| Summary: | 碩士 === 國立高雄第一科技大學 === 電腦與通訊工程所 === 96 === If there is a velocity difference between transmitter and receiver during transmitting signal, the Doppler shift effect will occur in this transceiver system. If carrier frequency of local oscillators between transmitter and receiver are unsynchronized, frequency offset will result in inter-carrier interference (ICI) which will degrade the system performance. Therefore, it is important to correct or compensate frequency offsets in OFDM systems.
The blind frequency offset estimation has the benefit of data capacity efficiency but conventional blind estimation method takes the expense of more complex algorithm. We can reduce the complexity by using signal power property for blind estimation method. In this thesis, we utilize the signal structure of inverse fast fourier transform (IFFT) and investigate the signals with frequency offset from time domain and frequency domain. An estimation method can be obtained from the relationship of the signals with different subchannels for PSK-systems which have equal power property. From that, we obtain a roughly frequency offset estimated value; then, we can modify the roughly estimated value more precisely by using iterative approaches. Subsequently, we proposed another estimation method based on referenced method. We divide the received signal into four parts and shift first part of the signal range. After we compare the MSE performance between shifted signal and non-shifted signal estimation method, we proposed the structure from a totally different estimation point of view. The method is still based on the butterfly signal structure of IFFT. The system transmits signal twice for offset estimation. The second transmitting signal is composed of first transmitting signal which is exchanged the odd and even data in frequency domain. We perform virtual transform operation (VTO) to obtain the second transmitted signal by using first signal in time domain. Observing the first signal passes channel with frequency offset after performing VTO and second signal passes channel with frequency offset directly, we can find out the difference between first and second signals, and then extract the frequency offset component. Finally, we compare the MSE performance for the referenced method and proposed method. Additionally, we also write down our viewpoints.
|
|---|
