One can think of only a few things that are unimpacted by the COVID-19 situation. But to be honest, even I struggle to put my finger on those few things. Let’s face it, most aspects of our lives have now changed, and it doesn’t look like we’ll return to normal anytime soon. Clearly, social distancing and self-isolation mean that telecommunications has become an elevated essential service which, of course, begs the urgent question: how will communications service providers (CSPs) meet the challenge of their new critical role in our changed world?
Think about the typical demands on a mobile network from a human behavioral perspective. In the morning, people wake up, check their emails and texts, get ready for work, drop their children at school and commute to work.
During busy mornings, while getting ready, there is little time for our favorite apps or content. However, during our commutes there is time for social interaction, whether it’s scrolling through our phones while on public transport or making phone calls while traveling by car. That is the time when mobile network operators (MNOs) see the first of the traffic peaks during a day. Then everyone gets to work, or school, and the mobile network traffic quiets down, as most people are now within WiFi coverage and are busy working or learning. The next spike comes during the lunch time when people try to catch up on their favorite social media platform. The third spike occurs during our commute home and this usually lasts a bit longer as people go out and socialize.
During weekends, mobile traffic is usually much less and flatter, as people go out and about at different times of day. Also, the weekend traffic levels tend to depend on the geographical location and the time of year. In places where people stay outside most of the weekend, mobile traffic will be larger than in places where weather prevents outdoor activities.
The biggest demand on mobile networks happens where there are large global live events that people want to follow and comment on through social media. These are big sporting events like the Olympics or World Cups that happen during people’s commute or weekends when they’re outdoors.
So, how does the demand on a mobile network change in light of COVID-19?
There is no daily commute right now. There is no going out in the evening at this time. There is no rushing out in the morning in the near term. There are no sporting events planned for months. Most people are inside utilizing WiFi coverage. So, one would expect that the demand on the data network has decreased or become flatter like during the weekend. However, there is a lot of news around the world - a lot of news updates. Social media channels are bustling with news. People communicate much more than in normal times.
Also, staying at home and not going out means people watch more TV and on-demand video. Such behavior puts more strain on cable operators than MNOs. However, even if people are at home using WiFi, they would still use their mobile phones for voice communications, and this is the area that MNOs need to keep a close eye on. With video conference services (Skype, Zoom and others) traffic going through the roof right now, the dial-in option is an alternative if video conferencing platforms are overloaded.
Also, with the need for social distancing, family and friends are apart for the time being, which will drive additional voice traffic, whether it’s on legacy circuit-switched systems or the omnipresent 4G network with Voice over Long Term Evolution (VoLTE) calls. The strain on voice services though, will be helped to some extent by the convergence of WhatsApp, Messenger, VoWiFi, Skype, Hangouts and FaceTime calling within the iOS and Android ecosystems. Nevertheless, some CSPs have already reported voice traffic growth in the range of 30-60%. This is against the regular year-on-year increase of only 5%.
However, with the potential congestion of fixed broadband service, there is a chance that devices will fall back onto mobile networks and this will cause a significant surge of mobile data traffic. Likewise, in areas where MNOs provide home broadband connectivity via MiFi devices, one would expect significant increases in data traffic, and in some parts of the world where unlimited data isn’t pervasive, an increase in subscriber costs.
Looking at that picture painted above, I would say that cable operators will definitely be more impacted by COVID-19 than MNOs due to the increased activity at home - working from home and heavier demand on streaming services; Netflix and Google have already announced that they would be reducing the quality of video streaming from high definition (HD) to standard definition (SD) in order to prevent network overload.
With significant changes in user behavior, not to mention the escalating use of telemedicine solutions, and the corresponding data traffic shifts, it is increasingly important for CSPs, whether it’s a mobile or cable operator, to keep vigilant and closely monitor network anomalies, congestion, availability and faults, as well as traffic trends. The problem is that manual traffic monitoring sometimes lacks sophistication and is prone to miss important information.
At Guavus, with our history of providing CSPs with network and operational analytics based on artificial intelligence (AI) and machine learning (ML), we understand the possible constraints and the importance of data analysis in both improving service availability and ensuring quality of experience (QoE).
Our AI/ML-based analytics software provides automated anomaly detection and network event data correlation. Our technology converts events into time-series datasets and not only alerts on anomalies, but also indicates which anomalies happened within the same timeframe. This enables MNOs to monitor overall network traffic and identify a change in the traffic profile (be it a sudden or potentially slow change over days). In addition, our technology enables MNOs to identify the reason behind the change or issue, whether it’s a network platform problem, user/device behavior or a change within the OTT (over-the-top) service, such as changes in the streaming resolutions that YouTube and Netflix are implementing in Europe right now.
Our technology can monitor anomalies in overall network traffic and correlate it with traffic aggregated at a platform level or an application level or even at the destination IP subnet level. It can also monitor any type of key performance indicator (KPI) and correlate identified anomalies with any other counters fed into the tool. As a result, the system will generate an alarm about the anomaly and provide insights about all other timeseries metrics that exhibited anomalies at the same time.
Bottom line, Guavus is committed to helping CSPs during this critical time, especially as many struggle to handle the increased impact on their mobile and traditional network infrastructure. After all, the network is our lifeline to communication and we all must do whatever we can to ensure that we keep it healthy, available and responsive to the needs of mobile and home subscribers. At no other time in our history have we depended more on the flow of information all over the world than right now.
