Search

Cost-optimization through the minimization of hardware and processing costs with minimal loss in performance is an interesting design paradigm in evolving and emerging Multiple-Input-Multiple-Output (MIMO) radar systems. This optimization is a challenging task due to the increasing Radio Frequency (RF) hardware complexity as well as the signal design algorithm complexity in applications requiring high angular resolution. Towards addressing these, the paper proposes a low-complexity signal design framework, which incorporates a generalized time multiplex scheme for reducing the RF hardware complexity with a subsequent discrete phase modulation. The scheme further aims at achieving simultaneous transmit beamforming and maximum virtual MIMO aperture to enable better target detection and discrimination performance. Furthermore, the paper proposes a low-complexity signal design scheme for beampattern matching in the aforementioned setting. The conducted performance evaluation indicates that the listed design objectives are met.

Mean square error optimal estimation requires the full correlation structure to be available. Unfortunately, it is not always possible to maintain full knowledge about the correlations. One example is decentralized data fusion where the cross-correlations between estimates are unknown, partly due to information sharing. To avoid underestimating the covariance of an estimate in such situations, conservative estimation is one option. In this paper the conservative linear unbiased estimator is formalized including optimality criteria. Fundamental bounds of the optimal conservative linear unbiased estimator are derived. A main contribution is a general approach for computing the proposed estimator based on robust optimization. Furthermore, it is shown that several existing estimation algorithms are special cases of the optimal conservative linear unbiased estimator. An evaluation verifies the theoretical considerations and shows that the optimization based approach performs better than existing conservative estimation methods in certain cases. Funding Agencies|Center for Industrial Information Technology at Linkoping University [17.12]; Industry Excellence Center LINK-SIC - Swedish Governmental Agency for Innovation Systems; Saab AB; Project Scalable Kalman filters - Swedish Research Council

In this paper, we propose a general framework to provide a desirable trade-off between inference accuracy and privacy protection in the inference as service scenario (IAS). Instead of sending data directly to the server, the user will preprocess the data through a privacy-preserving mapping, which will increase privacy protection but reduce inference accuracy. To properly address the trade-off between privacy protection and inference accuracy, we formulate an optimization problem to find the privacy-preserving mapping. Even though the problem is non-convex in general, we characterize nice structures of the problem and develop an iterative algorithm to find the desired privacy-preserving mapping, with convergence analysis provided under certain assumptions. From numerical examples, we observe that the proposed method has better performance than gradient ascent method in the convergence speed, solution quality and algorithm stability.

In this paper, we propose a robust beamforming (BF) scheme for an integrated satellite and high altitude platform network, where a multibeam satellite system shares the millimeter wave spectrum with a high altitude platform system. Specifically, we first exploit the weighted Tchebycheff method and formulate a multi-objective optimization problem to obtain the Pareto optimal trade-off between two conflicting yet desirable objectives, namely, sum rate maximization and total transmit power minimization, while satisfying the quality-of-service constraints of both earth stations and mobile terminals and per-antenna transmit power budget. Then, by using the angular information based imperfect channel state information, we propose a low-complexity discretization method to transform the non-convex objective function and constraints to the convex ones. Furthermore, a monotonic optimization scheme combined with iterative penalty function algorithm is presented to obtain the BF weight vectors with low computational complexity and fast convergence rate. Finally, numerical results are provided to confirm the effectiveness and superiority of the proposed approach in comparison to the existing related works.

© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works This paper addresses the energy-latency trade-off in distributed application offloading, in which an energy-limited handset offloads totally or partially an application to one or several virtual machines (VMs) located in remote locations or access points (APs) close to the mobile terminal (MT). One of the APs (the serving AP) provides radio access to the MT and is connected to the VMs through non-ideal backhaul (BH) links. In this setting, we optimize the offloading strategy (including the joint optimization of radio and computational resources) to minimize the energy consumption at the MT subject to a maximum latency constraint. In addition, we propose robust designs to cope with imperfect acquisition of the channel state information (CSI) and the BH parameters. Our findings show that, as far as the energy-latency trade-off is concerned, the optimal order of activation of the VMs does not depend on their processing capabilities but the delays of the BH links. However, once a VM is selected to participate in the processing, the optimal amount of processing allocated to such VM depends on its computational capabilities as well as on the features (capacity and delay) of the BH link. Additionally, offloading decisions become more conservative as the uncertainty in CSI and BH parameters increases.

