Radar ES Signal Processing

Radar ES Signal Processing. Various analog and digital approaches have been used in radar ES receivers to detect signals and measure their parameters. Descriptions and performance analyses of the more common ones have been published 25-27. The radar ES receivers used for current radar ES systems deployed for the self-protection of platforms such as aircraft and surface ships generate pulse descriptor words (PDWs) for each radar pulse that is received.

Each PDW consists of digital data that represents the principal signal parameters, typically frequency, power, time of arrival, pulse duration, and if available, angle of arrival and modulation type (phase or frequency). Early implementations made extensive use of analog techniques to generate PDWs, but more recent implementations are making increasingly extensive use of digital techniques.

Pulse train deinterleaving is required because the pulses that are received from the various radars in the signal environment will be interleaved in time (i.e., in a sequence of received radar pulses there is no certainty that for a given pulse in the sequence, the previous or next pulses in the sequence will be from the same radar). Deinterleaving is typically performed in a two-stage process.

First, clustering is performed as pulses are received to form clusters or groups of pulses having similar characteristics. A subset of the signal parameters contained in the PDWs, typically frequency, angle of arrival, and pulse duration, are used in this stage. The second stage involves analyzing the time relationships [Pulse Repetition Interval (PRI) deinterleaving] between the pulses collected in each cluster to identify patterns that are consistent with the hypothesis that they were transmitted by a single radar.

In addition to the radar PRI behavior, the radar scan pattern can be inferred by examining the time history of the measured power of received pulses in a deinterleaved pulse train. For example, a radar that is performing a circular scan will illuminate the platform carrying the ES system with its main beam response at uniform intervals in time.

Emitter identification involves comparing the various parameters that have been measured for each of the resultant deinterleaved pulse trains with those in an EW library and identifying the best match.

In practice, many potential difficulties may occur. The PDWs generated by the receiver will contain errors that result from various sources. At least some clusters formed in the first stage will have broad ranges. For example, a large frequency range may be needed to accommodate a frequency agile radar. Consequently, some clusters may overlap.

Accurate PRI deinterleaving can be very difficult to perform with limited signal data sets; many modern radars have complex PRI staggers (i.e., the time intervals between successive pulses transmitted by a radar vary randomly or follow patterns that repeat only over a long period). Deinterleaving errors can result in the pulse train transmitted by such a radar being fragmented into two or more partial pulse trains. Finally, EW databases can have errors, be incomplete, or as a result of ambiguities, may be unable to provide a unique identification.

More sophisticated approaches are being investigated for the extraction of features that can be used to provide additional information for the classification and identification of radar signals. For radars that use frequency or phase modulation to improve range resolution, knowledge of the type of modulation waveform and its parameters is useful for classification purposes.

Also, the waveforms transmitted by radar systems often have distinctive features, which are sometimes referred to as unintentional modulation on pulse (UMOP). Various techniques have been proposed for the extraction and processing of waveform features for signal identification.