Dr. Dobb's Journal April 2000
Speaking at the RSA Conference 2000, Vinton Cerf asserted that "we're still in a very infantile state when it comes to security on the Internet." Cerf, who with Robert Kahn invented TCP/IP and is considered one of the founders of the Internet, called for a "more widespread public key infrastructure" to improve Internet security. Cerf cited the rapid growth of "intercorporate exchange" (also called "business-to-business e-commerce") as an urgent reason to improve Internet security. Another reason is the expected explosion of Internet devices and appliances. "You're not going to want your 15-year-old neighbor reprogramming your refrigerator," Cerf quipped.
Cerf proposed that personal ATM cards might be the best vehicle for establishing a public key infrastructure. If banks agreed on a standard smartcard for ATM transactions, said Cerf, these smartcards could also carry digital signatures and other authentication information, which users could apply on the Internet.
Looking to the future, Cerf predicted that interplanetary networking is not that far away. Part of the Mars exploration program includes putting Internet-enabled space stations and vehicles on Mars and in orbit. Several satellites devoted to Internet communications are due to be launched by 2008, and, according to Cerf, "Internet outposts will be operational by 2010." For more information, see http:// www.wcom.com/cerfsup/.
Although Austrian-born actress Hedy Lamarr, who died recently at the age of 87, was famous for her acting career, it's less well-known that she was the coinventor of frequency hopping, also known as "spread spectrum" -- a fundamental technology used in wireless communications. Before WW II, Lamarr was married to an Austrian weapons manufacturing magnate who later supplied munitions and torpedoes to the German war effort. Lamarr fled Austria before the war, and brought with her to the United States information she had learned from her exhusband about torpedoes. She and pianist George Antheil, neither of whom had any formal training in electronics, came up with the idea of using frequency hopping on radio-controlled torpedoes to circumvent signal jamming by the Germans.
The idea stemmed from Antheil's work in synchronizing piano rolls and skipping from one frequency to another at random. Antheil and Lamarr were awarded a patent for the invention in 1942. Although the military did not use the invention during WW II, it did start using spread spectrum in the 1960s, most notably during the Cuban missile crisis. In the 1980s, after the patent had expired, companies such as Proxim and Agilis started using it for wireless communications. The Electronic Frontier Foundation awarded Lamarr and Antheil "Pioneer" awards in 1997.
The National Weather Service has put into operation a supercomputer with five times the speed of its predecessor. With processor upgrades expected in September, the new IBM Processor 786 System Parallel supercomputer will be 28 times faster than the Cray C-90 that it replaces. According to a prepared statement from the NOAA National Weather Service (http:// www.noaanews.noaa.gov/), the new machine is already "generating faster and more precise predictions of the atmosphere, resulting in more accurate forecasts for every city in the nation." Unfortunately, that didn't seem to include prediction of the big Nor'easter that slammed the East Coast in the last week of January. According to a CNN report, "Many forecasters -- and most of the public -- were not aware the bad weather was coming so quickly until late Monday, when the National Weather Service issued a winter storm warning for the Baltimore-Washington area. Until then, some forecasters had predicted only light snow, believing the bulk of the storm would head out to sea. Instead, it hugged the Atlantic coast." Gee, probably a hardware problem.
Teams relying on cooperation generally prevail over teams that rely on individual superstars. Researchers are finding that a teamwork approach works well with robotics, too. Canadian researchers have developed teams of inexpensive simple robots to perform complicated tasks normally reserved for more complex and expensive centrally controlled robots. In the team scenario, if some of the robots fail, the task still gets done, whereas the failure of a centrally controlled robot can doom the mission.
The so-called Collective Robotics Project is led by C. Ronald Kube, a researcher at the University of Alberta (http://www.cs.ualberta.ca/~kube/). The initial project involves a team of six robots whose task is to search an arena for food (in this experiment, a light source). If they locate the light source, which is situated in a heavy box, they are supposed to push the box (the food) to a second light source, which represents a nest. If they fail (because they're not facing the second light source), they simply wander around in a counterclockwise direction until they find the light source again and reevaluate their position. While the individual robots appear to be behaving chaotically, their actions are only dependent on the position of the light source. Eventually, some of the robots will push the food into the nest. They don't need to communicate with each other or with any external operators.
Speech-synthesis pioneer Ray Kurzweil predicts that the technology for uploading the human brain to a computer and recreating it electronically is only about 30 years away. In a recent article in Psychology Today (http://www.psychologytoday.com/), Kurzweil says the brain could be slightly frozen just before death, then scanned -- one thin slice at a time, with "every neuron, every connection and every neurotransmitter concentration represented in each synapse-thin layer." In this scenario, we could achieve a form of immortality and still carry on communications electronically with family and friends. Kurzweil went on to suggest that it may be possible to send billions of "nanobots -- blood-cell-size scanning machines -- through every capillary of the brain to create a complete noninvasive scan of every neural feature. A shot full of nanobots will someday allow the most subtle details of our knowledge, skills, and personalities to be copied into a file and stored in a computer."