-
Landscape Complexity for the Empirical Risk of Generalized Linear Models: Discrimination between Structured Data
Authors:
Theodoros G. Tsironis,
Aris L. Moustakas
Abstract:
We use the Kac-Rice formula and results from random matrix theory to obtain the average number of critical points of a family of high-dimensional empirical loss functions, where the data are correlated $d$-dimensional Gaussian vectors, whose number has a fixed ratio with their dimension. The correlations are introduced to model the existence of structure in the data, as is common in current Machin…
▽ More
We use the Kac-Rice formula and results from random matrix theory to obtain the average number of critical points of a family of high-dimensional empirical loss functions, where the data are correlated $d$-dimensional Gaussian vectors, whose number has a fixed ratio with their dimension. The correlations are introduced to model the existence of structure in the data, as is common in current Machine-Learning systems. Under a technical hypothesis, our results are exact in the large-$d$ limit, and characterize the annealed landscape complexity, namely the logarithm of the expected number of critical points at a given value of the loss.
We first address in detail the landscape of the loss function of a single perceptron and then generalize it to the case where two competing data sets with different covariance matrices are present, with the perceptron seeking to discriminate between them. The latter model can be applied to understand the interplay between adversity and non-trivial data structure. For completeness, we also treat the case of a loss function used in training Generalized Linear Models in the presence of correlated input data.
△ Less
Submitted 18 March, 2025;
originally announced March 2025.
-
Spectral Efficiency Expression for the Non-Linear Schrödinger Channel in the Low Noise Limit Using Scattering Data
Authors:
Pavlos Kazakopoulos,
Aris L. Moustakas
Abstract:
Transmission through optical fibers offers ultra-fast and long-haul communications. However, the search for its ultimate capacity limits in the presence of distributed amplifier noise is complicated by the competition between wave dispersion and non-linearity. In this paper, we exploit the integrability of the Nonlinear Schrödinger Equation, which accurately models optical fiber communications, to…
▽ More
Transmission through optical fibers offers ultra-fast and long-haul communications. However, the search for its ultimate capacity limits in the presence of distributed amplifier noise is complicated by the competition between wave dispersion and non-linearity. In this paper, we exploit the integrability of the Nonlinear Schrödinger Equation, which accurately models optical fiber communications, to derive an expression for the spectral efficiency of an optical fiber communications channel, expressed fully in the scattering data domain of the Non-linear Fourier Transform and valid in the limit of low amplifier noise. We utilize the relationship between the derived noise-covariance operator and the Jacobian of the mapping between the signal and the scattering data to obtain the properties of the former. Emerging from the structure of the covariance operator is the significance of the Gordon-Haus effect in moderating and finally reversing the increase of the spectral efficiency with power. This effect is showcased in numerical simulations for Gaussian input in the high-bandwidth regime.
△ Less
Submitted 24 February, 2025;
originally announced February 2025.
-
MIMO MAC Empowered by Reconfigurable Intelligent Surfaces: Capacity Region and Large System Analysis
Authors:
Aris L. Moustakas,
George C. Alexandropoulos
Abstract:
Smart wireless environments enabled by multiple distributed Reconfigurable Intelligent Surfaces (RISs) have recently attracted significant research interest as a wireless connectivity paradigm for sixth Generation (6G) networks. In this paper, using random matrix theory methods, we calculate the mean of the sum Mutual Information (MI) for the correlated Multiple-Input Multiple-Output (MIMO) Multip…
▽ More
Smart wireless environments enabled by multiple distributed Reconfigurable Intelligent Surfaces (RISs) have recently attracted significant research interest as a wireless connectivity paradigm for sixth Generation (6G) networks. In this paper, using random matrix theory methods, we calculate the mean of the sum Mutual Information (MI) for the correlated Multiple-Input Multiple-Output (MIMO) Multiple Access Channel (MAC) in the presence of multiple RISs, in the large-antenna number limit. We thus obtain the capacity region boundaries, after optimizing over the tunable RISs' phase configurations. Furthermore, we obtain a closed-form expression for the variance of the sum-MI metric, which together with the mean provides a tight Gaussian approximation for the outage probability. The derived results become relevant in the presence of fast-fading, when channel estimation is extremely challenging. Our numerical investigations showcased that, when the angle-spread in the neighborhood of each RIS is small, which is expected for higher carrier frequencies, the communication link strongly improves from optimizing the ergodic MI of the multiple RISs.We also found that, increasing the number of transmitting users in such MIMO-MAC-RIS systems results to rapidly diminishing sum-MI gains, hence, providing limits on the number of users that can be efficiently served by a given RIS.
△ Less
Submitted 9 October, 2024;
originally announced October 2024.
-
Reconfigurable Intelligent Surfaces and Capacity Optimization: A Large System Analysis
Authors:
Aris L. Moustakas,
George C. Alexandropoulos,
Mérouane Debbah
Abstract:
Reconfigurable Intelligent Surfaces (RISs), comprising large numbers of low-cost and almost passive metamaterials with tunable reflection properties, have been recently proposed as an enabling technology for programmable wireless propagation environments. In this paper, we present asymptotic closed-form expressions for the mean and variance of the mutual information metric for a multi-antenna tran…
▽ More
Reconfigurable Intelligent Surfaces (RISs), comprising large numbers of low-cost and almost passive metamaterials with tunable reflection properties, have been recently proposed as an enabling technology for programmable wireless propagation environments. In this paper, we present asymptotic closed-form expressions for the mean and variance of the mutual information metric for a multi-antenna transmitter-receiver pair in the presence of multiple RISs, using methods from statistical physics. While nominally valid in the large system limit, we show that the derived Gaussian approximation for the mutual information can be quite accurate, even for modest-sized antenna arrays and metasurfaces. The above results are particularly useful when fast-fading conditions are present, which renders instantaneous channel estimation extremely challenging. We find that, when the channel close to an RIS is correlated, for instance due to small angle spread, which is reasonable for wireless systems with increasing carrier frequencies, the communication link benefits significantly from statistical RIS phase optimization, resulting in gains that are surprisingly higher than the nearly uncorrelated case. Using our novel asymptotic properties of the correlation matrices of the impinging and outgoing signals at the RISs, we can optimize the metasurfaces without brute-force numerical optimization. Furthermore, when the desired reflection from any of the RISs departs significantly from geometrical optics, the metasurfaces can be optimized to provide robust communication links, without significant need for their optimal placement.
