Research On Synchronization Parameter Estimation For Distributed Multiple Antenna (MIMO) And MIMO-OFDM Systems

This paper researches the problem of synchronization parameter estimation in distributed Multiple-input Multiple-output (MIMO) and Frequency-Division Multiplexing (OFDM) systems based on statistical signal processing method. It regards lower bounds of channel fading and frequency offset estimation as evaluation indicators, and studies on optimal algorithms and a variety of sub-optimal iterative ones from the points of iterative convergence rate and algorithm complexity.The major work of this paper includes the following. It analyzes the two simplification algorithms based on the frame of expectation maximization (EM) algorithm, e.g., Expectation Conditional Maximization (ECM) algorithm and Space- alternating Generalized Expectation Maximization (SAGE) algorithm. It analyzes the correlation frequency estimation algorithm for acquiring the original value of iterative algorithms. It analyzes the Cramer-Rao bound (CRB) and asymptotic CRB for evaluation of the performance of all listed algorithms. To overcome the unstable convergence issues, it proposes improved parameter update method which combines the maximizing step of ECM and SAGE algorithm with the ML estimation. It has compared the algorithms performance on the self-established MIMO simulation platform. Simulation results show that the asymptotic CRB has a very high degree of approximation to CRB. Moreover, simulation results show that both iteration methods can achieve CRB accuracy without error platform at high signal to noise ratio. Simulation also shows that both improved algorithms have high stability, as well as modified SAGE algorithm has lower calculation complexity, higher convergence speed and more stability.The paper further investigates the synchronization application of the two kinds of improved algorithms on MIMO-OFDM system with Rayleigh channel. It derives the two improved algorithms for realization of synchronous problem in this system. It considered the algorithm convergence while equal or unequal power is distributed at the transmitter side. From the comparison and analysis of simulation results, we find that the two improved algorithms can also approximate CRB in this system. We also find that improved SAGE algorithm has the higher rate of convergence. Average noise allocation among the transmitter antennas has the higher rate of convergence.