It’s all related to the limitation of wireless as an access technology. So perhaps I should spend a few minutes on the realities of access technologies.
We have a few options to provide broadband services to the premise:
First, Copper using sophisticated Digital Signal Processing techniques embodied in DSL technologies to get the last bit of performance out of the limited bandwidth on copper ,
Second, wireless using various types of modulation and coding schemes to squeeze the most out of the available spectrum,
Third, coaxial cable which has a much higher bandwidth than copper but is shared across multiple users
Fourth, satellite technologies that use even more sophisticated coding and modulation schemes to overcome the issues of getting a satisfactory signal to and from a platform that is 36,000 km above the Earth, and
Fifth, Fibre.
Each of these access technologies has its own characteristics with respect to bandwidth (speed), distance dependence, symmetry, latency and coverage.
The technology options in Charles Todd’s day or even in the days when the copper CAN was first deployed were considerably simpler. The technology options available today make network design choices quite complex. It is a fact that communications is the field in which some of the most complex and sophisticated applications of modern physics, mathematics and statistics are used. That’s what makes it so fascinating, but also so challenging.
What are the pros and cons of the various access technologies:
Copper: with copper-based DSL we have now reached about the limit of what can be done. It is now a trade-off between distance and speed. We should show our venerable copper plant a great deal of respect. It has served us well for decades and provided us with more than we had any right to expect. It is now time to allow it to gracefully retire.
Wireless: we have a lot to thank James Clerk Maxwell for. He unified the physics of electricity and magnetism and predicted the propagation of electromagnetic waves on which wireless systems are based. Mobile technologies will continue to improve and mobile applications and devices will continue to proliferate. However, we need to be realistic. Every cellular base station is fed by fibre – for a good reason. Spectrum is a very scarce resource which is why, as I said earlier, you cannot buy a mobile broadband plan that will provide a 50GByte/month download capacity. Wireless speeds also depend on distance and load.
Coaxial cable: a “souped up” version of copper. A much higher bandwidth capability but because of the design of HFC systems it has symmetry problems. While it can provide high speeds it is not so good at providing symmetrical broadband services that you need for, for example, two-way videoconferencing. It is also a shared medium so it is load dependent. Of course you can continue to node split, in which case you get closer to a fibre PON (Passive Optical Network) architecture.
Satellite: great for broadcast coverage but needs careful engineering to ensure a good broadband service can be supported. It suffers from latency issues since there is not much we can do about the speed of light. and finally,
Fibre: is almost the ideal transmission medium. The cable itself is cheap, very robust these days, has negligible latency issues and has an incredible bandwidth capability. The limitation in speed achievable in fibre-based systems are related to the rates at which the electronics on either end of the fibre can switch. The information carrying capacity of the fibre itself is almost unlimited.
So, is fibre likely to be superseded in the next few decades by some new technology we have not yet even thought of?
I don’t think so.