Summary

2020

Session Number:D01

Session:

Number:D01-3

On an Alternative Approach to Congestion Detection in Ad Hoc Networks

Xiaojie Liu,  Ulrich Speidel,  

pp.529-533

Publication Date:2020/10/18

Online ISSN:2188-5079

DOI:10.34385/proc.65.D01-3

PDF download

PayPerView

Summary:
Ad hoc wireless networks without pre-existing infrastructure
depend on mutual collaboration among nodes. Congestion
in such networks presents more of a challenge than in
other types of networks. An essential task in this context is to
determine when a node is congested. This paper considers the
use of T-entropy to detect network congestion, and in particular
how to define an appropriate entropy threshold for congestion
detection. The paper first discusses how to recognise a congested
node from the perspective of (one of) its neighbours. We then
simulate ad hoc network scenarios, which we analyse for goodput
and entropy to determine the relationship between entropy and
the network congestion situation. As its main result, this paper
proposes an entropy threshold algorithm for congestion detection
in these scenarios for use in an AODV (Ad hoc On-Demand
Distance Vector Routing) derivate, RAODV (Relieving AODV),
that uses uncongested neighbours to detect and relieve congestion.