Channel Estimation for Mobile OFDM Systems Using CR Splines

Abstract

In OFDM systems, accurate channel estimation is the key to compensate the effect of the interchannel interference in a multipath Rayleigh fading channel. In this letter, a novel channel estimation method for mobile OFDM systems is proposed by using the interpolation technique based on Catmull-Rom spline function. The effectiveness of the proposed method is investigated by computer simulations.Keywords: channel estimation, OFDM, ICI, fast fading, CR splines, Doppler shift IntroductionOrthogonal frequency-division multiplexing(OFDM) is a potential technique for high-speed data transmissions over digital communications (1) . However, the time variation of a fading channel over one frame period causes a loss of subchannel orthogonality called the interchannel interference(ICI), resulting in an error floor that increases with the Doppler frequency. Accurate channel estimation is the key to compensate the effect of ICI in a multipath-fading channel.The channel estimation in mobile OFDM systems is generally based on the use of pilot symbols in the given positions of the time and the frequency domains. In the time domain, the channel fading envelope(CFE) can be considered as varying in a linear fashion and interpolation based methods have been proposed (2) (3) .In this letter, a novel channel estimation method for mobile OFDM systems is proposed by using the interpolation technique based on Catmull-Rom(CR) spline function (4) . As we know that the CFE is smooth and the interpolation scheme on basis of CR spline function can ensure both its continuity and the existence of its first order derivative (4) , we use CR spline function to approximate the CFE. This method can estimate the CFE with a slight computational load and two frames delay. OFDM System ModelThe OFDM transmitter converts modulated subcarriers into a time sequence by an Inverse Fast Fourier Transform(IFFT). After the IFFT, pilot symbol(PS) and Guard interval(GI) are inserted to estimate the CFE and to avoid the intersymbol interference(ISI), respectively. Assume that the multipath fading channel consists of L discrete paths. It is also assumed that the entire channel impulse response lies inside the GI of length N g , meaning 0 ≤ (L − 1)T s ≤ N g T s , where T s is the sampling interval. After removing the PS and the GI at the OFDM receiver, the m-th demodulated signal is obtained by the FFT as

A Novel Channel Estimation and Equalization for MIMO-OFDM Systems

“…To keep a balance between cost and performance, one interesting approximation method, CR-spline function (19) is introduced, which has been well studied in channel estimation for OFDM systems. (10) , (13) , (18) To reconstruct the CFE between p…”

“…For channel estimation and equalization in OFDM systems with doubly-selective fading channel, several frequency domain pilot (FDP) aided (8)- (10) and time domain pilot (TDP) aided (11)- (13) approaches have been proposed. In particular, these TDP aided approaches employ three different interpolation methods, the linear fashion interpolation (11) , low pass filter (LPF) upsampling interpolation (12) and spline function interpolation (13) to obtain the channel fading envelope (CFE). However, for the MIMO-OFDM systems, these methods cannot be applied directly because of the time domain cross interference from other transmitter antennas.…”