Paper Info
Title
On NACK-Based rDWS Algorithm for Network Coded Broadcast.
Abstract
The Drop when seen (DWS) technique, an online network coding strategy is capable of making a broadcast transmission over erasure channels more robust. This throughput optimal strategy reduces the expected sender queue length. One major issue with the DWS technique is the high computational complexity. In this paper, we present a randomized version of the DWS technique (rDWS), where the unique strength of the DWS, which is the sender's ability to drop a packet even before its decoding at receivers, is not compromised. Computational complexity of the algorithms is reduced with rDWS, but the encoding is not throughput optimal here. So, we perform a throughput efficiency analysis of it. Exact probabilistic analysis of innovativeness of a coefficient is found to be difficult. Hence, we carry out two individual analyses, maximum entropy analysis, average understanding analysis, and obtain a lower bound on the innovativeness probability of a coefficient. Based on these findings, innovativeness probability of a coded combination is analyzed. We evaluate the performance of our proposed scheme in terms of dropping and decoding statistics through simulation. Our analysis, supported by plots, reveals some interesting facts about innovativeness and shows that rDWS technique achieves near-optimal performance for a finite field of sufficient size.
Year
DOI
Venue
2019
10.3390/e21090905
ENTROPY
Keywords
Field
DocType
broadcast,drop when seen,feedback,integer partition,online network coding,probability of innovativeness,randomized drop when seen
Linear network coding,Network packet,Algorithm,Probabilistic analysis of algorithms,Principle of maximum entropy,Throughput,Decoding methods,Mathematics,Encoding (memory),Computational complexity theory
Journal
Volume
Issue
Citations 
21
9
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Sovanjyoti Giri100.68
Rajarshi Roy28018.39