Signal Analysis

This service specializes in processing Orthogonal Frequency Division Multiplexing (OFDM) signals.

Currently Supported Capabilities:

• Parameter Estimation: 

  1. 1. 1. Useful Symbol Period

        2. Guard Interval (GI).

• OFDM Receiver Engine: Feature extraction and limited demodulation optimized to prevent the decoding or exposure of private payload data.

  1. 1. 1. 802.11ah HaLow: Dedicated support for 1 MHz and 2 MHz operational modes, verifying whether the SIG control field is correctly decoded and confirming successful full-packet OFDM processing.

Deployment: https://radio-standards.fastapicloud.dev/

This service performs a basic lookup of known radio standards based on simple characteristics that are pulled from user selected annotations.  Currently, provided this information, a list of potential standards are returned.

• Radio Frequency

• Bandwidth

• Duration

Planned improvements

1. Add modulation to help identify the standard

2. Return full standard specification (this can then feed some generic receiver for the standard)

I recently identified the existence of TorchSig on GitHub project.  This supports the signal simulation and training of ML model for modulation classification.

I started a torchsig-gui project to provide a web-based UI for configuring the training set.  I extracted the XCiT Classifier from the example code so that the generated training set can be applied to train a classifier for the user's desired list of signals.

There are some limitations of this service that is not suitable for my use with the Spectral Analyzer annotations.

• AM/FM signals are generated with filtered white noise.  This will differ from real-world application of the modulation on speech input.