| Summary: | 碩士 === 國立成功大學 === 電腦與通信工程研究所 === 105 === In any wireless channel transmission, the receiver not only gets the signals with noise, but also suffers path loss, multipath interference, and fading effect during transmission. Different kinds of situations may cause distinct magnitudes of fading. To shorten the time of transmission, setting the relay station can help all the users in the coverage and have better communication quality. This is well-known as the cooperative communication. One advantage of setting the relay station is its low complexity and cost if the relay user is idle. Another is that it increases the spatial diversity. To improve the performance of bit-error rate, this thesis applies decode and forward to send the signal to destination. With the channel state information (CSI) at receiver, we propose combining techniques and give the comparison with the traditional diversity combining like Selection Combining (SC) and Maximum Ratio Combining (MRC).
The receiver makes use the characteristic of channel responses and feeds forward these estimated amplitudes of CRs to the transmitter. It exists some methods about simplifying the CSI and related feed forward/back algorithms. To efficiently feed forward estimated CSI, we use some schemes which feedback estimated CRs to transmitter based on Karhunen-Loève Expansion (KLE). After the KL transform, the number of KLE coefficients is much fewer than the size of CRs. We can feed forward/back these coefficients of KLE such that the other terminal can rebuild CRs. Because these coefficients are of Gaussian distribution, we can further apply the Gaussian Quantization (GQ) to reduce the quantization error. As simulation results show, by applying these algorithms, we can obtain better error performance in cooperative-OFDM systems.
|
|---|
