Indicative results for the obtained average PSNR of each method are given in Table
5.2 (25 runs per sequence/method/topology). Two representative cases of medium and low
average transmission bandwidth were chosen.
Table 5.2: Average PSNR results (Y-channel ??“ 25 runs with 300 video frames per run) for video
streaming in the multi-hop networks of Figure 5.2.
Medium bandwidth case PSNR (dB) Low bandwidth case PSNR (dB) Method/Topology
T1 T2 T3 T1 T2 T3
End-to-end 35.42 34.28 32.89 32.11 30.56 31.54
ETX optimization 34.15 31.89 31.58 30.33 29.74 29.55
Highest Bandwidth 33.18 30.51 30.45 28.66 27.22 27.00
Localized 34.08 32.48 30.86 29.67 28.55 28.19
Estimation based 33.21 30.11 29.81 29.31 27.45 27.12
In order to understand better the relationship between the obtained PLR for each case
and the derived PSNR, the percentage of losses for the video packets when clustered into
eight distinct distortion categories is presented in Figure 5 (example for the sequence
???Foreman???). The second topology of Figure 5.2 was used for these results; similar results
have been obtained for the remaining topologies and the remaining video sequences.
Notice that our choice of eight distinct categories is only performed for illustration
purposes, since each packet is associated with its own distortion-reduction.
Pages:
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342