Ethernet Header Compression via Connection Identity Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ethernet headers have substantial overhead, particularly when the payload size is small, which affects resource efficiency, reliability, and latency in wireless communication networks, especially in 3GPP 5G New Radio (NR) systems supporting Time Sensitive Networking (TSN).
Innovation Solution
The method involves compressing Ethernet headers using connection identities mapped in the user plane protocol stack layer, specifically in the Packet Data Convergence Protocol (PDCP) or Service Data Adaptation Protocol (SDAP) layer, and using Robust Header Compression (ROHC) for IP headers, reducing the need to transmit certain fields like destination and source addresses, and incorporating cyclic redundancy checks for verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Ethernet headers are transmitted in full form, then complete address information is provided for reliable delivery, but header overhead increases substantially reducing resource efficiency
Solution Approach 1:
The patent creates a simplified copy of the Ethernet header called a connection identity (CID) that contains only essential identification information. Instead of transmitting full source and destination MAC addresses (48 bits each), the CID uses compressed representations (e.g., 12 bits per address), providing a lightweight copy that maintains delivery reliability while dramatically reducing overhead from 100 bits to approximately 24-36 bits per header.
Solution Approach 2:
The patent changes the parameter representation of address fields from full-length MAC addresses to compressed connection identities. By mapping full addresses to shorter CID values and using techniques like triangular encoding for address differences, the system transforms the parameter format to achieve both reliability (through maintained address identification) and efficiency (through reduced bit length).
2Measurement precision
If full Ethernet headers are transmitted for each packet, then address accuracy is maintained, but transmission time increases due to larger packet size
Solution Approach 1:
The patent uses connection identities as compressed copies of full MAC addresses. The CID maintains sufficient accuracy for address identification through techniques like maintaining a mapping table at both transmitter and receiver, using triangular encoding to preserve address difference information, and incorporating cyclic redundancy checks to verify address accuracy without transmitting full address bits.
Solution Approach 2:
The patent transmits partial address information in the CID rather than complete MAC addresses. By sending only the most significant bits and using compression techniques, the system achieves sufficient address accuracy for routing and delivery while reducing transmission time. The partial information is supplemented by maintaining context at the receiver end through mapping tables.
Data Source
AI summary
Disclosed are methods, systems, apparatus, and computer programs for compressing an Ethernet packet. In one aspect, a method involves mapping a portion of the Ethernet packet to a connection identity to compress the Ethernet packet. Additionally, the method involves encoding, in a user plane protocol stack layer, the connection identity in a data structure for transmission.