arXiv subjects

Itsik Bergel

Publications and source records attributed to Itsik Bergel.

At least 19 recordsLinked to original sources

arXiv2026

Multi-layer MIMO Relay as Deep Physical Neural Networks: Power Amplifiers as Activation Functions

Wireless physical neural networks (WPNNs) embed neural computation directly into analog hardware, offering lower energy consumption and latency than conventional digital implementations. In this paper, we propose a deep WPNN in which nonlinear activations are realized by a multi-hop multiple-input multiple-output (MIMO) relay network, in which each relay implements a trainable complex linear gain and bias, followed by the power amplifier's intrinsic nonlinearity acting as an activation function. The cascade of multiple relays therefore realizes an over-the-air fully connected network whose parameters can be trained end-to-end. We develop two transceiver designs for different channel state information (CSI) availability scenarios: a least squares (LS)-based scheme requiring only receiver-side CSI, and a singular-value-decomposition (SVD)-based scheme requiring both transmitter-side and receiver-side CSI. Simulation results show that the proposed architecture enables accurate over-the-air inference for image classification. In particular, the results highlight the advantage of exploiting hardware nonlinearity for enhanced inference capability.

cs.LG↗

arXiv2026

Wireless Physical Neural Networks (WPNNs): Opportunities and Challenges

Wireless communication systems exhibit structural and functional similarities to neural networks: signals propagate through cascaded elements, interact with the environment, and undergo transformations. Building upon this perspective, we introduce a unified paradigm, termed \textit{wireless physical neural networks (WPNNs)}, in which components of a wireless network, such as transceivers, relays, backscatter, and intelligent surfaces, are interpreted as computational layers within a learning architecture. By treating the wireless propagation environment and network elements as differentiable operators, new opportunities arise for joint communication-computation designs, where system optimization can be achieved through learning-based methods applied directly to the physical network. This approach may operate independently of, or in conjunction with, conventional digital neural layers, enabling hybrid communication learning pipelines. In the article, we outline representative architectures that embody this viewpoint and discuss the algorithmic and training considerations required to leverage the wireless medium as a computational resource. Through numerical examples, we highlight the potential performance gains in processing, adaptability, efficiency, and end-to-end optimization, demonstrating the promise of reconfiguring wireless systems as learning networks in next-generation communication frameworks.

eess.SP↗

arXiv2023

Asymptotic Analysis of the Downlink in Cooperative Massive MIMO Systems

We consider the downlink of a cooperative cellular communications system, where several base-stations around each mobile cooperate and perform zero-forcing to reduce the received interference at the mobile. We derive closed-form expressions for the asymptotic performance of the network as the number of antennas per base station grows large. These expressions capture the trade off between various system parameters, and characterize the joint effect of noise and interference (where either noise or interference is asymptotically dominant and where both are asymptotically relevant). The asymptotic results are verified using Monte Carlo simulations, which indicate that they are useful even when the number of antennas per base station is only moderately large. Additionally, we show that when the number of antennas per base station grows large, power allocation can be optimized locally at each base station. We hence present a power allocation algorithm that achieves near optimal performance while significantly reducing the coordination overhead between base stations. The presented analysis is significantly more challenging than the uplink analysis, due to the dependence between beamforming vectors of nearby base stations. This statistical dependence is handled by introducing novel bounds on marked shot-noise point processes with dependent marks, which are also useful in other contexts.

cs.IT↗

arXiv2023

Non-Linear Relay Optimization using Deep-Learning tools

Widespread deployment of relays can yield a significant boost in the throughput of forthcoming wireless networks. However, the optimal operation of large relay networks is still infeasible. This paper presents two approaches for the optimization of large relay networks. In the traditional approach, we formulate and solve an optimization problem where the relays are considered linear. In the second approach, we take an entirely new direction and consider the true non-linear nature of the relays. Using the similarity to neural networks, we leverage deep-learning methodology. Unlike previous applications of neural networks in wireless communications, where neural networks are added to the network to perform computational tasks, our deep relay optimization treats the relay network itself as a neural network. By exploiting the non-linear transfer function exhibited by each relay, we achieve over 15dB gain compared to traditional optimization methods. Moreover, we are able to implement part of the network functionality over the relay network. Our findings shed light on the potential of deep relay optimization, promising significant advancements in future wireless communication systems.