△ Less
Submitted 20 August, 2022;
originally announced August 2022.
-
On the Uplink Performance of Finite-Capacity Radio Stripes
Authors:
Ioannis Chiotis,
Aris L. Moustakas
Abstract:
Cell-Free (CF) Massive MIMO (mMIMO) is a technology which can potentially augment not only the deployment of 5G, but also the deployment of beyond 5G (B5G) wireless networks. However, the cost for rolling out such systems may be significant. Radio stripes form a promising solution which offers the potential of scalability at a reduced price. This paper investigates the uplink scenario of a CF mMIM…
▽ More
Cell-Free (CF) Massive MIMO (mMIMO) is a technology which can potentially augment not only the deployment of 5G, but also the deployment of beyond 5G (B5G) wireless networks. However, the cost for rolling out such systems may be significant. Radio stripes form a promising solution which offers the potential of scalability at a reduced price. This paper investigates the uplink scenario of a CF mMIMO system, implemented with a limited-capacity radio stripe which integrates a novel arrangement of access points (APs), fully exploiting macro-diversity benefits. We also analyze a heuristic Compare-and-Forward (CnF) strategy, which, by comparing normalized linear minimum mean square error (N-LMMSE) soft estimates, enables optimal dynamic cooperation clustering, thus leading to a user-centric radio stripe network approach. Aiming at maximizing the per-user uplink spectral efficiency (SE), we ensure that, under finite capacity constraints, our solution can guarantee better performance than existing radio stripe architectures, especially when system size increases.
△ Less
Submitted 31 July, 2022;
originally announced August 2022.
-
Capacity Optimization using Reconfigurable Intelligent Surfaces: A Large System Approach
Authors:
Aris L. Moustakas,
George C. Alexandropoulos,
Mérouane Debbah
Abstract:
Reconfigurable Intelligent Surfaces (RISs), comprising large numbers of low-cost and passive metamaterials with tunable reflection properties, have been recently proposed as an enabler for programmable radio propagation environments. However, the role of the channel conditions near the RISs on their optimizability has not been analyzed adequately. In this paper, we present an asymptotic closed-for…
▽ More
Reconfigurable Intelligent Surfaces (RISs), comprising large numbers of low-cost and passive metamaterials with tunable reflection properties, have been recently proposed as an enabler for programmable radio propagation environments. However, the role of the channel conditions near the RISs on their optimizability has not been analyzed adequately. In this paper, we present an asymptotic closed-form expression for the mutual information of a multi-antenna transmitter-receiver pair in the presence of multiple RISs, in the large-antenna limit, using the random matrix and replica theories. Under mild assumptions, asymptotic expressions for the eigenvalues and the eigenvectors of the channel covariance matrices are derived. We find that, when the channel close to an RIS is correlated, for instance due to small angle spread, the communication link benefits significantly from the RIS optimization, resulting in gains that are surprisingly higher than the nearly uncorrelated case. Furthermore, when the desired reflection from the RIS departs significantly from geometrical optics, the surface can be optimized to provide robust communication links. Building on the properties of the eigenvectors of the covariance matrices, we are able to find the optimal response of the RISs in closed form, bypassing the need for brute-force optimization.
△ Less
Submitted 16 September, 2021;
originally announced September 2021.
-
Gallager Bound for MIMO Channels: Large-N Asymptotics
Authors:
Apostolos Karadimitrakis,
Aris L. Moustakas,
Romain Couillet
Abstract:
The use of multiple antenna arrays in transmission and reception has become an integral part of modern wireless communications. To quantify the performance of such systems, the evaluation of bounds on the error probability of realistic finite length codewords is important. In this paper, we analyze the standard Gallager error bound for both constraints of maximum average power and maximum instanta…
▽ More
The use of multiple antenna arrays in transmission and reception has become an integral part of modern wireless communications. To quantify the performance of such systems, the evaluation of bounds on the error probability of realistic finite length codewords is important. In this paper, we analyze the standard Gallager error bound for both constraints of maximum average power and maximum instantaneous power. Applying techniques from random matrix theory, we obtain analytic expressions of the error exponent when the length of the codeword increases to infinity at a fixed ratio with the antenna array dimensions. Analyzing its behavior at rates close to the ergodic rate, we find that the Gallager error bound becomes asymptotically close to an upper error bound obtained recently by Hoydis et al. 2015. We also obtain an expression for the Gallager exponent in the case when the codelength spans several Rayleigh fading blocks, hence taking into account the situation when the channel varies during each transmission.
△ Less
Submitted 27 November, 2017; v1 submitted 23 November, 2017;
originally announced November 2017.
-
Boltzmann meets Nash: Energy-efficient routing in optical networks under uncertainty
Authors:
Panayotis Mertikopoulos,
Aris L. Moustakas,
Anna Tzanakaki
Abstract:
Motivated by the massive deployment of power-hungry data centers for service provisioning, we examine the problem of routing in optical networks with the aim of minimizing traffic-driven power consumption. To tackle this issue, routing must take into account energy efficiency as well as capacity considerations; moreover, in rapidly-varying network environments, this must be accomplished in a real-…
▽ More
Motivated by the massive deployment of power-hungry data centers for service provisioning, we examine the problem of routing in optical networks with the aim of minimizing traffic-driven power consumption. To tackle this issue, routing must take into account energy efficiency as well as capacity considerations; moreover, in rapidly-varying network environments, this must be accomplished in a real-time, distributed manner that remains robust in the presence of random disturbances and noise. In view of this, we derive a pricing scheme whose Nash equilibria coincide with the network's socially optimum states, and we propose a distributed learning method based on the Boltzmann distribution of statistical mechanics. Using tools from stochastic calculus, we show that the resulting Boltzmann routing scheme exhibits remarkable convergence properties under uncertainty: specifically, the long-term average of the network's power consumption converges within $\varepsilon$ of its minimum value in time which is at most $\tilde O(1/\varepsilon^2)$, irrespective of the fluctuations' magnitude; additionally, if the network admits a strict, non-mixing optimum state, the algorithm converges to it - again, no matter the noise level. Our analysis is supplemented by extensive numerical simulations which show that Boltzmann routing can lead to a significant decrease in power consumption over basic, shortest-path routing schemes in realistic network conditions.
