arXiv subjects

M. Cenk Gursoy

Publications and source records attributed to M. Cenk Gursoy.

At least 37 records · Page 2Linked to original sources

arXiv2022

How to Attack and Defend NextG Radio Access Network Slicing with Reinforcement Learning

In this paper, reinforcement learning (RL) for network slicing is considered in NextG radio access networks, where the base station (gNodeB) allocates resource blocks (RBs) to the requests of user equipments and aims to maximize the total reward of accepted requests over time. Based on adversarial machine learning, a novel over-the-air attack is introduced to manipulate the RL algorithm and disrupt NextG network slicing. The adversary observes the spectrum and builds its own RL based surrogate model that selects which RBs to jam subject to an energy budget with the objective of maximizing the number of failed requests due to jammed RBs. By jamming the RBs, the adversary reduces the RL algorithm's reward. As this reward is used as the input to update the RL algorithm, the performance does not recover even after the adversary stops jamming. This attack is evaluated in terms of both the recovery time and the (maximum and total) reward loss, and it is shown to be much more effective than benchmark (random and myopic) jamming attacks. Different reactive and proactive defense schemes (protecting the RL algorithm's updates or misleading the adversary's learning process) are introduced to show that it is viable to defend NextG network slicing against this attack.

cs.NI↗

arXiv2022

Communication-Efficient and Privacy-Preserving Feature-based Federated Transfer Learning

Federated learning has attracted growing interest as it preserves the clients' privacy. As a variant of federated learning, federated transfer learning utilizes the knowledge from similar tasks and thus has also been intensively studied. However, due to the limited radio spectrum, the communication efficiency of federated learning via wireless links is critical since some tasks may require thousands of Terabytes of uplink payload. In order to improve the communication efficiency, we in this paper propose the feature-based federated transfer learning as an innovative approach to reduce the uplink payload by more than five orders of magnitude compared to that of existing approaches. We first introduce the system design in which the extracted features and outputs are uploaded instead of parameter updates, and then determine the required payload with this approach and provide comparisons with the existing approaches. Subsequently, we analyze the random shuffling scheme that preserves the clients' privacy. Finally, we evaluate the performance of the proposed learning scheme via experiments on an image classification task to show its effectiveness.

cs.LG↗

arXiv2021

Dynamic Channel Access and Power Control in Wireless Interference Networks via Multi-Agent Deep Reinforcement Learning

Due to the scarcity in the wireless spectrum and limited energy resources especially in mobile applications, efficient resource allocation strategies are critical in wireless networks. Motivated by the recent advances in deep reinforcement learning (DRL), we address multi-agent DRL-based joint dynamic channel access and power control in a wireless interference network. We first propose a multi-agent DRL algorithm with centralized training (DRL-CT) to tackle the joint resource allocation problem. In this case, the training is performed at the central unit (CU) and after training, the users make autonomous decisions on their transmission strategies with only local information. We demonstrate that with limited information exchange and faster convergence, DRL-CT algorithm can achieve 90% of the performance achieved by the combination of weighted minimum mean square error (WMMSE) algorithm for power control and exhaustive search for dynamic channel access. In the second part of this paper, we consider distributed multi-agent DRL scenario in which each user conducts its own training and makes its decisions individually, acting as a DRL agent. Finally, as a compromise between centralized and fully distributed scenarios, we consider federated DRL (FDRL) to approach the performance of DRL-CT with the use of a central unit in training while limiting the information exchange and preserving privacy of the users in the wireless system. Via simulation results, we show that proposed learning frameworks lead to efficient adaptive channel access and power control policies in dynamic environments.

