Tempest fax: standards fiction? Will the 161B Standard make a difference?
Will the 161B Standard Make a Difference?
Why is nothing easy in the Tempest world?
While the Tempest industry in general recovers from the twin hurricanes of new certification procedures and a new national policy almost guaranteed to reduce demand (see "Tempest Tension," JED, February 1990, and "Tempest in a Teapot?" JED, July 1989), smaller storms continue to rage in the community's outlying areas. One of these squalls clouds the Tempest fax marketplace, where attempts to mandate interoperability among machines and among user groups has led to contract protests, government investigations and general confusion. As appears endemic in the Tempest world, the long-term benefits of new methodology are being traded off the short-term disadvantages of increased costs to the government purchaser.
SHRINKING THE WASHING MACHINE
The present turmoil comes at a time when the demand for Tempest faxes has exploded.
Tempest fax units have been around since the mid 1970s, when a company called Dacom (which was subsequently purchased by current market leader Ricoh Corp.) introduced the 412. The 412 operated at 9,600 bps half-duplex, possessed its own modem, was compatible with KG30 encryption equipment--and was the size of a washing machine. It also cost $18,000. Following the lead of commercial units, secure faxes gradually shrunk, both in size and cost. Competitors to Dacom/Ricoh entered the market, adding a variety of equipment and giving purchasers a choice of machines.
The variety of equipment and the quest for smaller size are the roots of the present interoperability dilemma. Tempest technology does a fine job of preventing stray emissions from the fax units themselves; however, it does nothing to enhance the security of the transmissions leaving the machines on the data line. That task is the responsibility of encryption units like the KG series and other security devices like Secure Telephone Unit (STU) II and III equipment. Thus, few if any Tempest faxes were far from an encryption/security device, which linked them to the communications network and completed the security loop.
Since encryption units acted as modems, fax vendors removed the internal modems from their machines and further lessened their units' sizes. However, this left manufacturers with the problem of interacting with the security device, different generations of which were not compatible with each other. Therefore, each fax machine had to carry a variety of protocols for an equal variety of security equipment -- a task each vendor approached in a unique way. The result was that even if you could match encryption equipment, there was no guarantee that machines from, say, Ricoh, Valutec and SSTI would communicate with each other.
The expense of encryption equipment and secure lines added to the fax community's problems. STU-II units, for example, could run up to $20,000 apiece. Even if one could afford several sets, they were difficult to get, required two phone lines and provided relatively poor voice quality. The result was that most military installations had only one or two secure lines (most often found in the communications center and perhaps in the commander's office). Necessarily, that meant only one or two secure fax machines per base as well, a hardship for users who had to fight for transmission time and for vendors trying to sell equipment.
To the rescue rode STU-III. The latest generation of secure equipment (each unit carries a phone, encryption device and a modem) represents a significant improvement over its predecessor, particularly in price. Present STU-III units sell for approximately $3,000 each; the government expects this price to drop by another $1,000 in the near future. Such a deal has proven hard to pass up: while approximately 11,000 STU-II units were shipped during that generation's lifetime, about 200,000 STU-III devices have already been delivered.
"The STU-III has made all the difference as far as the booming of the [Tempest fax] marketplace goes," said Robert Johnson, government marketing manager of Ricoh's Communication Products Group. "What is happening now is that people who never had crypto before, never had a secure telephone before, etc., are getting them on their desks instead of comm centers." With a secure line so close, users can now enjoy the advantages of desktop fax units and still meet security requirements.
The influence of STU-III on the Tempest fax market is so pervasive that when asked to estimate the market's size, each vendor JED contacted resorted to taking a percentage of the STU-III universe. Common wisdom says there will be one fax machine for each 10 STU-III lines, making the potential market approximately 20,000 machines. While not all of these will be Tempest units (the new national policy that narrowed Tempest requirements will be felt in these applications too), the sales potential is certainly larger than the 5,000 units one vendor estimated were in place before STU-III.
Yet while STU-III has allowed military and civilian agencies to be able to use secure fax machines more frequently and cheaply, it only dented the interoperability problem. Certainly STU-III users faxing to other STU-III users did not have to worry about encryption compatibility (even though faxing to non-STU-III locations remained just as difficult as before). However, the same machines that were unlikely to talk to each other previously were being hooked into STU-III networks, with similar incompatibility results.
