5G PHY Simulator

A real-time Physical Layer (PHY) simulator implementing core 5G/LTE air interface DSP.

What It Simulates

Input Bits → Channel Coding → Modulation (QPSK/16-QAM)
          → OFDM Modulator → AWGN Channel
          → OFDM Demodulator → Demodulation → Bit Recovery → BER

Components

File	What it does
src/bit_generator.cpp	Random bit generation
src/fft.cpp	Cooley-Tukey radix-2 DIT FFT/IFFT
src/modulation/qpsk.cpp	QPSK modulation/demodulation
src/modulation/qam16.cpp	16-QAM (Gray-coded) mod/demod
src/ofdm_tx.cpp	OFDM transmitter (IFFT + cyclic prefix)
src/ofdm_rx.cpp	OFDM receiver (remove CP + FFT)
src/channel.cpp	AWGN channel model
src/ber.cpp	Bit Error Rate measurement
src/main.cpp	CLI entry point
visualize.py	BER curve + constellation + spectrum plots

OFDM Parameters

Parameter	Value
FFT Size	64
Cyclic Prefix	16 samples
Data Subcarriers	48
Guard Subcarriers	16

Build

make

Requirements: g++ with C++17 support. No external libraries needed (FFT is built-in).

Usage

# Single SNR point
./phy_simulator --mod qpsk  --snr 10
./phy_simulator --mod qam16 --snr 15

# Sweep SNR 0–20 dB
./phy_simulator --mod qpsk  --sweep
./phy_simulator --mod qam16 --sweep

# Generate plots
python3 visualize.py

Expected Output

=== 5G PHY Simulator ===
Modulation : QPSK
OFDM       : 64-FFT, 48 data subcarriers, CP=16
OFDM Syms  : 100

SNR        : 10.000000 dB

Bits transmitted : 9600
Errors           : 2
BER              : 0.000208

BER Reference Values (QPSK, AWGN)

SNR (dB)	Typical BER
0	~0.12
4	~0.03
8	~0.002
10	~0.0002
12+	~0

Theory

QPSK maps 2 bits per symbol onto 4 points on the unit circle. BER (theoretical) = ½ · erfc(√SNR)

16-QAM maps 4 bits per symbol onto a 4×4 grid. BER (approx) = (3/8) · erfc(√(SNR/10))

OFDM uses IFFT to transmit N symbols simultaneously on orthogonal subcarriers. The cyclic prefix (CP) prevents inter-symbol interference from multipath.

AWGN channel adds complex Gaussian noise calibrated to achieve the specified SNR relative to the actual transmitted signal power.