Tomorrow's Insta-Weapons
This is Nicholas Weaver's second article on the military impact of the spread of technology.
America owes a big chunk of its military superiority to what it can make – the tools, facilities, and expertise needed to put together sophisticated planes, ships, and weapons. So what happens when much of the high-precision manufacturing behind Predators and F-22s can belocated anywhere and owned by anyone?
The day isn't as far off as it might seem. Twenty years ago, if a designer wanted a new, high-precision part, he constructed a design and handed it to a skilled (and expensive) machinist, who would produce a prototype. If more than a handful of handmade parts were needed, an even more expensive set of tooling would be created.
Today, the same designer develops his plan on a computer and then emails the design to the machine shop, which uses an assortment of CNC (Computer Numeric Controlled) machine tools to produce the prototype. CNC lathes can turn effectively arbitrary radially symmetric parts, cutters can create 2D shapes, and CNC mills can cut a 3-dimensional part to extremely tight tolerances. And as long as the designer only needs a few dozen (or even a few hundred) parts, the shop just feeds more material into the CNC machines and out pops more parts.
All it takes is a few shipping containers, a power hookup, and a source of refined metal ingots to produce high precision parts and designs from a high-technology mobile factory. And if that doesn't seem like it's got much to do with the military, think again. A major reason why the AK-47 is a "Weapons of Slow Mass Destruction" is it's easy-to-make design. Any country with a factory base up to the low standards of Russia circa 1947 can stamp them out en masse – and wreak havoc. Think of the possibilities when small jet turbines or piston engines for cheap unmanned planes become so simple that anyone can stamp them out and build their own drone air force.
CNC systems have become almost ubiquitous for manufacturing, from $2500 Sherline desktop mills (for a complete system including a linux computer and GPL software) to million dollar high-throughput systems with automatic material feeds.
Anyone watching American Chopper has seen the CNC controlled water-jet used to cut custom wheels, with rims limited almost solely by imagination.
The US Army is starting to exploit these fantastic tools, prototyping and deploying a mobile repair yard, the Mobile Parts Hospital, which, instead of keeping a large inventory of spare parts, is able to produce replacements on demand.
The biggest limitation is one of design. Parts produced by forging, stamping, casting, or extruding can't necessarily be replaced using pieces made on a CNC mill. Clean-slate designs don't have this problem. If all the high-precision components were designed to use CNC-produced parts (with a conversion to higher-volume production techniques if necessary), CNC-replacements, and even entire CNC-based manufacturing are now straightforward.
Which brings us back to transportable factories. Write a (large but reasonable) check, place a few CNC machines in a shipping container, add a couple of containers of refined raw material, and now anywhere the containers go a factory resides. Be it critical spare parts for a broken well, a replacement piece for an automobile or the critical components of a rocket motor, the same factory can make all three. A huge revolution for the majority of the globe which remains largely unreachable by FedEx.
In the CNC world, proliferation becomes a matter of design, software, and materials, rather than finished systems. What happens when North Korea or Iran starts selling missiles as digital files rather than on ships which can be intercepted? When private designers and companies create designs which anyone can produce? Two words: Watch out.
-- Nicholas Weaver
Red Teaming Tomorrow's Radars
Nicholas Weaver is a researcher at the International Computer Science Institute in California. This is the first in an occasional series for Defense Tech.
In the past, military technology might have consistently outpaced civilian gear. Not any more.
Civilian electronics, manufacturing, and development cycles have radically shortened and improved. The computer which runs the F-22 is an absolute design marvel for its time, for example: 700 MIPS (Millions of Instructions per Second), approximately 300 Megabytes of memory, and some 20 billion DSP [digital signal processing] style operations.
Yet its time was the late 80s and early 90s, when much of the hardware was finalized. Today, a Playstation 3 meets or exceeds this performance, for $600 instead of perhaps $30,000,000. (Of course, the F22's avionics are considerably more robust and presumably more reliable.)