△ Less
Submitted 4 May, 2016;
originally announced May 2016.
-
Modern Telecommunications: A Playground for Physicists?
Authors:
Aris L. Moustakas
Abstract:
This report summarizes some of the material that was presented by the author during the 2015 Les Houches Summerschool on "Random Matrices and Stochastic Processes". In these Lectures, various applications of Random Matrix Theory in modern telecommunications are reviewed. The aim is to introduce the Physics community to a number of relevant problems that can be analyzed using such tools, while at t…
▽ More
This report summarizes some of the material that was presented by the author during the 2015 Les Houches Summerschool on "Random Matrices and Stochastic Processes". In these Lectures, various applications of Random Matrix Theory in modern telecommunications are reviewed. The aim is to introduce the Physics community to a number of relevant problems that can be analyzed using such tools, while at the same time briefly describing the way these methods are applied. More specifically, two applications on wireless communications and two on optical communications are presented.
△ Less
Submitted 4 May, 2016;
originally announced May 2016.
-
Optical Fiber MIMO Channel Model and its Analysis
Authors:
Apostolos Karadimitrakis,
Aris L. Moustakas,
Hartmut Hafermann,
Axel Mueller
Abstract:
Technology is moving towards space division multiplexing in optical fiber to keep up the trend in rate increase over time and to avoid an imminent capacity crunch. Thus, it is of paramount interest to estimate the potential gains of this approach. As more spatial channels are being packed into a single fiber, the increased crosstalk necessitates the use of MIMO to guarantee reliable operation. In…
▽ More
Technology is moving towards space division multiplexing in optical fiber to keep up the trend in rate increase over time and to avoid an imminent capacity crunch. Thus, it is of paramount interest to estimate the potential gains of this approach. As more spatial channels are being packed into a single fiber, the increased crosstalk necessitates the use of MIMO to guarantee reliable operation. In this paper, we exploit the analogy between an optical fiber and a model from mesoscopic physics - a chaotic cavity - to obtain a novel channel model for the optical fiber. The model captures both random distributed crosstalk and modedependent loss, which are described within the framework of scattering theory. Using tools from replica theory and random matrix theory, we derive the capacity of the fiber optical MIMO channel model.
△ Less
Submitted 11 April, 2016;
originally announced April 2016.
-
Outage and Capacity Comparisons For Ground Relaying Systems Using Stochastic Geometry
Authors:
Spyridon Evangelatos,
Aris L. Moustakas,
Andreas Polydoros
Abstract:
Concurrent cooperative transmission for relaying purposes in mobile communication networks is relevant in current institutional systems with limited infrastructure, and and may be viewed as a potential range-extension mechanism for future commercial networks, including vehicular autonomous networking. The complexity of the overall system has encouraged certain abstractions at the physical layer wh…
▽ More
Concurrent cooperative transmission for relaying purposes in mobile communication networks is relevant in current institutional systems with limited infrastructure, and and may be viewed as a potential range-extension mechanism for future commercial networks, including vehicular autonomous networking. The complexity of the overall system has encouraged certain abstractions at the physical layer which are critically analyzed in the present paper. We show via analytic stochastic geometry tools that the receiver structure plays a crucial role in the outage behavior of the relays, particularly for realistic flooding protocols. This approach aims to help understand the cross-layer aspects of such networks.
△ Less
Submitted 20 July, 2015;
originally announced July 2015.
-
In an Uncertain World: Distributed Optimization in MIMO Systems with Imperfect Information
Authors:
Panayotis Mertikopoulos,
Aris L. Moustakas
Abstract:
In this paper, we introduce a distributed algorithm that optimizes the Gaussian signal covariance matrices of multi-antenna users transmitting to a common multi-antenna receiver under imperfect and possibly delayed channel state information. The algorithm is based on an extension of exponential learning techniques to a semidefinite setting and it requires the same information as distributed water-…
▽ More
In this paper, we introduce a distributed algorithm that optimizes the Gaussian signal covariance matrices of multi-antenna users transmitting to a common multi-antenna receiver under imperfect and possibly delayed channel state information. The algorithm is based on an extension of exponential learning techniques to a semidefinite setting and it requires the same information as distributed water-filling methods. Unlike water-filling however, the proposed matrix exponential learning (MXL) algorithm converges to the system's optimum signal covariance profile under very mild conditions on the channel uncertainty statistics; moreover, the algorithm retains its convergence properties even in the presence of user update asynchronicities, random delays and/or ergodically changing channel conditions. In particular, by properly tuning the algorithm's learning rate (or step size), the algorithm converges within a few iterations, even for large numbers of users and/or antennas per user. Our theoretical analysis is complemented by numerical simulations which illustrate the algorithm's robustness and scalability in realistic network conditions.
△ Less
Submitted 5 February, 2015;
originally announced February 2015.
-
Cost-Efficient Throughput Maximization in Multi-Carrier Cognitive Radio Systems
Authors:
Salvatore D'Oro,
Panayotis Mertikopoulos,
Aris L. Moustakas,
Sergio Palazzo
Abstract:
Cognitive radio (CR) systems allow opportunistic, secondary users (SUs) to access portions of the spectrum that are unused by the network's licensed primary users (PUs), provided that the induced interference does not compromise the primary users' performance guarantees. To account for interference constraints of this type, we consider a flexible spectrum access pricing scheme that charges seconda…
▽ More
Cognitive radio (CR) systems allow opportunistic, secondary users (SUs) to access portions of the spectrum that are unused by the network's licensed primary users (PUs), provided that the induced interference does not compromise the primary users' performance guarantees. To account for interference constraints of this type, we consider a flexible spectrum access pricing scheme that charges secondary users based on the interference that they cause to the system's primary users (individually, globally, or both), and we examine how secondary users can maximize their achievable transmission rate in this setting. We show that the resulting non-cooperative game admits a unique Nash equilibrium under very mild assumptions on the pricing mechanism employed by the network operator, and under both static and ergodic (fast-fading) channel conditions. In addition, we derive a dynamic power allocation policy that converges to equilibrium within a few iterations (even for large numbers of users), and which relies only on local signal-to-interference-and-noise measurements; importantly, the proposed algorithm retains its convergence properties even in the ergodic channel regime, despite the inherent stochasticity thereof. Our theoretical analysis is complemented by extensive numerical simulations which illustrate the performance and scalability properties of the proposed pricing scheme under realistic network conditions.
