Channel Estimation Based On Implicit Pilots For Ultra-wideband Wireless Communications

Ultra Wideband wireless communication method plays a more and moreimportant role. The main reason is that Ultra Wideband UWB system has numerousadvantages, such as large channel capacity, high transmission rate. UWB system nowmainly has two methods to realize:MB-OFDM and DS-CDMA, MB-OFDM methodcan be the first scheme because of its good existing wireless communicationcoexistence.In the MB-OFDM system, channel estimation will finally decide theperformance of the whole system. Now the UWB system is still widely used in pilotchannel estimation for its mature technology and simple structure auxiliary. Howeverauxiliary pilot has the defect of serious occupation of wideband and lowlytransmission rate. Compared to it, the hide pilot hardly occupies wideband and allowsthe system to obtain a higher transmission rate. The estimation of hidden pilot at thetime-frequency two-dimensional is the estimation of the channel, but the hide pilotchannel estimation accuracy is so poor that the system has a higher bit error rate.In order to solve the defect of the common hide pilot, data pretreatment methodhas been proposed in this paper. With data pretreatment algorithm the data and thepilot can have good separation characteristics by a slight alteration data originalformation. The receiver can separate relatively pure pilot information by receiving datato improve the channel estimation accuracy. Though the data pretreatment methoddestroys the original formation of the data, it can still be able to guarantee the accuracyof the data after the judgment of the receiver. In this paper, it proposes four datapretreatment algorithms: Unipolar threshold algorithm, unipolar error algorithm,bipolar threshold algorithm and bipolar error algorithm. The data and the pilot have astrong separation characteristic after the unipolar algorithm pretreatment. However, itwill deteriorate the peak-to-average power ratio of the system; the bipolar algorithmcan’t deteriorate the system’s peak-to-average power ratio, but its data separation characteristic is weaker than the unipolar algorithm. The threshold algorithm has asimple structure, and also has a poor pre-processing data formation. The erroralgorithm is slightly more complicated than the threshold algorithm, but it canmaintain well the consistency of the data formation. At the same time, this papercombined auxiliary pilot with hidden pilot, proposed complex pilot channel estimationalgorithm. It is the crystallization of the two algorithms. It not only maintains highaccuracy advantage of the auxiliary pilot channel estimation, but also inherits the hidepilot which doesn’t occupy the bandwidth.Simulation results show that four algorithms above can effectively improve thesystem performance, the bipolar threshold has the best result and low complexity. Theerror rate of hide pilot method with the data pre-processing LS algorithm was lowerthan the auxiliary pilot, and can bring2.5dB improvement even in signal-to-noise ratioenvironments. The error rate with pre-processing algorithm system under LMMSEalgorithm improves7.5dB system performance than common hide pilot. The data pre-processing method within less computation can effectively improve the accuracy ofhide pilot channel estimation. Its error rate characteristics have also been close to orabove the auxiliary pilot method. The simulation results of the composite pilot showthat the composite pilot has more significantly improvement than the hide pilotchannel estimation accuracy in accurate performance. And it also has obviousadvantage compared to auxiliary pilot in bandwidth occupancy. So with the channelestimation pilot, the composite pilot is an ideal channel estimation method.