| Summary: | 碩士 === 國立中央大學 === 通訊工程研究所 === 100 === Nowadays, wireless communications systems transmit data required for increasing the amount of required bandwidth is also growing. In the limited bandwidth, the use of resources achieve maximum throughput. Because the orthogonal frequency division multiplexing is provided with high data rate and effective against frequency selective fading, it is widely used in communication systems. However, the peak to average power ratio (PAPR) of the orthogonal frequency division multiplexing (OFDM) is too high and cannot resist the frequency offset.
Single-carrier frequency division multiplexing (SC-FDM) is using a single carrier modulation. Orthogonal frequency division multiplexing architecture is similar to single carrier, so the performance is also similar. The difference is that single carrier frequency division in orthogonal frequency division multiplexing added before discrete Fourier transform pre-coding system. The method can reduce the peak to average power ratio, and effectively increase the transmission power efficiency. Therefore, 3GPP LTE uses single-carrier frequency division multiplexing architecture for the uplink. However, the single-carrier frequency division multiplexing is like the orthogonal frequency division multiplexing, and the oscillator is easy to instability. Doppler frequency shift effect produced is very sensitive, resulting in ICI and reduce system performance.
In order to improve system performance, channel estimation is required. In fact, the signal (pilot signals) is used in the single-carrier frequency division multiple access system for channel estimation. In transmission, every symbol insert into a fixed cycle or time. One of the pilot signal display arrangement (block-type pilot arrangement). Based on this arrangement, we use the least-squares channel estimation method for improvement, which is a sliding window least squares channel estimation techniques to estimate the channel states in this paper. Further according to the different signal to noise ratio (SNR), we estimate the SNR in the receiver and select the optimal sliding windowing length for comparison with the traditional frequency domain least squares estimation and minimum mean squares estimation channel estimation techniques and other low complexity channel estimation techniques. The proposed method performance can improve performance and the computational complexity is not too high.
|
|---|
