distribute traffic in advance.

On the 27th, LG Uplus joined hands with AWS and Samsung Electronics at MWC2024 to develop an automation function that optimizes the timing of 5G equipment expansion using AI. With this technology, users can use Samsung Electronics' cloud native core software built on the AWS cloud to predict the optimal equipment expansion time with AI and automatically expand it to provide seamless services.

Also, at MWC2024, major companies such as Ericsson, Nokia, Samsung Electronics, MS, and Arm have established the 'AI-RAN' alliance to advance RAN (Radio Access Network) technology and mobile networks by integrating AI into cellular technology. AI-RAN has the advantage of applying AI to wireless RAN and efficiently utilizing RAN's radio resources.

The AI-RAN Alliance said it aims to “paving the way for new economic opportunities for telecom companies through AI, facilitated by 5G and 6G systems.” Participating companies will advance RAN capabilities through AI, improve spectral efficiency, integrate AI with RAN processes, and deploy AI services at the network edge through RAN.

Meanwhile, Nvidia said in the report, “The application of AI in the communications field is still in its early stages, and joint development of solutions with AI experts will be important for successful application of AI.”