Infoscience

Conference paper

Polar Codes are Optimal for Lossy Source Coding

Korada, Satish Babu; Urbanke, Rudiger
  • Presented at: ITW, Taormina, october 11-16

We consider lossy source compression of a binary symmetric source with Hamming distortion function. We show that polar codes combined with a low-complexity successive cancellation encoding algorithm achieve the rate-distortion bound. The complexity of both the encoding and the decoding algorithm is O(N\log(N)), where N is the blocklength of the code. Our result mirrors Arikan's capacity achieving polar code construction for channel coding.

Keywords: lossy source coding ; polar codes ; low-complexity algorithms ; NCCR-MICS ; NCCR-MICS/CL1

Note: Invited Paper

Reference

  • LTHC-CONF-2009-005

Record created on 2009-08-11, modified on 2012-11-08

Fulltext

Contacts

EPFL authors