Editor's Note: An article on the DoD phone company in Chips? I'm sure some computer purists are scratching their heads and wonder- ing if I've
lost mine. However, when Alexander Graham Bell said, "Come here, Watson. I need you." What he meant was, "Hook up your modem and dial
my BBS." Obviously, hoping Watson could get a clear circuit. Corny? You're right - now that I have your attention....
 

Anyone who has served in the military or who has worked in a DoD office since the early 1960s has had experiences with the mili- tary's
long-distance phone system, universally called Autovon. Usually, these encounters involve frustration, long connection waits, frequent
cutoffs (referred to as being preempted) and occasionally poor reception. The only saving grace of the Autovon system was that it was free.
Right? Well, not really.

Autovon's notoriety grew as its coverage expanded. But, just where did Autovon come from? I wasn't surprised that nobody has ever
researched Autovon's history. It's like writing about the Q-tip. We take such a mundane, everyday tool for granted and never think about its
heritage or development. There's very little specific recorded history on Autovon's birth. The story is part of a corporate memory, currently
residing with members of the Defense Information Systems Agency (DISA), formerly the Defense Communications Agency. DISA manages
DoD's primary commu- nications worldwide.

Autovon had its beginnings with the Army's Switch Communications Automatic Network (SCAN), a three-switch system developed for their
own use. (A switch is a basic unit of an overall network and is usually an individual telephone system.) At this time, each service strung up its
own private networks according to requirements. Logistics bases would work circuits between them- selves. Quartermaster sections had
circuits to their counterparts throughout the country. Sometimes, one service would let another service use a few of its circuits to call a base, if
the sister service had an ongoing need.

At the height of the Cold War, DoD began looking for a common- user, long-distance telephone system that would survive enemy attack and
still give command and control capabilities to appro- priate levels of the government and the military. DoD selected the Army's SCAN as the
basis for a worldwide communications link, eventually listing SCAN as a DoD resource in 1963 and renaming it the Automatic Voice Network
(Autovon). By the mid 1970s, Autovon had been deployed in the European and Pacific theaters.

Jim Sage, Chief of DISA's Voice Network Operations Directorate, likened Autovon's structure to the public phone system. _In your system at
home, you dial 1, then a 10-digit number. You actually dial into a local system which then switches you into a long-distance network, passing
your call along until it reaches its destina- tion.

"We did essentially the same thing with Autovon. Post-camp stations had a small system that served all the users on the installation. If you
want to call downtown, dial 9, then the number. If you want to call long distance, your local phone system can be connected to the DoD
long-distance system, Autovon, by various methods. This is the Autovon long-distance network; it doesn't give you local service. It can be
compared roughly to AT&T, MCI or SPRINT long-distance telephone networks."

Autovon had some features that public service telephones lacked. Above all, it was a military communications network. The Joint Chiefs
wanted their command and control capability in a crisis or war. They wanted their phone system to be able to survive enemy attack - even if its
human users didn't - so they buried some of the Autovon switches underground.

To further ensure survivability, the system was so interconnect- ed that the loss of a few switches wouldn't affect the overall network.
Robustness was the watchword. Another feature of Autovon was multi-level precedence preemption (MLPP). There are various degrees of
importance regarding military phone calls: flash override, flash, immediate, priority and routine. People who might be calling from one finance
center to another to check on a serviceman's pay record would be classified as routine users. However, someone directing troop movements or
high-level security matters where decisions must get through, has flash override capability. When Autovon is saturated with calls, if the
support- ing trunklines are tied up, selected users with higher precedence will get their calls through by using MLPP.

As the far-ranging Autovon network grew, it became obvious that its ancient analog technology was out of date. Digital technology had made
tremendous strides, and DoD wanted to incorporate these advances into its long-distance phone system. By the mid-1970s, planning was
underway to replace Autovon. The new system was called the Defense Switched Network (DSN). The replacement cost was high, and the move
to DSN couldn't occur overnight. There were many switches involved in building and deploying DSN, while phasing Autovon out and
maintaining operational standards. DSN deployment continued through the early 1980s, mainly in the European and Pacific theaters. However,
the archaic Autovon was growing old and more difficult to maintain in CONUS.

