Cyclic Shift Network Coding for Low-Complexity Packet Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network coding methods face high encoding and decoding complexity and transmission delays due to the use of upper triangular and Vandermonde matrices, especially in wireless communication where packet loss is common, leading to inefficient data transmission.
Innovation Solution
The proposed method employs a generator matrix based on cyclic shift network coding, converting elements from a multi-element field to a binary vector field, and uses a low-density parity check code (LDPC) basis matrix to reduce complexity and delay, with the ability to dynamically adjust retransmissions based on channel erasure probability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If random linear network coding with upper triangular matrix and Vandermonde matrix is used, then linear independence between encoded data packets is ensured, but encoding and decoding complexity is high and transmission delay is large
Solution Approach 1:
The patent changes the field parameter from multi-element field to binary vector field in the generator matrix. This parameter change simplifies the encoding and decoding operations while maintaining the linear independence property of encoded data packets, thereby reducing computational complexity without sacrificing reliability
Solution Approach 2:
The patent applies different matrix structures to different parts of the encoding process. Specifically, it uses a generator matrix with elements from binary vector field converted from multi-element field, which has different properties from the traditional upper triangular and Vandermonde matrices. This local structural optimization reduces overall complexity while preserving essential coding properties
2Reliability
If random linear network coding with upper triangular matrix and Vandermonde matrix is used, then linear independence between encoded data packets is ensured, but transmission delay is large
Solution Approach 1:
The patent changes the field parameter from multi-element field to binary vector field in the generator matrix. This parameter change simplifies the encoding and decoding operations while maintaining the linear independence property of encoded data packets, thereby reducing computational complexity without sacrificing reliability
Solution Approach 2:
The patent enables faster encoding and decoding operations by using the simplified binary vector field structure, allowing the system to rush through the coding operations more quickly. This reduces the time required for encoding and decoding while still ensuring linear independence of the encoded packets
3Reliability
If ARQ feedback for each data packet is used, then packet loss is resolved, but network resource overhead is high and delay increases when channel performance is poor
Solution Approach 1:
The patent combines multiple data packets into encoded data packets using the generator matrix before transmission. This merging approach allows the system to handle packet loss more efficiently by transmitting coded combinations rather than individual packets, reducing the need for frequent ARQ feedback and thereby reducing network resource overhead
Solution Approach 2:
The encoded data packets generated by the generator matrix serve multiple functions: they carry information from multiple source packets, provide error protection, and enable efficient retransmission. This multi-functionality reduces the need for separate ARQ feedback mechanisms, thereby reducing network resource overhead
Data Source
AI summary
A network coding method includes obtaining a data packet group in to-be-sent data packets, processing the data packet group based on a generator matrix of cyclic shift network coding thereby generating encoded data packets, and sending the encoded data packets to a receiving end. The to-be-sent data packets include a plurality of data packet groups. Each data packet group of the plurality of data packet groups includes a plurality of data packets. An element comprised in the generator matrix is a corresponding element in a binary vector field converted from a multi-element field.