Research On Channel Estimation Of SC-FDE In The Next Generation Mobile Communication System

The fading characteristics of wireless channel causes performance degradation inthe single carrier frequency domain equalization system, so channel estimate isessential to ensure that the receiver can correctly recover send signal. For the problemthat the rate of pilot utilization is lower and channel estimate is not precise enough inthe single carrier frequency domain equalization system, the thesis is completed thefollowing work:Firstly, the fading characteristics of the radio channel and wireless channelmodel are presented, and SUI-3channel is simulated. Due to fading characteristics ofwireless channel, introducing SC-FDE technology is necessary. Then, the principleand performance of the SC-FDE technology are explained, and compared withorthogonal frequency division multiplexing.In the SC-FDE systems, multipath effect constrains increase of transmissionbandwidth, while block transmission is effective against multipath interference. Tosolve the problem that data blocking method has a lower rate of pilot utilization andpilot blocking method is not precise enough, an improved pilot frame structure isproposed, which divides data block and pilot block into multiple subblockssimultaneously, and adjacent pilots are used for channel estimation repeatedly,without increasing the overhead. The algorithm can improve bandwidth efficiency andchannel estimation accuracy, and effectively reduce the bit error rate.For the fast fading radio channel has severe Doppler frequency shift, SC-FDEsystem performance seriously decline using the fast Fourier transform. An improvedchannel estimation algorithm of SC-FDE is proposed that using fractional Fouriertransform (FRFT). Using pilot signal to obtain the initial channel estimate with theminimum mean square error criterion. The frequency spectrum of noise and pilotsignal is separated by FRFT operation. After fractional Fourier transform, thefrequency spectrum of noise is the stationary distribution in the whole frequencyband, while the frequency spectrum of pilot signal is concentrated in the lowfrequency and high frequency band, so the noise is filtered and signal is remained,through band stop filter. Finally, the whole channel impulse response function isobtained by interpolation. Filtering reduce effect on channel estimation by the noise.The method can improve the accuracy of channel estimation and system performance.