Delay-universal channel coding with feedback

In this paper, the design of error-correcting or channel codes for delay-universal/anytime communication is shown while considering systems with and without a feedback link. We construct practical and low complexity anytime channel codes based on spatially-coupled repeat-accumulate (SC-RA) codes. Pe...

Full description

Bibliographic Details
Main Authors: Md. Noor-A-Rahim, Khyam, M. O., Guan, Yong Liang, G. G. Md. Nawaz Ali, Nguyen, Khoa D., Lechner, Gottfried
Other Authors: School of Electrical and Electronic Engineering
Format: Journal Article
Language:English
Published: 2018
Subjects:
Online Access:https://hdl.handle.net/10356/87580
http://hdl.handle.net/10220/45442
Description
Summary:In this paper, the design of error-correcting or channel codes for delay-universal/anytime communication is shown while considering systems with and without a feedback link. We construct practical and low complexity anytime channel codes based on spatially-coupled repeat-accumulate (SC-RA) codes. Performance and density evolution analysis are shown for the binary erasure channel (BEC) and the binary input additive white Gaussian noise (BIAWGN) channel. We observe that the erasure/error floors exist even at low decoding delay in the following cases: 1) when the code rate is close to the Shannon capacity; and/or 2) when the code parameters are chosen to target a high decaying rate of erasure/error probability. To mitigate erasure/error floors, we present feedback algorithms for BEC and BIAWGN channels. We show that the proposed feedback strategies can greatly enhance the performance of anytime SC-RA codes. Numerical results also show that the feedback strategies significantly reduce the decoder complexity. The proposed feedback approach is applied to an aircraft tracking application to track/calculate/estimate the state information of the aircraft. Based on comparisons of the results obtained from the traditional block and anytime coding scenarios, it is observed that the latter significantly outperforms the former in terms of tracking performance.