Design techniques for sparse regression codes
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Prof. Ramji Venkataramanan.
Sparse regression codes (SPARCs) are a recent coding scheme for the additive white Gaussian noise channel, which achieve rates approaching the Shannon capacity of the channel with polynomial time decoding algorithms. This talk will introduce the codes and decoding algorithms, and then discuss techniques to improve their empirical performance and implementation efficiency.
In addition to optimising key code parameters such as power allocation, a novel decoder which combines a SPARC with an outer LDPC code is presented. This construction achieves excellent error performance which can exceed that of LDPC codes alone. Finally a new code structure called modulated SPARC is described. This structure also efficiently achieves the Shannon capacity and represents an interesting direction for future work.
This talk is part of the Probabilistic Systems, Information, and Inference Group Seminars series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|