Digital Signal Processing Reference
In-Depth Information
Fig. 6.6
Markov channel model used for indoor 802.11a wireless communication (
a
) BlER versus
SINAD and (
b
) histogram for steady state Markov state probabilities
6.2.3 Objectives for Efficient Energy and QoS Management
We determine the set of quality, energy cost and resource dimensions that are rel-
evant to the considered scenario where users transmit video data towards a central
AP (Fig.
6.2
).
1.
Cost Function
(
C
i
): The optimization objective considered in the design case in
this chapter is to minimize the total energy consumption of all users in terms of
Joules/Job. For example, in a video context, a job is the timely delivery of the
current frame of the video application.
2.
QoS Constraint
(
Q
i
): The optimization has to be carried out taking into account
a minimum performance or QoS in order to satisfy the user. As delivery of real-
time traffic is of interest, we describe the QoS in terms of a single QoS metric
which is in this design case defined to be the job failure rate (JFR) [85]. JFR is
defined as the ratio of the number of frames not successfully delivered before
the deadline to the number of frames issued by the application over the lifetime
of the flow. In this specific delay-sensitive video context, the QoS constraint is
hence specified at run-time by user
i
as a target
JFR
i
. This QoS definition could
