Communication networks involving the overlay of spread-spectrum systems an narrower band services are of increasing interest as a means of producing greater efficiencies and flexibility in the use of the radio spectrum. Although spread-spectrum systems enjoy a natural immunity to interference from narrowband sources, their performance in the presence of such interference can be significantly enhanced by active suppression techniques. The study of this problem has elicited a very rich body of methodology, which has progressed over nearly 25 years from some of the simplest signal processing paradigms to some of the most advanced. This paper provides an overview of a number of these techniques, most of which have been developed over the past decade. In particular, a progression of techniques is described, in which successively more information about the spread-spectrum signal and interference is used to make improvements on the interference suppression capability via more advanced signal processing methods. These include linear predictive methods that make use of the spectral properties of the spread-spectrum and narrowband signals, nonlinear predictive methods that make use of the spectra and first-order probability distribution of these signals, linear code-aided methods that make use of the spreading codes of the signals of interest and the second-order statistics of the narrowband interference, and finally, a maximum-likelihood code-aided technique that makes use of essentially all that is known about the useful signals and interference. Performance comparisons show that moving up this progression of improved modeling is rewarded with performance gains that can be quite significant.

Active interference suppression in CDMA overlay systems

The push toward more efficient and flexible use of the radio spectrum has led to the consideration of the overlay of spread-spectrum (SS) communication networks on preexisting narrow-band networks. Such systems are already in widespread and rapidly expanding use in the lightly regulated spectrum where personal networking is largely centered and they are of increasing interest in the more tightly regulated spectrum due to a dearth of spectrum for new services, the desire to incorporate multi-rate (e.g., multimedia) traffic in the same network and the survival of legacy systems. Even though SS signals are inherently robust to the effects of narrower bandwidth cochannel signals, it has been shown that the use of additional processing aimed at interference suppression can result in substantial performance improvement. Motivated by this consideration, the past quarter century has seen the development of a very large body of techniques for improving the performance of SS communications systems in the presence of narrow-band interference (NBI). Early techniques (up to late 1980s) have been reviewed in the survey by Milstein (1988). Since that time, more sophisticated strategies have been developed, making use of advances from the fields of beamforming, multiuser detection (MUD), and adaptive filtering. Also, the focus of interest has shifted from techniques aimed primarily at the suppression of NBI from single-user SS systems to systems in which the SS signaling is being used to implement a code-division multiple-access (CDMA) protocol. This paper provides a tutorial overview of the progress made in this area over the past 15 years. The focus of the paper is on direct-sequence CDMA (DS/CDMA) systems and on the so-called "code-aided" techniques for NBI suppression, a term coined to indicate those strategies in which knowledge of the spreading code of a SS signal of interest is explicitly exploited in suppressing NBI. Particular attention is devoted to the case in which the CDMA signals are subject to frequency-selective fading and to the issue of blind adaptive MUD in the presence of external NBI. In particular with regard to the former issue, the effects and implications of channel-state information on system design and performance are discussed. With regard to the latter issue, it is observed that the external NBI may introduce the need for a periodically time-varying detection rule, which has significant implications in the design of blind adaptive MUD algorithms for overlaid DS/CDMA systems. The performance of the techniques discussed is compared through analysis and simulation, as well as through considerations of their relative computational complexity and required prior information. Finally, the paper is concluded by a discussion of several challenging open problems in this area.

Code-aided interference suppression for DS/CDMA overlay systems

This paper proposes a new structure of decision feedback equalizer that exploits the cyclostationary properties of digitally modulated signals to mitigate interference. In the proposed structure, the forward filter of the conventional DFE is replaced by a cyclic filter. It is assumed that the desired and the interference signals use some mutually different signaling attributes, for example symbol rates, centre frequency etc. The resulting structure is evaluated in the presence of up to six strong interfering signals, a scenario that is typically found in wireless cellular systems. The proposed structure provides performance gains for some modulation formats but it reduces to the conventional DFE for other signal formats.

