#cybersecurity | #hackerspace | Network Operators Must Automate Their Way to 5G

Source: National Cyber Security – Produced By Gregory Evans

Artificial intelligence (AI) has crept into many aspects of our lives, from Google Maps to Alexa to our cars. At the same time, we rely more and more on internet of things (IoT) devices such as smart locks and dishwashers. So much so that the trinity of AI, IoT and 5G continues to be the main topic of conversation at many telecom industry events. 5G, with its increased connectivity, practically zero latency, network slicing and virtualization, promises to support the prolific increase in network traffic and expanding number and types of mobile devices and applications.

That increase inevitably will be accompanied by the accelerated growth of cyberthreats. This means that 5G’s promise will come with a cost: increased requirements for security—To make sure devices aren’t compromised or misbehave, to protect the very infrastructure that IoT relies on, to protect customers online and to ensure that the network delivers when and where it’s needed.

So how can network operators meet these demands? The answer lies in automation, powered by machine learning—more specifically, in closed-loop automation (CLA).

CLA is not a perfect solution. But as it learns, it offers an increasingly effective and continuous assessment of real-time network conditions, resource availability and traffic demands to identify network congestion and malicious traffic, whether it comes from within the network or from external sources. CLA uses all this data to learn how to optimize its tools and processes and implement solutions to combat damaging situations in real-time.

According to the recent Telecom Trends Report on CLA, the network operators surveyed indicate that closed-loop automation is a required tool in the 5G network operator’s kit.

CLA for Improved IoT Security

Recent IoT-based attacks have illustrated that hackers are getting craftier in their attack methods and are always looking for new ways to infiltrate consumer devices. In addition, most consumer IoT devices lack built-in security. It is, therefore, no surprise that it takes less than 43 seconds on average for an IoT device to be compromised once it is exposed to the internet.

Furthermore, some of the largest known distributed denial of service (DDoS) packages have used malware that targets and controls IoT devices, turning them into bots used for coordinated attacks. For instance, going back to 2016, the infamous “1Gbps” DDoS attack was initiated by an IoT botnet in the Mirai malware package. It took down the DNS provider’s server and resulted in large portions of the internet being unavailable on the East Coast of the U.S., including Twitter, Netflix and CNN and other sites. Since then, IoT botnet attacks have become even more sophisticated.

With the number of vulnerable IoT appliances growing in every aspect of life (home, health, transportation), soon to be enabled by 5G, the scope of attacks is primed to expand. To meet the demand for real-time detection and resolution, it is vital that network operators tap into the power of automation to deliver comprehensive IoT security.

CLA for Enhanced Quality of Experience

Today, network operators are starting to look beyond key performance metrics and familiarize themselves with the different quality indicators for various services. For example, quality of experience (QoE) for video service is based on stalls and resolution drops, but for car-to-car communication, QoE is based on message delays and successful message delivery.

CLA based on machine learning is imperative for analyzing complex network behavior and identifying the factors that affect the quality of experience. Once these factors are identified, automated corrective measures are taken to ensure that service quality meets the required levels depending on the type of service that has been impacted. For example, the video quality of a cellphone may drop for a number of reasons, such as low compute resources, radio congestion or congestion on the server-side. Machine learning-based CLA can identify the exact cause for the adverse video quality and apply the correct enforcement measure.

The Current CLA Adoption Landscape

As operators prepare for the transition from 4G to 5G, some are placing more focus on automation during the transition phase, while others feel CLA will be most effective only once 5G is fully implemented. Either way, the benefits of CLA are apparent to network operators. However, according to the aforementioned report on CLA, many still aren’t automating their way to 5G just yet because they either lack the in-house skills needed for CLA implementation or simply don’t understand how to employ CLA.

The good news is that network operators can realize the potential of CLA by tapping into a variety of CLA tools and expert guidance directly from independent software vendors.

Today, network operators have the unique opportunity to fortify their subscribers’ IoT security and QoE at the network level. Only operators that choose to proactively leverage CLA and automated decision-making will be positioned to best reap their efficiency and security benefits in the dawning 5G age.

Source link

The post #cybersecurity | #hackerspace |<p> Network Operators Must Automate Their Way to 5G <p> appeared first on National Cyber Security.

Leave a Reply

Your email address will not be published. Required fields are marked *