HTML5 Webook

153/182

depends on the amplifying characteristics of the commu-nication path. As compared to BER when implementing the optimal DPD for each Eb/No, BER when changing the LUT by 0.5 dB from the optimal operating condition de-teriorates. erefore, it was observed that the increasing case and the decreasing case for the input power were dierent. It is assumed that not only the change of the input power but also the dierence of non-linearities be-tween the input and output characteristics of the ampliers which are not compensated arise as the dierence of the degradation quantity. From these ndings, we found it necessary to evaluate BER in consideration of the signal at the input point of the amplier and the stability of the amplier characteristics in actual operation.4.3BER characteristic improvement by the DPD on 1.8 Gbps transmission systemFinally, the measurement of BER by the satellite loop-back system was made using a 450 Msps symbol rate, 0.4 roll-o rate and the 16 QAM signal to conrm the improvement eect of BER by the DPD at 1.8 Gbps data transmission speed which is expected to be realized for the next-generation Earth observation satellite. From the ex-perimental result in Subsection 4.1, the BER degradation due to the transmission path characteristics other than the non-linearity also exists and becomes large as the transmis-sion rate increases. For these reasons, the BER degradation quantity of the whole experiment system was reduced by reducing the transmission loss due to cables, etc. before this measurement. Figure 6 shows the experimental result of the BER measurement. e BER characteristics of the translator loopback path (Fig. 6: solid lines) as reference are improved over the time of the experiment. In addition, Figure 6 tells us that the result obtains very close charac-teristics compared to the prediction result for the BER characteristic of the translator path calculated by the characteristics of the device used in the ground station, and that the experimental system is able to be constructed as expected. e comparison of the result (Fig. 6: blue dashed line) before applying the DPD, obtained in the satellite loopback path, and the result aer applying the DPD (Fig. 6: blue solid line) allows the conrmation of the high-est dissolution of the error oor, which arises due to the non-linearity of the amplier, by applying the DPD.In addition, Figures 7 and 8 show the received spectrum and constellation before and aer applying the DPD when FiF5　Compensation sensitivity verification for the DPD1.00E‐061.00E‐051.00E‐0417.017.518.018.519.0BEREb/No[dB]DPDsensitivity48.248.547.9*Earth station HPA output power in dBm+0.5dBInput power‐0.5dBInput powerFiF6　Effect of DPD at 1.8 Gbps transmission1.E-041.E-031.E-0281012141618BEREb/No[dB]TheoryTranslator Route(Meas.)Satellite Routewith DPD(Meas.)Satellite Routew/o DPD(Meas.)46.0*48.0*47.0*46.8*45.8*47.6*43.2*44.9**Earth station HPA output power in dBmTranslator Route(Calc.)5.E‐04(Requirement)FiF7　Spectrum when applying DPD at 1.8 Gbps transmission(a)DPD=OFF(b)DPD=ONFiF8　Constellation when applying DPD at 1.8 Gbps transmission(a)DPD=OFF(b)DPD=ON1493-9 Experiment of Non-Linear Compensation on Satellite Channel