FRESH-DFE: A New Structure for Interference Cancellation

The performance of conventional receivers for wireless communications may severely deteriorate in the presence of unaccounted interference. The effectiveness of methods for mitigating these effects greatly depends on the knowledge that is available about the interference and signal-of-interest (SOI), therefore making the design of robust receivers a great challenge. This thesis focuses on receiver structures for channel coded systems that exploit different levels of knowledge about the SOI and interference in an iterative fashion. This achieves both robustness and overall performance improvement compared to non-iterative receivers. Code division multiple access (CDMA) and spatial division multiple access (SDMA) systems are considered. The overlay of a turbo coded direct-sequence spread-spectrum (DS-SS) system and strong digitally modulated tone interference is studied. An iterative receiver, which is capable of blind cancellation of both wideband and narrowband interference is proposed based on the adaptive selfreconfigurable -filter scheme. Asymptotic performance analysis of the iterative receiver shows that significant iteration gains are possible if the signal-to-interference-plus-noise-ratio (SINR) is relatively large and the processing gain (PG) of the SOI is relatively small. Robust diversity detection in turbo-coded DS-SS system with statistically modeled interference is studied. A non-parametric type-based iterative receiver that estimates the probability density function (PDF) of interference-plus-noise is proposed. Its performance is shown to be rather robust to the number of interferers and their distances from the victim receiver and very similar to the performance of a clairvoyant receiver. Amazingly, this is achievable with no prior knowledge about the interference parameters. Furthermore, iteration gain is shown to significantly reduce the length of the pilot sequence needed for the PDF estimation. A family of iterative minimum-mean-squared-error (MMSE) and maximum-likelihood (ML) receivers for convolutionally and space-time coded SDMA systems is proposed. Joint iterative multiuser-detection (MUD), equalization and interference suppression are proposed to jointly combat co-channel interference (CCI), inter-symbol-interference (ISI) and unknown CCI (UCCI) in broadband single-carrier systems. It is shown that both in convolutional and space-time coded systems the ISI and CCI interference can be completely eliminated if UCCI is absent. This is achievable with a number of receive antennas equal to the number of users of interest and not to the total number of transmit antennas. In case UCCI is present, the effectiveness of CCI and ISI cancellation and UCCI suppression depends on the effective degrees of freedom of the receiver. Receiver robustness can be significantly preserved by using hybrid MMSE/ML detection for the signals of interest, or by using estimation of the PDF of the UCCI-plus-noise. A low complexity hybrid MMSE/ML iterative receiver for SDMA is proposed. It is shown that its performance is not significantly degraded compared to the optimal ML receiver. Its sensitivity to spatial correlation and a timing offset is assessed by using field measurement data. It was shown that the hybrid MMSE/ML receiver is robust against spatial correlation. The sensitivity to the timing offset is significantly reduced if the receiver performs UCCI suppression.

/pdf/iterative-receivers-for-interference-cancellation-and-2e2lt85hdv.pdf

Iterative receivers for interference cancellation and suppression in wireless communications

Hidden Markov model algorithms for narrowband interference suppression in CDMA spread spectrum systems

This paper presents a novel Kalman filtering approach to the suppression of narrowband interference from direct sequence spread spectrum communications systems. The approach is based on the digital phase-locked loop Kalman filter. Because the interference is assumed to be much stronger than either the signal or noise, the Kalman filter locks onto a function related to the interference. The net result is an estimate of the phase of the interference and its envelope. Though more optimum for interferer's of constant envelope, the algorithm is characterized as a function of the Kalman filter parameters through computer simulation for narrowband Gaussian noise interference. Examples of the phase- and envelope-tracking capabilities of the algorithm are presented, followed by bit error rate curves for interference bandwidths ranging from 0.5% to 5.0% of the chip rate of the spread spectrum signal. >

B.W. Kozminchuk

Papers

A Kalman filter-based architecture for interference excision