Redesign of BATS Code With Improved Decoding Rate Based on Coupled Batch Size and Degree Distribution

Batched sparse (BATS) code is a promising technique in multi-hop transmission networks with the benefits of both network coding and rateless code. In the conventional BATS design, the batch size and degree distribution are often separately considered, while they are highly correlated in the decoding process. To study the relationship between the degree distribution and batch size, and to analyze how they impact the decoding performance, this letter proposes to redesign the encoding and decoding processes of BATS code by jointly considering the degree distribution and batch size. Simulation results demonstrate that the decoding rate of the new BATS code can achieve around 21% enhancement compared with benchmark schemes when the channel is error-free.