📡 QPSK-Demo

Demonstration of a QPSK modulator / demodulator in the audio range (with the ability to run it through a real speaker / mic)

---

C-library

Also, I've written this QPSK modulator-demodulator as a C library

This library can run 150 times faster than Python code

---

Block diagram of modulator section:

Block diagram of demodulator section:

As you can see, I'm using a Costas Loop with a PI (Proportional-Integrator controller) to control the local oscillator

---

The screenshot at the top shows an example of a "Hello!" message.

TX_DATA:

# [-1] - Silence,
# [0] * 60 - PLL sync burst,
# [0, 1, 2, 3] - phase (IQ) sync (MUST BE EXACTLY THESE VALUES IN THIS ORDER),
# string_to_qpsk_array("Hello!") - Data (payload),
# [-1] * 20 - Silence
TX_DATA = [-1] + [0] * 60 + [0, 1, 2, 3] + string_to_qpsk_array("Hello!") + [-1] * 20

Program output:

Generated 1729 samples. Total time: ~0.236s
Added ~0.005s (40 samples) of delay
Carrier detected @ ~0.025s. RMS volume: -18.4dBFS
PLL locked @ ~0.112s. VCO: ~998.80Hz (error between 950.0Hz: 51366ppm). I: 1, Q: 1
Detected phase change @ ~0.148s. Sampling position: 4 samples after zero-cross
IQ values of phase 1 recorded @ ~0.149s. I: -1, Q: 1
IQ values of phase 2 recorded @ ~0.151s. I: 1, Q: -1
IQ values of phase 3 recorded @ ~0.153s. I: -1, Q: -1
Decoded data @ ~0.155s: 1 (0b01)
Decoded data @ ~0.157s: 0 (0b00)
Decoded data @ ~0.159s: 2 (0b10)
Decoded data @ ~0.161s: 0 (0b00)
Decoded byte: 0b01001000, char: H
Decoded data @ ~0.163s: 1 (0b01)
Decoded data @ ~0.165s: 2 (0b10)
Decoded data @ ~0.167s: 1 (0b01)
Decoded data @ ~0.169s: 1 (0b01)
Decoded byte: 0b01100101, char: e
Decoded data @ ~0.171s: 1 (0b01)
Decoded data @ ~0.173s: 2 (0b10)
Decoded data @ ~0.175s: 3 (0b11)
Decoded data @ ~0.177s: 0 (0b00)
Decoded byte: 0b01101100, char: l
Decoded data @ ~0.179s: 1 (0b01)
Decoded data @ ~0.181s: 2 (0b10)
Decoded data @ ~0.183s: 3 (0b11)
Decoded data @ ~0.185s: 0 (0b00)
Decoded byte: 0b01101100, char: l
Decoded data @ ~0.187s: 1 (0b01)
Decoded data @ ~0.189s: 2 (0b10)
Decoded data @ ~0.191s: 3 (0b11)
Decoded data @ ~0.193s: 3 (0b11)
Decoded byte: 0b01101111, char: o
Decoded data @ ~0.195s: 0 (0b00)
Decoded data @ ~0.197s: 2 (0b10)
Decoded data @ ~0.199s: 0 (0b00)
Decoded data @ ~0.201s: 1 (0b01)
Decoded byte: 0b00100001, char: !
Decoded data @ ~0.203s: 2 (0b10)
Decoded data @ ~0.205s: 1 (0b01)
Decoded data @ ~0.207s: 2 (0b10)
Decoded data @ ~0.209s: 1 (0b01)
Decoded byte: 0b10011001, char: �
Decoded data @ ~0.211s: 1 (0b01)
Decoded data @ ~0.213s: 2 (0b10)
Decoded data @ ~0.215s: 0 (0b00)
Decoded data @ ~0.217s: 0 (0b00)
Decoded byte: 0b01100000, char: `
Decoded data @ ~0.219s: 3 (0b11)
Carrier lost and PLL unlocked @ ~0.220s. RMS volume: -30.1dBFS
Processing finished. Took 0.282s

As you can see the message is received and decoded successfully:

Decoded byte: 0b01001000, char: H
Decoded byte: 0b01100101, char: e
Decoded byte: 0b01101100, char: l
Decoded byte: 0b01101100, char: l
Decoded byte: 0b01101111, char: o
Decoded byte: 0b00100001, char: !

---

To test transmission using your speaker and microphone:

1. Install the PyAudio package
2. Ensure that your sound card supports the float32 format and the values of CHUNK_SIZE and SAMPLING_RATE
3. Set USE_LINE_SIMULATION to False
4. Make sure your speaker and microphone are selected as the default devices in your system
5. Place the microphone near the speaker and create a quiet environment
6. Run the script. You should hear the transmission. If the sound level is set correctly, you will see a message indicating the start and end of recording