Fed up, the Defense Communications Agency (DCA) gave the Joint Tactical [C.sup.3] Agency ([JTC.sup.3] A) at Ft. Huachuca, AZ, the task of defining a standard that would meet users' needs and ensure interoperability. The result was the MIL-STD 188-161A, issued July 4, 1988. This standard was a near cousin to MIL-STD 188-161, which was issued several years previously as a standard for tactical operations.
Unfortunately, two problems arose immediately, according to industry critics. The first concerned the tactical nature of the standard; it appeared that the [JTC.sup.3] A did little more than stretch the tactical specifications to include strategic applications, critics allege. This forced strategic users to pay for a transmission rate more than three times as fast as they were likely to use, as well as scan rates and operating modes that were unnecessary.
According to Johnson, "What the strategic guys were saying is ...[you] put out a Mercedes [spec], and I want a Chevrolet."
"To [[JTC.sup.3] A] the whole communications world was tactical; they didn't know the strategic market was even out there -- the `long haul' market, as they call it," said the manager of federal government marketing at Ricoh's Communication Products Group, David Shaver. "They took everything and tried to make everything a tactical piece of equipment, and that was the wrong approach."
Charles Zumbaugh, VP of government operations at Time and Space Processing, Inc. (TSP), agreed. "They erroneously assumed that the new MIL standard should incorporate all the other faxes, which they thought were a minor part of the overall fax world -- when in fact the strategic or the dial-up world will probably dwarf the tactical world by a ratio of somewhere [close to] 5:1 or 10:1."
What struck closer to home, however, was an alleged lack of coordination with industry in setting the standard. The Electronic Industries Association is a prominent communications channel between the government and industry on standards matters. Shaver claimed that the first time his chief engineer, who is part of the association's TR 29 Committee on facsimile, saw the 161A standard was a month after it was released. In fact, only one vendor, Cryptek Inc., was in a position to offer a 161A-compliant machine when the standard was issued.
"They put the standard out... and they said, here it is, it's mandatory," complained Johnson. "It caught Ricoh, it caught TSP, it caught [fax vendors] Ilex [and] SST...totally off guard. If you caught one of those guys off guard, it's because they didn't do their homework. [However,] the only person it didn't catch off guard was Cryptek."
The DCA turned up the heat when its Defense Commercial Communications Office (DECCO) issued a huge RFP two months later for 161A-compliant fax machines. While most vendors, still uncompliant with 161A, were uncomfortable, the RFP contained a clause which would allow the winning company to ask for a one-year waiver of compliance to allow it to catch up. However, that waiver was removed the following February -- and all hell broke loose.
Battle lines were quickly drawn, with the noncompliant vendors on one side and the DCA and Cryptek (which was rather happy with the status quo) on the other. The noncompliant camp quickly fired their biggest weapons, filing complaints with the Government Accounting Office (GAO) and the Inspector General's Office that the lack of a waiver created a de facto single-source procurement (and therefore was anticompetitive). They also claimed Cryptek had received an unfair advantage from DCA, including being granted meeting privileges with the [JTC.sup.3] A's Technical Standards Office not accorded to other companies during the solicitation and being tipped off about the standard a few days before it was released. The DCA and Cryptek denied these allegations and said dropping the waiver was the only way to ensure 161A-compliant machines would be acquired.
While the body count has not yet been completed, it appears the noncompliant army can claim victory. The protest to the GAO caused the immediate suspension of the RFP. The GAO then ruled that DECCO could arrange for sole source (i.e., Cryptek only) 161A procurements only for the waiver year it would take the rest of industry to develop 161A machines, and only in the quantities necessary to meet immediate needs.
More spectacularly, the Inspector General's Office jumped on the procurement with both feet. On January 2 of this year, it upheld an overwhelming majority of the noncompliant group's complaints -- including partially substantiating the charges of unfair advantage to Cryptek. It recommended, among other points, that the RFP be canceled and reissued with the waiver provision intact and the 161A standard itself be revisited through the formation of a [JTC.sup.3] A working group. It agreed with the GAO, however, that immediate needs could be met through sole sourcing.
Naturally, Cryptek was less than pleased. "The IG report is fraught with inaccuracies and illogical conclusions and bad premises," said Neel Price, Cryptek's VP of marketing and sales. Allegations that the DCA showed favoritism to his company are "clearly not true"; he pointed out that several rebuttals to the IG report (including one from Cryptek) were in the process of being filed.