△ Less
Submitted 2 February, 2015;
originally announced February 2015.
-
Effects of Mobility on User Energy Consumption and Total Throughput in a Massive MIMO System
Authors:
Aris L. Moustakas,
Luca Sanguinetti,
Mérouane Debbah
Abstract:
Macroscopic mobility of wireless users is important to determine the performance and energy effciency of a wireless network, because of the temporal correlations it introduces in the consumed power and throughput. In this work we introduce a methodology that obtains the long time statistics of such metrics in a network. After describing the general approach, we present a specific example of the up…
▽ More
Macroscopic mobility of wireless users is important to determine the performance and energy effciency of a wireless network, because of the temporal correlations it introduces in the consumed power and throughput. In this work we introduce a methodology that obtains the long time statistics of such metrics in a network. After describing the general approach, we present a specific example of the uplink channel of a mobile user in the vicinity of a massive MIMO base-station antenna array. To guarantee a fixed SINR and rate, the user inverts the path-loss channel power, while moving around in the cell. To calculate the long time distribution of the consumed energy of the user, we assume his movement follows a Brownian motion, and then map the problem to the solution of the minimum eigenvalue of a partial differential equation, which can be solved either analytically, or numerically very fast. We also treat the throughput of a single user. We then discuss the results and how they can be generalized if the mobility is assumed to be a Levy random walk. We also provide a roadmap to use this technique when one considers multiple users and base stations.
△ Less
Submitted 11 July, 2014;
originally announced July 2014.
-
Large System Analysis of the Energy Consumption Distribution in Multi-User MIMO Systems with Mobility
Authors:
Luca Sanguinetti,
Aris L. Moustakas,
Emil Bjornson,
Merouane Debbah
Abstract:
In this work, we consider the downlink of a single-cell multi-user MIMO system in which the base station (BS) makes use of $N$ antennas to communicate with $K$ single-antenna user equipments (UEs). The UEs move around in the cell according to a random walk mobility model. We aim at determining the energy consumption distribution when different linear precoding techniques are used at the BS to guar…
▽ More
In this work, we consider the downlink of a single-cell multi-user MIMO system in which the base station (BS) makes use of $N$ antennas to communicate with $K$ single-antenna user equipments (UEs). The UEs move around in the cell according to a random walk mobility model. We aim at determining the energy consumption distribution when different linear precoding techniques are used at the BS to guarantee target rates within a finite time interval $T$. The analysis is conducted in the asymptotic regime where $N$ and $K$ grow large with fixed ratio under the assumption of perfect channel state information (CSI). Both recent and standard results from large system analysis are used to provide concise formulae for the asymptotic transmit powers and beamforming vectors for all considered schemes. These results are eventually used to provide a deterministic approximation of the energy consumption and to study its fluctuations around this value in the form of a central limit theorem. Closed-form expressions for the asymptotic means and variances are given. Numerical results are used to validate the accuracy of the theoretical analysis and to make comparisons. We show how the results can be used to approximate the probability that a battery-powered BS runs out of energy and also to design the cell radius for minimizing the energy consumption per unit area. The imperfect CSI case is also briefly considered.
△ Less
Submitted 5 January, 2015; v1 submitted 23 June, 2014;
originally announced June 2014.
-
Outage Capacity for the Optical MIMO Channel
Authors:
Apostolos Karadimitrakis,
Aris L. Moustakas,
Pierpaolo Vivo
Abstract:
MIMO processing techniques in fiber optical communications have been proposed as a promising approach to meet increasing demand for information throughput. In this context, the multiple channels correspond to the multiple modes and/or multiple cores in the fiber. In this paper we characterize the distribution of the mutual information with Gaussian input in a simple channel model for this system.…
▽ More
MIMO processing techniques in fiber optical communications have been proposed as a promising approach to meet increasing demand for information throughput. In this context, the multiple channels correspond to the multiple modes and/or multiple cores in the fiber. In this paper we characterize the distribution of the mutual information with Gaussian input in a simple channel model for this system. Assuming significant cross talk between cores, negligible backscattering and near-lossless propagation in the fiber, we model the transmission channel as a random complex unitary matrix. The loss in the transmission may be parameterized by a number of unutilized channels in the fiber. We analyze the system in a dual fashion. First, we evaluate a closed-form expression for the outage probability, which is handy for small matrices. We also apply the asymptotic approach, in particular the Coulomb gas method from statistical mechanics, to obtain closed-form results for the ergodic mutual information, its variance as well as the outage probability for Gaussian input in the limit of large number of cores/modes. By comparing our analytic results to simulations, we see that, despite the fact that this method is nominally valid for large number of modes, our method is quite accurate even for small to modest number of channels.
△ Less
Submitted 1 December, 2013; v1 submitted 4 February, 2013;
originally announced February 2013.
-
Transmission of information via the non-linear Scroedinger equation: The random Gaussian input case
Authors:
Pavlos Kazakopoulos,
Aris L. Moustakas
Abstract:
The explosion of demand for ultra-high information transmission rates over the last decade has necessitated the usage of increasingly high light intensities for fiber optical transmissions. As a result, the fiber non-linearities need to be treated non-perturbatively. Similar analyses in the past have focused on the effects of non-linearities on existing transmission technologies, e.g. WDM. In this…
▽ More
The explosion of demand for ultra-high information transmission rates over the last decade has necessitated the usage of increasingly high light intensities for fiber optical transmissions. As a result, the fiber non-linearities need to be treated non-perturbatively. Similar analyses in the past have focused on the effects of non-linearities on existing transmission technologies, e.g. WDM. In this paper we take advantage of the fact that, under certain assumptions, light transmission through optical fibers can be described using the non-linear Schroedinger equation, which is exactly integrable. As a particular example, we show that in the low Gaussian noise limit, the Gaussian input distribution has a higher mutual information than the transmission using WDM over the same available bandwidth.
△ Less
Submitted 30 October, 2012;
originally announced October 2012.