cs.IT↗

arXiv2021

Dynamic Channel Access via Meta-Reinforcement Learning

In this paper, we address the channel access problem in a dynamic wireless environment via meta-reinforcement learning. Spectrum is a scarce resource in wireless communications, especially with the dramatic increase in the number of devices in networks. Recently, inspired by the success of deep reinforcement learning (DRL), extensive studies have been conducted in addressing wireless resource allocation problems via DRL. However, training DRL algorithms usually requires a massive amount of data collected from the environment for each specific task and the well-trained model may fail if there is a small variation in the environment. In this work, in order to address these challenges, we propose a meta-DRL framework that incorporates the method of Model-Agnostic Meta-Learning (MAML). In the proposed framework, we train a common initialization for similar channel selection tasks. From the initialization, we show that only a few gradient descents are required for adapting to different tasks drawn from the same distribution. We demonstrate the performance improvements via simulation results.

cs.NI↗

arXiv2021

Joint Activity Detection and Channel Estimation in Cell-Free Massive MIMO Networks with Massive Connectivity

Cell-free massive MIMO is one of the key technologies for future wireless communications, in which users are simultaneously and jointly served by all access points (APs). In this paper, we investigate the minimum mean square error (MMSE) estimation of effective channel coefficients in cell-free massive MIMO systems with massive connectivity. To facilitate the theoretical analysis, only single measurement vector (SMV) based MMSE estimation is considered in this paper, i.e., the MMSE estimation is performed based on the received pilot signals at each AP separately. Inspired by the decoupling principle of replica symmetric postulated MMSE estimation of sparse signal vectors with independent and identically distributed (i.i.d.) non-zero components, we develop the corresponding decoupling principle for the SMV based MMSE estimation of sparse signal vectors with independent and non-identically distributed (i.n.i.d.) non-zero components, which plays a key role in the theoretical analysis of SMV based MMSE estimation of the effective channel coefficients in cell-free massive MIMO systems with massive connectivity. Subsequently, based on the obtained decoupling principle of MMSE estimation, likelihood ratio test and the optimal fusion rule, we perform user activity detection based on the received pilot signals at only one AP, or cooperation among the entire set of APs for centralized or distributed detection. Via theoretical analysis, we show that the error probabilities of both centralized and distributed detection tend to zero when the number of APs tends to infinity while the asymptotic ratio between the number of users and pilots is kept constant. We also investigate the asymptotic behavior of oracle estimation in cell-free massive MIMO systems with massive connectivity via random matrix theory.

cs.IT↗

arXiv2021

Gibbs Distribution Based Antenna Splitting and User Scheduling in Full Duplex Massive MIMO Systems

A Gibbs distribution based combinatorial optimization algorithm for joint antenna splitting and user scheduling problem in full duplex massive multiple-input multiple-output (MIMO) system is proposed in this paper. The optimal solution of this problem can be determined by exhaustive search. However, the complexity of this approach becomes prohibitive in practice when the sample space is large, which is usually the case in massive MIMO systems. Our algorithm overcomes this drawback by converting the original problem into a Kullback-Leibler (KL) divergence minimization problem, and solving it through a related dynamical system via a stochastic gradient descent method. Using this approach, we improve the spectral efficiency (SE) of the system by performing joint antenna splitting and user scheduling. Additionally, numerical results show that the SE curves obtained with our proposed algorithm overlap with the curves achieved by exhaustive search for user scheduling.

cs.IT↗

arXiv2021

Scalable and Decentralized Algorithms for Anomaly Detection via Learning-Based Controlled Sensing

We address the problem of sequentially selecting and observing processes from a given set to find the anomalies among them. The decision-maker observes a subset of the processes at any given time instant and obtains a noisy binary indicator of whether or not the corresponding process is anomalous. In this setting, we develop an anomaly detection algorithm that chooses the processes to be observed at a given time instant, decides when to stop taking observations, and declares the decision on anomalous processes. The objective of the detection algorithm is to identify the anomalies with an accuracy exceeding the desired value while minimizing the delay in decision making. We devise a centralized algorithm where the processes are jointly selected by a common agent as well as a decentralized algorithm where the decision of whether to select a process is made independently for each process. Our algorithms rely on a Markov decision process defined using the marginal probability of each process being normal or anomalous, conditioned on the observations. We implement the detection algorithms using the deep actor-critic reinforcement learning framework. Unlike prior work on this topic that has exponential complexity in the number of processes, our algorithms have computational and memory requirements that are both polynomial in the number of processes. We demonstrate the efficacy of these algorithms using numerical experiments by comparing them with state-of-the-art methods.

