Blog | Sandvine

What if 5G networks were actually used as a fixed line replacement? Many operators (TMobile and Verizon among them) have announced that they will offer 5G modems for home use to compete with fixed line services with the higher throughput supported by 5G. Is building that network to deliver good QoE different than a normal mobile network?

The first mass market 5G phones from Samsung were announced yesterday, and it appears 5G might start to be available in a useful way soon for mobile consumption. But the initial 5G service announcements were about a fixed replacement service because of the lack of phones with 5G. When I ran the initial data for the Mobile Report, one of the datasets looked wrong. When I dug into it, it suddenly made sense.

The dataset was from a mobile operator, but it was an operator that offers a fixed mobile service rather than a traditional mobile offering. The data below is what came out of the analytics on the traffic share for that network type:

That data would not be out of place on a high capacity fiber network! Notice that seven of the top 10 are video, and they account for more than 62% of the traffic by themselves. The usual suspects are in the top five, including Disney+, with Netflix assuming a dominant share, and YouTube doing better than on traditional fixed networks (likely due to lower resolution requirements, encouraging usage versus other OTT services). Also notice on the upstream, Nest Thermostat and the Ring Doorbell, not something that you are likely to see on many mobile device networks in the top 10!

Building a network to deliver a high QoE for this mobile network would not be the same as building one to support a great experience for all mobile users - there is a distinct lack of social networking, and more gaming traffic (as you will see) than on a normal mobile network, so latency (which 5G promises to improve) are more important. 

What does a normal mobile network traffic distribution look like? Stay tuned....