Effect of Carrier Frequency Offset on Channel Estimation for SISO/MIMO-OFDM Systems

• Published in: IEEE transactions on wireless communications Vol. 6; no. 5; pp. 1854 - 1863
• Main Authors: Weng, L., Au, E.K.S., Chan, P.W.C., Murch, R.D., Cheng, R.S., Wai Ho Mow, Lau, V.K.N.
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01-05-2007; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Antennas; Applied sciences; Carrier frequencies; Channel estimation; Channels; Errors; Exact sciences and technology; Frequency division multiplexing; Frequency estimation; Gold; Interference; Least squares approximation; Least squares method; Mean square error methods; Mean square values; MIMO; OFDM; Offsets; Orthogonal Frequency Division Multiplexing; Pilots; Radiocommunications; Systems, networks and services of telecommunications; Telecommunications; Telecommunications and information theory; Transmission and modulation (techniques and equipments); Upper bound; Performance evaluation; MIMO system; Parameter estimation; System design; Mean square error; Upper bound; Carrier frequency; Orthogonal Frequency Division Multiplexing(OFDM); Least squares method; carrier frequency offset (CFO); Channel estimation; Analytical method; Orthogonal frequency division multiplexing; SISO system; Frequency offset; Antenna array
• ISSN: 1536-1276; 1558-2248
• DOI: 10.1109/TWC.2007.360387

Few works have addressed the effect of CFO (carrier frequency offset) on channel estimation performance. In this paper, compact analytical expressions on the mean square error of channel estimation in the presence of CFO are derived for OFDM (orthogonal frequency division multiplexing) systems employing single and multiple antennas. Concise upper bounds are also derived for SISO-OFDM systems. It is revealed that channel estimation MSE (mean square error) increases at the rate of approximately the square of the CFO and increasing the number of pilots does not contribute to better suppression of the MSE caused by CFO. In addition, it is observed that LMMSE (linear minimum mean square error) channel estimation is more resistant to CFO compared to LS (least square) in terms of channel estimation MSE. Furthermore, from the results derived herein, a clue for the pilot design of OFDM systems with CFO can be obtained