Digital Signal Processing Reference
In-Depth Information
21.5 Adaptive FEC and Interleaving Techniques
Researchers have been working for a long time to improve FEC-based error
control mechanisms. The major research interest is still how to make the FEC
code size adaptive instead of using a fixed FEC code under all communication
environments.
Several works proposed adaptive FEC schemes that adjust the code size
according to an optimization model based on the assumption that the packet loss
in a network follows a Bernulli process [ 9 ], a Gilbert-Elliott model [ 10 ], etc.
However, the method of employing fixed models to determine the characteristics
of wireless channels works reasonably well for an environment where the end nodes
are fixed or have low mobility. For an environment that changes dynamically in
time and speed, finding an appropriate model is still a major research issue. So we
propose an alternative solution, that is to use a feedback loop to determine the
changing channel conditions and consequently to adjust the strength of the FEC
code depending on the notification of corrupted packets at the receiver end. By
means of RTCP reports the loss pattern at the network level is regularly sent back to
the receiver, thus giving the streaming server the possibility to adapt the FEC
strength along the video stream. The schematic implementation of the proposed
closed-loop FEC and Interleaving Real-time protection technique (FIR) and its
optimized version (FIRO) is shown in Fig. 21.7 .
In the following of the chapter, the FIR and FIRO techniques are first presented;
then, their performance is evaluated with respect to isolated adaptive FEC, isolated
adaptive interleaving, and plain transmission. In all the simulations, the channel has
been modeled with a 2-state Gilbert-Elliott model.
Fig. 21.7 Implementation of the FEC and interleaving real-time protection technique
Search WWH ::




Custom Search