-
SINR Statistics of Correlated MIMO Linear Receivers
Authors:
Aris L. Moustakas,
Pavlos Kazakopoulos
Abstract:
Linear receivers offer a low complexity option for multi-antenna communication systems. Therefore, understanding the outage behavior of the corresponding SINR is important in a fading mobile environment. In this paper we introduce a large deviations method, valid nominally for a large number M of antennas, which provides the probability density of the SINR of Gaussian channel MIMO Minimum Mean Squ…
▽ More
Linear receivers offer a low complexity option for multi-antenna communication systems. Therefore, understanding the outage behavior of the corresponding SINR is important in a fading mobile environment. In this paper we introduce a large deviations method, valid nominally for a large number M of antennas, which provides the probability density of the SINR of Gaussian channel MIMO Minimum Mean Square Error (MMSE) and zero-forcing (ZF) receivers, with arbitrary transmission power profiles and in the presence of receiver antenna correlations. This approach extends the Gaussian approximation of the SINR, valid for large M asymptotically close to the center of the distribution, obtaining the non-Gaussian tails of the distribution. Our methodology allows us to calculate the SINR distribution to next-to-leading order (O(1/M)) and showcase the deviations from approximations that have appeared in the literature (e.g. the Gaussian or the generalized Gamma distribution). We also analytically evaluate the outage probability, as well as the uncoded bit-error-rate. We find that our approximation is quite accurate even for the smallest antenna arrays (2x2).
△ Less
Submitted 13 September, 2012;
originally announced September 2012.
-
Power Optimization in Random Wireless Networks
Authors:
Aris L. Moustakas,
Panayotis Mertikopoulos,
Nicholas Bambos
Abstract:
Consider a wireless network of transmitter-receiver pairs where the transmitters adjust their powers to maintain a target SINR level in the presence of interference. In this paper, we analyze the optimal power vector that achieves this target in large, random networks obtained by "erasing" a finite fraction of nodes from a regular lattice of transmitter-receiver pairs. We show that this problem is…
▽ More
Consider a wireless network of transmitter-receiver pairs where the transmitters adjust their powers to maintain a target SINR level in the presence of interference. In this paper, we analyze the optimal power vector that achieves this target in large, random networks obtained by "erasing" a finite fraction of nodes from a regular lattice of transmitter-receiver pairs. We show that this problem is equivalent to the so-called Anderson model of electron motion in dirty metals which has been used extensively in the analysis of diffusion in random environments. A standard approximation to this model is the so-called coherent potential approximation (CPA) method which we apply to evaluate the first and second order intra-sample statistics of the optimal power vector in one- and two-dimensional systems. This approach is equivalent to traditional techniques from random matrix theory and free probability, but while generally accurate (and in agreement with numerical simulations), it fails to fully describe the system: in particular, results obtained in this way fail to predict when power control becomes infeasible. In this regard, we find that the infinite system is always unstable beyond a certain value of the target SINR, but any finite system only has a small probability of becoming unstable. This instability probability is proportional to the tails of the eigenvalue distribution of the system which are calculated to exponential accuracy using methodologies developed within the Anderson model and its ties with random walks in random media. Finally, using these techniques, we also calculate the tails of the system's power distribution under power control and the rate of convergence of the Foschini-Miljanic power control algorithm in the presence of random erasures. Overall, in the paper we try to strike a balance between intuitive arguments and formal proofs.
△ Less
Submitted 17 December, 2013; v1 submitted 28 February, 2012;
originally announced February 2012.
-
Tails of Random Matrix Diagonal Elements: The Case of the Wishart Inverse
Authors:
Aris L. Moustakas
Abstract:
We analytically compute the large-deviation probability of a diagonal matrix element of two cases of random matrices, namely $β=[\vec H^\dagger\vec H]^{-1}_{11}$ and $γ=[\vec I_N+ρ\vec H^\dagger\vec H]^{-1}_{11}$, where $\vec H$ is a $M\times N$ complex Gaussian matrix with independent entries and $M\geq N$. These diagonal entries are related to the "signal to interference and noise ratio" (SINR)…
▽ More
We analytically compute the large-deviation probability of a diagonal matrix element of two cases of random matrices, namely $β=[\vec H^\dagger\vec H]^{-1}_{11}$ and $γ=[\vec I_N+ρ\vec H^\dagger\vec H]^{-1}_{11}$, where $\vec H$ is a $M\times N$ complex Gaussian matrix with independent entries and $M\geq N$. These diagonal entries are related to the "signal to interference and noise ratio" (SINR) in multi-antenna communications. They depend not only on the eigenvalues but also on the corresponding eigenfunction weights, which we are able to evaluate on average constrained on the value of the SINR. We also show that beyond a lower and upper critical value of $β$, $γ$, the maximum and minimum eigenvalues, respectively, detach from the bulk. Responsible for this detachment is the fact that the corresponding eigenvalue weight becomes macroscopic (i.e. O(1)), and hence exerts a strong repulsion to the eigenvalue.
△ Less
Submitted 11 April, 2011;
originally announced April 2011.
-
Distributed Learning Policies for Power Allocation in Multiple Access Channels
Authors:
Panayotis Mertikopoulos,
Elena V. Belmega,
Aris L. Moustakas,
Samson Lasaulce
Abstract:
We analyze the problem of distributed power allocation for orthogonal multiple access channels by considering a continuous non-cooperative game whose strategy space represents the users' distribution of transmission power over the network's channels. When the channels are static, we find that this game admits an exact potential function and this allows us to show that it has a unique equilibrium a…
▽ More
We analyze the problem of distributed power allocation for orthogonal multiple access channels by considering a continuous non-cooperative game whose strategy space represents the users' distribution of transmission power over the network's channels. When the channels are static, we find that this game admits an exact potential function and this allows us to show that it has a unique equilibrium almost surely. Furthermore, using the game's potential property, we derive a modified version of the replicator dynamics of evolutionary game theory which applies to this continuous game, and we show that if the network's users employ a distributed learning scheme based on these dynamics, then they converge to equilibrium exponentially quickly. On the other hand, a major challenge occurs if the channels do not remain static but fluctuate stochastically over time, following a stationary ergodic process. In that case, the associated ergodic game still admits a unique equilibrium, but the learning analysis becomes much more complicated because the replicator dynamics are no longer deterministic. Nonetheless, by employing results from the theory of stochastic approximation, we show that users still converge to the game's unique equilibrium.
Our analysis hinges on a game-theoretical result which is of independent interest: in finite player games which admit a (possibly nonlinear) convex potential function, the replicator dynamics (suitably modified to account for nonlinear payoffs) converge to an eps-neighborhood of an equilibrium at time of order O(log(1/eps)).
