Innovation & Strategies
AI, a tool to improve 5G services
AI is a tool to improve 5G services Artificial intelligence (AI) represented by “machine learning” and “deep learning” has done remarkable work in sectors such as industry, robotics, healthcare, and Internet protection. I am convinced that AI can also help telecom operators to optimize their investments, reduce costs and improve the efficiency of operation and maintenance, including precise 5G network planning, forecasting of capacity expansion, self-optimization of coverage, and planning of cloud network resources. 5G, the next generation mobile communication technology, will dramatically improve the processing capabilities of AI and narrow the performance gap between different service platforms like the cloud. Its deployment will also improve the bandwidth, latency and density of wireless data communications. For example, “Applications in the age of AI and 5G: The high throughput, low latency and high density of devices enabled by 5G technology allow the use of AI in connected devices. This technology will emerge not only to support low-latency localized applications, but also to enable new AI-based services. AI inference on devices, such as smartphones, cars and surveillance, will become increasingly powerful. 5G and AI with this symbiotic relationship will together bring unprecedented services and amenities into our daily lives. Products and chip designs with AI functions will continue to grow rapidly.
Telecom experts, MNOs and operators are convinced that the convergence between AI and 5G will have a huge impact on the development of future generation networks. Previously, many researchers were afraid to use AI-based algorithms for lack of understanding of the machine learning process. Those days are over. Today, with much more information available, many researchers are adapting their knowledge. Thus, they enrich their tools with models, algorithms and practices based on AI. This is especially true in the 5G world, where even a few milliseconds of latency can make a difference.
AI in 5G networks
It is expected that 5G will improve data collection, automate information flows and reduce latency in addition to enabling more IoT devices. In industrial automation, 5G networks may eventually replace wired connections, even in the most demanding applications such as motion control and high-speed vision systems. Another major benefit of 5G for the AI industry is the flexibility to implement new applications and services. Thanks to the IoT networks of sensors and the interconnection of devices which will allow operators to be made aware of problems, failures, etc. Thanks to the collection and analysis in real time of data ranging from performance to activity personnel, logistics – as well as predictive analytics.
Optimizing 5G with AI
Optimizing 5G with AI
Minimize latency by minimizing storage in base stations
A reliable 5G network requires extremely low latency. However, latency increases with distance and congestion of network links. Today base stations have a limited storage size, so they must learn to predict user needs by applying a variety of artificial intelligence tools. With these tools, each base station will be able to store a small but adequate set of files or content. This is an example for which our future networks must be predictive. Artificial intelligence is becoming crucial to optimize this type of problem in the networks of telecom operators.
Many applications can be made possible by the combination of AI and the faster data rates supported by 5G networks. These include:
- Improved mobile communication network and services.
- The provision of mobile communication services is becoming more and more complex.
- Deploying 5G networks is much more difficult than previous generations as it requires upgrades to radio, edge, transport, backbone and cloud infrastructure that would ideally be optimally managed with AI to meet the complexity associated with 5G networks and the billions of IoT Devices that these networks can support.
IA and 5G represent a symbiotic combination and activate each other. 5G is based on automated optimization based on AI algorithms. This allows us to infer that 5G is too complex to function with the static and manual planning optimization of previous generations of mobile networks. 5G networks change their topology dynamically, responding to changes in traffic. The more the optimization works, the better the network will perform in terms of spectrum and energy use. For example, 5G relies on SON (Self-Organizing Network) technology. SON continuously measures radio parameters such as path loss and throughput for different frequencies, drawing conclusions about optimal performance and automatically adapting network parameters, therefore mobile network operators are currently testing AI due to the increased complexity of the network which requires an intelligent and automated approach. Finally, it is expected that more than half of mobile operators will integrate AI into their networks by the end of 2020. According to the main conclusions of a study conducted by Ericsson, with 132 service providers worldwide, the AI is already being integrated into cellular networks mainly because of a potential reduction in capital expenditure. Service providers testing AI are focusing on several areas for potential improvements.
We can conclude that the application of AI in telecommunications is still in its infancy. The coming years will be a critical period for an intelligent transformation of operator networks. With its gradual maturity, AI will be introduced in various telecom scenarios to help operators move from the current human management model to the autonomous automatic management mode and truly achieve a smart transformation in the operation and maintenance of the network. However, with rapid development, AI is addressing these challenges as it becomes a promising potential support for issues associated with the 5G era, and will lead to revolutionary concepts and capabilities in communications.
Innovation & Strategies
by Lou CHEVREUX the 02/10/2020Read more