OVERVIEW

In the discussion and examples of DF systems so far, noise and its effect on system false alarm rate and DF accuracy has been deliberately ignored. Here we will address noise and where it comes from in various receiver configurations.

Broadly speaking, noise can be considered a source of error whose effect is tolerable as a function of acceptable sensitivity, accuracy, and false alarm rate. If a DF system is to be capable of receiving signals with maximum sensitivity, noise becomes a limiting factor. A noisy signal or a signal that is below the noise level can be detected by exchanging time and/or bandwidth for sensitivity to provide detection. This solution is called processing gain, which is achieved by signal averaging (time), system adaptation (changing bandwidths) and by applying interactive mathematical methods, such as auto-correlation. The processing gain solution, however, takes time, during which intercept probability is reduced, not always an important factor in ELINT systems but essential for threat warning.

The loss of time during which a system is disconnected or unable to respond to signals is defined as shadow time. This time is important because nonrecurring individual pulse trains or short burst signals can be missed during this period. The time that is required to measure and present a new set of data either to an operator or to a decision-making system is defined as refresh time, the rate at which the operator of the receiver sees the display change or, in an automatic system, the time...