eess.SP↗

arXiv2023

Network Optimization -- Using Relays as Neurons

We consider the optimization of a network with amplify-and-forward relays. Observing that each relay has a power limit, and hence a non-linear transfer function, we focus on the similarity between relay networks and neural networks. This similarity allows us to treat relays as neurons, and use deep learning tools to achieve better optimization of the network. Deep learning optimization allows relays to work in their non-linear regime (and hence increase their transmission power) while still avoiding harmful distortion. Moreover, like neural networks, which can implement almost any functionality, we can take advantage of the non-linearities and implement parts of the received functionalities over the relay network. By treating each relay element as a node in a deep neural network, our optimization results in huge gains over traditional relay optimization, and also allows the use of simpler receivers.

eess.SP↗

arXiv2023

Pseudo channel reciprocity in FDD satellite channels

Channel reciprocity can significantly reduce the overhead of obtaining channel-state information at the transmitter (CSIT). However, true reciprocity only exists in time division duplex (TDD). In this paper, we propose a novel tracking method that exploits implicit reciprocity in FDD line-of-sight (LOS) channels as in low-earth-orbit (LEO) satellite communication (SatCom). This channel reciprocity, dubbed pseudo-reciprocity, is crucial for applying multiple-user multiple-input multiple-output (MU-MIMO) to SatCom, which requires CSIT. We consider an LEO SatCom system where multiple satellites communicate with a multi-antenna land terminal (LT). In this innovative method, the LT can track the downlink channel changes and use them to estimate the uplink channels. The proposed method achieves precoding performance that is comparable to precoding with full CSIT knowledge. Furthermore, the use of pseudo reciprocity typically requires only initial CSIT feedback when the satellite rises. Over very long periods of time, pseudo reciprocity can fail as a result of phase ambiguity, which is referred to as cycle slip. We thus also present a closed-form approximation for the expected time until cycle slip, which indicates that in normal operating conditions, these cycle slips are extremely rare. Our numerical results provide strong support for the derived theory.

eess.SP↗

arXiv2021

Asymptotic Performance of TDOA Estimation using Satellites

We present novel lower bounds on the localization error using a network of satellites randomly deployed on a sphere around Earth. Our new analysis approach characterizes the localization performance by its asymptotic behavior as the number of satellites gets large while assuming a dense network. Using the law of large numbers, we derive closed-form expressions for the asymptotic Cramer Rao bound (CRB) from which we draw valuable insights. The resulting expressions depend solely on the network statistics and are not a function of a particular network configuration. We consider two types of estimators. The first uses the exact statistical model, and hence employs both timing and amplitude information. The second estimator ignores the amplitudes and hence uses only time difference of arrival (TDOA) information. The asymptotic CRB indicates that for practical system setup, a TDOA estimator approaches the performance of the ideal estimator. For both estimators, the localization accuracy improves as satellites get closer to Earth. The latter finding is essential in light of the proliferation of low-Earth-orbit (LEO) satellites and motivates a further study of localization-performance in such networks. Besides, we show that the vertical localization accuracy is lower than the horizontal accuracy and is also more sensitive to the receiver field-of-view.

eess.SP↗

arXiv2020

Variable pool testing for infection spread estimation

We present a method for efficient estimation of the prevalence of infection in a population with high accuracy using only a small number of tests. The presented approach uses pool testing with a mix of pool sizes of various sizes. The test results are then combined to generate an accurate estimation over a wide range of infection probabilities. This method does not require an initial guess on the infection probability. We show that, using the suggested method, even a set of only $50$ tests with a total of only $1000$ samples can produce reasonable estimation over a wide range of probabilities. A measurement set with only $100$ tests is shown to achieve $25\%$ accuracy over infection probabilities from $0.001$ to $0.5$. The presented method is applicable to COVID-19 testing.

stat.AP↗

arXiv2018

Cache Aided Communications with Multiple Antennas at Finite SNR

We study the problem of cache-aided communication for cellular networks with multi-user and multiple antennas at finite signal-to-noise ratio. Users are assumed to have non-symmetric links, modeled by wideband fading channels. We show that the problem can be formulated as a linear program, whose solution provides a joint cache allocation along with pre-fetching and fetching schemes that minimize the duration of the communication in the delivery phase. The suggested scheme uses zero-forcing and cached interference subtraction and hence allow each user to be served at the rate of its own channel. Thus, this scheme is better than the previously published schemes that are compromised by the poorest user in the communication group. We also consider a special case of the parameters for which we can derive a closed form solution and formulate the optimal power, rate and cache optimization. This special case shows that the gain of MIMO coded caching goes beyond the throughput. In particular, it is shown that in this case, the cache is used to balance the users such that fairness and throughput are no longer contradicting. More specifically, in this case, strict fairness is achieved jointly with maximizing the network throughput.

cs.IT↗

arXiv2018

Large Matrix Asymptotic Analysis of ZF and MMSE Crosstalk Cancelers for Wireline Channels

