Batch-Based Cross-Code Block Coding for Low-Latency Wireless Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network coding solutions for multicast and multiple source cooperation-based applications face challenges such as high latency, sub-optimal performance in non-erasure channels, and high decoding complexity, especially for large file sizes or large number of retransmissions.
Innovation Solution
The proposed batch-based cross-code block (CB) rateless coding solution combines a joint Fountain-like erasure code and cross-CB coding design, maintaining optimal global decoding and allowing each batch to contribute to decoding even when not all physical layer packets are fully decoded.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Fountain codes are used for multicast applications, then retransmission-related inefficiency is reduced, but decoding delay increases compared to physical layer HARQ schemes
Solution Approach 1:
The patent segments the higher-layer encoded packets into batches and maps each batch to a separate set of physical layer packets. This allows parallel processing of multiple batches at the physical layer, reducing overall decoding delay while maintaining the retransmission efficiency benefits of Fountain codes through the outer code structure.
2Loss of time
If cross-code block code-based HARQ scheme is used as a rateless code, then latency is reduced and decoding performance is improved, but decoding complexity increases for large file sizes
Solution Approach 1:
The patent divides the encoding and decoding process into batch-level outer code operations and packet-level inner code operations. This segmentation allows the receiver to perform simpler inner code decoding on individual batches first, then apply outer code decoding only when needed, significantly reducing overall decoding complexity for large files while maintaining low latency.
Solution Approach 2:
The patent enables partial decoding where the receiver can successfully decode information from some batches without needing to decode all batches completely. This partial action approach reduces the total computational burden and decoding complexity while still achieving the required throughput, as not all batches need to be fully decoded to recover the original data.
3Adaptability or versatility
If RLNC is applied to encode packets, then network coding is achieved, but decoding complexity becomes high when the number of packets is large
Solution Approach 1:
The patent segments the large set of packets into multiple smaller batches, each processed by RLNC independently. This segmentation reduces the decoding complexity for each batch while maintaining overall network coding capability. The outer code structure coordinates these segmented batches to achieve the desired network coding functionality for the entire data set.
Data Source
AI summary
Code blocks (CBs) are transmitted by a transmitting device and received by a receiving device in a wireless communication network. The CBs are generated based on a batch of physical layer packets to which a subset of higher-layer encoded packets are mapped, and the higher-layer encoded packets are generated from higher-layer information packets. Packets of the higher-layer information packets that are used to generate the subset of higher-layer encoded packets for the batch overlap with packets of the higher layer information packets that are used to generate a further subset of the higher-layer encoded packets for mapping to a further batch of physical layer packets. Check blocks are also transmitted and received, and each check block is generated based on portions of multiple CBs.