Worst-case jamming of legitimate communications over multiple-input multiple-output Gaussian channels is studied in this paper. A worst-case scenario with a ‘smart’ jammer that knows all channels and the transmitter's strategy and is only power limited is considered. It is shown that the simplification of the system model by neglecting the properties of the jamming channel leads to a loss of important insights regarding the effects of the jamming power and jamming channel on optimal jamming strategies of the jammer. Without neglecting the jamming channel, a lower-bound on the rate of legitimate communication subject to jamming is derived, and conditions for this bound to be positive are given. The lower-bound rate can be achieved regardless of the quality of the jamming channel, the power limit of the jammer, and the transmit strategy of the jammer. Moreover, general forms of an optimal jamming strategy, on the basis of which insights into the effect of jamming power and jamming channel are exposed, are given. It is shown that the general forms can lead to closed-form optimal jamming solutions when the power limit of the jammer is larger than a threshold. Subsequently, the scenario in which the effect of jamming dominates the effect of noise (the case of practical interest) is considered, and an optimal jamming strategy is derived in closed-form. Simulation examples demonstrate lower-bound rates, performance of the derived jamming strategy, and an effect of inaccurate channel information on the jamming strategy.

In this paper, we create a framework for training-based channel estimation under different channel and interference statistics. The minimum mean square error (MMSE) estimator for channel matrix estimation in Rician fading multi-antenna systems is analyzed, and especially the design of mean square error (MSE) minimizing training sequences. By considering Kronecker-structured systems with a combination of noise and interference and arbitrary training sequence length, we collect and generalize several previous results in the framework. We clarify the conditions for achieving the optimal training sequence structure and show when the spatial training power allocation can be solved explicitly. We also prove that spatial correlation improves the estimation performance and establish how it determines the optimal training sequence length. The analytic results for Kronecker-structured systems are used to derive a heuristic training sequence under general unstructured statistics. The MMSE estimator of the squared Frobenius norm of the channel matrix is also derived and shown to provide far better gain estimates than other approaches. It is shown under which conditions training sequences that minimize the non-convex MSE can be derived explicitly or with low complexity. Numerical examples are used to evaluate the performance of the two estimators for different training sequences and system statistics. We also illustrate how the optimal length of the training sequence often can be shorter than the number of transmit antennas. © 2010 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. QC 20100525

Distributed consensus algorithms have recently gained large interest in sensor networks as a way to achieve globally optimal decisions in a totally decentralized way, that is, without the need of sending all the data collected by the sensors to a fusion center. However, distributed algorithms are typically iterative and they suffer from convergence time and energy consumption. In this paper, we show that introducing appropriate asymmetric interaction mechanisms, with time-varying weights on each edge, it is possible to provide a substantial increase of convergence rate with respect to the symmetric time-invariant case. The basic idea underlying our approach comes from modeling the average consensus algorithm as an advection-diffusion process governing the homogenization of fluid mixtures. Exploiting such a conceptual link, we show how introducing interaction mechanisms among nearby nodes, mimicking suitable advection processes, yields a substantial increase of convergence rate. Moreover, we show that the homogenization enhancement induced by the advection term produces a qualitatively different scaling law of the convergence rate versus the network size with respect to the symmetric case.

Neural networks (NNs) have been extensively applied to many signal processing problems. In particular, due to their capacity to form complex decision regions, NNs have been successfully used in adaptive equalization of digital communication channels. The mean square error (MSE) criterion, which is usually adopted in neural learning, is not directly related to the minimization of the classification error, i.e., bit error rate (BER), which is of interest in channel equalization. Moreover, common gradient-based learning techniques are often characterized by slow speed of convergence and numerical ill conditioning. In this paper, we introduce a novel approach to learning in recurrent neural networks (RNNs) that exploits the principle of discriminative learning, minimizing an error functional that is a direct measure of the classification error. The proposed method extends to RNNs a technique applied with success to fast learning of feedforward NNs and is based on the descent of the error functional in the space of the linear combinations of the neurons (the neuron space); its main features are higher speed of convergence and better numerical conditioning w.r.t. gradient-based approaches, whereas numerical stability is assured by the use of robust least squares solvers. Experiments regarding the equalization of PAM signals in different transmission channels are described, which demonstrate the effectiveness of the proposed approach.