- The first computers
- Safety first
- Improving encryption techniques
- What's in the picture?
- Cost-effective communication
- Random questions provide security and privacy
- New languages that speed computers
- Pictures speak more clearly than words
The first computers
WEIZAC was one of the world's first electronic computers, and the first in Israel. It was designed and constructed at the Weizmann Institute in 1954, and operated until the early 1960s. WEIZAC was used for mathematical research, such as solving problems related to the computation of ocean tides; this entailed complex calculations which could not reasonably be performed manually. The calculations carried out using WEIZAC took hundreds of computer hours, and enabled scientists to chart maps giving a very close approximation of high and low tide fluctuations throughout the world. As a result, the Weizmann researchers predicted the precise location of an amphidromic point (at which high and low tides never occur) in the southern Atlantic. Measurements performed in the wake of the discovery confirmed the existence and location of this point.
In another research study, Weizmann scientists used WEIZAC to calculate the spectrum of a helium atom (which involves three particles: the nucleus and the two electrons moving around it). Solving the dynamic relationships between three bodies is considered a highly complex mathematical task, and despite the development of modern, powerful computers, we still have no general, complete solution to such problems. However, the problem of the helium atom has been solved completely and yielded results that were experimentally confirmed by the Brookhaven National Laboratory in the United States.
In other projects, WEIZAC was used to carry out calculations to examine various theoretical models of the internal structure of the Earth, taking into account its different strata. These studies were based on calculations of shock wave propagation through different strata; comparisons with actual measurements were carried out after the occurrence of several earthquakes around the world.
Two successors to WEIZAC - GOLEM I (1964) and GOLEM II (1972) - which were also designed and constructed at the Institute, were innovative and extremely advanced in their class. They were used for research at the Institute and by scientists from other Israeli research centers. In one area, the GOLEM I held a world record: It was capable of handling words 75-bits long, which was longer than the words any other computers at that time could hold.
Weizmann Institute scientists are continually developing sophisticated methods for verifying the correctness and reliability of computer hardware and software. Computer systems are becoming increasingly complex. A fault in these systems could lead to major damage and even loss of human life, particularly when they are monitoring crucial processes: controlling a nuclear reactor, an industrial plant or flight paths for aircraft, space rocket launches, or even monitoring medical instrumentation at hospitals and overseeing modern communications systems.
To prevent such faults, the systems can be tested in a long series of simulations. However, this method cannot cover every eventuality and could miss some potentially fatal errors and flaws. In response, the Institute's scientists developed ways to verify the accuracy of a system using logic, in the manner that a mathematical theorem would be proved.
The scientists used a mathematical language called temporal logic, which makes it possible to formulate and prove theorems which contain statements relating to time. These verification methods are already being used to check computer systems in large-scale international projects.
Improving encryption techniques
A Weizmann Institute scientist, with colleagues from research institutes in other countries, developed several original methods for encrypting and decrypting information based on multiplying two very large prime numbers (a prime number can only be divided by one and itself). This encryption method is very safe because the person who is performing the encryption determines how long it would take to crack the code. For example, the use of specific parameters can mean that the time it would take to crack the code would probably be a few thousand years.
One of the present day applications of this method is in "smart cards," which are installed in home television sets to prevent anyone who is not a paid-up customer to receive and decode commercial satellite broadcasts. The smart card enables the company operating the satellite to charge customers solely for the programs and movies they actually watch. This encryption and decryption method is also applied in economic, banking, and governmental communication.
What's in the picture?
Weizmann Institute scientists developed an original method for encrypting and decrypting visual information, without relying on computers or other advanced technology for the decryption process. The method is based on splitting the visual information into a number of parts or "shares," in a way that a small set of shares does not yield any information regarding the original picture; if enough shares are stacked together, the original image can be reconstructed. The method allows the determination of the desired threshold for reconstruction. For example, a picture can be split into five shares and it can be stipulated that a combination of at least three shares is needed to recognize what is in the picture and indeed any three shares would be sufficient for this purpose.
A simple way of setting up a point-to-point communications network is to lay direct lines between every two points in the network. However, in large networks, this method involves laying a vast number of lines, which is difficult and expensive. So planners have to find a compromise between a dense network, based on numerous direct lines (allowing high-speed communication but at a high cost), and a sparse network, where there is no direct communication between many vertices. This reduces costs but tends to slow down communication.
Mathematicians from the Weizmann Institute have developed a method for producing sparse subnetworks while preserving the communication quality between the main vertices. Sparse networks are based on dividing the network into a number of local regions, and then facilitating communication between these areas. The Institute's mathematicians found a way of determining how far the network can be thinned out without any undue adverse effect on the efficiency of communication between the various vertices. They also proposed an efficient structure for a communications network between computer components in a parallel computer; it is called multibutterfly because it looks like a number of intertwined butterflies.
This research today provides the basis worldwide for planning large communications networks, such as telephone systems and decentralized computer setups.
Random questions provide security and privacy
How can you carry out an identity check without the person concerned disclosing any information about himself? How do you allow personnel in economic, legal, political and other circles to identify themselves and log into a database, gain access and carry out operations in it - without their operations leaving behind any trail or documentation?
The answer lies in a method for proving mathematical arguments, called interactive proofs with zero knowledge, developed by a Weizmann Institute scientist. The method allows identification to be carried out via a form of dialogue between the persons identifying themselves and the person carrying out the check. The examiner asks random questions, and it is this randomness that foils any possibility of reconstructing the identification questionnaire. If the person being checked out answers all the questions successfully, his or her identity has been demonstrated. In contrast, if someone is trying to pass himself off as an authorized user, it is very likely that the imposter would fail to give the correct answer to at least one of the random questions.
The system is also applied to data security in decentralized computer systems, where it enables a number of individuals or parties to contribute information to a specific database, while limiting their ability to extract data. Specific examples include counting votes accurately, reliably and honestly in an election, without being able to identify individual voters or how they voted; deciding who has won an economic tender, without revealing to unauthorized parties the details of the bids submitted; and calculating the average wage without revealing the salaries of individual employees.
New languages that speed computers
The time taken by computers to carry out tasks has been reduced by a computer language, Concurrent Prolog, developed by Weizmann Institute scientists. The language is logic-based and can provide instructions to computers performing parallel processing (carrying out a number of operations simultaneously). The researchers subsequently developed a new operating system, called Logix, which enabled Concurrent Prolog to be used in developing shared systems of electronic mail and multiuser computer conferences. This technology is now used to construct virtual meeting places on the Internet.
Pictures speak more clearly than words
A Weizmann Institute computer scientist has created a new computer language based on the use of visual structures. This language, called Statecharts, is intended to provide a clear, accurate, and intuitive description of the behavior of reactive systems (complex, computerized systems which must respond to changing situations).
One version of Statecharts is associated with the functional breakdown of systems. A second version is designed to be used for object-oriented systems. The language is suitable for designing, analyzing and maintaining reactive systems, on the basis of a set of behavioral requirements. Computer tools constructed around this language enable the system to be examined via simulations. Tools have also been developed to translate the visual language into a standard programming language, enabling a simulation to be run on any computer. In this way, once the system's main attributes have been defined by a human being, the computer can actually program itself. Today, Statecharts is used around the world in the aircraft, automotive and chemical industries, and in communications systems.
The creation of Statecharts opened up a wave of new research areas in computer science, which focused on the formation of visual languages in a range of engineering and computer-based activities. This area is developing rapidly worldwide as it is believed to have a wide range of potential uses.