cs.LG↗

arXiv2021

Performance Analysis of Cell-Free Massive MIMO Systems with Massive Connectivity

In this paper, we investigate the performance of cell-free massive MIMO systems with massive connectivity. With the generalized approximate message passing (GAMP) algorithm, we obtain the minimum mean-squared error (MMSE) estimate of the effective channel coefficients from all users to all access points (APs) in order to perform joint user activity detection and channel estimation. Subsequently, using the decoupling properties of MMSE estimation for large linear systems and state evolution equations of the GAMP algorithm, we obtain the variances of both the estimated channel coefficients and the corresponding channel estimation error. Finally, we study the achievable uplink rates with zero-forcing (ZF) detector at the central processing unit (CPU) of the cell-free massive MIMO system. With numerical results, we analyze the impact of the number of pilots used for joint activity detection and channel estimation, the number of APs, and signal-to-noise ratio (SNR) on the achievable rates.

cs.IT↗

arXiv2021

Adversarial Reinforcement Learning in Dynamic Channel Access and Power Control

Deep reinforcement learning (DRL) has recently been used to perform efficient resource allocation in wireless communications. In this paper, the vulnerabilities of such DRL agents to adversarial attacks is studied. In particular, we consider multiple DRL agents that perform both dynamic channel access and power control in wireless interference channels. For these victim DRL agents, we design a jammer, which is also a DRL agent. We propose an adversarial jamming attack scheme that utilizes a listening phase and significantly degrades the users' sum rate. Subsequently, we develop an ensemble policy defense strategy against such a jamming attacker by reloading models (saved during retraining) that have minimum transition correlation.

cs.LG↗

arXiv2021

Anomaly Detection via Controlled Sensing and Deep Active Inference

In this paper, we address the anomaly detection problem where the objective is to find the anomalous processes among a given set of processes. To this end, the decision-making agent probes a subset of processes at every time instant and obtains a potentially erroneous estimate of the binary variable which indicates whether or not the corresponding process is anomalous. The agent continues to probe the processes until it obtains a sufficient number of measurements to reliably identify the anomalous processes. In this context, we develop a sequential selection algorithm that decides which processes to be probed at every instant to detect the anomalies with an accuracy exceeding a desired value while minimizing the delay in making the decision and the total number of measurements taken. Our algorithm is based on active inference which is a general framework to make sequential decisions in order to maximize the notion of free energy. We define the free energy using the objectives of the selection policy and implement the active inference framework using a deep neural network approximation. Using numerical experiments, we compare our algorithm with the state-of-the-art method based on deep actor-critic reinforcement learning and demonstrate the superior performance of our algorithm.

cs.LG↗

arXiv2021

A Scalable Algorithm for Anomaly Detection via Learning-Based Controlled Sensing

We address the problem of sequentially selecting and observing processes from a given set to find the anomalies among them. The decision-maker observes one process at a time and obtains a noisy binary indicator of whether or not the corresponding process is anomalous. In this setting, we develop an anomaly detection algorithm that chooses the process to be observed at a given time instant, decides when to stop taking observations, and makes a decision regarding the anomalous processes. The objective of the detection algorithm is to arrive at a decision with an accuracy exceeding a desired value while minimizing the delay in decision making. Our algorithm relies on a Markov decision process defined using the marginal probability of each process being normal or anomalous, conditioned on the observations. We implement the detection algorithm using the deep actor-critic reinforcement learning framework. Unlike prior work on this topic that has exponential complexity in the number of processes, our algorithm has computational and memory requirements that are both polynomial in the number of processes. We demonstrate the efficacy of our algorithm using numerical experiments by comparing it with the state-of-the-art methods.

