In this paper, we utilize a graph neural network to perform joint user association and access point activation/deactivation to optimize network blockage. Instead of using theoretical models to characterize the distribution of network traffic, we use a real-world network traffic dataset. In order to train the graph neural network, our method leverages reinforcement learning, specifically employing the deep deterministic policy gradient (DDPG) algorithm. This approach allows us to benefit from the advantages of both value-based and policy-based reinforcement learning methods. A simple but flexible reward function is defined to capture the trade-off between the fraction of active access points and network blockage. The graph neural network's awareness of the network topology gives the proposed method a clear performance advantage over the commonly used greedy heuristic optimization method, reducing blockage by up to more than 50% on real-world network traffic data while also reducing computational complexity from$\mathcal {O}(N^{3})$to$\mathcal {O}(N)$. The proposed user association and access point activation method is not limited by the network architecture or technology, and can be applied to different generations of cellular networks.

Distributed Denial-of-Service (DDoS) attacks are expected to continue plaguing service availability in emerging networks which rely on distributed edge clouds to offer critical, latency-sensitive applications. However, edge servers increase the network attack surface, which is exacerbated with the massive number of connected Internet of Things (IoT) devices that can be weaponized to launch DDoS attacks. Therefore, it is crucial to detect DDoS attacks early, i.e., at the network edge. In this paper, we empower the network edge with intelligent DDoS detection by learning from similarities between different data and DDoS attacks available across the edge servers. To this end, we develop a novel Outlier Exposure (OE)-enabled cross-silo Federated Learning framework, namely FedOE. FedOE enables distributed training of OE-based ML models using a limited number of labeled outliers (i.e., attack flows) experienced at edge servers. We propose a novel OE-based Autoencoder (oAE) that can better discriminate anomalies in comparison to the widely adopted traditional Autoencoder, using a tailored, OE-based loss function. We evaluate oAE in FedOE and demonstrate its ability to generalize to zero-day attacks, with just 50 labeled attack flows per edge server. The results show that oAE achieves a high F1-score for most DDoS attacks, outclassing its non-OE counterpart.

Cognitive network management is becoming quintessential to realize autonomic networking. However, the wide spread adoption of the Internet of Things (IoT) devices, increases the risk of cyber attacks. Adversaries can exploit vulnerabilities in IoT devices, which can be harnessed to launch massive Distributed Denial of Service (DDoS) attacks. Therefore, intelligent security mechanisms are needed to harden network security against these threats. In this paper, we propose Chronos, a novel time-based anomaly detection system. The anomaly detector, primarily an Autoencoder, leverages time-based features over multiple time windows to efficiently detect anomalous DDoS traffic. We develop a threshold selection heuristic that maximizes the F1-score across various DDoS attacks. Further, we compare the performance of Chronos against state-of-the-art approaches. We show that Chronos marginally outperforms another time-based system using a less complex anomaly detection pipeline, while out classing flow-based approaches with superior precision. In addition, we showcase the robustness of Chronos in the face of zero-day attacks, noise in training data, and a small number of training packets, asserting its suitability for online deployment.

Many studies investigated fair dynamic spectrum access in distributed wireless networks (DWNs). To limit the required communication between nodes, several schemes based on reinforcement learning (RL) have been proposed and their results are promising. One missing point in these studies is that they assume that all users are in saturated mode (always have data to transmit) which is not a correct assumption in reality. In this paper, we remove this assumption and propose a multi-agent RL scheme that incorporates the current need of a user to transmit (its buffer level) in deciding if a user should transmit or not. The complexity though each user is not aware of the buffer level of other users except some delayed information that it can receive from the previous behavior of those users. As simulation results show, the proposed method maximizes the total throughput while trying to make fair resource allocation by first serving the user with the highest level of packets in its queue.

Summary One of the best methods for estimating the softmax layer in neural network language models is the noise‐contrastive estimation (NCE) method. However, this method is not proper for word embedding applications compared with some other robust methods such as the negative sampling (NEG) method. The NEG method implements the pointwise mutual information (PMI) relation between the word‐context space in the neural network, and the NCE method implements conditional probability. Both the NCE and NEG methods use dot‐product‐based mapping to map words and contexts vector to the probabilities. This article presents the parametric objective function, which uses the mapping function as the parameter. Also, we obtained a parametric relation between word‐context space according to the mapping parameter. Using the parametric objective function, we identify conditions for a mapping that make it a proper selection for both softmax estimation and word embedding. The article also presents two specific mapping functions with the required conditions, and we compared their performance with that of the dot‐product mapping function. The performance of the new mapping functions is also reported over common word embedding and language models' benchmarks.

