Digital Signal Processing Reference
In-Depth Information
Figure 15-5.
Lostandlate speech packets as a function of the bit error rate; adaptive vs. not-
adaptive technique, with and without interfering FTP traffic.
The second simulation scenario tests the performance of the adaptive trans-
mission in presence of interfering FTP traffic. Besides the VoIP transmission
two other terminals are sending FTP traffic to the wired host through the AP.
The FTP packet size was set equal to the higher rate of the voice communica-
tion and its TCP congestion window was increased to make the source more
aggressive in terms of used bandwidth against the VoIP connection. Figure 15-5
illustrates the case of two concurrent FTP sources with a maximum number of 4
retransmissions at the MAC level: the adaptive technique with interfering traffic
performs better then the non-adaptive solution without interfering traffic. This
is an important result because it shows that, in presence of network congestions,
reducing the packet size by diminishing the source bit-rate is quite effective in
improving the quality of interactive speech communications: a reduced source
bit-rate decreases the congestion caused by packet retransmissions.
Regarding end-to-end delay, the proposed adaptive transmission scheme re-
duces the average delay because less retransmissions are needed. Figure 15-6
