Digital Signal Processing Reference
In-Depth Information
Figure 13-4.
Average delay of UDP packets as a function of the error probability over the
802.11b wireless channel. The results are plotted for a varying number of TCP traffic sources
and by setting the maximum number of transmission attempts for video traffic to be equal to 2.
the MPDUs and, thus, longer average delays are experienced at the UDP level.
A similar behavior can be observed by letting the number of interfering TCP
sources increase. The effect of a large number of interfering sources is twofold.
On the one hand, it translates into a large collision probability which delays the
access to the radio channel. On the other hand, when the channel is shared by
a large number of sources, the channel capacity perceived by individual sources
is smaller.
Let us focus on the case where no interfering TCP sources are considered.
Depending on the radio channel conditions, the service time of video MPDUs
is either equal to one or to two MPDU transmission times (remember that is
equal to 2). The service time results to be small enough that there is no queue at
the MAC layer and the delay perceived by UDP packets is extremely small.
These results suggest some criteria for the choice of the set of services which
can be provided by the system. Consider, for example, the case of interactive
video services, for which an important QoS constraint consists in the average
delay being kept lower than 150 ms. In this case, the number of interfering TCP
sources should be limited. Up to 3 interfering sources are acceptable, while
4 sources can be admitted only if the average error probability over the radio
