Today telecommunications challenges
by Daniel Olteanu
We are facing our days big challenges in this amazing field of human telecommunications.
The need for instant and natural communication between peoples increase day by day, as new
services emerging in a very fast rhythm.
A big issue rise when we talking how to define "instant" and "natural".
In my opinion any human conversation has to be as natural and instant as the direct conversation is.
If we look at various technologies for voice and data transmission and switching we'll realize that the POTS, TDM and circuit switching are still almost the only ones which accomplish the request stated above.
ATM as well as IP based technologies rely on "data cells" respectively "data packets" as basic "bricks" of voice information.
That implies a radical change of how the human communication real time process is broach.
A person when is talking to another person directly or over the telephone (POTS)
expects that the words or sentences to be received by the other person instantly as he's speaking, and to receive in the same manner the other person message. Over the POTS (including TDM and end to end permanent and direct circuit) this goal is achieved.
In the ATM and IP networks, the "path" and possible "paths" of voice information "bricks" between A and B points implies lots of "gateways", where we can't talk anymore about "direct, permanent and instant communication circuit". This it's OK for computer like communication but not for humans.
Of course reasonable "delays" and "echoes" are accepted, but the so called "voice fragmentation effect" is very
annoying.
There are various theories on human communication process and how much these new technologies of data transmission and switching
approaches and even outmatch the POTS techniques, but I think that the reality contradicts most of them.
If we want to see how Nature , the best "designer", solve some real time processes it's not necessary to search very far. From human "audio receiver" (the ear) to the "CPU" (the brain) we have a well and direct definite "path", and the sound travels only on that "path", it's not sent along some "networks" of "paths" to travel from ear to the brain.
Some equivalent onset of this problem in telecommunications these days is the "splitting of voice and data" at the customer location, in the same subscriber interface, and further the voice "path" is handled in circuit switching manner, meanwhile the data is handled on separate IP networks, this way the goals stated above are accomplished.
The new MPLS and GMPLS label switching protocols, emerged as standards after 1997 are connection oriented switching protocols overlaid over IP networks, that means "virtual fixed packets paths" over
no connection oriented networks.
To be continued ... maybe :)