Late in March, the Defense Department ordered Boeing to stop work on its Cluster 1 version of the new software-defined and-driven Joint Tactical Radio System (JTRS). The order, scaling down initial JTRS capabilities, reflects important emerging issues. A conventional radio is analog, but the signals it carries are digital. It is tuned physically to a frequency, typically by generating an internal signal whose frequency matches the target frequency. Much of the signal processing within the radio is analog; digital content is extracted at the last possible stage. Until recently, digital technology could not cope with radio frequencies. The price paid was limited versatility. Not only did the Defense Department have to buy different types of radios, but many could not interconnect. As the commercial world becomes increasingly digital, it must have seemed that the production base for analog radios would disappear. Radios would become more difficult to maintain, and more expensive because they would be less similar to their commercial counterparts. Slashing the variety of military radios would buy large dividends; the current estimate is that the number can be cut by two-thirds or more.
The ideal was to adopt flexible digital operation. Tuning would be digital. The incoming signal would be converted to digital quicker. In that way, information could be extracted from the signal. That might be significant for data links, like Linklo, in which the timing of the signal provides important content. The signal would be converted into data format and interpreted by the radio, so that it could be processed and passed on to another radio in, possibly, a different format for a different net. For really flexible fast digital radios this sort of operation would present no problems. The technology also promised a new capability: every radio would have an inherent signal intelligence capacity. Given enough power, the same device might also become a communications jammer.
In effect, JTRS is not only a radio but also an associated computer that makes sense of what it receives. This sort of technology is entering the civilian market, where hand-held computers and communication devices are merging. However, the civilian systems are limited in the nets they can access. That reduces their complexity. JTRS had to do more, because it was conceived not only as a universal radio but also as a gateway between networks. It incorporates a computer powerful enough for decryption and encryption, and, in at least many forms, a GPS receiver, so it is also a means of maintaining situational awareness.
Whether all of this was practicable depended on whether enough computing power could be packaged within the dimensions required. Cluster 1 is a vehicle radio for the Army. It is not the most demanding form of JTRS. However, it is more difficult than a radio intended for ships and for large aircraft. The JTRS project was inspired by the growth of computer chip speed, which seems to be governed by Moore's Law-every 18 months processing speeds double. What Moore's Law omits is that the heat output of a chip also increases, as does the power it drains to produce that heat.
The computer represents numbers as small voltages. Every time it performs an operation it has to change some of the numbers in its memory, which means that capacitors in its memory must be charged and discharged. The voltages involved must flow from capacitors to processor and back again. The wires involved have resistance and they heat up. Each operation takes power and generates heat. The faster the computer runs, the greater the power and heat involved. The solution has been to reduce the voltage involved. Unfortunately, the entire system is subject to thermal noise. The lower the voltage associated with a one or a zero in the computer's memory, the better the chance some random voltage incorrectly resets a tiny capacitor, so that the machine slips. Similar random voltages come out of the power supply that also is affected by heat. The faster the chip, the more heat it produces, and the more noise that goes with that heat. Thus a key packaging issue is whether the chips inside can be kept cool enough to function. That is why the fastest computers are refrigerated. Another packaging issue is power itself. Anyone who has used a laptop knows about power limits, because batteries typically do not last through a whole airline flight. It is also clear that next-generation chips, working with smaller voltages, require less power, and may actually last through a flight.
The current stop-work order seems to have arisen out of size, weight, and power problems, which would be critical for a vehicle radio. A larger question is whether the JTRS project exceeded what could be done. To the extent that the program was evaluated before any contract was awarded, the problem is more likely it was too expensive to try to reach the size, weight, and power goals. The question then is, what JTRS is worth? How much should be spent to solve its problems?
Actually the question is a wider one. JTRS is part of an attempt to rebuild U.S. forces based on command and control rather than weaponry. There is only so much money to spend. The current expansion in command and control that is justified on the basis that it promotes network-centric warfare comes at a cost in platforms and weapons. It is difficult to say how much improvement in communications is worthwhile. Historically, any major jump ahead in communications capacity has been overtaken by demand, yet it has never been clear that all of the demand was warranted.
