Signal Exploitation Processing Using Distributed Generic Computers, 16-9374Printer Friendly Version
Inclusive Dates: 01/01/03 - 07/01/04
Background - The majority of SIGINT systems that are in use rely on expensive and specialized hardware components. Systems that have been recently developed often use proprietary, custom hardware in order to meet the processing requirements of the specific system design. There has been a recent trend from proprietary designs toward commercial off-the-shelf (COTS) equipment. The problem with these specialized COTS hardware solutions is the rapid rate of hardware obsolescence.
Approach - The primary objective of this effort was to investigate and test the application of high-speed general-purpose computers in modern signal exploitation systems. The goal of this investigation was to demonstrate that there are significant portions of a SIGINT system that can be successfully implemented in low-cost hardware. There were technical hurdles that were overcome to prove the feasibility of this approach. These primarily dealt with the selection and design of signal processing techniques easily distributed over multiple processors, high-speed data flow and the responsiveness of the operating systems resident on these machines.
This program focused on narrowband channelization, which is the process of selecting a section of RF spectrum commensurate with the bandwidth of an individual communications signal and isolating it from the rest of the wideband data input/output data stream. This channelization process is used in both the acquisition and DF portions of the SIGINT systems.
A detailed and thorough design effort was undertaken for this task. The design was broken into two main tasks: algorithm design and software implementation design with the following broad requirements:
Accomplishments - The primary objective of this effort was to investigate and test the application of high-speed general-purpose computers in modern signal exploitation systems. This objective was not only accomplished by the successful implementation of the channelizer but also in the actual deployment on three different projects. Each of these programs use 12 to 2,000 channelizers in each system delivered. The combined effort of algorithm design, software design and finally software implementation, though initially intimidating, resulted in a proof of concept that will become the basis for many systems in the near future.