Blazing Fast Erasure-Coding with Random Linear Network Coding (RLNC)
-
Updated
Oct 15, 2025 - Rust
Blazing Fast Erasure-Coding with Random Linear Network Coding (RLNC)
A library for encoding and decoding sparse network codes
Random Linear Network Coding
A high-performance, Byzantine-resilient RLNC implementation in Rust
Media Erasure-coded Live Delivery: a coded, media-aware transport for live and contribution-grade video in pure Go. Sliding-window RLNC erasure coding over a deterministic sans-I/O core. Experimental.
The implementation of Reliability of Multicast under RLNC
Implantation of the block-based Adjustable RLNC method (M.Sc. thesis simulation)
PyErasure is a Python module that implements erasure coding. It is meant for research and teaching so ease of use has been the primary goal of the design.
Sub-IP bearer bonding protocol using RLNC across LoRa, Iridium SBD, APRS, SMS, and GSM ring signals — draft-papadopoulos-hemb-00
SIMD-Accelerated RLNC Sidecar for Deterministic MEV & HFT Networking
Decentralized, high-throughput P2P file sharing engine in Go using badrlnc (RLNC over GF(2)) and badcrypt (post-quantum ChaCha20-Poly1305 AEAD). 100% WebAssembly compatible.
Production-grade, ultra-high-performance Sliding-Window Random Linear Network Coding (RLNC) over GF(2) with incremental Gauss-Jordan solver in Go
To associate your repository with the rlnc topic, visit your repo's landing page and select "manage topics."