Predicting the future isn't easy. Alec Reeves got it correct with regards to some aspects of mobile telephones, but the handset was rather antiquated compared to reality. This sketch made by Alec Reeves, with a real mobile telephone alongside, shows an instrument that used magnetic tape to dial numbers, instead of electronic memory.
He believed that by about the year 2020 "the wandering subscriber" will have to be catered for - that any person, anywhere, will need or want to be able to talk to any other person anywhere else. It will become perfectly feasible, said Reeves, with a combination of PCM, optical fibre waveguide transmission and conventional radio. In the optical wave bands seven thousand million good quality speech circuits per waveguide were considered to be theoretically available. These would be used in massive traffic local and trunk fixed networks. He thought that cheap, reliable, mass-produced optical fibre was likely by 1990, room temperature (probably Gallium Arsenide) lasers for the repeaters perhaps by 1971, giving optical power sufficient to allow land-line repeater spacings of about 3km for a 300 Mb/s system and 2.25km for 10 Gb/s operation.
Another prediction: completely mobile world-wide personal telephone numbers. Assuming that world population in 2020 is about 16 thousand million, even if only one quarter of these become subscribers, 42 binary digits would be enough to identify each with reasonable reliability.
The frightening prospect of dialling 42 digits would not arise. Instead, about 100 42-digit numbers (enough for most people) would be stored on a narrow magnetic tape about 10 cms long contained in a personal pocket radiotelephone. Any number to be called would be selected using a pair of small wheels moved in decimal steps according to personal directory numbers. This would select the correct point on the tape and a head would then read off the 42 digits into the world network, probably through an intermediate store to allow read-out at an accurate, appropriate rate.
These personal pocket devices would plug into fixed sub-sets for immediate optical connection, or into vehicle systems, or be used independently when necessary. He had already clearly formulated methods of universal subscriber contact using the mobile units in self-forming "chains" with fixed radio stations, working into the fixed optical networks on land and under the sea. These were known as "SYNSOL" and "Synspec".
What if the number to be called is not known? Then said Reeves, speech recognition machines would almost certainly have been developed to the point when one will ask a store for the number and get it in a matter of milliseconds. Such stores he says, would very probably be electro-optical, using hologram storage in crystals with laser-beam access.
A store capable of holding all the world's worthwhile information (10^15 bits?) need not occupy more than a few cubic metres. [Nanotechnology was unknown then.]
So, the prospect arises not only of instant universal communication but also of almost instantly available knowledge.
And for Reeves there was of course one type of transmission in particular that he believed will be used in these future systems: pulse code modulation.