HTML5 Webook

134/182

of the transient waveform containing the boundary is smaller or almost the same as that not containing the boundary. e data set containing more waveforms of smaller amplitude is discarded as a false data set. An ex-ample of the results of averaging REV data is shown in Fig. 5. An example of a false data set with a small-amplitude transient phase is shown in Fig. 6.As the waveform of the real data of each element se-lected above is deformed due to atmospheric uctuation, the data set for each element is tted to a cosine function by the least-squares method. One example of the results of tting is shown in Fig. 7. e compensation phase (Δ0) that maximizes the relative amplitude of element n is derived from the tted cosine curve. en, the relative amplitude and relative phase of the element are derived as the follow-ing equations [1]. Here, we dene Γ = (r－1) / (r＋1) by expressing the dierence between the maximum and minimum receiving level as r.If the phase compensation for all elements is performed by retrieving the correction phase, the antenna gain at the REV measuring point reaches the maximum and the an-tenna beam of the APAA faces the measuring point.Relative amplitude of the element：√ (1)Relative phase of the element：tann (2)5Results of experiments5.1Waveform of received dataAs an example, we will look at the results of the received REV data performed on February 6, 2017. e level of the received signal was obtained by a real-time spectrum ana-lyzer for 2,048 points (1,024 points for 40 ms interval) by the update cycle of the REV of 80 ms interval. en, the data were averaged by 40 ms interval and are shown graphically in Figs. 8 to 11. Each gure shows antenna RX-1, RX-2, TX-1, and TX-2, respectively. In the gures, the initial part of the receiving level is a continuous wave without the REV, and the subsequent REV part corresponds to 128 elements, and the last part to the end is a continu-ous wave without the REV.e waveform of the REV part is a cosine function which implies that the element operates normally. Here, the waveforms of the cosine function are not the same because the initial phase of each element at the REV is randomized intentionally. is is because the level variation of the elements at the REV are relatively enlarged by de-creasing the peak gain of the antenna. e fact that the level variation of the transmitting system is smaller than that of the receiving system is caused by the setting of initial phase.115.00116.00117.00118.00119.00120.00121.00122.00123.00124.00125.000100200300400500Relative Power（ｄB)Measured PointsRX-1(False Data)2010/3/311st REV2ndREV3rdREVTransient PhaseBefore REV StartFiF6　Data processed sample (False Data)115.00116.00117.00118.00119.00120.00121.00122.00123.00124.00125.000100200300400500Relative Power（ｄB)Measured PointsRX-1(Correct Data)2010/3/31Transient PhaseBefore REV Start1stREV2ndREV3rdREVFiF5　Data processed sample (Correct Data)70075080085090095010000.0056.25112.50168.75225.00281.25337.50Relative Power (antilogarithm）Element Phase Rotation(deg)RX-1 Element#1MeasuredLeast Squares MethodFiF7　Data processed sample3 Ultra-High-Speed Satellite Communication Technology130　　　Journal of the National Institute of Information and Communications Technology Vol. 64 No. 2 (2017)