Sensor and data fusion architectures and algorithms are often utilized when multiple sensor systems gather and analyze data and information from some observation space of interest. Objects that may be difficult to differentiate with a single sensor are frequently distinguished with a sensor system that incorporates several sensors that respond to signatures generated from independent phenomena. Signatures generated by multiple phenomena also expand the amount of information that can be gathered about the location of vulnerable areas on targets. This is important in smart-munition applications where autonomous sensors, such as those that operate in the millimeter-wave (MMW) and infrared (IR) spectrums, guide weapons to targets without operator intervention. These wavelengths allow relatively compact designs to be realized to accommodate the volume and weight constraints frequently encountered in ordnance. By using operating frequencies that cover a wide portion of the electromagnetic spectrum, relatively high probabilities of object detection and classification, at acceptable false-alarm levels, can potentially be achieved in inclement weather, high-clutter, and countermeasure environments. Multiple-sensor systems are used in civilian applications as well, such as space-based sensors for weather forecasting and Earth resource surveys. Here, narrow-band wavelength spectra and multiple types of sensors, such as active radar transmitters, passive radar receivers, and infrared and visible sensors, provide data about temperature, humidity, rain rates, wind speed, storm tracks, snow and cloud cover, and crop type and maturity.
Because of the important role that MMW and IR sensors assume in these applications, much of this chapter is devoted to the operating characteristics of these sensors. Acoustic, ultrasound, magnetic, and seismic signature-generation phenomena are also exploited in military and civilian applications, but these are not addressed in detail in this chapter. However, their data can be fused with those of other sensors using the algorithms and architectures described in later chapters.