Synchronization in OFDM Systems

High data rate and high mobility are two major challenging requirements for future wireless communication systems. To meet such demanding requirements, orthogonal frequency division multiplexing (OFDM) has been considered a viable technique. High spectral efficiency, effective multipath and noise immunity, flexible resource allocation and being easy to integrate with other techniques are some of the major advantages of OFDM modulations. Nevertheless, OFDM also has its drawbacks that include high-peak-to-average power ratio and sensitivity to synchronization errors.

This dissertation addresses one of the critical issues in the development and implementation of OFDM systems, namely, synchronization. Focusing on continuous OFDM transmission, we utilize the acquired (i.e., either known or estimated) channel statistics to improve the synchronization performance and to simplify the process. For timing synchronization, we consider fine timing and propose a Maximum Likelihood (ML) timing estimation scheme. This proposed scheme is further developed for integer and real-valued precision implementations. Our analysis and simulation results showed that, compared to conventional methods, the proposed algorithm provides significantly improved performance, measured by timing error variance. The improved timing error variance will in turn benefit the OFDM system performance. For carrier synchronization, we have investigated the existing schemes for channel estimation and fine carrier estimation, and proposed a carrier frequency offset (CFO) estimation scheme based on time-domain channel estimates. The proposed method is, in essence, an ML estimate utilizing CFO information embedded in the phase rotation of the estimated channel multipaths. The ML CFO estimate is extended to the time-variant fading case to further explore the performance gain under fading channels. The proposed scheme yields excellent performance with reduced complexity. It also has the advantage of less restriction on pilot patterns.

In recent years, multiple-input multiple-output (MIMO) wireless communication systems have received a lot of attention. In particular, MIMO-OFDM appears to be one of the most promising techniques for future 4G systems. We extended our proposed OFDM timing and carrier synchronization schemes to MIMO cases. Specially, we developed CFO estimation schemes for MIMO-OFDM. It is shown that the CFO estimation can benefit from receiver diversity gain under the setting of MIMO systems. A comb-pilot based MIMO-OFDM system is studied in detail to illustrate the idea of estimating the CFO based on channel estimates. Again, the resulting CFO estimate has the attractive feature of not requiring specific pilot design.