In this work, we consider direction-of-arrival (DoA) estimation in the presence of extreme noise using Deep Learning (DL). In particular, we introduce a Convolutional Neural Network (CNN) that predicts angular directions using the sample covariance matrix estimate. The network is trained from multi-channel data of the true array manifold matrix in the low signal-to-noise-ratio (SNR) regime. By adopting an on-grid approach, we model the problem as a multi-label classification task and train the CNN to predict DoAs across all SNRs. The proposed architecture demonstrates enhanced robustness in the presence of noise, and resilience to a relatively small number of snapshots. Moreover, it is able to resolve angles within the grid resolution. Experimental results demonstrate significant performance gains in the low-SNR regime compared to state-of-the-art methods and without the requirement of any parameter tuning in both cases of correlated and uncorrelated sources. Finally, we relax the assumption that the number of sources is known a priori and present a training method, where the CNN learns to infer their number and predict the DoAs with high confidence. The increased robustness of the proposed solution is highly desirable in challenging scenarios that arise in several fields, ranging from wireless array sensors to acoustic microphones or sonars.

This paper considers the problem of power minimization based robust orthogonal frequency division multiplexing (OFDM) radar waveform design, in which the radar coexists with a communication system in the same frequency band. Recognizing that the precise characteristics of target spectra are impossible to capture in practice, it is assumed that the target spectra lie in uncertainty sets bounded by known upper and lower bounds. Based on this uncertainty model, three different power minimization based robust radar waveform design criteria are proposed to minimize the worst-case radar transmitted power by optimizing the OFDM radar waveform, which are constrained by a specified mutual information (MI) requirement for target characterization and a minimum capacity threshold for communication system. These criteria differ in the way the communication signals scattered off the target are considered: (i) as useful energy, (ii) as interference or (iii) ignored altogether at the radar receiver. Numerical simulations demonstrate that the radar transmitted power can be efficiently reduced by exploiting the communication signals scattered off the target at the radar receiver. It is also shown that the robust waveforms bound the worst-case power-saving performance of radar system for any target spectra in the uncertainty sets.

We explore the performance of multiple input multiple output (MIMO) transmitters in correlated channels where increasing numbers of antenna elements are fitted in a fixed physical space. As well investigated in the literature, two main effects emerge in such a design: transmit spatial correlation and mutual antenna coupling. In contrast to the literature however, here we investigate the combined effect of reducing the distance between the antenna elements with increasing the number of elements in a fixed transmitter space. In other words, towards the implementation of large-scale MIMO transmitters in limited physical spaces, we investigate the joint effect of two contradicting phenomena: the reduction of spatial diversity due to reducing the separation between antennas and the increase in transmit diversity by increasing the number of elements. Within this context, we analytically approximate the performance of two distinct linear precoding designs. The theoretical analysis and simulations show the somewhat surprising result that for a given number of receivers the improved transmit diversity dominates the performance of practical linear precoders. Consequently, important benefits in the system sum rate can be gleaned by fitting more antenna elements in a fixed space by employing separations smaller than the wavelength of the transmit frequency.

Interference in wireless communications is traditionally treated as a cause of performance degradation. Whilst from a statistical viewpoint this is entirely justified, this article discusses scenarios where the interfering signal enhances the desired signal's useful power on an instantaneous basis and provides an unexplored source of additional signal power. The potential of harvesting this energy, which naturally exists in the communication system, is shown to be readily available in downlink systems where interference can be predicted. The main concept discussed is that instead of using knowledge of the interference to cancel, eliminate or avoid it, it is potentially more fruitful to use this knowledge to manipulate and make use of the interference towards the system's advantage. A significant source of useful signal power, which with conventional transmission techniques is left unexploited, can be used to improve the radio systems' performance and achieve reliable power-efficient communications where performance benefits are yielded without the need to increase the average transmitted power.

In this paper, the distribution of the ratio of extreme eigenvalues of a complex Wishart matrix is studied in order to calculate the exact decision threshold as a function of the desired probability of false alarm for the maximum-minimum eigenvalue (MME) detector. In contrast to the asymptotic analysis reported in the literature, we consider a finite number of cooperative receivers and a finite number of samples and derive the exact decision threshold for the probability of false alarm. The proposed exact formulation is further reduced to the case of two receiver-based cooperative spectrum sensing. In addition, an approximate closed-form formula of the exact threshold is derived in terms of a desired probability of false alarm for a special case having equal number of receive antennas and signal samples. Finally, the derived analytical exact decision thresholds are verified with Monte-Carlo simulations. We show that the probability of detection performance using the proposed exact decision thresholds achieves significant performance gains compared to the performance of the asymptotic decision threshold.

We propose a low-complexity vector precoding (VP) scheme for the downlink of multi-user multiple input single output (MU-MISO) systems with limited feedback. Conventional VP requires the use of modulo operation and knowledge of the scaling factor used at the transmitter in order to remove the perturbation quantity at the receiver. The latter may be problematic in certain limited feedback scenarios where only quantized versions of the scaling factor can be made available at the receiver. To circumvent this shortcoming, we propose a modified VP technique where the search of perturbing vectors is limited to the area in the symbol constellation which is constrictive to the information symbols, i.e., the area where the distances from the decision thresholds are increased with respect to a distance threshold. By doing this, the perturbation quantities can only enhance detection and need not be removed at the receiver. Successful detection can therefore be done without the use of the modulo operation and the scaling factor, which makes the proposed VP scheme applicable to limited feedback scenarios. Analytical and simulation results show that the performance of the proposed scheme is free of the error floor typically encountered with conventional VP in the presence of limited feedback.