Since the turn of the 21st century, the job of network administrators has become increasingly difficult. With the advent of things like cloud computing, virtual machines, the internet of things (IoT), live video streaming, etc., monitoring and controlling networks is more and more complex. To add to that, users have higher expectations than ever before. The good news is that many of these tasks can be simplified, automated, and organized using artificial intelligence (AI) and machine learning (ML).
A Brief History of AI
The notion of artificial intelligence (AI) in the form of advanced robots was first introduced to the world in 1863 by Samuel Butler in an article he wrote entitled, Darwin Among the Machines.
During the 20th century, the idea was further popularized by celebrated science fiction writers such as Issac Asimov, Karel Capek, Arthur C. Clarke, and Robert Heinlein.
By the 1950s, the concept of artificial intelligence was so culturally assimilated that there was a whole new generation of scientists, mathematicians, and philosophers ready to bring AI out of the realm of fiction and into reality. A seminal paper entitled Computing Machinery and Intelligence was written in 1950 by a young British polymath named Alan Turing, in which he asks the question: Can machines think?¹
Although this technology has moved slowly over the past 71 years, AI and ML are starting to have real-world applications. In this article, we will discuss how AI/ML is enhancing network management.
What is AI?
Back in the 1950s, artificial intelligence was considered to be any task performed by a machine that was thought to require human intelligence. The modern definition has become much more specific. Francois Chollet, an AI researcher at Google and creator of the machine-learning library Keras, stated in a YouTube podcast:
Intelligence is the efficiency with which you acquire new skills at tasks you didn’t previously prepare for. Intelligence is not skill itself. It’s not what you can do; it’s how well and how efficiently you can learn things.²
Investopedia defines AI as follows:
Artificial intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. The term may also be applied to any machine that exhibits traits associated with a human mind, such as learning and problem-solving.³
AI systems typically show at least some of the following traits associated with human intelligence:
- Problem solving
- Social intelligence
What is ML?
Machine learning is a branch of AI that centers on the use of data and algorithms to emulate the way humans learn. MI is commonly used for the following:
- Banking software (to detect unusual transactions)
- Email filters (to expose spam)
- Internet search engines
- Phone apps (such as voice recognition)
- Websites (to provide personalized recommendations)
For AI to be successful, it needs ML to use algorithms to analyze data, learn from it, and make predictions without any explicit instructions. ML has recently advanced into more complex models such as deep learning (DL) and natural language processing (NLP).
Image Source: metal-am.com
The Role of AI and ML in Network Management
Good network management demands quick responses to any issues that may arise. While there are many ways an AI/ML system enhances a network, we will discuss three major areas. They are:
- Performance Monitoring
- Machine maintenance and Healing
- Cyberattack Detection
AI/ML is beginning to be integrated into network performance monitors (NPMs). While this technology is not widespread, organizations are becoming more and more aware of the benefits. AI/ML is beneficial in the following:
- Data processing and analysis: Networks deal with scads of data every day. Systems need to be in place to process and understand all this information. AI can sort through these large amounts of data and analyze it in real-time. ML can examine historical data to define network trends and then instantly recognize any problematic issues in the future without an in-depth analysis.
- Automatic problem solving: AI/ML can learn the common problems that affect a network. This technology can gather more information as a problem happens repeatedly and will be able to figure out how to solve it without human input. AI also is incredibly adept at eliminating harmful issues before they damage a network.
- Response customization: It is important to note that AI/ML does not work on its own; it needs to be trained. The plus side of this is that AI can be customized to analyze and respond to specific types of events. Thus, if something in particular you want it to handle on its own, it can be trained to do so.
Machine Maintenance and Healing
A network must be secure, with all resources available when needed; therefore, ongoing maintenance is crucial. Many organizations arrange for brief windows of downtime for maintenance during the workday—or add a third work shift for the technical teams. Both of these “solutions” are expensive and disruptive. Other organizations simply are unable to support regular maintenance and put it off or deploy redundant systems to ensure network availability.
An AI/ML program tackles maintenance in two ways.
- AI/ML intelligently automates patching and updates. It can understand and measure peak and non-peak usage and uses predictive analytics to establish a customized maintenance window that avoids service disruption. It will bring the system down, deploy the patch, and bring the system back up again.
People can do this too, of course. But AI/ML can be scaled up to carry out scheduled maintenance on thousands of devices across a large enterprise network without regular attention by the IT department.
- AI/ML systems closely monitor the performance of every device in the network. It can detect when specific services begin to hang or use more processing power than usual—way before users notice it. AI/ML can detect these problems and solve them before they cause any real impact on the network.
An AI/ML system learns its way around the network. It can detect things like denial of service attacks, attempts to force passwords, and other attempts at intrusion. According to Balbix, a leading cybersecurity company based in San Jose, CA:
AI and machine learning (ML) have become critical technologies in information security, as they can quickly analyze millions of events and identify many different types of threats—from malware exploiting zero-day vulnerabilities to identifying risky behavior that might lead to a phishing attack or download malicious code. These technologies learn over time, drawing from the past to identify new types of attacks now. Histories of behavior build profiles on users, assets, and networks, allowing AI to detect and respond to deviations from established norms.”⁴
As networks continue to become more extensive, more specialized, and more complex, efficiency management will become harder. The AI science fiction of the last century is steadily developing into reality. While our world may not have humanoid robots or self-learning spaceships yet, the current developments in AI and ML are certainly intriguing.
Check out the full range of Planet Technology last-mile IT networking support solutions, including extenders, switches, home automations, PoE, VoIP and wireless technologies.
1 A.M. Turning (1950) Computing Machinery and Intelligence. Mind. 49: 433-460
2 YouTube: Francois Chollet: Measures of Intelligence/Lex Fridman Podcast #120
3 Investopedia: Artificial Intelligence (AI)
4 Balbix: Using Artificial Intelligence in Cybersecurity