We present asymptotic expressions for user throughput in a multi-user wireline system with a linear decoder, in increasingly large system sizes. This analysis can be seen as a generalization of results obtained for wireless communication. The features of the diagonal elements of the wireline channel matrices make wireless asymptotic analyses inapplicable for wireline systems. Further, direct application of results from random matrix theory (RMT) yields a trivial lower bound. This paper presents a novel approach to asymptotic analysis, where an alternative sequence of systems is constructed that includes the system of interest in order to approximate the spectral efficiency of the linear zero-forcing (ZF) and minimum mean squared error (MMSE) crosstalk cancelers. Using works in the field of large dimensional random matrices, we show that the user rate in this sequence converges to a non-zero rate. The approximation of the user rate for both the ZF and MMSE cancelers are very simple to evaluate and does not need to take specific channel realizations into account. The analysis reveals the intricate behavior of the throughput as a function of the transmission power and the channel crosstalk. This unique behavior has not been observed for linear decoders in other systems. The approximation presented here is much more useful for the next generation G.fast wireline system than earlier digital subscriber line (DSL) systems as previously computed performance bounds, which are strictly larger than zero only at low frequencies. We also provide a numerical performance analysis over measured and simulated DSL channels which show that the approximation is accurate even for relatively low dimensional systems and is useful for many scenarios in practical DSL systems.

eess.SP↗

arXiv2018

Optimal and Suboptimal Routing Based on Partial CSI in Random Ad-hoc Networks

In this paper we consider routing in random wireless-adhoc-networks (WANETs), where each node is equipped with a single antenna. Our analysis uses a proper model of the physical layer together with an abstraction of higher communication layers. We assume that the nodes are distributed according to a Poisson-point-process and consider routing schemes that select the next relay based on the geographical locations, the channel gains of its neighbor nodes and the statistical characterization of all other nodes. While many routing problems are formulated as optimization problems, the optimal distributed solution is rarely accessible. In this work, we present the exact optimal solution for the scenario analyzed. The optimal routing is given as a maximization of a routing metric which depends solely on the known partial channel state information (CSI) and includes an expectation with respect to the interference statistics. The optimal routing scheme is important because it gives an upper bound on the performance of any other routing scheme. We also present sub-optimal routing schemes that only use part of the available knowledge and require much lower computational complexity. Numerical results demonstrate that the performance of the low complexity schemes is close to optimal and outperforms other tested routing schemes.

cs.IT↗

arXiv2018

Uplink performance of multi-antenna cellular networks with co-operative base stations and user-centric clustering

We consider a user-centric co-operative cellular network, where base stations (BSs) close to a mobile co-operate to detect its signal using a (joint) linear minimum-mean-square-error receiver. The BSs are at arbitrary positions and mobiles are modeled as a planar Poisson Point Process (PPP). Combining stochastic geometry and infinite-random-matrix theory, we derive a simple expression for the spectral efficiency of this complex system as the number of antennas grows large. This framework is applied to BS locations from PPPs and hexagonal grids, and are validated through Monte Carlo simulations. The results reveal the influence of tangible system parameters such as mobile and base-station densities, number of antennas per BS, and number of co-operating BSs on achievable spectral efficiencies. Among other insights, we find that for a given BS density and a constraint on the total number of co-operating antennas, all co-operating antennas should be located at a single BS. On the other hand, in our asymptotic regime, for the same number of co-operating antennas, if the network is limited by the area density of antennas, then the number of co-operating BSs should be increased with fewer antennas per BS.

cs.IT↗

arXiv2017

Signal Processing for Gigabit-Rate Wireline Communications

Signal processing played an important role in improving the quality of communications over copper cables in earlier DSL technologies. Even more powerful signal processing techniques are required to enable a gigabit per second data rate in the upcoming G.fast standard. This new standard is different from its predecessors in many respects. In particular, G.fast will use a significantly higher bandwidth. At such a high bandwidth, crosstalk between different lines in a binder will reach unprecedented levels, which are beyond the capabilities of most efficient techniques for interference mitigation. In this article, we survey the state of the art and research challenges in the design of signal processing algorithms for the G.fast system, with a focus on novel research approaches and design considerations for efficient interference mitigation in G.fast systems. We also detail relevant VDSL techniques and points out their strengths and limitations for the G.fast system.