cs.LG↗

arXiv2021

Jamming-Resilient Path Planning for Multiple UAVs via Deep Reinforcement Learning

Unmanned aerial vehicles (UAVs) are expected to be an integral part of wireless networks. In this paper, we aim to find collision-free paths for multiple cellular-connected UAVs, while satisfying requirements of connectivity with ground base stations (GBSs) in the presence of a dynamic jammer. We first formulate the problem as a sequential decision making problem in discrete domain, with connectivity, collision avoidance, and kinematic constraints. We, then, propose an offline temporal difference (TD) learning algorithm with online signal-to-interference-plus-noise ratio (SINR) mapping to solve the problem. More specifically, a value network is constructed and trained offline by TD method to encode the interactions among the UAVs and between the UAVs and the environment; and an online SINR mapping deep neural network (DNN) is designed and trained by supervised learning, to encode the influence and changes due to the jammer. Numerical results show that, without any information on the jammer, the proposed algorithm can achieve performance levels close to that of the ideal scenario with the perfect SINR-map. Real-time navigation for multi-UAVs can be efficiently performed with high success rates, and collisions are avoided.

cs.LG↗

arXiv2021

Learning-Based UAV Trajectory Optimization with Collision Avoidance and Connectivity Constraints

Unmanned aerial vehicles (UAVs) are expected to be an integral part of wireless networks, and determining collision-free trajectories for multiple UAVs while satisfying requirements of connectivity with ground base stations (GBSs) is a challenging task. In this paper, we first reformulate the multi-UAV trajectory optimization problem with collision avoidance and wireless connectivity constraints as a sequential decision making problem in the discrete time domain. We, then, propose a decentralized deep reinforcement learning approach to solve the problem. More specifically, a value network is developed to encode the expected time to destination given the agent's joint state (including the agent's information, the nearby agents' observable information, and the locations of the nearby GBSs). A signal-to-interference-plus-noise ratio (SINR)-prediction neural network is also designed, using accumulated SINR measurements obtained when interacting with the cellular network, to map the GBSs' locations into the SINR levels in order to predict the UAV's SINR. Numerical results show that with the value network and SINR-prediction network, real-time navigation for multi-UAVs can be efficiently performed in various environments with high success rate.

cs.RO↗

arXiv2021

Uplink Coverage in Heterogeneous mmWave Cellular Networks with Clustered Users

A K-tier heterogeneous mmWave uplink cellular network with clustered user equipments (UEs) is considered in this paper. In particular, UEs are assumed to be clustered around small-cell base stations (BSs) according to a Gaussian distribution, leading to the Thomas cluster process based modeling. Specific and practical line-of-sight (LOS) and non-line-of-sight (NLOS) models are adopted with different parameters for different tiers. The probability density functions (PDFs) and complementary cumulative distribution functions (CCDFs) of different distances from UEs to BSs are characterized. Coupled association strategy and largest long-term averaged biased received power criterion are considered, and general expressions for association probabilities are provided. Following the identification of the association probabilities, the Laplace transforms of the inter-cell interference and the intra-cluster interference are characterized. Using tools from stochastic geometry, general expressions of the SINR coverage probability are provided. As extensions, fractional power control is incorporated into the analysis, tractable closed-form expressions are provided for special cases, and average ergodic spectral efficiency is analyzed. Via numerical and simulation results, analytical characterizations are confirmed and the impact of key system and network parameters on the performance is identified.

cs.IT↗

arXiv2021

Channel Estimation for Intelligent Reflecting Surface Assisted Wireless Communications

