Fairness in MIMD Congestion Control Algorithms
We study fairness among sessions sharing a common bottleneck link, where
one or more sessions use a multiplicative increase multiplicative decrease
(MIMD) algorithm. Losses or congestion signals occur when the capacity is
reached but could also be initiated before that. Both synchronized as well as
non-synchronized losses are considered. In the non-synchronized case, only one
session suffers a loss at a time. Two models are then considered to determine
which source looses a packet: a rate dependent model in which the loss
probability of a session is proportional to its rate at the congestion
instant, and the independent loss rate model. We first study how two MIMD
sessions share the capacity in the presence of general combinations of
synchronized and non-synchronized losses. We show that, in the presence of
rate dependent losses, the capacity is fairly shared whereas rate independent
losses provides high unfairness. We then study inter protocol fairness: how
the capacity is shared in the presence of synchronized losses among sessions
some of which use additive increase multiplicative decrease (AIMD) protocols
whereas the others use MIMD protocols.
Philippe Nain
Last modified: Tue Mar 15 15:10:56 MET 2005