Home Wi-Fi technologies have followed a predictable cycle of upgrades to support new radio technologies and remote management capabilities. But upgrade cycles are accelerating, as service providers are realizing that improving their customer’s in-home Wi-Fi experience is a critical requirement in a world where broadband competition is increasing and margins are gradually beginning to decline.
A major part of recent and ongoing upgrades to home networking devices has been the integration of centralized, cloud-based management, along with machine learning and Artificial Intelligence (AI) capabilities to more efficiently understand home networking requirements and consumption and optimize services within the home to ensure consistent user experience.
At the same time as machine learning and cloud-based management of CPE are being introduced, Wi-Fi 6 is quickly being integrated into the next generation of home gateways and routers to improve physical layer throughput for the gigabit age. With a focus on providing a theoretical maximum of 10Gbps of throughput, one of the primary goals of Wi-Fi 6 is to ensure that a customer’s Wi-Fi network will not impede the delivery of high-bandwidth, latency-sensitive services such as cloud gaming, 8k video, and cloud VR services.
This combination of machine learning, AI, and Wi-Fi 6 is giving service providers a toolset they’ve never had before to not only improve how they deliver Wi-Fi services to home broadband subscribers but also how they can tailor and customize their broadband and Wi-Fi offerings to each user, based on their consumption requirements. The net result is a powerful new revolution in-home Wi-Fi.
Using AI to Reduce Latency
One of the areas where the combination of machine learning and Wi-Fi 6 is most effective is in reducing latency throughout home Wi-Fi networks, thereby enabling high-value services such as cloud gaming and cloud VR services.
Wi-Fi 6 introduces OFDMA (Orthogonal Frequency Division Multiple Access), which allows routers and access points to divide multiple channels into smaller allocations called resource units (RUs.) Each RU can then be divided into yet smaller channels, with traffic earmarked simultaneously for multiple devices. The net result is a reduction in latency for connected devices and an increase in the aggregate throughput across the entire wireless network.
With the granularity of OFDMA, service providers can apply the network slicing capabilities they’ve had on their own carrier and enterprise Wi-Fi networks and now apply them to home networks. In this scenario, a VLAN can be allocated to a specific flow of packets associated with a particular user, device, or service. Using machine learning, the VLAN, or network slice, can be dynamically altered based on the upstream and downstream packet flows, as well as the latency requirements not only of the end device but also of the particular service. The VLAN can also be prioritized in terms of packets and security, so that very latency-sensitive traffic from services such as cloud gaming always receive priority. Additionally, the VLAN slice can be used to provide complete traffic isolation between other network slices using the same Wi-Fi network, or between different Wi-Fi devices using the same network slice.
Some operators are already taking advantage of this type of service by promoting a dedicated gaming WAN, with a network slice that is guaranteed traffic prioritization and minimal latency not only throughout the home network but also across the carrier’s network, as well. It is a form of WAN acceleration provided specifically for gaming services, for which operators can charge an additional fee.
These same operators are also delivering constant improvements to the performance of their gaming service through the use of machine learning. As bandwidth consumption ebbs and flows throughout the day or during specific hours of the day, the operator can re-route the high priority traffic automatically and dynamically. This can occur both within their own network and also within the home, where less-congested Wi-Fi channels can be quickly identified and used to deliver the latency-sensitive traffic.
These capabilities go above and beyond existing DPI (Deep Packet Inspection) features included in most home routers today. While DPI gives routers the ability to quickly identify and mitigate security threats to the home or IoT environment, it typically doesn’t offer the ability to characterize packets into specific services or applications. Artificial Intelligence features integrated directly into the home CPE can add an additional layer of security by anticipating security threats before they occur, based on learned traffic models. AI on the home gateway can identify devices on the home network, and detect incoming threats and identify the websites and server addresses from which they originate.
AI can also be used to provide enhanced parental controls that go beyond the baseline features of managing Internet access based on a user profile or device MAC address. AI can be used to provide Web content filtering and text analysis to identify emails, social media posts and also text messages (when on the home Wi-Fi network) that include explicit content or are potentially harmful or threatening.
All of these AI-based features can help change the business model of home broadband in positive ways for service providers. First, the enhanced security and service identification capabilities help create operational efficiencies in the form of reducing service calls and expensive truck rolls to diagnose and troubleshoot common in-home Wi-Fi issues. Second, the advanced security, service and application identification and prioritization, and content filtering capabilities give the service provider a clear differentiator from their competitors. Also, as broadband price competition intensifies, service providers with a menu of advanced services designed to maximize the customer’s experience can hold fast on their pricing by pointing to the benefits of these unique services.
Using Slicing and AI to Improve IoT Security
While speed, throughput, and minimal latency are the primary requirements for high-end gaming and video services, security and stability are becoming equally critical in-home networks as the number of sensors and home security devices increases, and with them the threat of security breaches. Thus, just as service providers can deliver a Wi-Fi network slice-based on overall performance, they can provide one for mission-critical home IoT devices, including home security and monitoring systems. This network slice would rely heavily on machine learning and AI to identify usage and data consumption patterns for all devices and to anticipate when sensors and devices would require software and firmware upgrades to ensure those occur without issue.
