8
$\begingroup$

See also How were Intelsat 1 “Early Bird” and Telstar 1's "hundreds" of simultaneous telephone conversations multiplexed/demultiplexed?


Ouch! The 1960's

The Intelsat 1 “Early Bird” communications satellite, built by Hughes Aircraft Co., was able to relay 240 simultaneous phone conversations between Europe and North America.

From This Is What Broadband Satellite Communication Looked Like in 1965

The Intelsat 1 “Early Bird” communications satellite, built by Hughes Aircraft Co., was able to relay 240 simultaneous phone conversations between Europe and North America. But having all 240 lines apparently originate in one room in Los Angeles, as this Hughes publicity shot seems to suggest, would undoubtedly have created a slight bottleneck.


I have read that in the early 1960's Telstar 1 and 2 were the first satellites to route long distance telephone calls.

I seem to remember that long, long ago, long distance (overseas?) phone calls did have noticeable latency and sometimes annoying echoes, and I suspect that this was due in part to the light time from Earth to a satellite and then back to Earth, and the echoes were because these were analog signals.

These days, long distance phone calls are generally routed through fiber optic cables, and I assume that ground telephone traffic if still relayed by satellite, would be digital.

Questions:

  1. When did they stop routing conventional analog telephone calls through satellites as analog signals as opposed to digitized data?
  2. At its peak, what was the volume of telephone traffic that was routed through satellites? It can be expressed either in total number of concurrent calls, or calls per year, whichever is easier to find.

note: I'm asking about calls made between conventional telephones, and not satellite phones.


enter image description here

Mounting Telstar Satellite to the Thor-Delta rocket 1962 (Source: Nokia Bell Labs and AT&T Archives)

Source: The First Active Orbiting Communications Satellite

Then, for the first time, live television transmissions and phone signals could be relayed between the US and Europe by means of this simple looking, spherical black and white satellite. Its iconic exterior held within it 170 pounds of some of the most complex electronics known to humankind. It featured 3,600 solar cells for power and a traveling-wave tube for amplifying the radio signals. Back in 1962, the key task of the Telstar 1 was to receive signals beamed from the USA, amplify them 10 billion times and rebroadcast them to live audiences in Europe, and vice versa. TV and telephone communication signals were relayed and boosted to get back down to Earth. The Telstar 1 circled the planet every two and half hours. It was only in the right position to beam transmissions between the USA and Europe for 20 minutes in each orbit before dropping out of contact. Future satellites were designed to work in tandem with each other, seamlessly passing the broadcast to keep transmission live at all times. At launch, the Telstar 1 facilitated over 400 telephone, facsimile and television transmissions.

$\endgroup$
12
  • 1
    $\begingroup$ fyi I found a giant list of communications satellites at Colorado State University, though not sure how helpful it is i this case. $\endgroup$
    – uhoh
    Commented Sep 1, 2019 at 1:11
  • 4
    $\begingroup$ There's a fascinating paper here oecd.org/sti/ieconomy/2091239.pdf that sheds some light, but not a full answer. I learned a lot about comsats and COMSAT from reading through it. $\endgroup$ Commented Sep 1, 2019 at 1:24
  • 1
    $\begingroup$ @OrganicMarble thanks for that. This might be a challenging question to answer accurately. For example per Table 16, in 1991 there were 565 US transponders. Assuming 36 MHz equivalent (which the table doesn't) that's 20+ GHz of bandwidth. However customers can divvy up each transponder's bandwidth between phone, TV, and other things any way they like. I think a rough answer will be fine, though it's possible somewhere there's a quantitative estimate already published. $\endgroup$
    – uhoh
    Commented Sep 1, 2019 at 1:32
  • 2
    $\begingroup$ @OrganicMarble: re: satellite comms in Australia. Because of low population rates in rural & remote parts, they still can have satellite comm. Generally Australia uses both undersea cables & satellites for international telecommunications (NBN Slymuster) $\endgroup$
    – Fred
    Commented Feb 28, 2021 at 16:12
  • 2
    $\begingroup$ Nice "archeology" type question. May I comment with: when the term "call" will stop being used? We are definitely entering the era of "always on" and "video chats" (multi-party virtual meeting"). Before we paid according to distance and duration. Today, some of us pay according to "volume" used (per month), and some even don't). Tomorrow, we will pay according to network coverage alone ("do you cover Europa?"). $\endgroup$
    – Ng Ph
    Commented Dec 10, 2021 at 9:57

1 Answer 1

1
+400
$\begingroup$

There are two ways that the premise of this question can be challenged. The first is that truly analog satellite communications never really started, and the second is that it never ended. While these seem to be opposite points of view (and while the answer you are looking for probably falls somewhere in between) it helps to illustrate that the answer will depend a lot on how you choose to define an analog (versus a non-analog) phone call.

