Dario Fertonani

Time-frequency packing for linear modulations: spectral efficiency and practical detection schemes

We investigate the spectral efficiency, achievable by a low-complexity symbol-by-symbol receiver, when linear modulations based on the superposition of uniformly time- and frequency-shifted replicas of a base pulse are employed. Although orthogonal signaling with Gaussian inputs achieves capacity on the additive white Gaussian noise channel, we show that, when finite-order constellations are employed, by giving up the orthogonality condition (thus accepting interference among adjacent signals) we can considerably improve the performance, even when a symbol-by-symbol receiver is used. We also optimize the spacing between adjacent signals to maximize the achievable spectral efficiency. Moreover, we propose a more involved transmission scheme, consisting of the superposition of two independent signals with suitable power allocation and a two-stage receiver, showing that it allows a further increase of the spectral efficiency. Finally, we show that a more involved equalization algorithm, based on soft interference cancellation, allows to achieve an excellent bit-error-rate performance, even when error-correcting codes designed for the Gaussian-noise limited channel are employed, and thus does not require a complete redesign of the coding scheme.

Spectrally-efficient continuous phase modulations

We investigate the spectral efficiency of continuous phase modulations (CPMs). To this end, we need an effective bandwidth definition for a CPM signal, whose power spectral density has in principle an infinite support. The definition we adopt is based on the spacing between adjacent carriers in a frequency division multiplexed CPM system. We consider the inter-channel interference, which depends on the channel spacing, and we evaluate the spectral efficiency achievable by a single-user receiver in the considered multi-channel scenario. We then optimize the channel spacing with the aim of maximizing the spectral efficiency, showing that impressive improvements with respect to the spectral efficiencies reported in the literature and obtained by heuristic approaches can be achieved.

SISO Detection Over Linear Channels With Linear Complexity in the Number of Interferers

We consider detection over linear channels impaired by additive white Gaussian noise. For this general model, which describes a large variety of scenarios, novel detection algorithms are derived by applying the sum-product algorithm to a suitably designed factor graph. Being soft-input soft-output (SISO) in nature, the proposed detectors can be adopted in turbo processing without additional modifications. Among various applications, we focus on channels with known intersymbol interference, on frequency-division-multiplexed systems where adjacent signals are allowed to overlap in frequency to increase the spectral efficiency, and on code division multiple access systems. When compared with the existing interference-cancellation algorithms, the proposed schemes result very appealing in terms of tradeoff between performance and computational complexity. Particularly, the proposed schemes can approach or even outperform the performance provided by much more complex algorithms.

On reliable communications over channels impaired by bursty impulse noise

Digital communications over channels impaired by impulse noise are addressed. We adopt a two-state Markov model that allows to describe the typical bursty nature of the impulse noise, in contrast to the memoryless models generally considered in the literature. For this channel, we evaluate the achievable information rate and propose a couple of practical communication systems based on powerful codes and iterative receivers. Moreover, we discuss the effectiveness of the considered receivers in terms of performance/latency tradeoff as well as in terms of robustness to erroneous channel estimations. The proposed schemes are shown to perform fairly close to the theoretical limits, and significantly better than the conventional schemes employing memoryless detection.

Reduced-Complexity BCJR Algorithm for Turbo Equalization

We propose novel techniques to reduce the complexity of the well-known Bahl–Cocke–Jelinek–Raviv (BCJR) algorithm when it is employed as a detection algorithm in turbo equalization schemes. In particular, by also considering an alternative formulation of the BCJR algorithm, which is more suitable than the original for deriving reduced-complexity techniques, we describe three reduced-complexity algorithms, each of them being particularly effective over one of the three different classes of channels (minimum-phase, maximum-phase, and mixed-phase channels) affected by intersymbol interference. The proposed algorithms do not explore all paths on the trellis describing the channel memory, but they work only on the most promising ones, which are chosen according to the maximum a posteriori criterion. Moreover, some optimization techniques for improving the effectiveness of the proposed solutions are described. Finally, we report the results of computer simulations showing the impressive performance of the proposed algorithms, and compare them with other solutions in the literature.

A robust metric for soft-output detection in the presence of class-A noise

Digital communications over channels impaired by impulse noise are considered. We first address the problem from an information-theoretical viewpoint, discussing the performance limits imposed by the channel model. Then, we describe and compare a couple of practical communication schemes employing powerful channel codes and iterative decoding, with focus on a very simple and robust detection scheme that does not require the estimation of the statistics of the impulse noise.

Novel Graph-Based Algorithms for Soft-Output Detection over Dispersive Channels

We address the design of low-complexity algorithms for soft-output detection over channels impaired by intersymbol interference. Unlike most works with similar aims, which assume the presence of the whitened matched filter at the receiver (Forney approach), algorithms that can directly work on the matched filter output (Ungerboeck approach) are considered. We introduce a novel (cyclic) factor graph describing the channel and, by applying the sum-product algorithm to it, we derive soft-output detection schemes that can provide impressive complexity reductions with respect to the benchmark algorithms, since their complexity is linear, instead of exponential, in the channel memory. Finally, we report simulation results proving that the performance of the proposed algorithms makes them appealing for turbo equalization in various practical scenarios.

Improving the spectral efficiency of linear modulations through time-frequency packing

We investigate the spectral efficiency, achievable by a low-complexity symbol-by-symbol receiver, when linear modulations based on the superposition of uniformly time- and frequency-shifted replicas of a base pulse are employed. Although orthogonal signaling with Gaussian inputs achieves capacity on the additive white Gaussian noise channel, we show that, when finite-order constellations are employed, by giving up the orthogonality condition (thus accepting interference among adjacent signals) we can considerably improve the performance, even when a symbol-by-symbol receiver is used. We also optimize the spacing between adjacent signals to maximize the achievable spectral efficiency. Moreover, we propose a more involved system model, consisting of the superposition of two independent signals and a receiver based on successive interference cancellation, showing that it allows a further increase of the spectral efficiency.

A Simplified Metric for Soft-Output Detection in the Presence of Impulse Noise

We consider iterative decoding over channels affected by impulse noise. Since the hypothesis of perfect knowledge on the statistics of the impulse noise is often unrealistic, we describe a novel metric which results to be both very simple from a computational viewpoint and very robust in the case of a mismatch. The effectiveness of the proposed solution is proved by means of computer simulations, showing that it outperforms the existing suboptimal alternatives and ensures a very limited performance degradation with respect to the optimal detector even in the case of inaccurate information on the statistics of the impulse noise.

Performance evaluation of Viterbi decoders by multicanonical Monte Carlo simulations

We propose a novel simulation-based method to evaluate the performance of Viterbi decoders. In particular, we address scenarios where the error probability is very low, that is, scenarios where classical Monte Carlo simulations would require impractical execution times before producing reliable results. As other recent fast-simulation approaches, the proposed method relies on the multicanonical Monte Carlo technique, but, unlike the existing general-purpose methods, it is specifically designed for Viterbi decoders, the algorithm being driven by a control variable that depends on the state metrics of the various survivors over the trellis. In simple scenarios for which analytical tools are available, the simulation results agree with them, while, in the most common scenarios, no analytical tool is available and the proposed method gives the fastest way for the estimation of low error probabilities.