Intelligent Resource Allocation for D2D Communication in 5G/6G Networks

---

Table of Contents

• Overview
• Objectives
• System Architecture
• Simulation Setup
• Algorithm
• Results
• How to Run
• Future Enhancements
• License
• Contributors

---

Overview

This project presents a reinforcement learning–driven approach using PPO to optimize device-to-device (D2D) communication in 5G/6G networks. Designed for simulation in edge computing environments, it improves throughput, energy efficiency, and adaptability, offering a scalable solution for next-gen wireless systems.

---

Objectives

• Optimize spectrum and power allocation in D2D communication
• Minimize interference while ensuring high spectral efficiency
• Improve energy efficiency and provide robust QoS (Quality of Service)
• Demonstrate scalability of AI-driven methods in 5G/6G networks

---

System Architecture

Network Model

• Simulates a cellular base station with multiple D2D pairs sharing spectrum.

Resource Allocation Framework

• Multi-agent PPO (Proximal Policy Optimization) dynamically assigns channels and power.
• State: wireless channel quality, interference, traffic load
• Action: channel selection, power level assignment

Reward Function

• Designed to boost throughput, reduce power use, and ensure fairness.

---

Simulation Setup

• Language: Python
• Libraries: TensorFlow / PyTorch, NumPy, Pandas, Matplotlib, Seaborn
• Environment: Simulated 5G/6G and edge-computing infrastructure
• Metrics:
  • Throughput (Mbps)
  • Energy efficiency (bits/Joule)
  • Fairness
  • Adaptability under varied conditions

---

Algorithm

Proximal Policy Optimization (PPO)

• A stable, policy-gradient RL algorithm
• Optimized for training multi-agent scenarios over iterative simulation episodes

---

Results

• PPO outperforms baseline (random/heuristic) allocation methods
• Achieves higher throughput and spectral efficiency, with improved energy metrics
• Demonstrates strong adaptability in dynamic network and traffic conditions

---

How to Run

Prerequisites

• Python 3.8+
• pip

Installation

git clone https://github.com/your-username/AI-D2D-Resource-Allocation.git
cd AI-D2D-Resource-Allocation
pip install -r requirements.txt