The solution was what Jim Sage called "a technological shot in the arm." The Defense Commercial Telecommunications Network (DCTN)
included some of DSN's advances as well as the new capa- bility of video-teleconferencing. DCTN interconnected with Auto- von via a variety
of circuit arrangements, including one- and two-way links. DISA expanded DCTN throughout the late 1980s. AT&T, the prime contractor for
DCTN, as well as the original Autovon system, agreed to take out many of the old analog switch- es and replace them with new No. 5 Electronic
Switching Systems at no cost to the government. DISA could also take out more switches and further reduce the communication system's cost.
From 1988 to 1991, DISA claims to have saved $49 million in moderniz- ing the Autovon-DSN-DCTN system.

A common misconception is that DSN service allows free long- distance calls. In fact, DoD's overall annual budget for long- distance
communication is $289 million worldwide. This amount doesn't reflect the fact that much of the hardware is already bought and paid for. Much
of the money goes toward financing the cost for individual post/camp/station access and backbone trunk- ing.

When a user in Norfolk calls another office, say in California, the cost of that call is part of the overall budget and expense of communications.
Household phone consumers pay two rates for their services: a flat rate for local service and a call-by-call rate for long distance service. The
military setup is basically the same, with a little variation. DoD offices pay a flat rate for the local lines - the numbers you call by first dialing 9 -
and a user fee for DSN lines. However, the Navy, and the rest of the military, tailors its individual phone service to the local budget and
requirements of the particular military base. Using a shopping list supplied by DISA, a particular base may select two or three overseas lines, 10
transcontinental hookups and a simi- lar number of local lines. Each of the hookups is charged at a particular rate and makes up that office's
annual communications budget. Thus, each military installation has a specific number of DSN lines based on the available funds in its budget.

DISA uses a "P" (for percentage) factor to describe the success or failure rate of connections on DSN. Usually, the desired rate is P-10. That is,
for every 100 calls within a geographic area (referred to as a theater), 10 are blocked. Considering how many DSN calls are being made at any
one time, it's easy to see why we have so many failures, one of the most frustrating and time- consuming aspects of DSN. P-10 is included in the
linkage between the originator and destination. For instance you want to call California from your office in Virginia, there may be only 10 DSN
lines available from your base, which block three out of every 100 calls. After getting onto one of those 10 local DSN lines, you must now get
across the backbone network, which will block four out of every 100 calls, to the funnel of, perhaps, another 10 lines, at your destination, which,
in turn, will block three out of every 100 calls. At any stage along the road, your call could fail to complete. Adding up the numbers of blocked
calls (3+4+3), you arrive at the P-10 factor.

To further confuse things, some areas may enjoy a P-0! In Novem- ber 1991, Norfolk had an overall P-47 rating for DSN access. However, during
the same timeframe, NAS Alameda was rated at P-0, no trouble getting onto the DSN. In some cases, a rating of P-60 is not uncommon.

The current top five high-blocking DCTN (Navy) Access Areas are NAS Moffett Field, NAVSTA San Diego, NAS Lemoore, MCAS El Toro and
NAS Oceana. The P-factors for these areas range from 48 to 65. Funding will probably not allow the necessary increase in circuits to relieve the
congestion.

OK, so that may explain some of the difficulty in using DSN, but what about the cost? Again, the military pays a flat rate for DSN service. Thus,
the more you use DSN, the cheaper each call is. If your base pays $1,000 a month for a DSN line, and you make only two calls, then each call is
$500! Hardly economical. But, if you make 1,000 calls on the same circuit each month, the individual cost is only $1.

What about using commercial service when the DSN is uncoopera- tive? While it might not seem at first that substituting commer- cial calls for
DSN is wasteful, particularly on routine matters, it is. Consider the office worker in Norfolk who decides to check on his buddy in California, just
a short 5-minute DSN call to see how he's doing.