Abstract In this paper, distributed heterogeneous wireless networks are studied and a spectrum access scheme that fairly allocates channels among users based on their request levels is investigated. To be closer to what happens in practice, users' request levels to have different priorities are also considered, i.e. some packets are more important for the user to be transmitted than the others. Due to the distributed nature of the problem and since users are not able to coordinate with each other before transmission, a scheme based on reinforcement learning is proposed in which the network state is estimated based on the previous successful transmissions of packets in the network. The reinforcement learning method's adaptive nature lets our method work in heterogeneous settings where there exist users with other medium access control protocols. The performance of the proposed scheme is evaluated in different network settings, and it is shown that it tries to implement a fair spectrum access policy (considering the packets' priorities) while maximising the total throughput of the network.

Low-rank matrix completion has been exploited in many signal processing problems over the past few years. It has been shown that nuclear norm minimization can retrieve such low-rank matrices via uniform sampling under coherence conditions. However, in many applications such as the Netflix problem, the low-rank matrix is not coherent, hence nuclear norm minimization is unable to solve the problem because of uniform sampling. To overcome this issue, we propose two algorithms to recover both coherent and incoherent low-rank matrices. In the first algorithm, it is assumed that we have a limited sampling budget that is associated with the available number of samples. Using a portion of the sampling budget, we propose an Uncertainty Information (UI) metric to quantify the uncertainty that we have for each element and consider the elements with the most uncertainty as the most informative elements to sample. Finally, the simulations results are performed to investigate the performance of the proposed algorithms and show their superiority to conventional approaches such as nuclear norm minimization Candès and Recht, 2009 and two-phase sampling algorithm Chen et al. 2015.

Distributed Denial of Service (DDoS) attacks continue to draw significant attention, especially with the recent surge in cyber attacks that targeted the healthcare, education and financial sectors, during the COVID-19 pandemic. The expansion of virtualization and softwarization technologies, and the surge in Internet of Things (IoT) devices, increase the attack surface and the impact of attacks on networks. In this paper, we present a novel time-based anomaly detection system that leverages an Autoencoder. We explore the impact of different time-windows on detecting multiple DDoS attacks that are difficult to detect via the widely used flow-based features. We train and evaluate our Autoencoder on the recent CICDDoS2019 dataset, and show that our approach achieves an anomaly detection F1-score of over 99% for most attacks and greater than 95% for all attacks.

This study considers the problem of indoor localisation in which the authors try to generate the radio map of the environment based on both labelled and unlabelled measurements (i.e. measurements with and without information about the location that they have been collected). In such problems, the method that has been used for radio‐map generation as well as the quality of the measurements (the received signal strength from different environment's WiFi Access Points) are very critical in the overall accuracy of the generated radio map. In this study, they apply two of the well‐known graph‐based signal processing schemes (namely iterative least square reconstruction and graph‐based label propagation) for polishing the input data, determining the outliers and radio‐map generation. Experimental results show the superior performance of the proposed method compared to previous schemes.

A number of recent source coding techniques compress the source signal into multiple layers such that a destination is able to reconstruct the original signal (with some distortion) even if it has not received all the layers. Implementation of such source coding techniques in wireless networks requires the application of coding mechanisms (such as multilayer coding) which allow unequal error protection for different layers of the transmitted data. In this paper, we study the performance of multilayer coding for quasi-static fading channels where the source and the destination are equipped with multiple antennas and the Channel-State-Information (CSI) is only known at the destination. We limit the study to the scenarios where the destination is only able to perform successive-decoding (joint-decoding is not possible) and the objective is to find the design of a multilayer code which maximizes the average data rate received at the destination. To this end, we first propose a design rule for constructing a proper multilayer code for Multiple-Input-Multiple-Output (MIMO) networks. Furthermore, the paper presents a procedure which uses the proposed design rule to determine the parameters of the multilayer code. The performance of the designed multilayer coding scheme is then studied for different network setups.

A distributed random access network with K users and one Access Point (AP) is considered. It is assumed that users and the AP are equipped with M and N antennas, respectively. Each user independently decides whether to transmit in a time slot or not. We initially focus on two-user random access networks and characterize the network average Degrees of Freedom (DoF)1. For the K-user networks, an upper-bound on the network average DoF is first proposed. Then, it is shown that the proposed upper-bound can be achieved using single stream data transmission for many network configurations. Finally, we show through a few examples that there exist some network configurations where multi-stream data transmission in conjunction with interference alignment is necessary in order to achieve the upper-bound.

The growing interest for location-based services motivates many researchers to study different localization techniques for indoor environments. The main objective of these studies is to find a balance point between the accuracy of the scheme and its deployment/training cost. RSS-based schemes and in particular Graph-based Semi-Supervised Learning (G-SSL) constitute a group of techniques which has low setup cost and good localization accuracy. In this paper, we analyze the G-SLL scheme and show that, despite its high performance, the G-SSL method (in its original format) is not a very accurate model for a localization problem. Based on this observation and to improve the accuracy of localization, we propose an alternative approach which incorporates our knowledge of wireless signal propagation into the label propagation mechanism. Experimental results are then used to evaluate the performance of the proposed scheme compared to the original G-SSL.

