By Dinesh Rao, EVP & Co-Head of Delivery, Infosys
Generative AI will disrupt the way communication networks are built, tested, managed, and monetized in near future. As many know, Generative AI conceptually uses a Machine Learning and Large Language Model (LLM) technique called Generative Adversarial Network (GAN) which has two types of neural networks, Discriminator and Generator
While, Discriminator evaluates, filters and improve the data , Generator keeps improving its output based on the data it receives as feedback. It is clear that, any industry which deal with data extensively can benefit the maximum from Generative AI and hence Gartner forecasts an up of 30% in the Generative AI usage for multiple industries in the extremely near term of 2 years. Telecom industry uses petabytes of data daily and it is obvious that the impact of Generative AI would be multi fold there.
Driving new features ,customer experience and revenue
Hyper personalization in Media leading to premium services: Media services in future will be extremely preference driven. Present method of content being consumed by many will change to same content being personalized for individual by taking in customer preferences and personal choices with LLM’s reasoning capability advancements . The marketing ecosystem around media, like advertisements, product placement, branding etc can be hyper personalized and dynamic using Generative AI . This will help service providers create new and premium services.
Elevating Collaboration, Contact Center(CC) experience and generating Up-sell : Contact Center(CC) experience can be enhanced through contextual and insightful information in every interaction. Some examples are, real time processing of speech and related data for sentiment analysis, dynamic summary of conversation intent for better actions etc. Present chat-bots will be replaced by Generative AI driven Chat GPT like bots which can give accurate answers to the queries and bring the conversational capability in the interactions . The insightful conversation and data analysis can lead to more contextualized and personalized up-sell.
AI driven marketplaces : Modern day generalized catalogue marketplaces will yield to more intent driven marketplaces. Generative AI and AI can help in identifying user preference based on available data ,personal choices and search patterns. AI driven marketplaces can thus create dynamic services catalogue , generated content and even personalized UIs. Users can choose products, services like personalized mobile and home connectivity plans, personalized upgrades, usage pattern based service changes etc and in future even dynamically self-design products and services.
5G & Edge Network Monetization powered by AI : 5G and Edge IoT usecases like smart spaces, Smart Farming etc requires multiple data analysis like images , videos which can be done efficiently using AI and Generative AI’s use of synthetic data creation capability to improve accuracy . Edge Video Gaming contents will be AI created and metaverse will be powered by Generative AI. The usecases which requires closed loop feedback like autonomous vehicles, robotics and tele robotics can be accelerated using Generative AI based modelling and training.
Improving communication product engineering and network life cycle
While its evident that AI & Generative AI can boost the experience and revenue, it can also be a catalyst for improving product engineering and network life cycle as well.
Accelerating Chip and Radio Frequency component designs : Generative AI can produce optimal visualization for component placements in chip by generating multiple combinations and recommending best option faster than any known process. It can also generate synthetic data sets which can test the components better, for E.g.: generating signal data for Radio Frequency (RF) components for 6G component validation.
Improving quality through better test data generation : Testing the network for all conditions is difficult as you can never predict manually all real life scenarios which are dependent on the number of users, devices, and the data they are consuming. Generative AI can produce multitude of synthetic data and scenarios and choose the right ones for testing easily.
Fault prediction and efficient operations : AI can use historical data to create fault prediction models which can be applied on real time data for fault prediction. Fault prediction can lead to automated trouble shooting, proactive action and self-healing which will improve the network stability and reduce operational cost. The patterns can be machine learned (ML) and archived for future knowledge model creation.
Code generation : AI & Generative AI can take the current automated rule based code generation method to a more intelligent ,machine learned pattern based method. This will help to improve the quality of the code with every iteration and can also generate more optimal code.
AI & Network Optimization : Deep Learning capability of AI can help in optimizing the planning through analysing the infrastructure data and predicting the optimal capacity enhancements. . This can be achieved through AI analysing the data from multiple data sources like maps, dimensions of constructions, urban data , outside plant material costs and create the most cost effective design.
Security and AI : With software defined and cloud natïve method of network build, security is a critical factor to be considered. Incidentally, 3GPP V.18 focuses on virtualized network products and Management Function (MnF) , Mission critical enhancements like Phase 3 Security and privacy aspects of RAN & SA features. AI can enable better security management using dynamic control of role based and attribute-based policies. Generative AI can help to simulate the static and dynamic security vulnerabilities during the development phase itself and secure the network.
Moving Forward
We are moving into an AI driven era where data is analysed and used in every walk of life. Communication, Media, and Technology industry will also have to inevitably embrace this. As discussed, right use of AI & Generative AI in Network Product Engineering, Deployment, 5G & Edge Services ,Enterprise network services like Unified Communication & Contact Center will bring both efficiency and revenue growth. Engineering services providers who can bring in AI based solutions coupled with Communication domain expertise and software skills will play a pivotal role in realizing this vision and journey.