The Flat Fast Fading Channel Prediction Based on the Fourth-order Cumulant and APA
|Course||Communication and Information System|
|Keywords||flat fast feeding channels the ESPRIT based on the fourth-order cumulant least mean square (LMS) method recursive least squares (RLS) algorithms affine projection algorithm (APA)|
The wireless mobile communication channel is very complex. It limits theperformance of the wireless communication system a lot. The multiparty problembetween the transmitter and the receiver and the fast moving terminal which causethe Doppler affect result in the amplitude and phase of the receive signal vary fast.We call it “feeding”. It is the influence of the deep feeding that limit theperformance of the communication system very much. We need the bettermodulation, coding and power control methods to be used in the feeding channelefficiently. A.J. Goldsmith and S.G..Chua have investigated some new adaptivetransmitter technique, for example adaptive modulation, adaptive coding, adaptivepower control and adaptive transmitter antenna diversity and so on. Theseadaptive transmitter schemes all vary the constellation size, symbol rate, codingrate, transmitted power level weights of transmission antennas or any combinationof these parameters by instantaneously monitoring channel conditions. They aretrying to use both power and spectrum more efficiently without sacrificing the biterror rate performance to realize the higher information transmitter rate.To implement adaptive transmission methods in practice, the channel state(CSI) must be available at the transmitter. With the sampling data of the receiversamples in the past interval, we use the arithmetic of channel predictions topredict the fading channel coefficient. Afterwards, transmit the fading channelcoefficient to transmitter in the feedback channel. The transmitter will decide thetransmission power, modulation methods, coding methods and the transmissionantennas, to fit the transmission conditions of the time. Thus, transmitter isoptimized, and the communication performance is improved at the same time.The thesis first introduces the significance of this subject, development statusand some basic theory knowledge. The important part and main task of this thesisare the improvement of the current channel prediction arithmetic and the instructionof some new method in channel prediction. It has two parts: the first part gave usseveral arithmetic such as ESPRIT, LRP and MEM that have been used forprediction. From the comparison between them we can get the truth that theESPRIT has the best performance among them. So in this paper I bring forward animproved ESPRIT arithmetic: the ESPRIT based on the fourth-order cumulant. Thesecond part is the investigation of the adaptive arithmetic. First we show two basicadaptive methods: the least mean square (LMS) method and the recursive leastsquares (RLS) algorithm and compared their convergence speed. At last the paperproposed a new method – the affine projection algorithm (APA) for channelprediction.1.Prediction method for fading channels based on the fourth-order cumulantThe basic objective of the method is to use fourth-order cumulant in ESPRITarithmetic to decrease the Gaussian noise. The ESPRIT arithmetic predicts thesignal poles and the channel coefficient exactly in high SNR but it has poorperformance in very low SNR. The character of the fourth-order cumulant is that itcan weaken the Gaussian noise which is the chief noise in wireless mobilecommunication channel. We can prove that the fourth-order cumulant of theGaussian signal equal to zero, so we can decrease the Gaussian noise added to thesignal which is not the Gaussian one.2. The affine projection algorithm (APA) for channel predictionThe affine projection algorithm (APA) proposed by Ozeki K and UmedaT is a generalization of the normalized least mean square (NLMS) algorithm. Ithas properties that lay between those of the least mean square (LMS) method andrecursive least squares (RLS) algorithms. The LMS -a typical adaptive noisecancellation technique -has characters such as sluggish convergence andcomplexity .The RLS has fast convergence rate and small mean square error but isnearly the most complex one in those of adaptive algorithms. After the analysis ofthe two typical methods I introduce a new adaptive algorithm -affine projectionalgorithm (APA) -for channel prediction. The simulation result shows that theperformance of this algorithm is better than LMS and RLS. It has fastconvergence speed than LMS while more simple operation than RLS.At last the paper summarized the contents of the work and pointed out theachievement and the insufficiency of the method in channel prediction that thepaper given. We should do more about the channel prediction for its fastdevelopment in the future. The working emphases of the future channel predictionarithmetic are given also.