Research On Channel And Interference Prediction In LTE Uplink

Due to the enhancement of throughput,data rate and spectral efficiency,link adaptation has been adopted as the key technology of long term evolution(LTE)uplink.It has a significant impact on the performance of the adaptive technology that whether the feedback information is timely and what the interference condition is.On the one hand,in order to reduce the influence of inherent delay,it is necessary to predict the channel quality to make sure the timeliness of channel state information(CSI).On the other hand,in order to reduce the influence of inter cell interference(ICI)and improve the performance of signal to interference plus noise ratio(SINR)estimation,we can predict interference by means of inter cell cooperation.We deeply study the background and significance of LTE system,and make a summary of the key technologies in LTE uplink.Then starting from the LTE uplink system,the frame structure and the cell distribution are analyzed,and the corresponding channel and interference prediction schemes are presented.The main contents and achievements of this paper are as follows:Channel prediction research is carried out in both high speed scenario(60~120km/h)and low speed scenario(20~50km/h).In high speed scenario,firstly variable coefficient auto regressive(AR)model is proposed to use AR model to adjust the result of polynomial fitting algorithm,whose prediction performance is slightly improved than polynomial.Secondly,adaptive filtering algorithm(Kalman,Recursive Least Squares)achieves better performance(3ms)by continuously updating the model coefficients with a large number of training samples.In low speed scenario,radial basis function(RBF)neural network prediction algorithm,which based on phase space reconstruction,is proposed.It use chaos attractor to predict channel coefficients in high dimensional space.The proposed scheme can predict channel coefficient accurately for 5ms.The inter cell cooperative interference prediction(CIP)algorithm proposed by Muller A is applied to the LTE uplink system,and the simulation and improved algorithm are carried out.Firstly,the solving process and some parameters of CIP algorithm are simplified in consideration of the complexity and accuracy.Secondly,to solve the problem that SDEdecision exchange(SDE)and equal probability exchange(EPS)has similar performance,improvement is proposed as follows: reducing the specified transmission rate,adjusting the interference power of the EPS,and selecting the main interference source of nearby cells.The simulation results show that SDE is the best CIP scheme,and its average cell rate is 0.2 bpcu(10 ~ 60 km / h)higher than the EPS scheme,CIP can effectively reduces the influence of ICI.