Research On Channel Estimation Of WPDM-CDMA System

Wavelet Packet Division Multiplexing Code Division Multiplexing Access(WPDM-CDMA) has been widespread concern as a new kind of multi-carriercommunication system. It uses wavelet packet transform/inverse wavelet packet to achievethe modulation/demodulation of signal. WPDM-CDMA system supports high spectrumefficiency, strong anti-interference, and multirate transmission. For WPDM-CDMAsystems, the wireless channel is a vital part of the system and wireless channel parametersis important for the equalization and adaptive modulation. So how to design a simplegood performance is the key research issue of the WPDM-CDMA system.The purpose of this paper is to discuss the pilot-aided channel estimation techniquesfor the WPDM-CDMA system. The implementation of two typical channel estimationmethods is analyzed in the proposed system and the performance of the WPDM-CDMAsystem under Rayleigh Channel and Gaussian Channel is simulated and analyzed. Anoptimal LS channel estimation algorithm is presented based on the original LS algorithm.A brief introduction Firstly, the basic theory of wavelet packet is introduced simply,leading to the concept of wavelet packet. Then several multi-carrier modulationtechnologies as well as the basic principles of the WPDM-CDMA system are introduced,and the WPDM-CDMA system simulation platform is established. Next, based on thewireless radio model, two pilot-aided channel estimation algorithms are analyzed andcompared by Matlab. Simulation results show that the MMSE algorithm is superior to theLS algorithm, however, MMSE algorithm requires the operation of inverse matrix,increasing the complexity of the algorithm, and LS algorithm is relatively simple and usedwidely in practical applications. Finally, in order to solve the problem that LS channelestimation is sensitive to noise, an improved method base on the wavelet packet isproposed to reduce noise. Simulation results show that the influence of noise has beendecreased and the performance of LS channel estimation algorithm through reducing noisebased on wavelet packet is superior to the original LS algorithm significantly.