Paper Info
Title
On the Capacity of 1-to- Broadcast Packet Erasure Channels With Channel Output Feedback
Abstract
This paper focuses on the 1-to-K broadcast packet erasure channel (PEC), a generalization of the broadcast binary erasure channel from the binary symbol to a finite field G F(q) with sufficiently large q. We consider the setting in which the source node has instant feedback of the channel outputs of the K receivers after each transmission. The main results of this paper are: (i) The capacity region for general l-to-3 broadcast PECs and (ii) The capacity region for two types of 1-to-K broadcast PECs: the symmetric PECs, and the spatially independent PECs with one- sided fairness constraints. This paper also develops (iii) A pair of outer and inner bounds of the capacity region for arbitrary 1-to-K broadcast PECs, which can be easily evaluated by any linear programming solver. The proposed inner bound is proven by a new class of intersession network coding schemes, termed the packet evolution schemes, which is based on the concept of code alignment in GF(q) that is in parallel with the interference alignment techniques for the Euclidean space. Extensive numerical experiments show that the outer and inner bounds meet for almost all broadcast PECs encountered in practical scenarios and thus effectively bracket the capacity of general 1-to-K broadcast PECs with COF.
Year
DOI
Venue
2012
10.1109/TIT.2011.2173723
IEEE Transactions on Information Theory
Keywords
Field
DocType
channel output,Channel Output,broadcast PECs,spatially independent PECs,broadcast binary erasure channel,1-to-K broadcast packet erasure,capacity region,symmetric PECs,1-to-K broadcast PECs,l-to-3 broadcast PECs,Broadcast Packet Erasure Channels,inner bound
Linear network coding,Topology,Discrete mathematics,Telecommunications,Packet erasure channel,Computer science,Network packet,Communication channel,Binary erasure channel,Packet switching,Decoding methods,Erasure
Journal
Volume
Issue
ISSN
58
2
0018-9448
Citations 
PageRank 
References 
40
1.75
16
Authors
1
Name
Order
Citations
PageRank
Chih-Chun Wang179555.20