Some people may define a non-analog phone call as one that involves the process of time-slicing and quantizing the analog waveform.

But the idea of slicing up a phone call's audio signal, discarding unimportant bits of audio, and then shuffling around the rest to efficiently route them over a communication channel perhaps started with time-assignment speech interpolation (TASI) which was invented by Bell Labs in the early 1960s to increase the capacity of transatlantic telephone cables.

It's not fundamentally all that different from what happens in modern systems. The original audio signals are discretized, redundant information is identified and discarded, and the rest is reorganized and sent over a communication channel which - even today - one could argue is "analog". In modern systems its just done at a much finer granularity and more sophisticated techniques are used to identify and remove redundant information and reorganize and route the rest.

Consider, for example, Quad Amplitude Modulation. Here digital information is encoded by modulating the phase and amplitude of a sine wave. The recovery of the digital information from the raw signal requires analyzing and discretizing the analog waveform.

So if you wanted to argue that analog satellite communications never really started, you could say that even with the earliest satellites, it would not have been possible to tune into the transmitted signal and hear to a coherent phone conversation. You would have needed TASI equipment, at the very least, to convert the received signal into coherent conversations.

On the other hand, if you wanted to argue that it never ended, you simply need to list of all of the MODCODs that are used by modern satellites. It will quickly become apparent that at the physical layer they still make use of analog waveforms and the techniques of analog electronics.

The use of backwards compatibility also makes it difficult to answer the question. It is likely that a pure analog phone call could be transmitted in an unadulterated manner on the communication channel supported by many satellites, but it is unlikely that this would ever be done in practice. It would be supported, because telecom engineers love backward compatibility, but it is unlikely that it would be "used" in practice.

To help emphasize the degree to which communications engineers love backward compatibility, the Connections Museum in Seattle has a demonstration where they route a phone call through telecom equipment spanning from 1923 to 1976. That said, nothing under the roof of the Connection Museum has anything to do with satellite communications.

Satellite image of Connections Museum

$\endgroup$
5
  • $\begingroup$ Wow! While it is true that when pushing the limits of bandwidths, the transmission of 1s and 0s (bits) requires circuits, analysis, and signal processing that looks "analog", the goal in those cases is still to recover the digital 1s and 0s. I don't think it's productive to consider everything to be some kind of analog. Your answer is fascinating and thought provoking and I will need to read it a few times on a few different days to take it all in. but is it possible to add something that addresses analog as not-based-on-1s-and-0s, and digital as based-on-1s-and-0s? Did the former ever stop? $\endgroup$
    – uhoh
    Commented Jul 29 at 5:31
  • $\begingroup$ For purposes of eye candy, scrolling down through Eye patterns shows all kinds of "analog" waveforms which occur during the process of encoding and transmitting 1s and 0s. $\endgroup$
    – uhoh
    Commented Jul 29 at 5:36
  • $\begingroup$ note to future readers, the bounty message (now hard to find) says "At this point, the best answer may be to simply challenge the premise that it ever completely stopped, and provide a few examples to lend credence to the challenge. I'd certainly accept a (well written) answer along those lines." Since this meets that criteria, I happily awarded the bounty! $\endgroup$
    – uhoh
    Commented Jul 29 at 5:38
  • $\begingroup$ I would tend to think of "digital" or "non-analog" signals as implying the ability to store data or handle it out of order, perhaps? This is not impossible in analog, but it is trivial in most digital systems. $\endgroup$
    – ikrase
    Commented Jul 29 at 11:49
  • $\begingroup$ Analog systems such as early color televisions used a surface acoustic wave filter which converted an electronic signal into a sound wave to create a delayed version of the signal. This would be an early version of storage - as the signal was temporarily "stored" in a sound wave. You could probably shuffle the order of analog signals with such devices, although I don't know if it was tried in practice. $\endgroup$
    – phil1008
    Commented Jul 29 at 12:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.