Chastened by the criticism of the 161A standard and the outside scrutiny the resulting protests attracted, the DCA decided to go back to the interoperability drawing board. The result is a compromise between tactical and strategic requirements -- MIL-STD 188-161B.
Many observers describe 161B as a subset of 161A, and indeed the former looks like a stripped-down version of the latter (see Table 1). The standard, which superseded 161A on March 30 of this year, mandates much lower performance levels than the old tactical standard. The high-performance specs are now available as options -- as are capabilities not contained in either standard.
"The government customers can now have more say in specifying what they actually need to support their missions," said Shaver. "In other words, they can take this standard, they can tailor some of the requirements in that standard to fit their needs and they don't have to buy the whole ball of wax."
Yet while most of the non-161A-compliant companies are happier with 161B, they are not turning cartwheels. Said Zumbaugh, "Its objective -- to assure interoperability -- while that's a good objective and we applaud it and we support it, the standard that is in place right now still has some serious flaws which will almost guarantee that there won't be interoperability."
The principle causes of concern are transmission parameters not covered by either standard, particularly multipage transmission, the time between the transmission of pages and transmitting terminal identification. "It really leaves to interpretation how [these parameters] should be implemented," complained Zumbaugh, "and you can bet money that the key players are going to implement [them] in a different way."
Even without these problems, some doubt the usefulness of the standard. "Even though users now have to buy a 161B-compliant terminal, I don't foresee anyone operating in that mode, at all, for at least three to five years," predicted Shaver. "Obviously, the existing base of terminals that are in the field right now don't have 161 protocols. So the only mode [fax units] can operate in and be able to talk to the existing base of terminals -- which certainly is a mission requirement -- is to operate in the bypass mode."
"The only thing the MIL standard is really going to do for the next couple of years is increase the price [of equipment]," said Zumbaugh. "Everybody is going to have to talk to all these units that are already out there, and they're modified Group 3 [a commercial transmission standard] machines. There isn't going to be one solicitation that's going to come out and say meet the new MIL standard without also saying you also have to meet the modified Group 3."
Another problem, in the words of Shaver, is that 161B "is not a very efficient protocol," because it is based on older technology. "If I ever use it, and I want to move a document from here to there, I'm going to do it the fastest, most expeditious way that I can do it -- and that certainly isn't 161B."
However, hope is not entirely lacking. "I think that industry is trying to work with DOD on this standard, which was not the case in the past," offered George Kinzel, government marketing manager of the fourth and final major company in the Tempest fax market, Ilex Systems Inc. "So I think things will be done in the standard to make it more user friendly." Kinzel pointed out that the TR29 Committee is currently working on the missing parameters. "My feeling is that a lot of these concerns we have will be worked out over the next year and recommendations will be made and, hopefully, implemented."
An important point for both users and vendors, however, is the realization that in some cases interoperability is not only impossible, but undesirable. Security concerns demand at least a measure of exclusivity, which differing communications technologies only serve to buttress. "There are going to be, by design, nets that can't communicate with each other," Zumbaugh explained. "The largest common net will be STU-III, and then the next one will be the tactical world. Each one of those will be a little bit different because not only do they use different crypto equipment, but they use different communications devices. Some NATO forces use HF and others use UHF/VHF and others use satellites -- and they're not going to talk to each other."
PICK AND CHOOSE
Because of the vendors' sometimes conflicting viewpoints on the events that have led to the 161B and the standard's relative merits, users can expect to be bombarded with a potentially bewildering array of sales pitches. DECCO is planning on rereleasing the RFP which started the standards controversy, but this time couched in 161B terminology; while it had not hit the streets at the time this issue went to press, industry observers expect to see it before summer begins. The jockeying, lobbying and finger pointing this procurement inspires should be instructional for future purchases.
In the meantime, a review of the present offerings might be educational for our readers planning on purchasing a Tempest fax. Each of the four companies has a flagship machine (in some cases, it is their only machine): Cryptek has the Transcrypt 10, Ilex its 750T, Ricoh the 2112T and TSP the 9100A. All four are said to be complaint with 161B; however, only Cryptek has had its claims verified by an outside agency. (The Transcrypt 10's 161A compliance was certified by an arm of the DCA -- another example of favoritism in the eyes of rival vendors.) Unlike the 161A regime, however, the [JTC.sup.3] A will make certification facilities for 161B available to vendors -- at a price. Users can expect their suppliers to pay that price and earn certification as quickly as possible.
