Skip to content

Latest commit

 

History

History
56 lines (47 loc) · 2.7 KB

README.md

File metadata and controls

56 lines (47 loc) · 2.7 KB

V2M-DGO-Point-Cloud-Video-Multicasting-Optimization

Point Cloud Optimization Using PPO and V-PCC

Project Overview

This project optimizes the transmission of point cloud data in a multi-user environment. It leverages Proximal Policy Optimization (PPO) and V-PCC (Video-based Point Cloud Compression) technologies to dynamically optimize user grouping, layered transmission, and video quality allocation. The aim is to balance bandwidth utilization and user experience (QoE - Quality of Experience).

Features

  1. Compression: Uses V-PCC to compress point cloud data into Base Atlas and Auxiliary Atlas.
  2. Dynamic Optimization: Applies PPO reinforcement learning to optimize:
    • User grouping based on FoV, device performance, and bandwidth.
    • Base Atlas resolution and Auxiliary Atlas quality allocation.
  3. Baseline Comparison: Includes three baselines:
    • Static Transmission: Fixed quality for all users.
    • Bandwidth Priority: Allocates quality based on bandwidth.
    • No Optimization: Directly transmits data based on user requirements.
  4. Metrics: Evaluates strategies based on:
    • QoE (Quality of Experience).
    • Bandwidth Utilization.
    • Transmission Delay.
    • System Throughput.

Project Structure

PointCloudOptimization/ ├── README.md # Project description and setup instructions ├── data/ │ ├── input_point_clouds/ # Original point cloud data │ └── compressed_atlases/ # Compressed Base and Auxiliary Atlas files ├── env/ │ └── point_cloud_env.py # Custom multi-user simulation environment ├── models/ │ ├── ppo_agent.py # PPO reinforcement learning model │ └── distillation_model.py # Optional knowledge distillation model ├── compression/ │ └── v_pcc_compression.py # V-PCC compression logic ├── baseline/ │ ├── static_transmission.py # Static transmission strategy │ ├── bandwidth_priority.py # Bandwidth-priority baseline │ └── no_optimization.py # No optimization baseline ├── tests/ │ ├── test_compression.py # Unit tests for compression │ ├── test_ppo.py # Unit tests for PPO │ ├── test_baselines.py # Unit tests for baseline methods ├── main.py # Main script to run the experiment └── requirements.txt # List of Python dependencies

Setup Instructions

1. Clone the repository

git clone https://github.com/your-repo/point-cloud-optimization.git
cd point-cloud-optimization