The US National Aeronautics & Space Administration’s Jet Propulsion Laboratory in Pasadena, California has applied for a patent on a computer security system that is based on a system of challenge and password, or sign, challenge and countersign. All responses are correlated with random alphanumeric codes in a matrix of two or more dimensions. The codes are stored on a disk or plug-in card, and are subject to frequent change. A computer or terminal attempting to access the protected system must have the identical code set to the system to be accessed. In a two-dimensional matrix the host system might transmits a challenge by choosing any two codes in the matrix. The codes will not be located in the same row or column and they define two of the corners of a rectangle sized at random from the matrix. The terminal seeking access has to respond by sending the codes for the opposite corners of the same rectangle. If the host agrees that the same rectangle has been defined, it grants access. The system is designed to discard codes that have been used in a successful access and it never allows the same set of codes to be used more than once, so that anyone attempting illegal access can’t get in by simply repeating the code sequence entered by a valid user. But if a two-dimensional matrix is considered too limiting, hypercube matrices can be used. The system provides for any caller attempting too many erroneous responses to be disconnected automatically – and the next caller will still be presented with a new set of codes from the matrix. A key claimed advantage is that since a potential intruder that places multiple calls will be greeted with a different set of challenge codes each time, computer-programmed break-ins that rely on flooding the host with passwords would be frustrated. The system was developed for the Laboratory by a researcher at the California Institute of Technology, and the Space Administration hopes to license it to commercial computer companies in the US.
