Mainly based on Ciscos forecasts, there seems to be a general consensus about that we need 1000 times more capacity in mobile broadband before 2020. In a recent post we discussed how various manufacturers and operators mean that this can be achieved. Network densification is an important and costly component in this equation – it’s not only new (5G)technology or new (elusive) spectrum options, more capacity is mainly provided by more base stations. An interesting question that arises is what exactly makes up this capacity. The “capacity” is roughly the product of bandwidth (bitrate) provided times the number of users. In the first mobile data revolution it has mainly been the bitrates that have gone up – rather few users have consumed most of the 100-1000 fold data rate increase we have seen. With flat rate, this spells disaster since there are so few to pay for the new infrastructure needed. But what happens in the next step ? There are two possible trajectories
- The Gigabit track – extrapolating the current trend, higher and higher data rates to a few users not willing to pay significantly more. The revenue gap opens even more .. the road to hell ?
- The “proliferation track” – keep the current data rates, but build capacity to allow for a significantly increased number of smartphone users. More users are paying to share the cost of infrastructure. Revenues grows more like the expenditures. Repeat the success story of mobile telephony.
Reality is likely somewhere in the middle, but it is clear that more users cannot be bad. In places where many already have the service it is of course a more difficult proposition. Does someone understand where we stand on this – how many more MBB users will there be here in Western Europe vs the rest of the world?
Honestly, what is the benefit from sometimes, somewhere getting 100 Mbit/s in a smartphone under current tariffs ( besides a cool marketing stunt)? It only means that you can consume your monthly “allowance” in roughly a minute. Making sure that I can get 2-3 Mbit/s anytime instead sounds a more reasonable (but difficult) proposition. High peak rates IS one way of achieving this, but not a very effective one.