The Air Force Can Learn a Lot from What It Has Already Seen in Cyberspace. Old Lessons
Total Page:16
File Type:pdf, Size:1020Kb
The Air Force can learn a lot from what it has already seen in cyberspace. Old Lessons, “New”By Rebecca Grant Domain Cape Cod radar tower in Massachusetts was a prototype for the SAGE air defense system. SAGE needed computers with memory, digital relays linking radar sites, and systems engineering to bring them together. 86 AIR FORCE Magazine / September 2013 ashington is once again strategic challenges, such as continental wrestling with how to tack- air defense in the 1950s and real-time le the military challenges command and control in the 1980s, of cyberspace. “The rise of fueled progress in the exploitation of Photos via MITRE Corp. cyber is the most striking cyberspace. development in the post-9/11 national In its infancy, the domain of cyber- Wsecurity landscape,” Chairman of the space did not look much like the clouds Joint Chiefs of Staff Army Gen. Martin and commons known around the globe E. Dempsey said in a June 27 speech at today. The special qualities of cyberspace the Brookings Institution. Dempsey said emerged only when computers gained about 4,000 new military cyber positions more memory and power and networks could be created. Perhaps 1,000 of those linked them together. may be within the Air Force. Responding to new growth in the In the (Cyber) Beginning cyber mission poses a challenge to the Back before social media, the World Air Force. Over the past decade, the Wide Web, the fi rst emails, and even USAF position has swung from taking before ARPAnet, the fi rst closed cy- a vigorous lead in the mission area to berspace system was the Air Force’s going slow on cyber—to avoid a po- Semi-automatic Ground Environment, tential “black hole” as Air Force Chief or SAGE. of Staff Gen. Mark A. Welsh III termed The Air Force bought and paid for it in late 2012. SAGE, which was arguably the fi rst true As the debate continues, it is important cyberspace environment. to recall that cyberspace is not new ter- SAGE’s intent was to direct continen- ritory for airmen. The Air Force made tal air defenses to intercept attacking its fi rst deliberate move to create a cyber Soviet bombers. From 1949 on, the entire force structure almost 20 years ago. United States was vulnerable to nuclear “The longer we think cyber confl ict is attack from Soviet bombers refueled near new, the more we will repeat the same the Arctic Circle. mistakes and relearn old lessons,” wrote The defensive problem had grown too Jason Healey, director of the Cyber State- complex and immense for the grease craft Initiative at the Atlantic Council, pencil and telephone line methods of in Air University’s Strategic Studies World War II to remain effective. Quarterly in fall 2012. MIT professor George E. Valley Jr. Cyber lessons mark out a heritage dove into the problem as a member of every bit as interesting as biplanes and the Air Force Scientifi c Advisory Board. bridge bombing. In fact, USAF can look Valley visited an air defense site in Mas- back at six decades of involvement in the sachusetts and was horrifi ed by the old domain now called cyberspace. Air Force equipment and procedures. An early “cyber warrior” in 1959 uses a light gun to target potential intercept coordinates. AIR FORCE Magazine / September 2013 87 Austin Mills photo via National Cryptologic Museum Museum of Science photo viaWikipedia Above: The disk containing the 99 lines of code comprising the Morris worm. Right: Frostburg, a supercomputer programmed to perform higher-level mathematical calculations for the National Security Agency, operated from 1991 to 1997. Valley briefed Air Force Chief Sci- receive radar data and entist Louis N. Ridenour, who then per- respond with an inter- suaded MIT President James R. Killian cept path in real-time, Jr. to establish a new laboratory at MIT thanks to its nascent and use Air Force money on air defense electrostatic random research. They also hoped to stimulate access memory and the information electronics industry. programmable read- SAGE was a unique step toward only memory. The Air Force took over line. This was one of the earliest work- cyberspace because the system was funding of the computer from the Offi ce able modems. conceived from the beginning as an in- of Naval Research. Airmen also got their fi rst taste of formation architecture. “SAGE was one Digital computer maturation in the working in a computer-driven, interactive of the fi rst systems to include immediate, SAGE project “laid the foundation for environment linking sites all over the interactive man-machine communication a revolution in digital computing, which country. This was an important early step via displays, light guns, and switches,” subsequently had a profound impact on in building up the cyberspace domain. noted a 1974 RAND report on future the modern world,” summed