The Gilbert-Elliott channel is used to evaluate the performance of Raptor codes for file download in the standard digital video broadcasting for handhelds (DVB-H). This allows for an analytical evaluation without assuming the log-normal fading affecting the different time-slices to be uncorrelated. The approach is also a complement to performing evaluation using prototypes in laboratory, which during the standardization work was found to be rather cumbersome when the channel was slowly varying. The choice of performance criterion is discussed, and some of the results that at first might seem counter intuitive are found to be a consequence of the used criterion. Finally, the coding gain obtained by using Raptor codes in addition to MPE- FEC is evaluated for various speeds of the receiver, showing a substantial gain, although noticeable reduced when the receiver is moving slowly.
[1]
Laurence B. Milstein,et al.
On the effect of imperfect interleaving for the Gilbert-Elliott channel
,
1999,
IEEE Trans. Commun..
[2]
E. O. Elliott.
Estimates of error rates for codes on burst-noise channels
,
1963
.
[3]
Rahim Tafazolli,et al.
Correlation model for shadow fading in land-mobile satellite systems
,
1997
.
[4]
Gerard Faria,et al.
DVB-H: Digital Broadcast Services to Handheld Devices
,
2006,
Proceedings of the IEEE.
[5]
M. Gudmundson.
Correlation Model for Shadow Fading in Mobile Radio Systems
,
1991
.
[6]
E. Gilbert.
Capacity of a burst-noise channel
,
1960
.