So the question becomes, what happens if America's opponents start massively adopting commercial technology and commercial design styles? In Iraq, insurgents are already using commercial gear to build and trigger bombs. But it's not hard to imagine absorption on a much broader scale. After all, the weapon business is a business, there are brilliant engineers around the world, and the basic building blocks continue to grow more sophisticated.
This occasional series of speculations will attempt to predict that future, by technological "red-teaming," sketching out what an opponent could do. This first article attempts to postulate what the future of air defense radar will be, and how it will force radical changes in US military operations.
The United States enjoys pure air superiority. No other nation can hope to match the USAF, and no other country will likely try. But an opponent doesn't have to match our fighters, they only need ground based air defenses, which starts with radars.
Today, they don't have much of a hope. Between stealth aircraft and anti-radar missiles, an opponent's air defenses will be destroyed within minutes of a conflict. , or simply remains offline in an attempt to preserve some capabilities. {Which is what the Serbs did in the 90s – keeping their radars off, mostly, and using ballistic firing.)
But there is a technology which might change this balance. And it's got its roots in the commercial world. Multipath radar would provide a defender with a robust radar system, able to detect and track many stealth aircraft, counter anti-radar missiles, and enable the defender to track all radio emitters within the country.
In a conventional radar, a radio signal is broadcast. When a plane or other object is in the path of this beam, it may be reflected back towards the radar station. By using timing, direction, and the size and intensity of the reflected signal, the radar site can track and identify objects. Yet it is this very radar signal which anti-radar missiles target, making the stations vulnerable to attack.
Stealth aircraft avoid radar by being made of materials that are either transparent to, or absorbing of, the radar's signal. Or, the planes scatter the radio signal so that it bounces away from the radar station. That's why stealth aircraft have such unusual shapes.
But there is another way to build a radar. If you scatter a bunch of radio sources around the countryside, each of which are broadcasting, the signals will scatter off any aircraft in the area. With a group of distributed receivers, these scattered signals can be received and analyzed. This is called "multipath radar", as the signals traverse multiple paths to receivers.
There are a few prerequisites for multipath radar. The broadcasters, although simple, need to transmit an identifier as part of their signals, and be at known locations. The receivers, on the other hand, need to be very sophisticated. This requires sophisticated radio antennas and, more importantly, "serious DSP magic," which, when networked together, can compute a cohesive picture of the defender's airspace.
Yet the hardware to perform such DSP operations is becoming commonplace and commercially prevalent. GNU radar and other designs can receive the signals, and conventional computers and DSPs can then process the results, extract the features, and create an overall picture. There have been prototypes built in the United Kingdom, able to track commercial aircraft by observing the reflected signals from cell-phone towers.
Why do I believe multipath radar will be a case where civilian technology may have a huge military impact? Simply because the "serious DSP Magic", the signal processing components and programming skills needed to make everything work, are the same principles behind spread-spectrum cellular basestations, software radios, and even MIMO antennas for 802.11N basestations.
If multipath radar is deployed by adversaries or potential adversaries, it could greatly affect US operations. Stealth aircraft based on scattering the signal are simply not stealthy to multipath radar. Worse, the transmitters are no longer co-located with the receivers and electronics. Thus anti-SAM and anti-radar tactics will need to be restructured, as simply blowing up the transmitters destroys valueless targets and an adversary could simply build more $500 transmitters than the US has anti-radiation missiles.
Finally, the same DSP processing and antenna infrastructure which forms a multipath radar also enables the defender to track radio sources, by detecting unique sources and using timing to triangulate their locations. Simple traffic analysis, knowing where your opponents are, can be invaluable for military strategists. Radio silence protocols would need to be strictly enforced and enhanced, which could also affect proposed "system of systems" technologies.
A new technology can change the world. Multipath radar might change how the US military needs to operate, both in the air and on the ground. And the building blocks are in catalogs, now.
-- Nicholas Weaver