up MIT’s “The primary responsibility for humans USAF command and control software offi cial history. To boost performance, in the SAGE system would be their in- requirements. researchers developed magnetic core teraction with computers through the use To work, SAGE needed computers memory and bolted it on to Whirlwind. of keyboards and other devices in order with memory, digital relays linking radar Magnetic core memory became the to specify which of the airplanes picked sites to command and control nodes, industry standard for the next 20 years. up and followed by radar and shown and systems engineering to bring them SAGE also utilized primitive mo- on the computer cathode-ray monitors together. dems—the skeletal structure of cy- should be targeted,” summarized Thomas Air Force requisites for SAGE carved berspace. Scientists at the Air Force P. Hughes in Rescuing Prometheus, his out many of the tools for cyberspace. Cambridge Research Lab in Cambridge, landmark book on innovation. First was a fast computer with program- Mass., also fi gured out how to convert SAGE also stressed systems engineer- mable memory. The MIT campus had a analog radar into digital code and ing skills because of the diffi culty of computer known as Whirlwind that could transmit over a dedicated telephone developing and exploiting new computer 88 AIR FORCE Magazine / September 2013 and communications technology. Instead, Baran envisioned a series of The National Security Agency went off In retrospect, “the military require- backup centers with commanders. As and built its own version of the ARPAnet ments for SAGE sytem placed it long as they all had good information, called Platform. beyond the leading edge of soft- nearly any one of the senior military For the Air Force, this had two cyber ware technology,” noted a 1974 offi cers in charge of the set of centers implications. Cyberspace would grow in RAND report. could make a good decision on how deep secrecy at NSA as signals and intel- It took until 1958 for SAGE to to cope with incoming attacks. Baran ligence analysis became closer entwined become fully operational. By then, described potential non-hierarchical with the cyber world. Airmen would be it was also obsolete. Still, SAGE network formats starting with a simple closely involved in that work, too. marked the fi rst major commitment “round robin” network. The key was However, cyberspace would also grow of USAF dollars and expertise and “distributed computation, or totally from the worlds of education and business. provided components of the future independent apparatus at each node” With commercial companies producing cyberspace domain. providing such routing “without reliance faster, more capable computers, USAF One of SAGE’s fl aws was its upon a vulnerable central computer.” turned its focus to tactical applications hierarchical communications de- In Baran’s concept, the message would for information technology. sign. What if a Soviet attack wiped travel over the shortest path, carry a Air Force Systems Command com- out communications links and security tag, and have its geographical missioned a study in March 1971 titled blinded SAGE and its successor point of origin authenticated. The system “Information Processing/Data Automa- air defense systems? Survivability as a whole would be set up to identify tion Implications of Air Force Command of command and control in time malfunctions right away. Baran’s work and Control Requirements in the 1980s.” of nuclear attack was a big preoc- on Strategic Air Command’s nuclear Its purpose was to scope the information cupation in the 1960s as the Soviet command and control problem laid out processing technology anticipated for Union increased its bomber and a crucial turn in the roots of networking command and control of Air Force combat missile forces. and the Internet to come. units in the next decade. A RAND Project Air Force Software was becoming the problem of researcher named Paul Baran took Platform via ARPAnet the 1970s; USAF was already spending on the problem and ended up with By the 1970s, much of the ground- almost $1.25 billion per year on soft- another big push in the develop- breaking work in cyberspace was tucked ware—three times more than hardware ment of cyberspace—a theory of under the Pentagon’s Advanced Research spending for automatic data processing. distributed communications. Projects Agency. ARPAnet was not con- And software progress was already at- Baran saw right away that ac- ceived as a military communications tracting complaints. “Software has yet curate Soviet intercontinental bal- project. Instead, the main motivation to live up to its potential in [command listic missiles spelled big trouble was to facilitate time-sharing by linking and control] systems,” noted the 1974 for the current system. together powerful computers that were RAND report. “The proven development of geographically separated.