An OFDM signal consists of N sinusoids (bins) with spacing 1/T, that are modulated by data symbols that have a duration T, equal to the inverse channel spacing. Hence, although the modulated bins spectrally overlap, they are orthogonal. The generation of the OEDM signals at the transmitter and the demodulation at the receiver can be pedormed efficiently by means of Fast Fourier Transforms. The use of OEDM signals has been proposed for various applications, such as broadcasting of digital audio and digital TV, and high bit rate transmission on twisted pair cables [ 1-41.During a symbol period T, the complex envelope of the transmitted OFDM signal can be expressed asThe data symbol o, modulates the sinusoid at frequency m/T during a period T. The total symbol rate R equals NIT. In the following O(t) denotes a time varying phase caused by either a carrier offset between the receiver and transmitter carrier, or the phase noise of these carriers. In the first case, O(t) is deterministic and equals 2aAFt + 00 where AP is the carrier offset. In the latter case, 8(t) is modqled as a Wiener process for which E[O(t)] =O and E[(8(t +to) -0(h)12 = 4 4 I t I, where /3 [Hzl denotes the one-sided 3 dB linewidth of the Lorentzian power density spectrum of the free-running carrier generator. During a symbol period, the signal at the FFT output corresponding to the k-th bin, can be written as Paper appmved by Desmond I? Taylor. the Editor for Signal Design, Mudulation and Detection of the DeEE Communications Socity. Manuscript received: October 13,1993;revised: June22, 1994. Port of this work hos been supported by Alcatel-Bell, Antwerp, Belgium and Barco, Komijk, Belgium. The authors are with the Communication Engineering lab, University of Ghent, B-9ooo. Ghent, Belgium.where I,, = f ~* , -j 2 r t~, p ( t ) d t .We observe that the presence of phase noise andor frequency offset affects the received useful signal. Both impairments rotate and attenuate the useful symbol ai. In addition, signal components originating from bins other than the considered one give rise to interbin interference (IBI). The third term in (1) denotes the thermal noise contribution of the AWGN channel.The receiver rotates clockwise the signal TI^ over an angle \y which is an estimate of arg(lo), and feeds the resulting signal to the decision device. In the following, we (optimistically) assume that the receiver can perfectly estimate u~g(10). i.e. Y = org(1o). This yields N-1 vk = 7'le-J' = all101 + ( 4nIk-m + N k ) e -J y . Writing 1 1 0 1 as EO + 6 with EO = E[lIol], the useful signal component in V k is akEo whereas ak6 is an additional noise component. The signal-to-noise-ratio (SNR) at the input of the decision device is given by EO2 SNR=-. NO -+ vo E, where Eo2 is the power of the useful component of U k , &/E, denotes the variance of the thermal noise contribution, and VO equals the variance of the other noise terms, i.e. N-I vo = E[l6IZI + E[lLnlzl * (3) m d m#h Comparing (2) with SNR = E,/& in absence of carrier phase impairments, we define the degr...

No abstract

No abstract

This paper is devoted to turbo synchronization, that is to say the use of soft information to estimate parameters like carrier phase, frequency offset or timing within a turbo receiver. It is shown how maximum-likelihood estimation of those synchronization parameters can be implemented by means of the iterative expectation-maximization (EM) algorithm [1]. Then we show that the EM algorithm iterations can be combined with those of a turbo receiver. This leads to a general theoretical framework for turbo synchronization. The soft decision-directed ad-hoc algorithm proposed in [2] for carrier phase recovery turns out to be a particular instance of this implementation. The proposed mathematical framework is illustrated by simulations reported for the particular case of carrier phase estimation combined with iterative demodulation and decoding [3]. 2933 0-7803-7802-4/03/$17.00

Abstract-In mobile radio communication, the fading channels generally exhibit both time-selectivity and frequency-selectivity. Orthogonal frequency-division multiplexing has been proposed to combat the frequency-selectivity, but its performance is also affected by the time-selectivity. In this paper, we investigate how various parameters, such as the number of carriers, the guard time length, and the sampling offset between receiver and transmitter, affect the system performance. Further, we determine the optimum values of the above parameters, which minimize the degradation of the signal-to-noise ratio at the input of the decision device.