Expert Systems with Applications: An International Journal
A Blind LCMV-Based CFO Estimation for MC-CDMA Systems Over Multipath Fading Channels
Wireless Personal Communications: An International Journal
| Hi-index | 35.69 |
This paper presents a training-based estimation method of carrier frequency offset (CFO) for multicarrier code-division multiple-access (MC-CDMA) uplink. The basic assumption is the same as in Morellis' work (OFDMA), i.e., active users in the network are classified into two categories: reference (synchronized) users (RUs) and new (asynchronized) users (NUs). But, we consider the system that RUs use CDMA to share the wireless medium. NUs employ time-division multiple-access (TDMA) for the training purpose. Our major idea is to perform the CFO estimation in the desired signal subspace that is orthogonal to the RU subspace. The proposed approach is investigated both in the subband and interleaved MC-CDMA systems. It is shown that implementation of the CFO estimation together with the training-pattern design could be relatively simple when users do not experience the considerable frequency-selective fade. In the highly frequency-selective fading scenario, an adaptive training-scheme is proposed to make the CFO estimator robust. We also consider NUs operating in the frequency-division multiple-access (FDMA) training mode. Due to the presence of inter-NU-interference, the estimation performance may be not as good as in the TDMA training mode. The performance degradation may be severe when RUs operate in the interleaved MC-CDMA. In this case, we propose an inter-NU-interference self-canceling training-pattern and use simulations to demonstrate its impact on the estimation performance