Channel estimation, data detection and carrier frequency offset estimation in OFDM systems

Abstract

Orthogonal Frequency Division Multiplexing (OFDM) plays an important role in the implementation of high data rate communication. In this thesis, the problems of data detection and channel and carrier frequency offset estimation in OFDM systems are studied. Multi-symbol non-coherent data detection is studied which performs data detection by processing multiple symbols without the knowledge of the channel impulse response (CIR). For coherent data detection, the CIR needs to be estimated. Our objective in this thesis is to work on blind channel estimators which can extract the CIR using just one block of received OFDM data. A blind channel estimator for (Single Input Multi Output) SIMO OFDM systems is derived. The conditions under which the estimator is identifiable is studied and solutions to resolve the phase ambiguity of the proposed estimator are given.A channel estimator for superimposed OFDM systems is proposed and its CRB is derived. The idea of simultaneous transmission of pilot and data symbols on each subcarrier, the so called superimposed technique, introduces the efficient use of bandwidth in OFDM context. Pilot symbols can be added to data symbols to enable CIR estimation without sacrificing the data rate. Despite the many advantages of OFDM, it suffers from sensitivity to carrier frequency offset (CFO). CFO destroys the orthogonality between the subcarriers. Thus, it is necessary for the receiver to estimate and compensate for the frequency offset. Several high accuracy estimators are derived. These include CFO estimators, as well as a joint iterative channel/CFO estimator/data detector for superimposed OFDM. The objective is to achieve CFO estimation with using just one OFDM block of received data and without the knowledge of CIR.