...
Ironically, as generative AI and LLM adoption becomes more widespread in many industries, telecom has lagged somewhat due to a number of valid factors. As was covered in the previous white paper, the overall industry challenges remain the same, that is the constant pressure to increase the efficiency and capacity of operators’ infrastructures to delivery deliver more services to customers for lower operational costs. The complexity and lack of a common standard understanding of network traffic data remains a barrier for the industry to speed the adoption of AI/ML to optimize network service delivery. Some of the challenges that are motivating continued research and adoption of intelligent networking in the industry include:
...
Data Privacy and Security: The ability to monetization monetize the data comes with a big caveat, which is that the use of customer data must be handles handled with care to ensure data privacy, security and regulatory compliance. This requires clear policies and security procedures to ensure anonymity, safety and privacy at all times. The good news is that AI can be used to address the growing threat of network vulnerabilities, zero day exploits and other security related issues with predictive analytics and threat analysis.
...
A natural question arises on how the the power of LLMs can be harnessed for problems and applications related to Intelligent Networking, network automation and for operating and optimizing telecommunication networks in general, at any level of the network stack. Datasets in telco-related applications have a few particularities unique to the industry. For one, the data one might encounter ranges from fully structured (e.g. code, scripts, configuration, or time series KPIs), to semi-structured (syslogs, design templates etc.), to unstructured data (design documents and specifications, Wikis, Github issues, emails, chatbot conversations).
...
- In-context-learning, prompting and retrieval augmentation techniques;
- Finetuning Fine tuning the models
- Hybrid approaches.
For finetuning fine tuning LLMs, unlike for regular neural network models, several specialized techniques exist in the general area of PEFT (Parameter Efficient Fine Tuning), allowing one to only fine tune a very small percentage of the many billions of parameters of a typical LLM. In general, the best techniques to achieve domain adaptation for an LLM heavily depends on:
...
In conclusion, while foundation models and transfer learning have been shown to work very well on general human language when pretraining pre-training is done on large corpuses of human text (such as Wikipedia, or the Pile [29]), it remains an open question to be answered whether domain adaptation and downstream task adaptation work equally well on the kinds of domain-specific, semi-structured, mixed modality datasets we find in the telecom industry. To enable this, telecoms should focus on standardization and data governance efforts, such as standardized and unified data collection policies and developing high quality structured data across a common definition and understanding.
...
Network Intelligence Enables Experience Monetization and Differentiated Operations
For a long time, operators have strived to realize traffic monetization on MBB(Mobile Broadband) networks. However, there are three technical gaps: non-assessable user experience, limited or no dynamic optimization, and no-closed-loop operations. To bridge these gaps, there is a need for an Intelligent Personalized Experience solution, designed to help operators add experience privileges to service packages and better monetize differentiated experiences. Typically in the industry, the user plane on the mobile core network processes and forwards one service flow using one vCPU. As heavy-traffic services increase, such as 2K or 4K HD video and live streaming, microbursts and extremely large network flows become the norm, it becomes more likely that a vCPU will become overloaded, causing packet loss. To address this issue, Intelligent AI supported 5G core network networks need to be able to deliver ubiquitous 10 Gbps superior experiences.
Service Intelligence Expands the Profitability of Calling Services
In 2023, New Calling was put into commercial use, based on a 3GPP specification, it can enhanced enhance intelligence and data channel (3GPP specification)-based interaction capabilities; it enabling enables users to use multi-modal communications, and helping helps operators construct more efficient service layouts. In addition, the 3GPP architecture allow allows users to control digital avatars through voice during calls, delivering a more personalized calling experience. One example where this can be seen as a business opportunity might be an enterprise using the framework to customize their own enterprise ambassador avatar to promote their brand.
O&M Intelligence Achieves High Network Stability and Efficiency
Empowered by the multi-modal large model, the Digital Assistant & Digital Expert (DAE) based AI technology could reduce O&M workload and improve O&M efficiency. It can reshape cloud-based O&M from "experts+tools" to intelligence-centric "DAE+manual assistance". Using the DAE, it is possible that up to 80% of telecommunication operator trouble tickets can be automatically processed, which is far more efficient than the manual processing it is for the most part today. DAE also enables intent-driven O&M, avoiding manual decision-making. Before, it commonly took over five years to train experts in a single domain, with the multi-modal large model it is now possible to for it to be trained and updated in merely mere weeks.
Thoth project - Telco Data Anonymizer
...
It is important to understand the current landscape of Open Source Software (OSS) projects and initiatives and how they came into existence to see how OSS is well positioned to address the challenges of Intelligent Networking. Several such initiatives have already laid down the ground work groundwork for building Network AI solutions, or are actively in the process of creating them. Building on these foundations, it is easy to envision the critical role open source software can play in unleashing the power of AI for the future generations of networking. Some of the required technologies required are unique to the Networking industry, and will have to be addressed by existing OSS projects on the landscape, or by the creation of additional ones. Some pieces of the technologies are more generic, and will need to come from the broader community of AI OSS. Here is a rough outline of the different layers of Networking AI and the sources for the required technology:
...
There is a lot of debate on the definition of what an "Open AI model" really means. While it is out of the scope of this paper to try and settle any of those debates, it is obvious that there is a clear need to create the definition of "open LLM". The sooner such definition is definitions are created and blessed by the industry, the faster innovation can happen.
...
In recent years The Linux Foundation Networking launched a set of Open Source Networking "Super Blueprints" that outline architectures for common networking use cases and are built using OSS technology. Those blueprints consists consist of collections of OSS projects and commercial products, integrated together by the open source community and documented for free use by any interested party. Several of these blueprints started incorporating AI technologies and that trend is expected to accelerate, as existing blueprints inspire additional AI-driven solutions. One promising area is intent-based network automation. There is current work on blueprints that use NLP and LLM to translate user intent, expressed in natural spoken language to network requirements, and generate full network configuration based on those requirements. Such an approach can significantly simplify existing network operation processes, and enable new automated services that can be directly consumed, commonly known as Network-as-a-Service (NaaS).
...
The future of Intelligent Networks and AI adoption in the telecom industry is in the hands of the individuals and organizations who are already contributing to projects and initiatives, and those who will join them. If you are involved in building and operating networks, developing network technology or consuming network services, you are most heartily encouraged to getting get involved. Engaging with OSS communities is a way to shape the future of networking. Your contribution could be small or large, and does not necessarily involve writing code. In fact the community is very much in need of contributors of white papers such as this one, evangelists and big thinkers who want to drive the realization of some really cool and useful leading edge technologies. Some of the ways to contribute include:
...