Keeping in mind that 161B requires less horsepower than 161A, users should not assume that just because all the machines are said to be 161B, they have similar capabilities. Again, Cryptek's machine is a 161A unit, with all the capabilities that standard required. Ilex and TSP have units more closely in line with the new standard, although they offer some (not all) of the 161A features as options.
Ricoh's position is somewhat ironic. At the same time it was fighting the DCA over 161A, Ricoh was scrambling to make its 2112T unit compatible with the standard. It succeeded by developing a protocol converter that brought it up to 161A -- just in time to see 161B supplant it. So after successfully clearing the way for Chevrolet users, Ricoh has only a Mercedes to offer.
The protocol converter also poses a Tempest problem for Ricoh, for while the 2112T is Tempest certified (courtesy of Mitek Systems Inc.), the protocol converter is not. Cryptek's unit is listed on the moribund Preferred Products List, while Ilex's 750T has made its way onto the new Endorsed Tempest Products List. TSP expected its 9100A to join Ilex's fax on the new list by the time this issue of JED was mailed.
The differences in capability have a definite correlation to cost. Cryptek and Ricoh offer their machines for similar prices -- roughly between $8,000 and $10,000, depending on options. These prices are greater than those charged for the lower-power Ilex and TSP units, which should cost between $6,000 and $8,000. Users should watch closely, however, because some companies -- particularly Ricoh and TSP -- indicated that their pricing strategies were not yet solidified.
The key for Ricoh and Cryptek will be how many RFPs call for higher-order machines. "If we've got the [161A], that will be a plus for us in some cases," said Shaver. "But in other cases, it won't be because we can't be cost competitive with the [161B] terminal."
Cryptek also is aware of the risk. "If RFPs come out and they don't need 32 kilobits [transmission speed] and Cryptek's product is too expensive, then probably someone else's machine will be acquired," Price predicted. "I think that a lot of users will not specify higher than 9.6 kilobits [the 161B minimum]. But the question will be will we lose procurements because of cost. And I think that remains to be seen."
Neither company will be stationary, however. "Our strategy was that early on, government users would be buying fax machines as shared resources," said Price. "And therefore those fax machines would be busy most of the time and would have to be high duty cycle types of machines. As more and more machines are shipped, the need for high duty cycle machines is somewhat supplanted by the need for medium and low duty cycle machines.... And so I would foresee us moving in that direction." Future Cryptek machines will be smaller and lighter, Price said.
Ricoh, meanwhile, is hard at work on a 161B machine -- one that does not require a protocol converter. Shaver said the company hopes to have the unit ready by spring of next year -- provided it can be Tempest and 161B certified by then.
TSP and Ilex will probably play to the present strengths of their machines, at least for the time being. Zumbaugh pointed to the 9100A's small size, light weight and ease of maintenance as its most positive features. Kinzel indicated his unit was ideal for the STU-III market and aggressively priced.
IT TAKES EDUCATION
The key for users is education -- learning exactly what capabilities they need and how to specify correctly to get them. Shaver, for one, is not optimistic. "Everyone is specifying [MIL standard] and I think, to a man, none of them know what they're specifying. They don't have a clue of what it is; all they know is they've been told to do it." Taking the time to research the standard, then research requirements, will be the users' challenge -- and the deciding factor in determining whether the effort to reach 161B was a waste of time. [Tabular Data Omitted]
PHOTO : Cryptek's Transcrypt 10 is the only machine to meet 161A without the use of external
PHOTO : devices.
PHOTO : TSP's 9100A Easy-Fax is a lightweight, portable 161B unit.
PHOTO : Ilex's 750T is one of the newest entrants into the Tempest fax market. The 161B-compatible
PHOTO : unit operates in both synch and asynch.
PHOTO : Ricoh's 2112T, Tempest certified by Mitek Systems, has added a protocol converter to meet
PHOTO : 161B up to 161A levels.
|Printer friendly Cite/link Email Feedback|
|Author:||Hardy, Stephen M.|
|Publication:||Journal of Electronic Defense|
|Date:||Jun 1, 1990|
|Previous Article:||MIL-SPEC computing and the rugged compromise.|
|Next Article:||Single station location HF direction finding.|