cs.IT↗

arXiv2017

Deep-LMS for gigabit transmission over unshielded twisted pair cables

In this paper we propose a rapidly converging LMS algorithm for crosstalk cancellation. The architecture is similar to deep neural networks, where multiple layers are adapted sequentially. The application motivating this approach is gigabit rate transmission over unshielded twisted pairs using a vectored system. The crosstalk cancellation algorithm uses an adaptive non-diagonal preprocessing matrix prior to a conventional LMS crosstalk canceler. The update of the preprocessing matrix is inspired by deep neural networks. However, since most the operations in the Deep-LMS algorithm are linear, we are capable of providing an exact convergence speed analysis. The role of the preprocessing matrix is to speed up the convergence of the conventional LMS crosstalk canceler and hence the convergence of the overall system. The Deep-LMS is important for crosstalk cancellation in the novel G.fast standard, where traditional LMS converges very slowly due to the ill-conditioned covariance matrix of the received signal at the extended bandwidth. Simulation results support our analysis and show significant reduction in convergence time compared to existing LMS variants.

cs.IT↗

arXiv2017

Lower Bound on the Localization Error in Infinite Networks with Random Sensor Locations

We present novel lower bounds on the mean square error (MSE) of the location estimation of an emitting source via a network where the sensors are deployed randomly. The sensor locations are modeled as a homogenous Poisson point process. In contrast to previous bounds which are a function of the specific locations of all the sensors, we present CRB-type bounds on the expected mean square error; that is, we first derive the CRB on the MSE as a function of the sensors' location, and then take expectation with respect to the distribution of the sensors' location. Thus, these bounds are not a function of a particular sensor configuration, but rather of the sensor statistics. Hence, these novel bounds can be evaluated prior to sensor deployment and provide insights into design issues such as the necessary sensor density, the effect of the channel model, the effect of the signal power, and others. The derived bounds are simple to evaluate and provide a good prediction of the actual network performance.

cs.IT↗

arXiv2015

A lower bound on the data rate of dirty paper coding in general noise and interference

Dirty paper coding (DPC) allows a transmitter to send information to a receiver in the presence of interference that is known (non-causally) to the transmitter. The original version of DPC was derived for the case where the noise and the interference are statistically independent Gaussian random sequences. More recent works extended this approach to the case where the noise and the interference are mutually independent and at least one of them is Gaussian. In this letter we further extend the DPC scheme by relaxing the Gaussian and statistical independence assumptions. We provide lower bounds on the achievable data rates in a DPC setting for the case of possibly dependent noise, interference and input signals. Also, the interference and noise terms are allowed to have arbitrary probability distributions. The bounds are relatively simple, are phrased in terms of second-order statistics, and are tight when the actual noise distribution is close to Gaussian.

cs.IT↗

arXiv2015

Uplink Downlink Rate Balancing and throughput scaling in FDD Massive MIMO Systems

In this work we extend the concept of uplink-downlink rate balancing to frequency division duplex (FDD) massive MIMO systems. We consider a base station with large number antennas serving many single antenna users. We first show that any unused capacity in the uplink can be traded off for higher throughput in the downlink in a system that uses either dirty paper (DP) coding or linear zero-forcing (ZF) precoding. We then also study the scaling of the system throughput with the number of antennas in cases of linear Beamforming (BF) Precoding, ZF Precoding, and DP coding. We show that the downlink throughput is proportional to the logarithm of the number of antennas. While, this logarithmic scaling is lower than the linear scaling of the rate in the uplink, it can still bring significant throughput gains. For example, we demonstrate through analysis and simulation that increasing the number of antennas from 4 to 128 will increase the throughput by more than a factor of 5. We also show that a logarithmic scaling of downlink throughput as a function of the number of receive antennas can be achieved even when the number of transmit antennas only increases logarithmically with the number of receive antennas.

cs.IT↗

arXiv2015

Uplink Downlink Rate Balancing in Cooperating Cellular Networks

Broadcast MIMO techniques can significantly increase the throughput in the downlink of cellular networks, at the price of channel state information (CSI) feedback from the mobiles, sent over the uplink. Thus, it creates a mechanism that can tradeoff some uplink capacity for increased downlink capacity. In this work we quantify this tradeoff and study the exchange ratio between the feedback rate (over the uplink) and the downlink rate. We study both finite and infinite networks, and show that for high enough (but finite) SNR, the uplink rate can be exchanged for increased downlink rate with a favorable exchange ratio. This exchange ratio is an increasing function of the channel coherence time, and a decreasing function of the number of measured base stations. We also show that devoting a constant fraction of the uplink to CSI feedback can increase the downlink multiplexing gain continuously from 0 to 1, in finite networks. On the other hand, in infinite networks (with infinite connectivity) our bounds can only show doubly logarithmic scaling of the rate with SNR. The presented results prove that the adaptation of the feedback rate can control the balance between the uplink and downlink rates. This capability is very important in modern cellular networks, where the operators need to respond to continuously changing user demands.

cs.IT↗