Pith. sign in

REVIEW

Beyond the Min-Cut Bound: Deterministic Network Coding for Asynchronous Multirate Broadcast

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1401.0543 v1 pith:PDPD746D submitted 2014-01-02 cs.IT math.IT

Beyond the Min-Cut Bound: Deterministic Network Coding for Asynchronous Multirate Broadcast

classification cs.IT math.IT
keywords networkreceiverscodingdeliverypacketbroadcastcodeddeterministic
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In a single hop broadcast packet erasure network, we demonstrate that it is possible to provide multirate packet delivery outside of what is given by the network min-cut. This is achieved by using a deterministic non-block-based network coding scheme, which allows us to sidestep some of the limitations put in place by the block coding model used to determine the network capacity. Under the network coding scheme we outline, the sender is able to transmit network coded packets above the channel rate of some receivers, while ensuring that they still experience nonzero delivery rates. Interestingly, in this generalised form of asynchronous network coded broadcast, receivers are not required to obtain knowledge of all packets transmitted so far. Instead, causal feedback from the receivers about packet erasures is used by the sender to determine a network coded transmission that will allow at least one, but often multiple receivers, to deliver their next needed packet. Although the analysis of deterministic coding schemes is generally a difficult problem, by making some approximations we are able to obtain tractable estimates of the receivers' delivery rates, which are shown to match reasonably well with simulation. Using these estimates, we design a fairness algorithm that allocates the sender's resources so all receivers will experience fair delivery rate performance.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.