△ Less
Submitted 16 November, 2011; v1 submitted 17 March, 2011;
originally announced March 2011.
-
Dynamic Power Allocation Games in Parallel Multiple Access Channels
Authors:
Panayotis Mertikopoulos,
Elena V. Belmega,
Aris L. Moustakas,
Samson Lasaulce
Abstract:
We analyze the distributed power allocation problem in parallel multiple access channels (MAC) by studying an associated non-cooperative game which admits an exact potential. Even though games of this type have been the subject of considerable study in the literature, we find that the sufficient conditions which ensure uniqueness of Nash equilibrium points typically do not hold in this context. No…
▽ More
We analyze the distributed power allocation problem in parallel multiple access channels (MAC) by studying an associated non-cooperative game which admits an exact potential. Even though games of this type have been the subject of considerable study in the literature, we find that the sufficient conditions which ensure uniqueness of Nash equilibrium points typically do not hold in this context. Nonetheless, we show that the parallel MAC game admits a unique equilibrium almost surely, thus establishing an important class of counterexamples where these sufficient conditions are not necessary. Furthermore, if the network's users employ a distributed learning scheme based on the replicator dynamics, we show that they converge to equilibrium from almost any initial condition, even though users only have local information at their disposal.
△ Less
Submitted 25 November, 2010;
originally announced November 2010.
-
Balancing Traffic in Networks: Redundancy, Learning and the Effect of Stochastic Fluctuations
Authors:
Panayotis Mertikopoulos,
Aris L. Moustakas
Abstract:
We study the distribution of traffic in networks whose users try to minimise their delays by adhering to a simple learning scheme inspired by the replicator dynamics of evolutionary game theory. The stable steady states of these dynamics coincide with the network's Wardrop equilibria and form a convex polytope whose dimension is determined by the network's redundancy (an important concept which…
▽ More
We study the distribution of traffic in networks whose users try to minimise their delays by adhering to a simple learning scheme inspired by the replicator dynamics of evolutionary game theory. The stable steady states of these dynamics coincide with the network's Wardrop equilibria and form a convex polytope whose dimension is determined by the network's redundancy (an important concept which measures the "linear dependence" of the users' paths). Despite this abundance of stationary points, the long-term behaviour of the replicator dynamics turns out to be remarkably simple: every solution orbit converges to a Wardrop equilibrium.
On the other hand, a major challenge occurs when the users' delays fluctuate unpredictably due to random external factors. In that case, interior equilibria are no longer stationary, but strict equilibria remain stochastically stable irrespective of the fluctuations' magnitude. In fact, if the network has no redundancy and the users are patient enough, we show that the long-term averages of the users' traffic flows converge to the vicinity of an equilibrium, and we also estimate the corresponding invariant measure.
△ Less
Submitted 27 February, 2010; v1 submitted 21 December, 2009;
originally announced December 2009.
-
Living at the Edge: A Large Deviations Approach to the Outage MIMO Capacity
Authors:
P. Kazakopoulos,
P. Mertikopoulos,
A. L. Moustakas,
G. Caire
Abstract:
Using a large deviations approach we calculate the probability distribution of the mutual information of MIMO channels in the limit of large antenna numbers. In contrast to previous methods that only focused at the distribution close to its mean (thus obtaining an asymptotically Gaussian distribution), we calculate the full distribution, including its tails which strongly deviate from the Gaussian…
▽ More
Using a large deviations approach we calculate the probability distribution of the mutual information of MIMO channels in the limit of large antenna numbers. In contrast to previous methods that only focused at the distribution close to its mean (thus obtaining an asymptotically Gaussian distribution), we calculate the full distribution, including its tails which strongly deviate from the Gaussian behavior near the mean. The resulting distribution interpolates seamlessly between the Gaussian approximation for rates $R$ close to the ergodic value of the mutual information and the approach of Zheng and Tse for large signal to noise ratios $ρ$. This calculation provides us with a tool to obtain outage probabilities analytically at any point in the $(R, ρ, N)$ parameter space, as long as the number of antennas $N$ is not too small. In addition, this method also yields the probability distribution of eigenvalues constrained in the subspace where the mutual information per antenna is fixed to $R$ for a given $ρ$. Quite remarkably, this eigenvalue density is of the form of the Marcenko-Pastur distribution with square-root singularities, and it depends on the values of $R$ and $ρ$.
△ Less
Submitted 14 October, 2010; v1 submitted 28 July, 2009;
originally announced July 2009.
-
The emergence of rational behavior in the presence of stochastic perturbations
Authors:
Panayotis Mertikopoulos,
Aris L. Moustakas
Abstract:
We study repeated games where players use an exponential learning scheme in order to adapt to an ever-changing environment. If the game's payoffs are subject to random perturbations, this scheme leads to a new stochastic version of the replicator dynamics that is quite different from the "aggregate shocks" approach of evolutionary game theory. Irrespective of the perturbations' magnitude, we find…
▽ More
We study repeated games where players use an exponential learning scheme in order to adapt to an ever-changing environment. If the game's payoffs are subject to random perturbations, this scheme leads to a new stochastic version of the replicator dynamics that is quite different from the "aggregate shocks" approach of evolutionary game theory. Irrespective of the perturbations' magnitude, we find that strategies which are dominated (even iteratively) eventually become extinct and that the game's strict Nash equilibria are stochastically asymptotically stable. We complement our analysis by illustrating these results in the case of congestion games.
△ Less
Submitted 21 October, 2010; v1 submitted 11 June, 2009;
originally announced June 2009.
-
Performance of MMSE MIMO Receivers: A Large N Analysis for Correlated Channels
Authors:
Aris L. Moustakas,
K Raj Kumar,
Giuseppe Caire
Abstract:
Linear receivers are considered as an attractive low-complexity alternative to optimal processing for multi-antenna MIMO communications. In this paper we characterize the performance of MMSE MIMO receivers in the limit of large antenna numbers in the presence of channel correlations. Using the replica method, we generalize our results obtained in arXiv:0810.0883 to Kronecker-product correlated c…
▽ More
Linear receivers are considered as an attractive low-complexity alternative to optimal processing for multi-antenna MIMO communications. In this paper we characterize the performance of MMSE MIMO receivers in the limit of large antenna numbers in the presence of channel correlations. Using the replica method, we generalize our results obtained in arXiv:0810.0883 to Kronecker-product correlated channels and calculate the asymptotic mean and variance of the mutual information of a MIMO system of parallel MMSE subchannels. The replica method allows us to use the ties between the optimal receiver mutual information and the MMSE SIR of Gaussian inputs to calculate the joint moments of the SIRs of the MMSE subchannels. Using the methodology discussed in arXiv:0810.0883 it can be shown that the mutual information converges in distribution to a Gaussian random variable. Our results agree very well with simulations even with a moderate number of antennas.
