Research On Frame Structure Designing For Massive MIMO Systems

With the rapid development of mobile broadband service and intelligent terminal technology,the demand for higher mobile network capacity and stabler service quality is also increasing exponentially.The 5th generation(5G)mobile communication system,which will be deployed on a large scale in the near future,also needs to be able to meet higher requirements in terms of spectral efficiency and energy efficiency.Largescale antenna system,also called massive multiple-input-multiple-output(MIMO),is a key technology of 5G,and has incomparable advantages over the traditional MIMO system.By deploying hundreds or thousands of antennas at the base station(BS)to serve a much smaller number of users simultaneously,massive MIMO can offer significant improvements in spectral efficiency and energy efficiency,as well as other huge performance gains.However,due to a large number of BS antennas and a tremendous growth in the number of connected mobile terminals,the current transmission frame structure of massive MIMO needs to consume plenty of pilot resources to obtain accurate channel state information(CSI).This not only reduces the transmission rate of payload data,but also results in the surcharge of valuable communication resources.In this paper,we focus on optimizing the frame structure of massive MIMO to improve its performance and ease the overwhelming CSI overhead.The main contribution of this work is as follows.1.We propose two novel frame structures for frequency-division-duplex(FDD)massive MIMO systems by exploiting the CSI feedback waiting phase directly.Based on the basic FDD frame structure,we consider fetching the estimated CSI of last transmission slot to compute precoding matrix and transmit downlink data in novel frame structures.In this case,the BS does not need to wait for the CSI feedback and can transmit data in both the CSI feedback waiting and the transmission phases,not only in the transmission phase,by which the proposed frame structures can achieve higher downlink rates than the conventional one.2.We provide a novel Bayesian neural network(BNN)-based channel prediction method by introducing Bayesian learning to neural network so as to incorporate uncertainties into predictions,which can not only deal with the overfitting issue but also control model complexity so that this procedure can be implemented online to track the rapidly-changing channels.Based on that,we propose a channel prediction-aided FDD frame structure to exploit the CSI feedback waiting phase more efficient.Numerical results show that our proposed channel prediction-aided FDD scheme can achieve remarkable performance gains in terms of either achievable downlink rates or bit error rate.Moreover,our proposed BNN-based channel predictor is much more effective and robust in contrast to the state-of-the-art channel prediction techniques such as autoregressive model and recurrent neural network.3.We propose a novel transmission frame structure for time-division-duplex(TDD)massive MIMO systems,in which the users with slow changing channels can estimate channels with lower frequencies(i.e.longer time interval)than users with fast changing channels by exploiting the temporal channel correlation underlying the channel aging effect.As compared to the conventional TDD scheme,the proposed one can fully utilize the diversity and redundancy among the actual user coherence times for transmitting payload data.We also derive a rigorous lower bound on the achievable rates,which takes into account the channel estimation error and the channel aging error.By maximizing the lower bound and solving the optimization problem,we determine the optimal time intervals of CSI estimation for all users.Numerical results show that the proposed TDD frame structure can offer great capacity gains over the conventional one,and provide insights on how to put the proposed scheme into practice.In conclusion,we study the optimization of transmission frame structure for massive MIMO systems in this paper.We propose algorithms and schemes respectively to alleviate the adverse effects caused by the consumption of a large number of pilot resources in the current massive MIMO frame structure,and realize a more efficient and flexible data transmission,which has been validated in the simulation experiments.Further,the research work in this paper provides some feasible and effective ideas for the optimization of massive MIMO frame structure in the future.