Research On Synchronization Estimation For OFDMA Uplink Transmissions

Because of higher speed of date transmission, higher spectrum efficiency andstronger anti-multipath interference, Orthogonal Frequency Division MultiplexedMultiple Access (OFDMA) has became an air interface standard of wirelesscommunication system and an important support of the Fourth-Generation broadbandwireless mobile communication systems (4G) TD-LTE. As the influence of thetime-varying characteristics of wireless channels, the frequency selectivity, additivenoise and the frequency inconsistency of crystal oscillator between transfer and receiver,it will causes Timing-Error (TE) and Carriers-Frequency-Offset (CFO), and destroy theorthogonality of sub-carriers for OFDMA uplink transformations. it even leads to blocksinterference, and multiple access interference at the same time. The communicationperformance of OFDMA system will degrades at last. So synchronization is particularlyimportant for OFDMA uplink transmissions. Synchronization parameters(for exampletiming offset, frequency offset, and channel informations) estimation is also animportant links of synchronization. Comparing with downlink transmissions, it greatlyincreases the difficulty of synchronization parameters estimation for OFDMA uplinktransmissions. Firstly, because at Basic Station (BS) the received signals is thesuperposition of multiple-users signal. Every user signals across different channels andsubject to different fading, its TE and CFO are difference. That is to say, the uplinksignal of different user accompany by signal feature with different TE and CFO. Theissue of single parameter estimation for traditional OFDMA becomes multipleparameters estimation problems for OFDMA uplink transformations. Secondly,parameters estimation closely associated with the subcarrier allocation scheme.Different subcarrier allocation scheme has different difficulty of separating differentuser signal, and increases the difficulty of synchronization parameters estimation. Tomeet the demand of people utilizing communication and data, OFDMA system must besynchronized. The technologies of synchronization parameters estimation with efficientand good performance for OFDMA uplink transmissions will promote the developmentof future mobile communication.During OFDMA uplink transformations with generalized subcarrier allocationscheme, there are problems that the traditional SAGE algorithm not only has high complexity, but also limits the number of users communicated simultaneously in thesystem. The traditional SAGE algorithm is improved by joint maximum likelihoodestimation algorithm, which makes the use of Minimum Residual Decomposition (MRD)to separate each user signal from the received signals for reducing signalsdecomposition at first, then calculates each user synchronization parameters estimationvalued as the initial value, finally fine-tune it with Alternating Projection(AP) methodimproving the estimation accuracy. The Improved algorithm avoids the iterativeoperation, reduces the estimation algorithm complexity, and increases the number ofusers in the OFDMAsystem uplink at the same time activation.In this paper, synchronization parameters estimation techniques are researched forOFDMAsystem uplink transmissions;First of all, the history of the formation and research status of OFDMA is recalled,the importance of timing and frequency synchronization for OFDMA uplinktransmissions is elaborated.Further more, OFDMA system model and some key technologies are described, aswell as give a brief introduction to the advantages and disadvantages of the system.Then, the synchronization principles of frame and frequency is explained, as wellas the corresponding OFDMA sub-carriers allocation and the damage of timing andfrequency offset to OFDMAsystem are analyzed.The last but not least, we focus on the research on synchronization parametersestimation for OFDMA uplink transmissions, then take use of simulation and analyzethe advantages and disadvantages of leading frequency sequence algorithm, MLalgorithms, CTY algorithm and SAGE algorithms. With high complexity and subjectingto the issue of limiting the number of users, SAGE algorithm is improved by a joint MLestimation technique with its basic idea and design process. By MATLAB simulation,joint ML estimation is compared performance with the SAGE algorithm and APalgorithm. The results show that the improved algorithm reduces algorithm complexity,improves estimation precision, and increases the number of users in the OFDMA systemuplink at the same time activation.