But beyond creating a dedicated network slice for high-priority home IoT devices, home Wi-Fi networks also must rely on machine learning to understand when, and with what frequency, IoT devices are accessed by outside cloud and mobile applications. Machine learning algorithms in the home CPE can monitor both incoming and outgoing device traffic to establish an overall device and/or ecosystem profile so that anomalous traffic to and from the IoT devices can be quickly flagged for the user. From there, the user can decide what to do with the potentially malicious traffic and compromised device. That particular decision can be learned and automatically applied in the future so that the user can be assured their device will remain secure in the face of potential hacks and denial-of-service attacks.
The onus of security is falling more and more on the application provider or the network operator, as IoT devices typically don’t have the processing power and storage capacity to maintain a library of malware signatures. As broadband subscribers increasingly become more reliant on IoT devices in their homes, their willingness to pay extra to ensure the security and reliability of those devices is likely to increase.
Once again, a network operator can combine the power of machine learning along with the additional capabilities of Wi-Fi 6 to package an IoT device management service, as part of, or in addition to, a managed Wi-Fi service offering. Specifically, Wi-Fi 6 incorporates a feature called target wake time (TWT). TWT allows the home gateway to set a schedule for IoT devices to ping it to report their status. Thus, devices do not have to fight for the channel spectrum to complete their communication. Each device can be guaranteed an optimal slot to ping the router, and it can remain in battery-saving sleep mode for a longer period of time.
The combination of machine learning, Wi-Fi 6, and cloud-based management of home networking and IoT devices give service providers a potent solution for ensuring the performance, reliability, and security of their customers’ growing IoT and home security and monitoring devices.
FTTH Networks a Key Focus Area for AI and Wi-Fi 6
FTTH network deployments continue to expand around the world. The reasons for this go beyond just increased speeds and future-proof networks. Those reasons also include service and application flexibility made possible because of the underlying stability of the network. Operators can deliver a much wider set of applications and services over their FTTH networks because they can allocate bandwidth dynamically to those individual applications when they are undergoing heavy usage or in anticipation of heavy usage.
But that flexibility isn’t available end-to-end without ONTs that are capable of processing, identifying, and anticipating packets by service or application type. In most FTTH networks today, service providers still rely on basic bridge ONTs for the physical fiber termination and then rely on a separate router to provide intelligent routing and management of packets and applications. But service providers are now realizing that integrating those functions along with AI capabilities into an intelligent ONT gives them the ability allocate bandwidth by application, identify, anticipate, and correct bottlenecks within the home network itself, and provide an extra layer of security for home users and all IoT devices on the home network.
The combination of AI and Wi-Fi 6 can help reduce the latency of specific services by more than 50%, which is critical for sensitive, high priority applications, such as online gaming, 4k/HDR and 8k video, and distance learning. Operators who have this equipment in their network and offer these services hold a key competitive advantage by providing guaranteed QoE (Quality of Experience.) That guarantee can result in a higher ARPU (Average Revenue per User) and greater customer satisfaction.
Though the industry as a whole remains in the early stages of improving connections and services both to and within the home, the list of operators rolling out or planning to roll out these capabilities continues to grow quickly. It makes sense, especially when subscribers are becoming more dependent on their home broadband service for critical applications and when these same subscribers are becoming savvier about how bandwidth, throughput, and latency impact each of those applications.
AI and Wi-Fi 6: The Future of Home Networking
Amidst the hype of 5G, there has been a renewed effort globally to strengthen Wi-Fi’s carrier-class capabilities and extend them into home environments. Nowadays, home broadband and Wi-Fi are synonymous with subscribers. You only rarely have one without the other. And just as service providers have improved the throughput, reliability, and security of their fixed broadband networks, their focus over the next few years is on doing the same for home Wi-Fi networks.
The combination of emerging Wi-Fi 6 gateway devices and technologies, along with cloud-based management and machine learning principles makes this goal of rock-solid home Wi-Fi networks an achievable reality. This not only improves the reputation of service providers among their broadband subscribers but also opens up new revenue opportunities through both an overarching managed Wi-Fi service offering and also through individual service tiers designed for specific user profiles.
For the home network scenario, Huawei already has done much innovation, Huawei provides the industry’s first series of eAI ONTs which can intelligently identify service types of home users by AI technologies. With innovative Wi-Fi 6 slicing and optimized technologies, the ONTs can reduce the latency of specific services by more than 50%, achieving zero frames freezing for high-priority services, such as gaming, e-learning, and home/SOHO office, By leveraging this advantage, operators can launch value-added services with a guaranteed experience to improve the average revenue per user (ARPU) of home broadband users.
And currently, Thailand’s 3BB project has adopted Huawei’s eAI ONT, helping 3BB build a home broadband network that offers the best gaming experience.