^{Democracy & Technology Blog} My Canonical Paradigm

A Chinese American entrepreneur engineer named Henry Gao has written a Chinese book paralleling, enriching and affirming the more far reaching propositions in Telecosm.

His theme is that the history of communications networks has passed through three eras: 1) the telegraph (data with delay and buffering); 2) the public switched telephone network (PSTN for real-time two-way voice), and 3) now back to the telegraph (the data-rich Internet protocols and layers, with many asynch buffers and best efforts and lost bits).

Today under the stress of an interactive video exaflood, there is a new fork in the road. On the one hand, the industry wants to continue on its current path back to a new video best efforts telegraph — an ever more complex Internet patched and epicycled and upgraded for interactive video. This is the current choice.

But Telecosm suggested a Wavelength Division Muxed path back to an optically circuit switched PSTN. In China Gao has experiments and prototypes for this preferred path back to a synchronous phone network optimized for realtime video teleconferencing and subsuming all else plus an afterthought Internet. Optimized for interactive video, from games and virtual worlds to conferences and software clouds — the new phone network will be founded on synchronous Ethernet, which is good for our companies, but it will be hugely simpler than the current Internet, which will incur many casualties.

In wireless, the fork is different and defined by the Shannon equation for channel capacity, which combines bandwidth with spectral efficiency (more potent signaling through high signal to noise ratios). The current solutions — QCOM’s Long Term Evolution etc. — pursue spectral efficiency and gain the ultimate Shannon capacity from the current system of cells upgraded with more spectrum. But spectrum remains scarce and political, auctioned and nationalized. There is little headroom.

So the current 4G strategy cannot handle the tsunami of two-way video portended by teleputers such as iPhones and iPads. The only Shannon-authorized route to wireless video teleconferencing is to focus on the signal to noise term in the equation. This requires moving the antenna closer to the receiver as in femtocells, gaining a power of three or four increase in signal to noise. It is overwhelmingly the way to go, but
no one has outlined a clear path to the goal.