△ Less
Submitted 20 February, 2009;
originally announced February 2009.
-
Asymptotic Performance of Linear Receivers in MIMO Fading Channels
Authors:
K. Raj Kumar,
G. Caire,
A. L. Moustakas
Abstract:
Linear receivers are an attractive low-complexity alternative to optimal processing for multi-antenna MIMO communications. In this paper we characterize the information-theoretic performance of MIMO linear receivers in two different asymptotic regimes. For fixed number of antennas, we investigate the limit of error probability in the high-SNR regime in terms of the Diversity-Multiplexing Tradeof…
▽ More
Linear receivers are an attractive low-complexity alternative to optimal processing for multi-antenna MIMO communications. In this paper we characterize the information-theoretic performance of MIMO linear receivers in two different asymptotic regimes. For fixed number of antennas, we investigate the limit of error probability in the high-SNR regime in terms of the Diversity-Multiplexing Tradeoff (DMT). Following this, we characterize the error probability for fixed SNR in the regime of large (but finite) number of antennas.
As far as the DMT is concerned, we report a negative result: we show that both linear Zero-Forcing (ZF) and linear Minimum Mean-Square Error (MMSE) receivers achieve the same DMT, which is largely suboptimal even in the case where outer coding and decoding is performed across the antennas. We also provide an approximate quantitative analysis of the markedly different behavior of the MMSE and ZF receivers at finite rate and non-asymptotic SNR, and show that while the ZF receiver achieves poor diversity at any finite rate, the MMSE receiver error curve slope flattens out progressively, as the coding rate increases.
When SNR is fixed and the number of antennas becomes large, we show that the mutual information at the output of a MMSE or ZF linear receiver has fluctuations that converge in distribution to a Gaussian random variable, whose mean and variance can be characterized in closed form. This analysis extends to the linear receiver case a well-known result previously obtained for the optimal receiver. Simulations reveal that the asymptotic analysis captures accurately the outage behavior of systems even with a moderate number of antennas.
△ Less
Submitted 18 February, 2009; v1 submitted 6 October, 2008;
originally announced October 2008.
-
Correlated Anarchy in Overlapping Wireless Networks
Authors:
Panayotis Mertikopoulos,
Aris L. Moustakas
Abstract:
We investigate the behavior of a large number of selfish users that are able to switch dynamically between multiple wireless access-points (possibly belonging to different standards) by introducing an iterated non-cooperative game. Users start out completely uneducated and naive but, by using a fixed set of strategies to process a broadcasted training signal, they quickly evolve and converge to…
▽ More
We investigate the behavior of a large number of selfish users that are able to switch dynamically between multiple wireless access-points (possibly belonging to different standards) by introducing an iterated non-cooperative game. Users start out completely uneducated and naive but, by using a fixed set of strategies to process a broadcasted training signal, they quickly evolve and converge to an evolutionarily stable equilibrium. Then, in order to measure efficiency in this steady state, we adapt the notion of the price of anarchy to our setting and we obtain an explicit analytic estimate for it by using methods from statistical physics (namely the theory of replicas). Surprisingly, we find that the price of anarchy does not depend on the specifics of the wireless nodes (e.g. spectral efficiency) but only on the number of strategies per user and a particular combination of the number of nodes, the number of users and the size of the training signal. Finally, we map this game to the well-studied minority game, generalizing its analysis to an arbitrary number of choices.
△ Less
Submitted 7 May, 2008;
originally announced May 2008.
-
Closed-Form Density of States and Localization Length for a Non-Hermitian Disordered System
Authors:
Pavlos Kazakopoulos,
Aris L. Moustakas
Abstract:
We calculate the Lyapunov exponent for the non-Hermitian Zakharov-Shabat eigenvalue problem corresponding to the attractive non-linear Schroedinger equation with a Gaussian random pulse as initial value function. Using an extension of the Thouless formula to non-Hermitian random operators, we calculate the corresponding average density of states. We analyze two cases, one with circularly symmetr…
▽ More
We calculate the Lyapunov exponent for the non-Hermitian Zakharov-Shabat eigenvalue problem corresponding to the attractive non-linear Schroedinger equation with a Gaussian random pulse as initial value function. Using an extension of the Thouless formula to non-Hermitian random operators, we calculate the corresponding average density of states. We analyze two cases, one with circularly symmetric complex Gaussian pulses and the other with real Gaussian pulses. We discuss the implications in the context of the information transmission through non-linear optical fibers.
△ Less
Submitted 19 June, 2008; v1 submitted 21 June, 2007;
originally announced June 2007.
-
Vector Precoding for Wireless MIMO Systems: A Replica Analysis
Authors:
Ralf R. Mueller,
Dongning Guo,
Aris L. Moustakas
Abstract:
We apply the replica method to analyze vector precoding, a method to reduce transmit power in antenna array communications. The analysis applies to a very general class of channel matrices. The statistics of the channel matrix enter the transmitted energy per symbol via its R-transform. We find that vector precoding performs much better for complex than for real alphabets. As a byproduct, we fin…
▽ More
We apply the replica method to analyze vector precoding, a method to reduce transmit power in antenna array communications. The analysis applies to a very general class of channel matrices. The statistics of the channel matrix enter the transmitted energy per symbol via its R-transform. We find that vector precoding performs much better for complex than for real alphabets. As a byproduct, we find a nonlinear precoding method with polynomial complexity that outperforms NP-hard Tomlinson-Harashima precoding for binary modulation on complex channels if the number of transmit antennas is slightly larger than twice the number of receive antennas.
△ Less
Submitted 8 June, 2007;
originally announced June 2007.
