Parameter estimation and error reduction for OFDM-based WLANs

Abstract

We consider parameter estimation and error reduction for orthogonal frequency-division multiplexing (OFDM) based high-speed wireless local area networks (WLANs). We devise or select algorithms that can provide benefit to the overall system performance and can be efficiently implemented in real-time. In particular, first, we give a channel model which is especially useful for assessing the channel parameter estimation methods devised for OFDM-based WLANs. Second, we provide a sequential method for the estimation of carrier frequency offset (CFO), symbol timing, and channel response by exploiting the structure of the packet preamble specified by the IEEE 802.11a standard. Finally, to correct the residue CFO induced phase error using the pilot tones, we consider maximum-likelihood phase tracking and least-squares phase fitting approaches; to improve the channel estimation accuracy using the decoded data, we present a semiblind channel estimation method; to mitigate the sampling clock induced time delay error, we provide a sampling clock synchronization approach that obviates the need of an automatic frequency control clock recovery circuit. The overall system performance of using our algorithms is demonstrated via several numerical examples.