Artificial Intelligence Enhances Network Automation and Security

2019-11-29

With the advent of the 5G era, communication service providers have begun to save money from network transformation investments and generate revenue, providing customers with a better network experience. However, everything has two sides, and the constant emergence of problems has made the industry eagerly anticipate artificial intelligence. Artificial intelligence applied to networks can improve many technologies and solve many problems, such as closed-loop automation, encrypted traffic analysis, and radio optimization for 5G beamforming. The increasing complexity of network infrastructure, coupled with low latency and determinism associated with next-generation services, makes network deployment and management based on traditional network management methods and static policies impossible. These methods are not scalable and have a significant impact on business costs. Networks must be dynamic and automated. The advancement of artificial intelligence and machine learning technologies makes it possible to achieve closed-loop automation for network functions virtualization (NFV), which is crucial for remotely monitoring and managing thousands of network edge locations and billions of connected devices. When integrated with operations support systems, AI-supported network monitoring and predictive network analysis can detect network anomalies and failures, analyze root causes, and repair or transfer fault sources before network failures occur. Self-monitoring, self-management, and self-healing networks make it possible to dynamically adjust resource allocation and power consumption. According to incomplete data statistics, in 2019, over 90% of Google's network traffic was encrypted, and 80% of enterprise network traffic in the United States was also encrypted, which is a significant improvement compared to last year. However, it is still insufficient to cope with highly skilled hackers. To enhance network security, researchers envision using artificial intelligence to create a new defense system. By populating telemetry data such as packet sequences, packet boundaries, computational operation characteristics, and memory access patterns into an AI model, real-time intrusion detection, network isolation, and preventive measures can be effectively applied to encrypted traffic. Furthermore, artificial intelligence can also help achieve wireless coverage, capacity optimization, advanced traffic management, dynamic distribution of users across frequencies, dynamic radio resource management, beamforming configuration, management of multiple radio access traffic control, and service-aware resource management for network slices. For networks, artificial intelligence is like a golden key that can unlock many previously difficult-to-open locks. Although there are still many technical challenges to overcome, compared to having no direction to move forward, the obstacles along the way are not really problems.

RECENT AI NEWS

RECENT AI TOOLS