| Summary: | 碩士 === 國立臺北科技大學 === 電子工程系 === 106 === In the process of wireless communication transmission, signal suffers from the effects of multipath and noise, causing the quality of the received signal to be degraded. Therefore, channel estimation can be used to greatly reduce the bit error rate of the received signal and increase the throughput of the whole system. Compressed sensing (CS) is a signal processing technique used to efficiently acquire and reconstruct a signal. The channel estimation method based on the compressed sensing technique can effectively increase the spectral efficiency while using fewer pilot symbols in transmission data, since it can make full use of the sparse characteristics of wireless channel.
Three channel estimation algorithms based on compressed sensing and neural network technique are proposed in this thesis. First, an orthogonal matching pursuit (OMP) algorithm, one of the CS reconstruction algorithms used to estimate the initial value of the channel frequency response, is applied. Then, further value estimation is performed by combining with neural networks. Finally, the actual channel impulse response is found by using least mean square (LMS) estimation method and recursive least square (RLS) estimation method respectively. Experimental results indicate that the proposed schemes can improve the system performance in terms of the bit error rate (BER), compared to the conventional channel estimation method based on compressed sensing technique.
|
|---|