-
Maximum Entropy MIMO Wireless Channel Models
Authors:
M. Guillaud,
M. Debbah,
A. L. Moustakas
Abstract:
In this contribution, models of wireless channels are derived from the maximum entropy principle, for several cases where only limited information about the propagation environment is available. First, analytical models are derived for the cases where certain parameters (channel energy, average energy, spatial correlation matrix) are known deterministically. Frequently, these parameters are unkn…
▽ More
In this contribution, models of wireless channels are derived from the maximum entropy principle, for several cases where only limited information about the propagation environment is available. First, analytical models are derived for the cases where certain parameters (channel energy, average energy, spatial correlation matrix) are known deterministically. Frequently, these parameters are unknown (typically because the received energy or the spatial correlation varies with the user position), but still known to represent meaningful system characteristics. In these cases, analytical channel models are derived by assigning entropy-maximizing distributions to these parameters, and marginalizing them out. For the MIMO case with spatial correlation, we show that the distribution of the covariance matrices is conveniently handled through its eigenvalues. The entropy-maximizing distribution of the covariance matrix is shown to be a Wishart distribution. Furthermore, the corresponding probability density function of the channel matrix is shown to be described analytically by a function of the channel Frobenius norm. This technique can provide channel models incorporating the effect of shadow fading and spatial correlation between antennas without the need to assume explicit values for these parameters. The results are compared in terms of mutual information to the classical i.i.d. Gaussian model.
△ Less
Submitted 20 December, 2006;
originally announced December 2006.
-
Capacity of Differential versus Non-Differential Unitary Space-Time Modulation for MIMO channels
Authors:
Aris L. Moustakas,
Steven H. Simon,
Thomas L. Marzetta
Abstract:
Differential Unitary Space-Time Modulation (DUSTM) and its earlier nondifferential counterpart, USTM, permit high-throughput MIMO communication entirely without the possession of channel state information (CSI) by either the transmitter or the receiver. For an isotropically random unitary input we obtain the exact closed-form expression for the probability density of the DUSTM received signal, w…
▽ More
Differential Unitary Space-Time Modulation (DUSTM) and its earlier nondifferential counterpart, USTM, permit high-throughput MIMO communication entirely without the possession of channel state information (CSI) by either the transmitter or the receiver. For an isotropically random unitary input we obtain the exact closed-form expression for the probability density of the DUSTM received signal, which permits the straightforward Monte Carlo evaluation of its mutual information. We compare the performance of DUSTM and USTM through both numerical computations of mutual information and through the analysis of low- and high-SNR asymptotic expressions. In our comparisons the symbol durations of the equivalent unitary space-time signals are both equal to T, as are the number of receive antennas N. For DUSTM the number of transmit antennas is constrained by the scheme to be M = T/2, while USTM has no such constraint. If DUSTM and USTM utilize the same number of transmit antennas at high SNR's the normalized mutual information of the differential and the nondifferential schemes expressed in bits/sec/Hz are asymptotically equal, with the differential scheme performing somewhat better, while at low SNR's the normalized mutual information of DUSTM is asymptotically twice the normalized mutual information of USTM. If, instead, USTM utilizes the optimum number of transmit antennas then USTM can outperform DUSTM at sufficiently low SNR's.
△ Less
Submitted 12 December, 2005;
originally announced December 2005.
-
On the Outage Capacity of Correlated Multiple-Path MIMO Channels
Authors:
Aris L. Moustakas,
Steven H. Simon
Abstract:
The use of multi-antenna arrays in both transmission and reception has been shown to dramatically increase the throughput of wireless communication systems. As a result there has been considerable interest in characterizing the ergodic average of the mutual information for realistic correlated channels. Here, an approach is presented that provides analytic expressions not only for the average, b…
▽ More
The use of multi-antenna arrays in both transmission and reception has been shown to dramatically increase the throughput of wireless communication systems. As a result there has been considerable interest in characterizing the ergodic average of the mutual information for realistic correlated channels. Here, an approach is presented that provides analytic expressions not only for the average, but also the higher cumulant moments of the distribution of the mutual information for zero-mean Gaussian (multiple-input multiple-output) MIMO channels with the most general multipath covariance matrices when the channel is known at the receiver. These channels include multi-tap delay paths, as well as general channels with covariance matrices that cannot be written as a Kronecker product, such as dual-polarized antenna arrays with general correlations at both transmitter and receiver ends. The mathematical methods are formally valid for large antenna numbers, in which limit it is shown that all higher cumulant moments of the distribution, other than the first two scale to zero. Thus, it is confirmed that the distribution of the mutual information tends to a Gaussian, which enables one to calculate the outage capacity. These results are quite accurate even in the case of a few antennas, which makes this approach applicable to realistic situations.
△ Less
Submitted 28 September, 2005;
originally announced September 2005.
-
Capacity and Character Expansions: Moment generating function and other exact results for MIMO correlated channels
Authors:
Steven H. Simon,
Aris L. Moustakas,
Luca Marinelli
Abstract:
We apply a promising new method from the field of representations of Lie groups to calculate integrals over unitary groups, which are important for multi-antenna communications. To demonstrate the power and simplicity of this technique, we first re-derive a number of results that have been used recently in the community of wireless information theory, using only a few simple steps. In particular…
▽ More
We apply a promising new method from the field of representations of Lie groups to calculate integrals over unitary groups, which are important for multi-antenna communications. To demonstrate the power and simplicity of this technique, we first re-derive a number of results that have been used recently in the community of wireless information theory, using only a few simple steps. In particular, we derive the joint probability distribution of eigenvalues of the matrix GG*, with G a semicorrelated Gaussian random matrix or a Gaussian random matrix with a non-zero mean (and G* its hermitian conjugate) . These joint probability distribution functions can then be used to calculate the moment generating function of the mutual information for Gaussian channels with multiple antennas on both ends with this probability distribution of their channel matrices G. We then turn to the previously unsolved problem of calculating the moment generating function of the mutual information of MIMO (multiple input-multiple output) channels, which are correlated at both the receiver and the transmitter. From this moment generating function we obtain the ergodic average of the mutual information and study the outage probability. These methods can be applied to a number of other problems. As a particular example, we examine unitary encoded space-time transmission of MIMO systems and we derive the received signal distribution when the channel matrix is correlated at the transmitter end.
△ Less
Submitted 27 September, 2005;
originally announced September 2005.
