Lte-A Downlink3D Mimo Channel Modeling And Precoding Research

In LTE using MIMO can increase the channel capacity. The performance evaluation of LTE is required to establish a standardized MIMO channel model. LTE-A introduces WINNER model. WINNER channel model represents the latest development of MIMO channel modeling, which can produce any wireless channel model scenarios, and it introduces3Dchannel model concept. With the development of antenna technology, antenna tilt adjustment and electronic vertical beam forming technology have developed rapidly.3D channel model can be combined with these new technologies, so as to further increase the channel capacity. However, as the new technologies been used, some problems have arisen. For example, when3D channel model is implemented in the LTE system, the precoding matrix needs to be redesigned to meet the higher dimensional beamforming. The user also needs to consider constructing higher dimensional feedbacks of channel state information.In order to solve the above problems, this paper works as follows. First we combine WINNER model and LTE-A network together by using modeling and simulation methods. And then we calculate3D MIMO channel capacity under different number of antennas and different SNRs. By the comparison with2D channel model, we observed that3D channel model improves the accuracy of the simulation, and3D model has wider range of application.The second part of this paper combines the3D channel scenarios and the LTE-A precoding techniques. We proposed two beamforming methods by considering the LTE-A user distribution. And based on the beamforming methods we propose the corresponding codebooks which are applicable to the3D channel model. The channel capacities under3D channel model when using precoding method show that the precoding method in3D channel model can still bring a considerable channel capacity increase, especially for cell edge users.Channel feedback in actual system is always restricted; a limited feedback beamforming method will inevitably bring the quantization error. To solve this problem, we propose a new precoding method. In the proposed method, without increasing the channel feedback, by using the original codebook and using the difference feedback method, the precoding matrix is gradually adjusted, which can make the channel capacity improvement. In this paper, we first use2D channel model to verify this optimization method, and then extend this method to the3D channel model. We both simulate the new precoding method in static2D/3D channel and time-variant2D/3D channel. Through observing the simulation results in2D and3D channel, we see that the new method brings a certain channel capacity increase.