Wireless Resource Management Based On Prediction Technology

The future mobile communication system is moving in the direction of high-speed, broadband, large capacity, more business, low cost, expected to obtain a higher data rate and better customer service quality, satisfied the demand of multi-media services, such as video and telephone conference. MIMO-OFDM is a key technology which is regarded as mainly physical layer technology of the fourth generation mobile communication system. It has been confirmed that the precoding/beamforming and adaptive modulation can improve the performance of MIMO-OFDM systems with real-time effective channel state information (CSI). However, the channel time-varying characteristics and delay often leads to the CSI of transmitter hysteretic and inaccuracy, in the channel estimation of traditional system can only get the current CSI, it will bring serious performance loss of wireless resource management scheme based CSI. Channel prediction technology is considered to be a way to overcome this effect, which can provide accurate and timely CSI for the system. A lot of research found that the current channel prediction algorithm is either high computational complexity without practical application significance, or is not fully considering the time, frequency, space channel correlation features.In this thesis, focusing on the above problems, The channel prediction technology and its application in MIMO-OFDM system are researched.First of all, the MIMO-OFDM system and its key technology are studied and analyzed. And then introduces the wireless channel propagation characteristics. The time-varying channel characteristic is analyzed. In MIMO system, CSI important role in the performance of the system are analyzed. The current wireless resource allocation and management technology are based on the perfect CSIT hypothesis, however, due to processing delays in practical systems, feedback delay and time-varying channel, which will cause CSIT inaccuracy. Then, we present the affect with unperfected CSI to the performance of the system with simulation analysis.Based on the present situation of channel prediction research, a method is proposed in this paper, which considering three dimensionalities of the channel correlation, time, frequency and space. It can bring a better prediction performance with low computational complexity. Tt is concluded that the proposed approach can achieve accuracy CSI with better application of channel3D channel correction features in under the premise of not obviously increase the computational complexity.On the basis of channel prediction technology research, this article also study the application of channel prediction in precoding/beamforming system. A robust linear precoding scheme is proposed in MU-MIMO system, which is based on channel prediction technology. And it has carried on the simulation analysis to the scheme. The prediction based compensation CSI for linear precoding scheme can bring system performance boost, overcome the serious effect of outdated CSI, which is caused by such factors of the time-varying channel and delay.The end of the paper is a summary of the paper, and further to the future research directions are discussed. The important research points in the future are pointed out. It is highlight that the application of channel prediction technology in future wireless communications, future research will pay more attention to adaptive technology with the real-time/predicted accurate CSI, in order to realize the optimization of wireless resource utilization.