Inmarsat recently announced that its first communications standard, the Inmarsat-A, would be retired at the end of this year (2007) after serving Inmarsat for more than 25 years. This announcement has provoked a wave of nostalgia among COMSAT's early mobile service veterans since Inmarsat-A is in fact the MARISAT communications system, introduced to the world in the summer of 1976 by COMSAT General. It was then adopted by Inmarsat, with minor modifications, to initiate its global service 5 ½ years later. What exactly is this system and why did it survive so long? As to its longevity, one should probably look to the caliber of the design team, which consisted of a large number of COMSAT General and COMSAT Labs staff, including Dave Lipke, Dan Swearingen, Tom Calvit, and Dick McClure, just to mention a few of the key players.
Following initial design studies, the basic parameters selected for the shipboard terminal were a G/T of -4 dB and an e.i.r.p. of 37 dBW, a 4-foot-diameter tracking dish, and a 40-W transmitter. A more detailed description can be found in the Fall 1977 issue of COMSAT Technical Review (Volume 7, No. 2)
MARISAT-A MARITIME SATELLITE COMMUNICATIONS SYSTEM
D. W. Lipke, D. W. Swearingen, J. F. Parker, E. E. Steinbrecher, T. O. Calvit AND H. Dodel
It is available in pdf format [7MB] on the Comsat Legacy web site:
In 1970 COMSAT began the sequence of events leading up to the MARISAT system by recruiting retired Major General John L. Martin Jr. to form a project office that would expand the company's activities into the domestic and mobile businesses. . The 1971 ITU Space Conference agreed that maritime satellite communications would be provided using L-band frequencies; this made it possible to begin looking at parameters like the appropriate size of shipboard antennas and the power of transmitters. In 1972, to promote interest in the possibilities offered by satellites to the mariatime community, COMSAT Labs installed an 8-foot converted W.W.II radar antenna to demonstrate two-way communications with the Queen Elizabeth 2 through an INTELSAT satellite over the Atlantic. Also, during 1972, Tom Calvit and Ed Martin, the initial members of the Mobile Project Office, collaborated on a parametric tradeoff study to determine the appropriate balance between size and cost of ship antennas and satellite power that would minimize the overall cost of communications. This study suggested ship antennas in the range of 4 feet in diameter. In Europe, and in Britain, in particular, it was argued that steering such antennas would be too complex and expensive. Those opinions proved wrong.
In spring 1972, the Project Office fielded an inquiry from the US Navy regarding the feasibility of COMSAT providing leased satellite capacity in the UHF band for fleet communications on an interim basis while the Navy awaited the availability of its much delayed FLEETSAT satellites. COMSAT concluded that it could meet these needs and find room on the satellite to provide L-band commercial maritime services. Because only a fraction of the satellite's capacity could be used for the L-band service, the design parameters for the communications system had to be chosen carefully. Time was also very short so that only proven and available technology could be used. The designers, worried about obsolescence, were instructed to think "10 years". A delay of over a year in production of the satellites gave designers some breathing space to get it right.
A number of sacred cows in the COMSAT culture derived from INTELSAT experience were skewered to eliminate any excess performance margins that might threaten commercial success. A major controversy broke out over voice modulation, with one side urging the digital deltamod approach, while the other advocated conventional analog FM.In the end, companded FM was selected for, among other reasons, its soft failure characteristic at the edge of satellite coverage. Also, FM fared better in comparative listening tests. The non-voice part of the system was all-digital with control signals and shore-to-ship telex traffic carried on a 1200-bps TDM carrier. For each outbound TDM carrier there was a corresponding inbound TDMA channel shared by all ships bursting at 4800 bps. This channel carried the ship-to-shore telex signals and, due to its higher speed and big time gaps between bursts, would eliminate any requirement for synchronization of transmission bursts among the ships.
After COMSAT signed contracts with the Navy for UHF service and with AT&T for domestic service via the separate COMSTAR satellite program, the Project Office evolved into COMSAT General, with John Johnson at the helm. Johnson decided to procure 100 ship terminals to get the service going quickly, a figure that was later doubled to 200. This was, numerically, the largest earth station procurement yet to be seen in the industry. The procurement was broken into two parts. Since that antenna development was thought to be the most challenging, the above-decks equipment was ordered first. Scientific Atlanta, in its own internal R&D program, had developed a fairly simple, reliable and inexpensive platform and it won the first procurement handily. The second procurement called for production of the communications equipment, both for the earth stations and the ships and for the integration of the below-decks equipment with the antenna system. Scientific Atlanta also won this procurement and the first terminals were in service in July 1976. Other manufacturers in the U.S., Japan, and Europe soon joined in the production of MARISAT ship terminals and COMSAT backed off from further procurements to let the marketplace reign. The service was well received in the maritime community and usage grew to 1000 vessels by the time the system was handed off to Inmarsat.