It's not uncommon for every one of the DSN circuits of a partic- ular base to be busy. But, perhaps one is open at the time the yeoman places
his call to his friend. At the same moment, another worker in another office has official business to negotiate. He picks up the phone, but the
vacant line is now carrying the yeoman's personal call. The second worker can't get through. He dials repeatedly, his frustration and sense of
urgency rising with each rapid busy signal. Finally, he gets permission from his boss to use commercial service. Now, that $10 commercial call,
probably made at the top mid-day rate, becomes an added expense that might have been saved. Of course, the usual reaction is that commercial
calls are figured into the operating budget, right along with DSN service. True, but in these times of drastic budget cuts, it is well to consider
how commercial calls can eat so far into the budget that there may come a time where the base commander tells his office heads, "Hey! I don't
have any money for outside long-distance calls. Tell your folks to use DSN."

Even with purely official calls, the DSN system is periodically saturated. Each November, AT&T notes a huge increase in the number of calls
coming into the Arlington area. All over the world, sailors know that this is the time when the advancement test scores are released. Detailers
and counselors are deluged with frantic inquiries about the caller's success or failure in making E-5 or E-6. (For the Air Force, this busy time is in
August, and the place is Texas.) In some respects, the military, beset with budgetary crunches and operational concerns, isn't worried about
easing the plight of the harried DSN consumer. Remember, the system was always intended as a command and control network for high level
government and DoD officials. Its use as a daily communications service for office workers was secondary.

Jim Sage talked about discussions between DISA and DoD. "We try to lean on the military users. We tell them, `Look, your circuits are
overloaded, and your people are angry.' We argue with them a lot. But the real story is simple: DoD is saying that they only have so much
money. `DISA,' they say, `you may be 100 percent right, but not only do we not have the money. But the money we thought we had has been
cut again.'"

"When the Navy in Norfolk says it can't afford the same services any more, we ask, `Well, what can you afford?' We try to tailor the service, but
usually end up taking out some of the circuits or services. And it's going to get worse. In DoD's defense, they're getting the best bang for their
buck. When the DSN lines are saturated during a busy day, they're getting their money's worth. And if a crisis arises, those authorized
precedence will be able to get through by pre-empting calls of lesser importance."

Will the service get better? What are the problems now? As in other areas of current military concern, economics play a large part in defining
DSN in the 1990s and beyond. DISA monitors traffic along its existing lines, much like those people on the side of a busy thoroughfare who
count cars during the rush hour. An internal program monitors DSN switches, samples call flow and tells system engineers what's happening. If
the number of calls rises dramatically in a particular area, DISA adds more trunk lines, although not immediately.

Outside the metropolitan Washington area, near Leesburg, Virgin- ia, in the small town called Dranesville, AT&T maintains a modern network
control center dedicated to monitoring CONUS DSN opera- tions. One of the minor phenomena of Autovon and DSN is the so called high and
dry connections. This abortive call occurs when, after getting on the DSN, and dialing your destination, the connection is completed but all you
hear is...nothing, dead air.

People usually hang up and try again. Eventually, they manage to complete their connection. What they don't realize is that the bad connection
- in reality, the bad circuit, much like a floppy disk's bad sector - is still there. Someone else will encounter it; maybe even the original caller if the
system is busy enough. DISA strongly recommends that consumers call the Dranesville control center and report a bad connection. The DSN
number to Dranesville is 550-1611.

While DISA and DoD have realized substantial savings in the last 15 years - $94 million, in fact - that money doesn't go back into the DoD
phone system. A JCS recent study revealed that with an extra annual $10 million, DISA could offer every CONUS military base a P-capability.
But DoD has other places to spend that money.

As we head toward the turn of the century, DSN will continue evolving into the planned integrated network its designers envi- sioned. Voice
and data services will combine into one network for local and long-haul transmissions, called the Integrated Services Digital Network (ISDN).

Acknowledgements: I would like to thank Mr. Jim Sage; LTC Stephen Kubiak, USA; LT Carlene Wilson, USN; and Ms. Beverly Sampson of
DISA; and CDR John Howard and Mr. Ron Olson of NCTC's Network Validation Department for their help.

About the Author: Mersky is the assistant editor of Approach, the Naval Aviation Safety Review. He has written or coauthored several books
on Navy and Marine Corps aviation. Mersky is a commander in the Naval Reserve. He can be reached at Commercial (804) 444- 7758 or DSN
564-7758.