KRS2: Signal Classification
This page is still under development and the experiment described on this page is not yet generally available.
1) Experiment Overview
Experiment Definition: The user logs remotely to open N6820ES Signal Surveyor 4D software and its extensions. The software is integrated with N6841A Sensor Management Tool.
Experiment Scenarios Applicable: S2, S3, S4, S5, and S6. These scenarios correspond to various different user experimentation scenarios in AERPAW.
Experiment Goals: The main goal of this experiment is to develop a basic understanding of the signal detection process, signal external parametric data extraction, and saving process to a SQL database. We will consider detecting some well-known transmitted signals.
Experiment Requirements: It is assumed that the users are familiar with the Surveyor 4D interface. For a brief introduction, users are recommended to check this page.
Energy vs Signal: Energy is either radio frequency (RF) noise, an RF signal, or part of a signal. and used by the surveyor software to describe unqualified RF energy. “Signal” is used to describe energy when more information has been discovered in the signal development process. A Signal indicates that more information has been determined for an “energy”. The information may help the operator determine what type of device emitted the signal. A “signal” describes a specific characteristic of the energy. Examples of information that contribute to identifying a signal’s type are:
Modulation type
Baud rate
Peak spacing/offset
Location of the emitter
Universal Signal Detection (USD): The Universal Signal Detection (USD) option provides a foundational capability and structure upon which Surveyor 4D signal detectors can be created to find Signals of Interest (SOI) without the need to program and/or compile code. The USD pre-filter uses selectable parameters to perform signal isolation, or signal classification [1].
2) Experiment Steps
First, open the USD from the tool bar icon (left figure), which contains the Setup tab, the Monitor tab, and the Design tab.
Check the 'Enable Signal Detection' box
Subsequently click the 'Add Detector' button as shown in the right-side figure below.
A total list of 58 available signal detection options will pop up and choose the required detectors. For this experiment, we choose 'BLUETOOTH', 'DECT 6_0', 'LoRa' and 'UMTS' detectors. Click on OK.
Note that the Surveyor software can use 23 detectors simultaneously and there are some dedicated built-in detectors for capturing the drone control signals. Users can also build a custom detector using the Design tab at the top of the USD window. Details can be found in the user manual [1].
Once the detectors are added, click the 'Frequency Plan' button at the bottom of the Setup tab. This brings up a display showing all active signal detectors and the frequency plan for each detector. The Frequency Plan display gives a rapid way to see if the active signal detectors are contributing to the search for an SOI. If the coverage indicator is red, the signal detector is not currently adding any signal detection capability to the USD system. Click the 'Optimize Tuning' button to automatically set the Surveyor 4D search setup's start and stop frequencies to maximize the amount of spectrum covered by the signal detectors. Click on 'Close' button to apply the settings.
In the main USD window, select the " Monitor " Tab that represents a block diagram of the complete Universal Signal Detection process as depicted below. This is a very powerful and interactive tool that includes graphs and colored arrows for immediate feedback on the system performance. A red arrow means the block is not working properly and one can hover on the red arrow to see the problem. Then click on the red arrow to fix the problem. For a properly working USD system, all arrows should be green.
Once all the arrows are green, the Surveyor window will be tuned to the frequency range of the added detectors. The detected and classified signals will be listed in the 'Signal Database' pane as shown below.
3) Exporting the Signal Data:
The detected signals can be easily exported to .XML format files and can be used for further analysis. Click on the 'Export' icon (Red circle) as shown below and save the file. The .XML files can be opened with Microsoft Excel. Note that the captured signal data can also be stored in the built-in SQL database. Saving data to the SQL database is presented in the spectrum monitoring experiment.
4) Setting Alarm for the Detected Signals:
Click on the 'Alarm' icon from the tool bar. After that, the left window (in the below figure) will open. There are already 100 alarms defined but all of them except one are inactive. You need to select one alarm and click on ‘Modify Signal Alarm’. Choose the first alarm described as 'AUTO IQ'. Users can also choose other alarm options too. Then the right window will open where you can define the alarm criteria according to your requirements.
For this experiment, we choose the following options as shown in the right figure:
Alarm Type as Signal
Status as Active
Trigger as Once
Feature as Signal id
Type as Any
Check Alarm associated tasks as modulation recognition (MR1), Visual alarm, Time (IQ samples), Audible alarm, and Freq List.
Users can refer to the user manual for the details of these alarm options [1]. Note that the alarm option can also be used for the detected energy as described in the spectrum monitoring experiment. The captured files will be saved in the 'C:\Snapshot' directory for further processing.
Resources
Understanding the Signal Processing Chain: Energy Detection and Databases
Techniques for collection and use of Big Data for Spectrum Management
While using the tool, click on the ‘About’ icon and Help will open. You can search the keyword there and get useful links.