Use of Diverse Delayed Correlation for an ML Carrier Frequency Offset Estimator in OFDM-Based IEEE 802.11a WLANs

Abstract

Estimation of carrier frequency offset (CFO) is an important issue in the design of a WLAN receiver that employs OFDM techniques. In this paper, a new and efficient CFO estimation scheme that uses the ten short training symbols specified in the IEEE 802.11a standard is proposed. This scheme, which we call DDC-ML, makes use of a diverse delayed correlation (DDC) skill in the ML estimator. We show that by using DDC-ML to estimate the CFO both large range and low variance of estimation error (VOER) can be attained. For AWGN channels under moderate signal to noise ratio (SNR) conditions, a mathematical analysis is developed to evaluate the VOER resulted from a CFO ML estimator that uses delayed correlation (DC-ML). The analysis was corroborated via simulations, and compared with the formulated Cramer-Rao lower bound (CRLB). An optimum parameter combination for DC-ML estimator that can achieve minimum VOER was obtained. VOER for the DDC-ML operated in a multipath environment is also investigated via simulations. In addition, a new ambiguity resolution algorithm (ARA) for the DDC-ML is introduced and probability of false resolution is presented. Performances of DC-ML and DDC-ML are compared in terms of probability of estimation error.