Physical Memristor Brain Network

A complete implementation for building and controlling a physical 8-memristor brain network.

Project Overview

This project implements a brain-like neural network using 8 physical memristors with the following specifications:

• Ron: 400Ω (low resistance state)
• Roff: 1000Ω (high resistance state)
• Switching voltages: ±3V
• Compliance current: 12.5mA

Architecture

Network Topology:

• 3 inputs (I0, I1, I2)
• 6 memristors (M0-M5) connecting inputs to junctions
• 2 wire junctions (H0, H1)
• 2 memristors (M6, M7) connecting junctions to outputs
• 2 outputs (O0, O1)

Files

Core Implementation

• improved_physical_brain_software.py - Main control software with adaptive learning
• customizable_memristor.py - Memristor device models
• corrected_physical_connections.py - Physical wiring diagrams and breadboard layout

Setup

• requirements.txt - Python dependencies
• .venv/ - Python virtual environment

Quick Start

1. Simulation Testing:

   python improved_physical_brain_software.py

2. Physical Implementation:

   • Follow wiring diagrams in corrected_physical_connections.py
   • Use control software in improved_physical_brain_software.py

Features

✅ Validated Learning Algorithm - Demonstrates actual memristor resistance changes
✅ Realistic Device Models - Based on your exact memristor specifications
✅ Physical Wiring Diagrams - Ready for breadboard/PCB implementation
✅ Adaptive Learning - Progressive pulse strengthening for effective switching
✅ Pattern Classification - Network learns to differentiate input patterns

Status

• Simulation: ✅ Complete and validated
• Physical Implementation: 📋 Ready to begin

Last successful learning demonstration showed resistance changes:

• M3: 984Ω → 407Ω
• M4: 1017Ω → 385Ω
• M5: 974Ω → 401Ω
• M7: 1075Ω → 397Ω

Next Steps for Physical Implementation

1. Hardware Preparation

   • Acquire 8 memristors with specified characteristics
   • Set up 3 programmable voltage sources
   • Prepare oscilloscope for monitoring

2. Assembly

   • Follow corrected_physical_connections.py wiring diagrams
   • Build network on breadboard or PCB
   • Verify all connections

3. Testing

   • Run improved_physical_brain_software.py
   • Apply learning pulses to physical network
   • Measure and compare results with simulation

Hardware Requirements

• 8 memristors (400Ω Ron, 1000Ω Roff, ±3V switching)
• 3 programmable voltage sources (for I0, I1, I2 inputs)
• 2-channel oscilloscope (for O0, O1 output monitoring)
• Breadboard or PCB for assembly
• Connecting wires and components

Project Structure

memristors-python/
├── .venv/                                    # Python environment
├── improved_physical_brain_software.py      # Main control software  
├── customizable_memristor.py               # Device models
├── corrected_physical_connections.py       # Wiring diagrams
├── requirements.txt                         # Dependencies
└── README.md                               # This file

License

This project is for research and educational purposes.