| As people’s demands for mobile communication service rates and quality is higher and higher, the organization of 3GPP brought up a program on 3G long term evolution(LTE) in 2004 to remain competitive. LTE is capable of increasing data transmission rates efficiently, decreasing system delay, enlarging system capacity and coverage and improving service quality. Therefore LTE will be a communication technical standard and trend for a long time in the future, which is important to doing some research about LTE. This paper is focusing on LTE uplink timing synchronization based on sounding reference signal(SRS).Firstly, this paper introduces LTE about historical background, current domestic and international development, performance requirement and system architecture. It also elaborates some concepts about LTE physical layer, including uplink and downlink transmission technology, structure of data frame, time-frequency resource elements, division of physical channel and reference signals. With respect to reference signals this paper specifically focuses on SRS’s ways of generation and mapping, transmission period and location.Next, this paper introduces briefly timing offset’s effect on received signals from the perspective of graph and mathematics,then LTE process of uplink synchronization is present. Some synchronization algorithms are studied in detail based on reference signal which has two same parts. These algorithms are classical Schmidl & Cox algorithm, and relevant improvement algorithms, including Minn algorithm, Adegbenga B algorithm and Zhang algorithm. Performance simulations are done in AWGN channel to check these algorithms’ timing metric curve. Further, their performance are compared to present advantages and disadvantages. The curves indicate that SC algorithm and Zhang algorithm have a poor performance because of the existence of plateau; Minn algorithm And Adegbenga B algorithm perform well at the cost of complexity.At last, simulation channel models of LTE are introduced. Then the paper brings up two kinds of improved algorithms based on sounding reference signals. The first is the improvement to SC algorithm and the second is to extract timing information from received signals’ rotated phase. Simulations with these two new algorithms are done in AWGN and multipath EVA channels while One user and two users are communicatingto check their performance in different kinds of communication situations. The comparison results show that SC improved algorithm has a better timing performance and can effectively resist frequency deviation and noise.. Frequency-domain improved algorithm has a advantage of lower computational complexity, and does well in timing performance compared with the mentioned algorithms. |