Simple Acoustic Sensor

---

Description

Sets up secure audio streaming, sampling, processing, FF transformation, and publishing via MQTT protocol

Dependencies and Package Requirements

The following installations and imports are required to run this application.

Helpful Guides

1. AWS IoT Development Guide: Describe the basic features and hierachy of AWS IoT
2. Git Guide: Helpful Git commands

Installs

1. Python 3.6 via python software foundation
2. AWSIoTPythonSDK via aws-iot-device-sdk-python on github
   • allows developers to access / control devices via the AWS IoT platform
3. RPI Audio Levels via rpi-audio-levels on github
   • this script also uses the GPU FFT library installed on RPi (see link)

Imports

1. Pyaudio
2. Numpy
3. Queue
4. Threading
5. Cython (if installing RPI Audio Levels library, above)

Security Credentials

1. Root Certificate Authority Certificate File
2. Private Key File
3. Activated Certificate File

Hardware Used

1. Raspberry Pi 3 Model B
2. Snopy Rampage SN-RM7X USB Microphone

Configuration

acu_sens_config.json

Keys

• aws_vars (mandatory): necessary to connect to Hala AWS endpoint broker

Values

• <clientId> in "*_topics" keys gets replaced with the actual 'clientId' value

acu_sens_prototype.py

ENVR VARS

• callback: will the script use the pyaudio stream callback feature or not
• recording_flag: used if callback == True, start recording immediately at runtime or wait
• NUM_THREADS: if callback == True, indicate how many workers to create for processing streaming audio data
• REC_SECONDS: if callback != True, run the script for a number of seconds
• iot_config_json: /path/to/config.json

Pyaudio and FFT VARS

• RATE: sample rate of the microphone in Hz
• DATA_SIZE: sets the sample size of the buffer (base 2 exponential --> 2^X)
• BUFFER_SIZE: actual number of samples in the buffer
• FORMAT: Pyaudio data format of samples