The Inmarsat agreements were opened for signature in the summer of 1976 just as MARISAT went into service and the international debate on selection of system design began in earnest within the Inmarsat preparatory committee and in other forums. Some in western Europe who had attempted to block progress of MARISAT pressed for a decision to junk MARISAT and develop a new all digital "Forward Looking System." This initiative died a well-deserved death and the MARISAT system was adopted by the time Inmarsat came into being in July 1979. As it turned out, the European satellite Marecs suffered some delays and so the initial Inmarsat service used all three MARISAT satellites as well. The earth stations on the air when Inmarsat started operations in 1982 were the three MARISAT stations at Southbury, CT; Santa Paula, CA; and Yamaguchi, Japan-so Inmarsat's initial system was in fact all MARISAT with an Inmarsat label. By chance, the 1000th MARISAT terminal was commissioned on the same day that the MARISAT system was turned obver to Inmarsat, February 1, 1982. Those who participated in the development and deployment of this system should take well-earned pride in their accomplishment.
* * *
I have always found it interesting that you all say "Inmarsat or MARISAT was a good program" rather than "the people that did Inmarsat or MARISAT did earth-moving things". Why don't you tell us what YOU did so we can record it??
And so they did-
I recall working on the signaling system and the multiple-access concept and taking part in selecting the voice processor that would be part of MARISAT. The first two were techy and interesting, but the last was memorable. At the time I was at COMSAT Laboratories, which was in the forefront of applying digital techniques to everything communications. When I became involved in the MARISAT work, there was no doubt in my mind that MARISAT should use a digital voice codec. The only question was, which one? After doing some work I decided it should use a delta modulation technique, which was then somewhat in vogue though it's long since disappeared into the communications history books. I took my case to the group at L'Enfant Plaza-Ed Martin, Tom Calvit, and perhaps Dave Lipke. (The associations have disappeared into history as well!)
Pretty soon, however, I got into discussions with Jack Dicks, who was equally convinced that low-index FM should be used and that digital techniques, including delta modulation, were the wrong choice for the MARISAT voice system. So the young digital upstart (me) and the experienced communications engineer (Jack) went at it, though with gloves on.. Finally, I believe, we organized a listening test. It was no contest. I had to concede that delta modulation had inadequate performance, given the link over which it would have to operate, and that FM should be used. History proved that to indeed be a wise choice. Since then, digital source coding techniques have made giant strides, and I think one would be hard-put to find an analog FM voice link nowadays. But in 1972, the state of the digital art was not sufficiently advanced, and FM won the day. Bravo, Jack!
Dick, thanks for the reminder of the interesting back and forth that all of us had back in the autumn of 1973 concerning the choice between FM and delta modulation. As I recall, we came at the comparisons with different concerns and had a rapid exchange of memos concerning the criteria for decision-making (e.g., quality of noise near threshold?). All in all, I remember the time as one of good collaboration between engineers at the Labs and in the MARISAT Program Office. I, for one, was grateful that you and the other Labs experts were in there pitching and helping with the program design choices.
Shortly after I left COMSAT (by then Lockheed) in 2000, I wrote down my recollections of how the MARISAT system design was rationalized and the rationales for the Inmarsat system evolution . Dave Lipke and Ed Martin provided feedback on my first draft, which I incorporated into a revised version of the article. Although the article has not been published, I did use it as an handout in a course I taught at GWU and part of it for a contribution I made to a the third chapter of a book published in 2004 (Communication Satellites-Global Change Agents, edited by Pelton, Oslund,& Marshall).
As Walt McKee notes, one of those old satellites is still providing a useful service after 31 years (launched 10-14-76). This seems to be a good time of the year for some positive nostalgic recollections and for being thankful that we had the opportunity to participate in a successful program.
Dan, thanks for the reminder. I have a strong recollection of sitting with you and Dave and, perhaps, others on this issue. We had decided that we could afford and needed transmission link margins for the shore to ship telex circuits but didn't have enough satellite power for the voice circuits. As in other design issues, solar panel margin, fuel margin, etc., we decided to dispense with the INTELSAT-type conservatism and include only margins that were clearly necessary. Otherwise, the program would never have gotten off the ground. You and Dave persuaded me that we could achieve softer degradation in the fringe reception areas with FM than with Deltamod.
Dick's interesting digital voice codec story set off a host of MARISAT networking memories.
Dr. Charyk's internal memo - more or less saying, if you are worth your salt, you should apply for a position on the MARISAT (Gapsat) program - threw a bunch of dissimilar (to say the least) engineers together to develop and implement a system, the like of which had never been seen before; aided and abetted by a small part time team of procurement specialists, lawyers, and imported folk with nautical experience, and overseen by a very patient program manager, Ed Martin.
