HTML5 Webook

175/182

the time variation of throughput of hperf and “Actual roughput” shows the eective throughput measured by the receiver (other throughputs are not discussed in this study). Figure 8 (b) shows the time variation of the PLR and RTT. e measured throughput is much less than 10 Mbps as shown in Fig. 8 (a), and the PLR is over 60% at maximum as shown in Fig. 8 (b). ere is a large discrep-ancy between this result and that in Fig. 4. is may be caused by network disturbance. In Subsection 4.2, we in-vestigate the cause of this result by checking the bit error using the “-crc” option.4.2Results of experiments4.2.1Case of no occurrence of network disturbancee time line of the experiment is shown in Fig. 9. e time of no occurrence of network disturbance is marked with green (as discussed in Subsubsection 4.2.1 and Subsection 4.4) and the time of network disturbance is marked with blue (as discussed in Subsubsection 4.2.2) to distinguish the condition over time. Note that the network condition during the experiment is not stable and the network disturbance happened multiple times.Figure 10 shows the result of measurement for the time of no occurrence of network disturbance. e formats of Figs. 10 (a) and (b) are the same as those of Figs. 8 (a) and (b). e data shown in Fig. 10 is taken 20 minutes aer the measurement of data shown in Fig. 8. e weather of this period is cloudy and the cloud distribution is almost the same, as shown in Fig. 6.FiF5　WINDS satellite communication experiment (Nakatsugawa, Gifu Prefecture) in December 2016 when the weather is almost always cloudy or rainyFiF7　4K video streaming experiment by WINDS (the noise is recognizable on the monitor)FiF6　Distribution of cloud on the day of experiment taken by Himawari-8 real-time web (December 27, 2016)中津川（Nakatsugawa）FiF8　Results of preliminary experiment: (a) throughput and (b) PLR and RTT(a)(b)1713-12 A Network Quality Check Tool via HpFP Protocol on WINDS Satellite Link