I’m on vacation at the moment and I’ve been thinking a bit about how network technology has evolved over the past few years and where we are headed. At least for the moment, it seems to be that we have reached an access network tech plateau, at least as far as infrastructure is concerned. Let me explain…
In the fixed line world, fiber to the home is the thing to have. In Paris, I’ve had it for more than half a decade now. In Germany, I’m still waiting for it. Most people with a fiber connection at home use GPON (Gigabit Passive Optical Network) with up to 1 Gbps in the downlink and several hundreds of megabits in the uplink direction. It’s a decade old technology and in some places XGPON with speeds up to 10 Gbps is already deployed. But most fixed and wireless network infrastructure at homes and in offices today is based on gigabit Ethernet, so few people actually know what do do with anything beyond the plain 1 Gbps GPON. I don’t see a trend for a significant change anytime soon. And if it is required, upgrading is simple, the fiber is there. My point here: Technology to increase data rates for subscribers by an order of magnitude is available for many years now, but there is no significant take-up, as current technology is faster than what most people need. But don’t get me wrong, there is still a lot of work to be done to push fiber to more places. No new technology is needed for this. However, price is always an issue, so deploying, running and maintaining networks for less is perhaps the avenue where innovation will be a driver for change in the next few years?
In the cellular domain, 5G started in 2019. In it’s incarnation as a speed booster alongside LTE, it is very well established and has done a tremendous job of increasing network capacity in cities. There is also the 5G standalone variant with a new core network based on Kubernetes and containers, but those who have started offering ‘real 5G’ service have a hard time pointing out what the noticeable advantage for end customers is over the 5G non-standalone service in combination with LTE access and an LTE core network.
Cellular data volumes keep rising however and while this can easily be absorbed in fixed line fiber networks with their almost infinite capacity reserves, things are much more tricky in wireless networks. Here, demand keeps rising in the order of 30% on a year over year basis. Capacity is directly connected to spectrum availability, which is limited. A significant part of the spectrum that has been assigned for cellular networks is now in use in cities. On the country side, much less spectrum is used, mainly because cells tend to be much larger, and particularly spectrum in the 2 and 3 GHz range doesn’t reach very far. In lower spectrum ranges, capacity is limited, so rural areas are the first to experience capacity shortages.
During the last World Radio Conference in 2023, a significant amount of additional spectrum was assigned for cellular use in the 6 GHz band, at least for Europe. That will help in cities to extend capacity for a while and it is your and my best guess if end-user capacity demand will flatten out before the capacity limit with the available spectrum is met.
So even with today’s technology and its evolution, there is a lot that can be done to improve cellular networks. If ways could be found to increase capacity in the country side while keeping cost in check, this would help a lot. Also, network deployments at train stations, along railway tracks, airports, indoor shopping centers and office buildings are often at their limit or even non-existent. Ikea buildings on a Saturday are a good example. And super-hotspots such as fair grounds, sports arenas and temporary festival areas, where tens of thousands of people come together for a limited amount of time continue to be a challenge to be served with sufficient capacity and at a reasonable cost.
As capacity demand keeps rising, ways have to be found to absorb the additional cost for adding more capacity, as each new band or channel in a band requires additional hardware at base station sites. Here, advances in technology can help to increase the number of bands that can be supported by integrated radio units. But the radio is only one part of the base station equipment, the baseband that does all the calculations is the other. And here, advances in hardware can also help to make things cheaper and hence absorb the cost for rising computational power. So compared to fixed line networks, even more evolution is required to keep up with capacity demand. To me, it looks like this will most likely be done by refining existing technology rather than by some revolutionary new technology.
All of this sounds a bit vague. Yes, it’s hard to predict the future, but I think the important point is to be able to imagine what could be improved in network technology in the next decade. The ‘how’ is the problematic bit. William Gibson once said: ‘The future is already here, it is just not evenly distributed’. I really like this quote but wonder if it helps very much. The quote seems to suggest that there is just a single ‘unevenly distributed future’ and once you find it, you know how things will play out. But from looking back into the past, I think it is safe to say that there were many future technologies around in the past that were not evenly distributed. And while many of them were great, only very few made it to a wider distribution.