Nils Holte

Optimal finite duration pulses for OFDM

This paper presents pulses of finite duration for use in orthogonal frequency division multiplexing (OFDM) using offset QAM modulation. In such a system the pulses will overlap in both frequency and time without intersymbol or interchannel interference. The pulses are generated by an optimization procedure which minimizes the out-of-band energy under the constraints of zero intersymbol and interchannel interference. Low out-of-band energy is important to achieve high bandwidth efficiency, especially when the number of channels is small. The optimization procedure leads to an integral equation which is solved by expanding the pulse in truncated prolate spheroidal wave functions. Examples of optimized pulses and their spectrum are presented for lengths of 2 and 4 symbol intervals.

Decision feedback sequence estimation for continuous phase modulation on a linear multipath channel

An approach to reduced-complexity detection of partial response continuous phase modulation (CPM) on a linear multipath channel is presented. The method, referred to as decision feedback sequence estimation (DFSE), is based on a conventional Viterbi algorithm (VA) using a reduced-state trellis combined with decision feedback (DF). By varying the number of states in the VA, the receiver structure can be changed gradually from a DF receiver to the optimal maximum-likelihood sequence estimator (MLSE). In this way different tradeoffs between performance and complexity can be obtained. Results on the receiver performance, based on minimum distance calculations and bit error rate simulations, are given for Gaussian minimum-shift keying modulation on typical mobile radio channels. It is shown that for channels with a long memory, a significant complexity reduction can be achieved at the cost of a moderate degradation in performance. >

Carrier frequency offset robustness for OFDM systems with different pulse shaping filters

A carrier frequency offset (CFO) in an orthogonal frequency division multiplexed (OFDM) system will create interchannel, or interbin interference (IBI). IBI gives a system performance degradation with decreased SNR and increased symbol error rate (SER). The amount of CFO which creates a given performance degradation depends on the number of channels, the SNR and the pulse shaping. In this contribution we will compare the common OFDM scheme, which uses M-QAM and square pulses, with an OFDM scheme which uses M-Offset QAM (M-OQAM) and pulses of finite duration with minimized excess band-width power. Both systems use an additive white Gaussian noise (AWGN) channel. Calculations and simulations show that the system with pulse shaping allows significantly higher CFO than the system utilising QAM and square pulses.

Fading and carrier frequency offset robustness for different pulse shaping filters in OFDM

A carrier frequency offset (CFO) or a time varying (fading/Doppler spread) channel in an orthogonal frequency division multiplexed (OFDM) system will create interbin interference (IBI). IBI causes decreased SNR and increased symbol error rate (SER). The amount of CFO or Doppler spread which creates a given performance degradation depends on the number of channels, the SNR and the pulse shaping. The common OFDM scheme, with M-QAM and square pulses, are compared with an OFDM scheme with M-offset QAM (M-OQAM) and pulses of finite duration. The comparison is made both for stationary and flat fading channels. For flat fading channels, analytical methods have been developed for calculation of the SER caused both by IBI due to CFO and by IBI due to Doppler spread. The models show a good agreement with simulations. Comparison of the two different OFDM systems show that for stationary channels the system with pulse shaping allows a higher CFO than the system using square pulses. For a flat fading channel the two systems have similar Doppler spread and CFO robustness.

Optimal Pulses Robust to Carrier Frequency Offset for OFDM/QAM Systems

Optimal pulses with minimum ICI power at a given CFO point for OFDM/QAM systems are found analytically. A lower bound of the ICI power is introduced as a measure of the robustness to CFO for a given pulse. Numerical comparison verifies that the optimal pulse is more robust to CFO than previously suggested pulses at the selected CFO point.

New Methods for Blind Fine Estimation of Carrier Frequency Offset in OFDM/OQAM Systems

Carrier frequency offset (CFO) in OFDM/OQAM systems can be estimated based on the conjugate correlation function of the received signals. Traditionally this is done before demodulation. We present new methods based on the subchannel signals. The proposed estimators are robust to multipath effects due to the narrowband property of the subchannels. The performance of the estimators is evaluated by asymptotic analysis and simulation results. Our results show that methods based on estimation in subchannels have better performance than estimators operating before demodulation

OFDM for broadcasting in presence of analogue co-channel interference

This paper treats the problems of digital TV terrestrial broadcasting in presence of co-channel interference from analogue TV services. The channel capacity is calculated for the optimum distribution of the transmitted power. An OFDM scheme is presented which is designed to be robust to the analogue TV interference. Power and number of bits per symbol are assigned to each OFDM channel to give the maximum data rate for a given signal to interference ratio. Simulations of the proposed scheme, for 8 MHz bandwidth and interference from a system G PAL signal, show that acceptable error rate can be achieved for e.g. 27 Mbit/s and 39 Mbit/s at signal to interference ratios of -6 dB and 0 dB, respectively. The proposed scheme is suitable for the next generation terrestrial TV networks which are going to coexist with the current analogue systems. >

Design of robust pulses to carrier frequency offset for OFDM/OQAM system

The high sensitivity to carrier frequency offset of OFDM systems compared to single carrier systems is well known. In this paper we study how this can be remedied for OFDM/OQAM systems by appropriate pulse shaping. Building on earlier methods for minimizing out-of-band energy, a procedure for pulse shape design is derived. The new pulses are optimal with respect to interference in OFDM/OQAM systems. It is also demonstrated that, in this respect, OFDM/OQAM outperforms OFDM/QAM systems using rectangular pulses.

An analytical model for the probability distributions of crosstalk power sum for a subset of pairs in a twisted pair cable

This work presents a new and more accurate crosstalk model that is used for calculating the probability distributions of crosstalk power sum for M randomly selected pairs in an N-pair twisted pair cable. The calculation is based on averaged measurements of crosstalk in each individual pair combination. Probability distributions are calculated both for near end crosstalk (NEXT) and far end crosstalk (FEXT) in one specific type of commonly used cable that contains one 10-pair binder group. The results for the 10-pair cable show the same type of variation for the 99% point of crosstalk power sum as in the standard model used for DSL systems, which is based on 50-pair binder groups. However, the parameters are different for the two cases. For the 10-pair cable, there is a smaller variation in crosstalk as a function of the active number of pairs. The results for FEXT in a 10-pair group are in close correspondence with the standard model, while the level of NEXT is 1 2.5 dB weaker in the 10-pair group than in the standard model.

MMSE equalization of OFDM/OQAM systems for channels with time and frequency dispersion

An MMSE per channel fractionally spaced equalizer is developed for an OFDM/OQAM system with a time varying channel, under the assumption of perfect channel knowledge at the receiver. The equalizer is analyzed for an example with a doubly dispersive channel with Jakes Doppler spectrum and an exponential delay profile, and where the transmitter and receiver filters are square root of cosine rolloff. The performance is measured in terms of ergodic channel capacity, which will be an upper bound for the throughput of the system, assuming perfect channel estimation and ideal channel coding. The results show that the channel capacity may be significantly increased by using an equalizer of finite length, instead of a one-tap equalizer. The channel capacity is for instance increased by 60 % for Es/N0 = 20 dB in a channel with spreading factor 0.01.