Research On Some Key Technologies Of Arrier Recovery In Wireless Ommunication

In wireless communication systems, noise and interference are two challenges forcarrier recovery. On one hand, the data aided estimators and phase lock loop wouldperform with a slow acquisition time and poor accuracy in low SNR environments. Onthe other hand, the cooperative MIMO systems have to deal with the problem ofestimation of multiple carrier offsets in low SINR scenarios. Thus, the design of carriersynchronizer with good performance can be difficult at low SNR and SINR. Theorganization and contributions of this thesis is as follows.The parameter design of frequency and phase detector was investigated in largefrequency offset and low SNR satellite systems. An analysis about the influence ofwindow type about the acquisition time and range was made. It is proved that thebigger window has a bigger range and slower speed. We also derived the closed-formexpression of the equivalent noise power spectrum density, which shows a less phasejitter and lower threshold can be achieved in bigger window. At last, a modified phaseand frequency detector is proposed by means of replacing the phase error signal withits polarity. The new phase and frequency detector has a better performance ofacquisition speed and phase jitter.The EM algorithm in deep space communication system employs optimal searchto perform ML estimation, and involves high complexity. Two simplified algorithmsare proposed. The first approach simplifies the status of the posterior mean andchanges the compensating way to reduce the complexity, the number of addition andmultiplication in EM algorithm is4times more than the first new algorithm. Thesimulations by the Turbo code with1/6code rate verify that the new algorithm canachieve perfect synchronization with only0.05dB loss when SNR exceeds-7.9dB.Inthe second method, a new equation is created by the equivalent change of thederivatives of objective function with respect to synchronization parameters. And then,a closed-form approximate solution of the new equation is developed by means ofTaylor approximation. Thus the second algorithm can avoid exhaustive search. Simulation and analysis results show the second simplified algorithm can achieve anenormous reduction in complexity without performance loss.In order to solve the contradiction between range and accuracy of estimation, acarrier recovery algorithm joint assisted by pilot and code symbols was researched.First, a coarse synchronizer with optimal pilot placement is designed based on the sumof autocorrelation functions of the pilots, whose frequency estimation variance isderived and can attain CRB. In addition, this estimator with low hardware complexitysufficiently exploits the correlation functions. Second, the coarse synchronizationalgorithm is applied in the equal spacing placement, and a unified formula for theestimation variance is derived. A fine synchronizer is implemented by means ofinserting pilot and data symbols together. In last, we discuss the problem of how toachieve the tradeoff between estimation range and accuracy by adjusting the parts ofpilots. Simulation results show the joint aided algorithm can avoid the drawbacks ofsingle data or code aided algorithm.The multiple carrier offsets in cooperative communication system not onlyintroduce performance loss, but also are a big challenge for the design of receiver. Wefirst analyzed the influence of residual carrier offsets on the SINR and capacity loss ofcooperative MIMO system. An expression for SINR is derived. It is proved as long asthe absolute value of the accumulated phase exceeds π, the SINR and capacity loss ismaximized. And then, we derived a closed-form expression for the CRB of multiplecarrier offsets in cooperative MIMO system. In last, a joint multiple frequency offsetsand channels estimator is proposed. This estimator consists of two stages: anapproximate ML coarse synchronization is implemented by correlation of receivedsignal with transmitted signal matrix. Parallel interference cancellation is deployed toremove multiuser interference firstly and generates a new signal. Then finesynchronizer searches frequency offset in the new signal gradient ascent. Simulationresults show the proposed estimator can attain CRB bound, and also achieve fullsynchronization rang and1.5dB improvement compared to the traditional correlationalgorithm in estimation accuracy.