Video Conferencing

Members:

1. Hoàng Minh Tuấn - B21DCVT444
2. Nguyễn Hoàng Long - B21DCCN497
3. Nguyễn Doãn Hoàng Giang - B21DCAT003

Description:

• (Focus on) video conferencing solution to handle online meetings for 1-to-1 and group interactions.
• Real-time messaging for both individual and group chats.

Technology:

• Java (Spring Boot)
• WebSocket
• WebRTC with Mesh architecture

Setup Instructions

1. Generate SSL Certificates

First, create SSL certificates using Git Bash. Follow these steps:

1. Create an SSL directory in your project root:

mkdir ssl

2. Generate SSL certificates using OpenSSL (replace <YOUR_LOCAL_IP> with your computer's local IP, e.g., 10.24.9.200):

openssl req -x509 -nodes -days 365 -newkey rsa:2048 \
-keyout ssl/private_key.pem \
-out ssl/certificate.pem \
-subj "//C=US//ST=California//L=San Francisco//O=MyOrganization//OU=MyDepartment//CN=<YOUR_LOCAL_IP>"

2. Configure NGINX

Update the nginx.conf file:

1. Locate your nginx.conf file
2. Replace all instances of 10.24.9.200 with your local IP address

3. Update Client Configuration

Modify the WebRTC client configuration:

1. Navigate to: src/main/resources/static/webrtc.js
2. Update the following variables with your local IP address:

const LOCAL_IP_ADDRESS = "10.24.9.200";
const signalingServer = new WebSocket(`wss://${LOCAL_IP_ADDRESS}/ws`);

Important Notes

• You must use your computer's actual local IP address throughout the configuration
• All IP addresses (in SSL certificates, NGINX config, and client code) must match
• The SSL certificates will be valid for 365 days

WebRTC Background Research

1. WebRTC (Web Real-Time Communication)

Audio and video communication system designed to work inside web pages.

Core Components

WebRTC consists of three main parts:

1. Signalling
2. Connecting
3. Communicating

Key Features

• P2P Protocol: Direct peer-to-peer communication with no need for traffic routing through a central server
• Signaling Server:
  • Used only for initial channel setup
  • Determines media format and peer communication preferences
  • Facilitates exchange of:
    • Session keys
    • Error messages
    • Media metadata
    • Codecs
    • Bandwidth information
    • Public IP addresses and ports

Network Components

NAT and ICE Framework

• NAT (Network Address Translation):

  • Maps private addresses inside a local network to public IP addresses
  • Enables transmission onto the internet

• ICE (Interactive Connectivity Establishment):

  • Framework that tests all connections in parallel
  • Selects the most efficient connection path
  • Two primary connection types:
  1. STUN (Session Traversal Utilities for NAT) Servers:
     • Detects NAT type in use
     • Provides public IP address for peer communication
     • Handles "Symmetric NAT" cases where routers only accept previously connected peers

2. TURN (Traversal Using Relays Around NATs):
   • Functions as relay servers for failed P2P connections
   • Maintains media relay between WebRTC peers

Technical Components

Video Codec (VP9)

• Compresses and decompresses audio/video
• Key features:
  • Packet loss concealment
  • Noisy image cleanup
  • Cross-platform capture and playback

JavaScript APIs

Three main APIs handle core functionality:

1. MediaStream:

   • Captures audio and video from user devices
   • Allows setting of constraints (frame rate, resolution)

2. RTCPeerConnection:

   • Manages real-time audio/video transmission
   • Handles connection lifecycle (connect, maintain, close)

3. RTCDataChannel:

   • Transmits arbitrary data
   • Integrated with RTCPeerConnection
   • Provides security and congestion control

2. WebRTC Mesh Architecture

Advantages

1. Cost Efficiency:

   • No media servers required
   • Reduced bandwidth usage
   • Lower overall costs

2. Privacy:

   • Direct media flow between users
   • Service provider has no access to data
   • Enhanced privacy protection

Disadvantages

1. Bandwidth and CPU Limitations:

   • Multiple media streams per device
   • Can overwhelm system resources
   • Performance degradation with increased load

2. Scalability Issues:

   • Less effective for 3+ user calls
   • Media server-based solutions preferred for larger groups
   • Performance and user experience challenges

Future Development

SFU (Selective Forwarding Unit)

A video conferencing architecture that optimizes data transmission between server and endpoints through the following process:

Data Flow Process

1. Server receives incoming video streams from all endpoints
2. Server distributes uncompressed video streams to each endpoint
3. Endpoints merge incoming video streams from other participants

Key Characteristics

• Stream centralization through server-based distribution
• Each participant:
  • Sends single outgoing signal
  • Receives multiple streams from other users via server
  • Reduces individual endpoint processing load