Development of an algorithm for the detection of coherency in radar signal waveforms

The estimation of the stability of radar emissions is of considerable interest in the evaluation of radar clutter rejection performance and also for the general knowledge of the waveform required for the design of threat simulators. It should be stressed that for the estimation of clutter rejection capability, it is the stability of the entire waveform that is of general importance, although the stability of parameters such as phase, Pulse Repetition Interval (PRI) and amplitude are typically specified because of the ease in instrumenting the measurement. The parametric estimates are indeed the most useful in describing the characteristics of the waveform but not necessarily for evaluating clutter rejection performance.

Two broad categories into which radar emissions can be subdivided are coherent and non-coherent RF. A great deal of confusion often surrounds the use of these terms, especially among those who measure radar emissions rather than those who build the radar sets. For the purposes of this paper, coherence will be defined in terms of the square root of the variance of the first pulse-to-pulse phase difference, Ï (Δθ ). For the case where Ï (Δθ) << 1 radian, the signal will be considered coherent. When the phase is uniformly distributed over 2Ï radians, the signal will be considered nonâ coherent. Since it is likely that, for most practical signals, the signal will be well within one of these two categories, ambiguity will be unlikely.

If a radar emission is observed to be coherent, it implies that the radar uses this property for Moving Target Indication (MTI) processing. The performance of the MTI will probably, but not necessarily, depend on the pulse-to-pulse phase stability as the most critical parameter for this type of system. Alternatively, if the radar emission is observed to be non-coherent, it implies that if the radar has an MTI processor, it is likely that it is of the stored reference variety. The performance of the MTI will probably, but again not necessarily, depend on the pulse-to-pulse RF stability as the most critical parameter.

The common thread between the two types of systems which indicates clutter rejection performance is the repeatability of adjacent pulse waveforms regardless of phase. This is not to imply that phase is not critical; it is important for determining the type of processor. The difference lies in the fact that for the internally coherent system, the phase information of the coherent reference oscillator is not observable as it is for the extremely coherent system. Hence, the only hint that such an emitter has an MTI processor is contained in the repeatability of adjacent pulse waveforms.

This paper addresses the general problems of detecting coherence, estimating MTI performance, and estimating the phase stability, frequency stability and PRI stability using sample data derived from a system based on the IBM-PC. Both the analysis and radar waveform generation systems were implemented in software utilizing Microsoft Fortran and Microsoft C compilers.