In this paper, the minimum mean square error (MMSE) channel estimation for intelligent reflecting surface (IRS) assisted wireless communication systems is investigated. In the considered setting, each row vector of the equivalent channel matrix from the base station (BS) to the users is shown to be Bessel $K$ distributed, and all these row vectors are independent of each other. By introducing a Gaussian scale mixture model, we obtain a closed-form expression for the MMSE estimate of the equivalent channel, and determine analytical upper and lower bounds on the mean square error. Using the central limit theorem, we conduct an asymptotic analysis of the MMSE estimate, and show that the upper bound on the mean square error of the MMSE estimate is equal to the asymptotic mean square error of the MMSE estimation when the number of reflecting elements at the IRS tends to infinity. Numerical simulations show that the gap between the upper and lower bounds are very small, and they almost overlap with each other at medium signal-to-noise ratio (SNR) levels and moderate number of elements at the IRS.

cs.IT↗

arXiv2020

Statistical Learning Based Joint Antenna Selection and User Scheduling for Single-Cell Massive MIMO Systems

Large number of antennas and radio frequency (RF) chains at the base stations (BSs) lead to high energy consumption in massive MIMO systems. Thus, how to improve the energy efficiency (EE) with a computationally efficient approach is a significant challenge in the design of massive MIMO systems. With this motivation, a learning-based stochastic gradient descent algorithm is proposed in this paper to obtain the optimal joint uplink and downlink EE with joint antenna selection and user scheduling in single-cell massive MIMO systems. Using Jensen's inequality and the characteristics of wireless channels, a lower bound on the system throughput is obtained. Subsequently, incorporating the power consumption model, the corresponding lower bound on the EE of the system is identified. Finally, learning-based stochastic gradient descent method is used to solve the joint antenna selection and user scheduling problem, which is a combinatorial optimization problem. Rare event simulation is embedded in the learning-based stochastic gradient descent method to generate samples with very small probabilities. In the analysis, both perfect and imperfect channel side information (CSI) at the BS are considered. Minimum mean-square error (MMSE) channel estimation is employed in the study of the imperfect CSI case. In addition, the effect of a constraint on the number of available RF chains in massive MIMO system is investigated considering both perfect and imperfect CSI at the BS.

cs.IT↗

arXiv2020

Anomaly Detection and Sampling Cost Control via Hierarchical GANs

Anomaly detection incurs certain sampling and sensing costs and therefore it is of great importance to strike a balance between the detection accuracy and these costs. In this work, we study anomaly detection by considering the detection of threshold crossings in a stochastic time series without the knowledge of its statistics. To reduce the sampling cost in this detection process, we propose the use of hierarchical generative adversarial networks (GANs) to perform nonuniform sampling. In order to improve the detection accuracy and reduce the delay in detection, we introduce a buffer zone in the operation of the proposed GAN-based detector. In the experiments, we analyze the performance of the proposed hierarchical GAN detector considering the metrics of detection delay, miss rates, average cost of error, and sampling ratio. We identify the tradeoffs in the performance as the buffer zone sizes and the number of GAN levels in the hierarchy vary. We also compare the performance with that of a sampling policy that approximately minimizes the sum of average costs of sampling and error given the parameters of the stochastic process. We demonstrate that the proposed GAN-based detector can have significant performance improvements in terms of detection delay and average cost of error with a larger buffer zone but at the cost of increased sampling rates.

cs.LG↗

arXiv2020

Multi-Agent Double Deep Q-Learning for Beamforming in mmWave MIMO Networks

Beamforming is one of the key techniques in millimeter wave (mmWave) multi-input multi-output (MIMO) communications. Designing appropriate beamforming not only improves the quality and strength of the received signal, but also can help reduce the interference, consequently enhancing the data rate. In this paper, we propose a distributed multi-agent double deep Q-learning algorithm for beamforming in mmWave MIMO networks, where multiple base stations (BSs) can automatically and dynamically adjust their beams to serve multiple highly-mobile user equipments (UEs). In the analysis, largest received power association criterion is considered for UEs, and a realistic channel model is taken into account. Simulation results demonstrate that the proposed learning-based algorithm can achieve comparable performance with respect to exhaustive search while operating at much lower complexity.

eess.SP↗