Abstract:The Symbol timing estimation in a high data rate mobile OFDM systems is a challenging task, this is due to the high delay spread. A conventional timing estimation based on autocorrelation experience performance degradation due to multipath channel distortion. To improve the performance of autocorrelation technique, we propose a normalized 3-rd order central moment for symmetric correlator as test statistics. When the training symbol is received, the distribution of the symmetric correlator is Rician distribution. This will be different from no training symbol is received, the distribution of the symmetric correlator is Rayleigh distribution. After we perform coarse timing estimation using a normalized 3-rd order central moment, fine timing estimation is done using the cyclic shift of a channel estimate. The simulations results have shown that our method is robust in the fading multipath channel and has better performance than the previous methods.Keywords: High data rate, mobile, multipath channel, OFDM, timing estimation.
IntroductionDemand for high data rate applications in mobile environment continues to rise. Robust mobile communication with high data rate is still challenging. High data rate transmission over mobile environment typically experience higher delay spread. OFDM (Orthogonal Frequency Division Multiplexing) systems is robust in environments with a high delay spread. OFDM is a digital multicarrier systems, wherein all channels are split into several sub-channels in order to transmit data at high rate.Compared with single carrier systems, OFDM systems offer high frequency efficiency and increase robustness against Intersymbol Interference (ISI) and fading caused by the multipath channel. Therefore, OFDM systems have been widely adopted for WLAN [1], DVB systems [2], and WMAN 802.16 [3]. The last two applications require high data rate transmission in mobile environment. OFDM is also a new technology for Cognitive radio systems [4]- [5]. However, the performance of an OFDM systems is influenced by timing synchronization. Tight timing synchronization is needed as the timing error introduces ISI and decrease the overall system performance [6]. In this paper, the timing synchronization refers to the detection of the start of an OFDM symbol.The symbol timing estimation in a high data rate mobile OFDM systems is a challenging task, this is due to the high delay spread. Moreover, in case high speed mobility, the channel fluctuate rapidly. A conventional timing estimation based on autocorrelation experience performance degradation due to multipath channel distortion.To improve the performance of autocorrelation technique, we propose a normalized 3-rd order central moment for symmetric correlator as test statistics. Their values are compared with a predefined threshold to obtain the first arriving path, which refers to the start of an OFDM symbol. After we perform coarse timing estimation using a normalized 3-rd order central moment, fine timing estimation is done using the cyclic shift of a channe...