This paper considers a two-hop network in which information is transmitted from a source via a relay to a destination. It is assumed that channels are quasi-static fading with additive white Gaussian noise and that all nodes are equipped with a single antenna. The channel state information (CSI) of each hop is available only at the corresponding receiver and relay is not capable of data buffering over multiple coding blocks. One commonly used design criterion in such configurations is the maximization of the average received rate at the destination. Considering infinite-layer coding at both the source and the relay, in conjunction with decode and forward strategy at the relay, we present a procedure to optimally distribute the available source and relay powers to different layers of their corresponding codes. Next, we demonstrate how this transmission technique can be generalized to a multihop setting. Assuming Rayleigh fading, the performance of the proposed coding scheme is evaluated for a two-hop network and compared with the performance of previously known strategies.

This paper studies a multiple access network with two users and one Access Point (AP). It is assumed that the users and the AP are equipped with M and N antennas, respectively. To access the network, each user independently decides whether to transmit in a time slot or not (no coordination between users). Focusing on the high SNR behavior of the system, this paper presents the optimal value of the network average Degrees of Freedom (DoF) in different network settings. To this end, after finding an upper-bound for the network average DoF, we propose a transmission scheme (based on interference alignment) which achieves this upper-bound. Some illustrative examples are also presented in the paper.

It is known that the achievable data rate per user can be increased when relays are deployed in wireless networks. However, the drawback of this solution is that some of the network's resources should be allocated to the relays. In this paper, we consider a two-tier network in which all users send or receive data in two hops. By applying vector quantization, we compute the relays' locations to improve network's average transmission rate. These locations are also computed analytically when the number of relays is less than six. Having determined the relays' locations, the network's average transmission rate is evaluated. Subsequently, we define the "neutrality-surface" such that the performance of any relay network operating below this surface is inferior to that of the same network without relays. Finally, we study the relative relaying gain for different network configurations.

We investigate how accurate Kronecker model is compared to Spatial Channel Model (SCM) for MIMO link layer simulations. The simulation results show that Kronecker model is a good approximation for MIMO in link layer simulations compared to SCM.

Given a slowly fading channel, the performance of multi-layer coding is studied for single-user scenarios. Both the source and destination are equipped with multiple antennas. The channel state information is perfectly known at the destination but not at the source. The objective is to maximize the average data rate received at the destination when the destination is able to perform successive decoding. This paper, first, proposes a design rule for constructing an infinite-layer code for Multiple Input Multiple Output (MIMO) channels. Furthermore, we present a procedure describing how the introduced design rule is applied to optimally determine the multi-layer code parameters. The achievable rate of the multi-layer coding and successive decoding for the Rayleigh fading 2×2 MIMO channel is also evaluated.

It is known that the achievable data rate per user can be increased when relays are deployed in wireless networks. However, the drawback with this solution is that some of the network resources should be allocated to the relays. In this paper, we consider a two-tier network where all users should send/receive data in two hops (via a relay). Applying vector quantization, we approximately find the location of the relays. These approximate relays' locations are also computed analytically when the number of relays is less than six. Having the relays' locations, the network average transmission rate is evaluated in terms of a set of network parameters. Then, in the multi-dimensional space of these network parameters, we introduce the concept of neutrality-surface. The neutrality-surface is defined such that the performance of any relay network operating below this surface is inferior to that of a simple no-relay network with the same parameters. Finally, we study the relative and differential relaying gain for different network configurations.

In this paper, we evaluate the impact of the multi-antenna schemes on the performance of the ad-hoc WLAN systems in a quasi-static Rayleigh fading environment. We use a Markov chain model to calculate the saturated throughput of an IEEE 802.11 network working in an ad-hoc mode. The throughput has a random nature due to the slowly random changing environment. Hence, the cumulative distribution function (CDF) of the throughput is considered as the performance measure. The effect of different multi-antenna schemes, such as the Alamouti scheme and the maximum ratio receiver combining (MRRC) technique, on this performance measure is evaluated based on the obtained analytical results. We also study the effect of different system parameters on the ad-hoc multi-antenna WLAN performance

In this paper we study the impact of using multi- antenna techniques, such as MIMO and diversity, on the performance of the IEEE802.11b network in a quasi-static Rayleigh fading channel. As a performance measure we consider the saturated throughput. Considering the saturated throughput as a random variable, its CDF for each transmission method is evaluated. Based on the obtained results, the effect of the environment parameters on the system throughput is also discussed.