 |
 |
 |
 |
| |
|||||
|
|
|
This article describes the novel application of signal processing to Dominant Battlefield Awareness. Two trends in signal processing are forcing a reconsideration of traditional approaches to situational awareness and decision making. Modern sensors are capable of gathering enormous amounts of data in a short time. In addition, demand for available communications bandwidth is increasing. MITRE is making unique contributions to this problem on two fronts. The Common View of the Battle Space (CVBS) provides a symbolic common view of the battlefield, assists an operator in target recognition through filtering for objects of interest, automates the time alignment and correlation of disparate sensor data, and selectively transmits images based on bandwidth availability. Smart Decision Directed Signal Processing (Smart DDSP) takes advantage of the dramatic increase in processing power to develop technologies for smart Intelligence, Surveillance, and Reconnaissance (ISR) platforms with integrated sensors and communication processing functions. MITRE has developed a unique adaptive, decision-directed processing approach that uses information from different sources to select areas of interest for cueing and tasking of sensor assets.
Areas of interest are depicted in this Common View of the Battle Space surveillance area. The tempo of military operations is increasing. Targets are mobile, easily concealed, and frequently time-critical. Achieving Dominant Battle Space Awareness means knowing everything going on in the air, space and on the ground; that is, the location and activity of all the enemy and friendly forces and knowing the environment. Dominant Battle Space Awareness is achievable only through efficient ISR data collection exploitation, and dissemination. Modern sensor platforms are capable of gathering enormous amounts of data in a short time. At the same time, demand for available communications bandwidth is increasing. The combination of these trends is forcing a reconsideration of traditional approaches to situational awareness and decision making. For example, 200 square kilometers of collected Synthetic Aperture Radar (SAR) imagery (see "A Hitchhiker's Guide to Radar Fundamentals" on Page 2) may require tens of hours or even days of processing by trained analysts. At the same time, it may be necessary to act on information in minutes or hours. Radar is an important technology because it provides an all-weather, day-and-night capability to detect and locate objects. SAR increases the azimuth resolution of a real aperture antenna by coherently summing signals transmitted and received from a sequence of antenna locations. CVBS uses Aided Target Recognition (AiTR) as a means to alert the analyst; AiTR is not used to identify objects. Thus, the AiTR module examines large numbers of (SAR) images and provides an estimate of the number of potential targets in each image. The analyst uses this information in conjunction with other information to decide whether a particular image requires more attention (see the illustration). We are developing additional AiTR techniques to aid the analyst, such as moving thresholds and images created from AiTR preprocessing. While still in the experimental stage, these techniques are enticing tools for the analyst. They attempt to divide labor between human and machine, assigning pattern recognition and broad-scale correlation tasks to the human being, and assigning display, organization, and processing tasks to the computer. In addition to AiTR technology, CVBS uses other sensors. Of particular importance is Moving Target Indicator (MTI). MTI is radar that detects only moving targets from the Doppler shift of reflected radar pulses. It is typically displayed as a dot on a map. By itself, an MTI dot, or even an hypothesized track is not enough to identify an object of military interest, but when combined with SAR imagery, signals intelligence, and map information, a valuable picture can be discerned. The techniques used in CVBS alone will not fully address signal processing challenges. Smart DDSP uses a combination of AiTR and automatic target recognition (ATR) functions to select areas of interest which then cue and task processors on board the ISR platform. ATR refers to use of computer processing algorithms to detect and recognize target signatures in sensor data. The main processing functions are front-end processing, signal processing and exploitation, fusion, and correlation. The extracted information is combined with other sensor data and information to increase its reliability; we call this fusion and correlation. Traditionally, signal processing used fixed functions that performed a repetitive set of calculations. Today we are moving more to doing adaptive signal processing where the algorithms use the signal data itself to determine how best to process the signal. Technology development in the area of field programmable gate arrays is enabling adaptive computing systems. The Defense Advanced Research Projects Agency (DARPA) has an Adaptive Computing Systems Program whose goal is to create unprecedented capabilities for dynamic adaptation of information systems to threats and rapidly evolving mission requirements of the Department of Defense. MITRE's Smart DDSP project plans to use the Adaptive Computing Systems technology. The focus of the Smart DDSP project is on processing directed by adaptive decision as opposed to current standard fixed processing. A key concept is using information from different sources to select the regions or areas of interest (AOI). The AOI information cues and tasks a multisensor or multimode ISR platform to select the optimum radar mode(s) or combination of sensors needed to collect the data, and to process, exploit, extract, and disseminate the information for the tasked mission. MITRE has developed a decision-directed, multiresolution image formation approach called multiscale planar sub-array processing (MS PSAP). The MS PSAP algorithm transitioned to DARPA's Intelligent Bandwidth Compression (IBC) program, and the concept has been demonstrated using advanced SAR data collected by reconnaissance aircraft (U2 ASARS). MS IBC incorporates dynamic, on-line AOI detection algorithms directly into SAR image formation process to perform decision-directed SAR imaging. The effect is to automatically focus the SAR imagery to fine resolution in regions where there is likely to be a target, while forming the remainder of the image at possibly coarser resolution. High-fidelity complex SAR data compression algorithms are used to preserve the coherent radar signature information over each AOI. Remaining image portions are transmitted at varying compression rates, depending on scene content. There is also potential decreased processor load since only a fraction of the image is formed to full resolution. The ongoing Smart DDSP work will raise decision-directed processing to a new level and is breaking grounds in terms of defining future processing and tasking of smart ISR platforms. Situational awareness requires a multidiscipline approach. It calls on signal processing for ATR, noise reduction, and other image processing techniques. It requires smart processing on the platform to reduce bandwidth and improve response times. It requires the integration of Geographic Information Systems with real-time sensor feeds. And it requires a flexible human computer interface and the creation of new user paradigms. It is important to remember that on today's and tomorrow's battlefield, sensors will be the eyes and ears of the decision maker. We must be able to process and disseminate information quickly to meet the shrinking timelines of battle. For more information, please contact Ward Evans or Danny Tromp using the employee directory. |
Solutions That Make a Difference.® |
|
|
