N6820USDE Universal Signal Detection (USD) Feature

Introduction

Keysight’s N6820USDE USD feature, of the N6820ES Surveyor 4D software, provides a foundational structure and capability to create detectors that find Signals of Interest (SOI) without the need to program or compile code.

USD wideband detection feature

USD wideband detection operates on spectral data and uses selectable parameters to perform signal isolation, or signal classification. With every sweep, USD compares detected energies with the USD criteria defined in the setup. A detected match sets a flag to “True” in the energy history data indicating that the criteria has been met for one of the detectors. The USD “Monitor” tab shows the USD workflow and process of sorting and filtering energy as shown below.

The USD wideband section uses external parameters to classify signal energy as explained below.

• Detector – the general detector description has a naming convention, author field, and note field. The name identifies the signal or device in Surveyor 4D. The note field keeps track of the detector status, and revisions, indicating whether it is ready for use or not. Additionally, the note field can be used to give practical advice for the user.

• Bandwidth Filter - Only signals meeting a bandwidth criterion are parsed and classified as different types of signals. The bandwidth filter is the first USD block after energy detection takes place as shown in the image above. If the user has prior knowledge of the signal of interest’s expected bandwidth, the bandwidth filter can be used to eliminate detected energy that doesn't meet the bandwidth requirement. This results in fewer signals passed from the bandwidth filter block into the frequency plan block. The bandwidth filter has a fixed value with a lower and upper tolerance setting (i.e., 1.25 MHz - 80% to 120%).  Additionally, it can pass a set bandwidth value to Surveyor 4D that represents the value of the filter and therefore, pass the real signal bandwidth to S4D. This is useful for downstream processes, like geolocation, IQ recording or modulation classification, that should use the real bandwidth value.

• Frequency Plan - Applied for signal types operating on a known frequency allocation and channel plan. The frequency plan can also specify frequency bands (multiple f1-f2) for signals that may not appear where expected due to drift, age, quality, or other issues. Aside from a pure bandwidth filter, the easiest form of USD detector is a simple frequency plan. If the frequency plan is published, as in the example, the data is imported directly into the USD frequency plan editing tool. If the device you want to create a detector for is accessible, the frequency plan and channelization scheme can be determined using Surveyor 4D’s environment threshold. However, if you are working from field data with only one or two intercepts of the signal of interest, the full frequency plan may not be known. In this case, development of an effective wideband detector may be used to help collect the rest of the signal activity across the full channelization plan.

• Wideband Detector - Provides development tools (spectral shape correlation, peaks, and limit lines) used to characterize transmitters and create a signal detector unique to a spectrum event. There are three types of wideband detectors, and each has strengths and weaknesses to consider.

   o The Spectral Shape Correlator is useful for signals that have a unique feature or signature in the spectrum. Resolution-bandwidth and trace-average settings have an important impact on the effectiveness of this kind of detection scheme. The Spectral Shape Correlator detector is not recommended for signals that have a flat, noise-like appearance, as shown with the Limit Line detector image below. A shape detector working on flat, noise-like signal will also correlate with the trace noise, making it of little use.

   o The Limit Line detector is good for short-duration bursted signals that have a wide, noise-like modulation or are swept/chirped. Limit Line detectors are also good for emission-mask testing. You can create specific profiles by defining the relative frequency, bandwidth, and amplitude on a point-by-point basis.

   o The Peaks detector is suitable for n-FSK signals and is effective for various kinds of paging and data link transmissions.