On Key Techniques For Data Transmission In LTE Uplink Systems

Due to the advantages of high spectral efficiency and low peak-to-average power ratio,single carrier-frequency division multiple access(SC-FDMA)has been applied as the transmission scheme in long term evolution(LTE)uplink.The trend of the future uplink is higher data transmission rate,lower transmission delay and greater access capacity.Limited by the shortage of spectrum resources and the time-frequency dynamic characteristics of wireless channels,the further increase of data transmission rate is facing challenges.In order to solve the contradiction between high transmission rate and the shortage of spectrum resources in the uplink,the employment of higher-order modulation serves as one possible approach.As the modulation order increases,the impairments of the phase noise to the SC-FDMA system will increase.As a kind of RF front-end impairment,phase noise destroys the orthogonality between the subcarriers of the SC-FDMA system and leads to self-interference and multiuser interference,which greatly reduces the uplink transmission performance.In order to solve the contradiction between the high transmission rate and the time-frequency dynamic characteristics of the wireless channel in the uplink,link adaptation technique has been widely studied.As one of the main techniques in link adaptation,adaptive modulation and coding(AMC)maximizes link throughput by selecting adaptive parameters such as the best modulation and coding scheme(MCS)under certain restrictions.This dissertation focuses on improving the performance of uplink data transmission,and conducts in-depth studies on the methods of suppressing uplink phase noise and improving the performance of AMC schemes.The main work and contributions are listed as follows:1? The method of suppressing phase noise at the transmitting end in uplink single-user scenarios is studied and a phase noise pre-compensation scheme is proposed.The phase noise prediction model is trained by using the past and present samples of the phase noise,and the future samples of the phase noise are predicted and pre-compensated.Compared with the existing methods,the proposed scheme has low hardware complexity and is easy to implement.Moreover,the proposed scheme can improve the quality of the transmitted signal and reduce the out-of-band radiation.In order to improve the performance of the proposed pre-compensation scheme,a phase noise adaptive prediction algorithm is proposed.The proposed algorithm uses an iterative approach to find the optimal model to predict phase noise.With a reasonable increase in complexity,the proposed algorithm achieves better performance than the existing prediction algorithms.2? The method of suppressing phase noise at the receiving end in uplink single-user scenarios is studied and a blind phase noise estimation and compensation scheme based on unscented Kalman filter(UKF)is proposed.The phase noise estimation is obtained by utilizing the UKF and the decision symbols.Then,the proposed scheme performs two-step phase noise compensation on the received symbols.Without using pilots,the proposed scheme achieves high spectral efficiency.In order to improve the performance of the proposed blind estimation scheme,a symbol decision method based on Mahalanobis distance is proposed.Compared with the traditional hard decision method using Euclidean distance,the proposed method improves the decision accuracy by exploiting the statistics of phase noise,and can significantly enhance system performance in scenarios with large phase noise and high signal-to-noise ratio(SNR).3? The method of suppressing phase noise in uplink multi-user scenarios is studied and a multi-user phase noise suppression scheme is proposed.The received signal of each user is extracted in descending order of SNR level in frequency domain,and phase noise estimation and compensation are performed through iterative methods.Then,the multi-user interference caused by phase noise is eliminated and the transmission performance of users with low SNR is improved.In order to improve the performance of phase noise estimation compensation module,an iterative algorithm of joint decoding and phase noise estimation compensation based on transform domain sub-block processing is proposed.By exploiting the low-pass nature of phase noise,the proposed iterative algorithm reduces its complexity by reducing the number of phase noise samples to be estimated,and improves its initial estimation accuracy by sub-block averaging process.In order to further reduce the complexity of the proposed algorithm,a modified soft information calculation method is proposed.The algorithm uses the statistical information of residual phase noise to improve the accuracy of soft information,thereby improving the performance of the decoder and reducing the number of iterations.In order to improve the performance of the proposed iterative algorithm,a phase noise estimation method based on EM algorithm is proposed.4? The AMC scheme and signal model in LTE uplink is studied.To solve the problem of channel state information loss in the channel quality measurement existing in traditional AMC schemes,a R REBER based AMC scheme is proposed.The proposed scheme uses the ratio of the average bit power and the bit error power of the transform domain reconstructed signal in place of the conventional signal-to-interference-plus-noise ratio(SINR)as a channel quality metric.The relationship between the proposed metric and the bit error rate is analyzed under high SNR scenario and low SNR scenario,and the performance of the proposed scheme under VehA and PedB channels is verified by simulations.Compared with the traditional AMC schemes,the proposed scheme has better performance.In order to further improve the performance of AMC schemes in LTE uplink,the application of machine learning methods in AMC schemes is explored.The k-nearest neighbor(k-NN)algorithm is applied to the uplink AMC scheme.The scheme uses a sliding window to update the feature samples for training,and classifies the target sample by using the k nearest feature samples,and then selects the optimal MCS to implement AMC transmission,thus solving the problem that the fixed look-up table in traditional AMC schemes cannot adapt to the dynamic change of the user.The support vector machine(SVM)algorithm is applied to the uplink AMC scheme.The scheme uses SVM mapping to map the low-dimensional channel quality information to high-dimensional to achieve optimal classification,thus reducing channel state information loss in the channel quality measurement and improving the total performance.At the same time,the introduction of the kernel function can greatly reduce the computational complexity.