The Research On Channel Estimation Of Orthogonal Wavelet Packet Division Multiplexing System

Orthogonal wavelet packet division multiplexing is a new modulation techniquecombing wavelet packet decomposition and reconstruction theory with multi-carriermodulation technique of communication. In orthogonal wavelet packet divisionmultiplexing system, inverse wavelet packet transform and wavelet packet transformis used to realize the modulation and demodulation of multicarrier signal, which canget higher spectrum utilization rate and high anti-interference capability thantraditional OFDM system. Because there have multipath fading and Doppler shift inthe wireless channel, the system performance is still not satisfactory. For therealization of high bit rate transmission in OWPDM system, the channel estimationtechnique is needed to combat multipath fading and Doppler shift. So the research ofOWPDM system and its key technologies such as channel estimation has importantsignificance and practical value.The paper summarized the principle of OWPDM system, simulated and comparedOFDM and OWPDM system from the aspects of the system structure diagram，anti-interference ability,the spectrum utilization,data security and system complexity.On the basis of it, the article first discusses whether the commonly used channelestimation algorithm (LS channel estimation algorithm, LMMSE channel estimationalgorithm) is available in our system, then the interpolation based LMMSE algorithmof OWPDM system is researched, and its performance is compared with LSalgorithm in different mapping mode. The simulation results show that the algorithmcan effectively improve the performance of OWPDM system. And in case of highorder mapping, more precise interpolation algorithm can improve the systemperformance more obviously. In the pilot-based channel estimation, the cubicconvolution interpolation algorithm is proposed, and in the OWPDM system, thealgorithm and other interpolation algorithms are compared by simulation. Thesimulation results show that the algorithm can improve the accuracy of channelestimation. According to the traditional LMS semi-blind channel estimation is easy to causeinfection of the error, a improved LMS semi-blind channel estimation algorithm isput forward. Simulation result shows that, when insert fewer pilot number, newalgorithm is outperforms LS channel estimation algorithm, when insert the samenumber of pilot, compared with LMS semi-blind channel estimation algorithm, thenew algorithm enables the system to obtain a better BER performance.