At L'Enfant Plaza, the ground equipment weekly program meetings were rightly called" the Friday morning fights" as we tried to reach agreement on such bizarre subjects as the size of a mariner's finger (gloved or not) as he pushed the SOS switch.
Extending the system to the commercial networks brought with it memorable, and in some cases hilarious, gaffes (just like Connecticut trees, Scottish rail-mounted dockside cranes are not transparent to L- & C-band frequencies; and operators do not address the King of Jordan on his yacht as "Mr. King").
Due to the configuration, testing the complete system was not easy; which resulted in the Scientific Atlanta contractor essentially throwing the switch and seeing what happened. "Good God, it works!" was all he said; and it did.
Ivor, I remember getting the name of one Madeline Cantin, a retired AT&T overseas operator from you. She was not happy that AT&T had retired her at age 65 and she became a consultant to train the MARISAT operators. What a wonderful lady she was, and she has to be credited with most of the success of the MARISAT operators, their hiring and training, and helping getting operating agreements with overseas telephone companies and their operators.
Remember how the ship radio operators fought the MARISAT system, thinking it would put them out of a job when exactly the opposite happened? They just didn't get to use their Morse Code any more for message transmission.
Ed [Martin], I don't remember any of the operators and the ship operators getting married but the ship operators did visit the MARISAT shore station operators frequently when in port and showered them with candy and flowers. They apparently shared a unique bond that went on for many years and of course the comrade[ship?] they shared helped the acceptance of the MARISAT system by the ship operators and others-a tribute to Madeline Cantin, who probably had more influence on the operators than anyone else, even though she was never an employee but a consultant.
What a great group of people we had in the MARISAT System! Not many people get the opportunity we all had to usher in a new era, and it certainly brings back good memories. How lucky we all were to have been a part of it!
Many of us made a good living on MARISAT for many years. Last I heard, MARISAT F-2 is still serving the South Pole with Dennis Boiter waving the baton now. It is 31 years old. Pretty good for a satellite with a 5 year design life!
Many people can take pride in their contribution to the MARISAT program, including the technical authors of the program, the COMSAT and Hughes engineers who worked diligently to ensure successful construction, test, launch and deployment of MARISAT F-1(AOR), F-3(POR), and F-2(IOR) in that order, all in 1976, and the COMSAT General TT&C earth stations personnel at Southbury, CT, for AOR and Santa Paula, CA, for POR. The IOR TT&C operation was performed via the Fucino earth station in Italy. Engineers and Orbital Mechanics personnel at the COMSAT General Control Center managed and controlled MARISAT satellites initially from L'Enfant Plaza and later from COMSAT Labs in Clarksburg throughout their useful lives.
Many interesting events during deployment and operation of the MARISAT program have been chronicled elsewhere. However, end of life activities as remembered below, proved to be almost as interesting. The MARISAT F-3 command receive path became less sensitive over time, making it more and more difficult to command the satellite from Santa Paula. After much discussion, it was decided to have Willie Walters take a Command Generator to the Brewster earth station and utilize their 10-kW transmitter and 100-m antenna (which the station donated to the cause) to generate a sufficiently strong uplink to reliably command F-3. With monitoring from Santa Paula and direction from the Clarksburg Control Center, F-3 was successfully deorbited and turned off on September 26, 1996, after more than 20 years of service.
The MARISAT F-1 C-band downlink failed, ending commercial C/L-L/C service and interrupting C-band telemetry normally required as feedback for satellite control. Guy White's Southbury team rigged up an L-band beacon monitor and were able to utilize the L-band beacon spin ripple as a synch pulse for satellite maneuvers. We were able to provide limited (UHF) service for another 18 months before the satellite was deorbited and turned off via commands from Southbury on April 9, 1997, after 21 years of service.
For many of us, MARISAT was an important part of our COMSAT careers. Many dedicated and innovative people worked on the MARISAT program and exhibited the "can do" attitude that made the program a success. That same attitude was demonstrated on many programs by COMSAT people who made the company a great place to work. It is enjoyable now to share in reminiscing about these accomplishments with others who contributed with such enthusiasm to the success of the MARISAT program.
One thing for sure, the MARISAT program was never boring.
* * *
Those of you who wish to read more should turn to the June 1986 issue of the ITU Journal, a special issue on Mobile Services prepared by Ivor Knight with some valuable networking content from global players including Charles Dorian and Ed Slack. This might be useful (and less reliant on memory) for the dedicated early marine satellite communications enthusiasts. More importantly, it has a great cover picture of the French Flag "Calypso" in the Potomac off L'Enfant Plaza.
In closing, the editors also want to congratulate Howard Feldman, Director, Network Operations and Engineering, Inmarsat, who has seen the program through its final days.
Copyright 2007 by COMSAT Alumni & Retirees Association. All rights reserved.