The network idea is powerful, and it is a much-expanded version of what the U.S. Navy and allied navies have been doing for years with their digital data links. The network-centric name is somewhat misleading, in that the most basic point of the network is to provide each subscriber with a usable shared tactical picture. If the tactical picture is good, then it extends the subscriber's vision beyond the horizon defined by his sensors. The enemy is always under some surveillance, and surveillance can turn into targeting, so virtually all attacks are surprise attacks. How well the system works depends on how good the communication is. For example, it is vital that all contributors to the shared tactical picture know where they are and how they are oriented so what they send clarifies the tactical picture. At least in theory, the more who contribute to the picture, the better it can be.
This costs money, translating into fewer platforms and weapons. In theory, the better the picture, the more precise the targeting. Fewer only needs to mean fewer wasted rounds. It also can be argued that under expeditionary conditions the number of weapons available is not the number bought, rather the smaller number available in a forward area. In that case, the apparent tradeoff between weapons and communications may be illusory. Communications will be present on board all platforms, but platform weapons capacity is limited.
Unfortunately, the case for emphasizing communications and sensors has not been made. It is obvious they are valuable, but that does not help when the choice has to be made between money to cure a radio problem and money to buy enough tactical missiles. JTRS offers greater bandwidth than an analog radio, because its limits are set by software. How many bombs is 5% more tactical bandwidth worth? In an army unit, the digital radio may be able to handle more frequency-hopping patterns, so that more subscribers can share a single net. How many missiles are 20% more subscribers worth? At what point can one say we have done enough to cure our bandwidth problems, and attention should turn towards weapons or airplanes or ships?
JTRS seems to be the tip of a large iceberg. There are rumors the other major network projects, Global Information Grid (GIG) among others, are all running behind schedule and cost. That should not be surprising, because all of them are part of a drastic shift in the way the United States fights. Unfortunately, the shift has been little discussed, and even the means of measuring what is needed and affordable are lacking. None of that means the shift is pointless; far from it.
We find ourselves in an increasingly expeditionary situation. That is the case in the war against terrorism. In this war, the main function of large military forces is to deny the terrorists sanctuary where they can group and train. Because the terrorists are mobile, it is difficult to predict where they will concentrate. Quick operations can uproot them; sustained buildups will warn them to move elsewhere.
The war in Afghanistan illustrated this point. Even the war in Iraq can be taken as a case in point, because the great hope in Iraq is that spawning a decent society will have impacts on neighboring societies. The suddenness of the attack on Iraq minimized its cost, because it limited what Saddam could do to prepare (he did have time to organize post-collapse resistance, which is what we currently face).
Quick operations mean that we cannot use massed fires. We have to be smarter. The theory of network-centric warfare is that sensors plus communications can economize on the expendables such as ammunition. We are learning how to use whatever leverage we can create in this way. Is it by identifying and hitting an enemy center of gravity? By predicting the effects of particular attacks so as to optimize? By moving so fast that the enemy leadership collapses? Agility requires quick reliable communications. What we do not know is how much is enough. For example, should every vehicle's radio be able to communicate with all nearby command levels? Or are some vehicles more important than others? To what extent is the structure of a ground force based on what is practicable for human beings to understand, and to what extent on what communications net has been practicable in the past? More cannot always be better, but often it is.
Up to a point, spending more money brings an enormous return. Beyond that point, the money ought to go somewhere else. The challenge is to apply professional judgment to understanding just what is needed. Some form of JTRS is undoubtedly going to be part of the communications mix that emerges.
National Security Cutter Keel Laid
Homeland Security Secretary Michael Chertoff and his wife and ship sponsor, Meryl, stand with Philip A. Dur, President of Northrop Grumman's Ship Systems, holding the ceremonial plaque for the first National security Cutter (WMSL-750). The vessel is the first new construction for the Coast Guard's Deepwater program.