This study investigates the use of Multi-Objective Particle Swarm Optimization (MOPSO) to optimize the hyper-parameters of Cooperative Adaptive Cruise Control (CACC) controllers, aiming to enhance performance in dynamic traffic scenarios. The proposed approach, named CACCpo, identifies hyperparameter combinations on Pareto fronts for a CACC controller based on the reduction of its accumulated error and overshoot, which are related to the system’s response time and stability, respectively. The proposed CACC controller, tuned with MOPSO, achieves an accumulated error of 21.21 and an overshoot of 6.38%, outperforming other controllers in speed and accuracy when responding to variations in the leader vehicle’s speed. In addition, it provides smoother acceleration and deceleration with minimal oscillations, ensuring a more stable and efficient system. Moreover, it maintains a constant safe distance between vehicles, which is essential for passenger safety and comfort.

Deep learning has been adopted for a wide range of wireless communication tasks, including modulation classification, because of its great classification capability. However, deep learning models have been shown to also introduce risks and vulnerabilities. For instance, adversarial attacks craft and introduce imperceptible perturbations that compromise the accuracy of deep learning-based modulation classifiers on wireless receivers. Therefore, in this article, we propose a novel wireless receiver architecture that enhances deep learning-based modulation classifiers to defend them against adversarial attacks. Our experimental results show that our defense technique significantly diminishes the accuracy reduction that is caused by adversarial attacks by protecting modulation classifiers at least 18% more than existing defense techniques.

Recently, there has been an interest from the automotive industry in a novel single-pair, half-duplex Ethernet multidrop provision (i.e., 10BASE-T1S), which relies on a coordinated channel access method provided by Physical Layer Collision Avoidance (PLCA). PLCA avoids frame collisions and provides bounded latency, while guaranteeing fairness among nodes and optimal bandwidth utilization. However, despite these advantageous features, PLCA is not aware of flow priorities and, consequently, does not provide priority-based Quality of Service (QoS). Thus, there is an opportunity for integrating QoS-aware schemes, such as the Time-Sensitive Networking (TSN)-based Time-Aware Shaper (TAS) algorithm, into PLCA. In this paper, we present modeling and analysis strategies using Network Calculus for integrating TAS with PLCA to support scheduled traffic, and we compare our modeling and analysis with a baseline work which previously modeled TAS for full-duplex point-to-point Ethernet only. We show that replacing full-duplex point-to-point links with half-duplex PLCA multidrop links still complies with end-to-end delay deadlines, while leading to a reduction of system costs due to the fewer number of Ethernet transceivers in a PLCA multidrop. Moreover, we also show that our results for TAS, when applied only on full-duplex point-to-point links, outperform the results of the baseline work in most use cases.

Cyber-attacks cyber-physical systems (CPSs) can lead to sensing and actuation misbehavior, severe damages to physical objects, and safety risks. Machine learning algorithms have been proposed for hindering cyber-attacks on CPSs, but the absence of labeled data from novel attacks makes their detection quite challenging. In this context, generative adversarial networks (GANs) are a promising unsupervised approach to detect cyber-attacks by implicitly modeling the system. However, the detection of cyber-attacks on CPSs has strict latency requirements, since the attacks need to be stopped before the system is compromised. In this article, we propose FID-GAN, a novel fog-based, unsupervised intrusion detection system (IDS) for CPSs using GANs. The IDS is proposed for a fog architecture, which brings computation resources closer to the end nodes and thus contributes to meeting low-latency requirements. In order to achieve higher detection rates, the proposed architecture computes a reconstruction loss based on the reconstruction of data samples mapped to the latent space. Other works that follow a similar approach struggle with the time required to compute the reconstruction loss, which renders them impractical for latency constrained applications. We address this problem by training an encoder that accelerates the reconstruction loss computation. Experiments show that the proposed solution achieves higher detection rates and is at least 5.5 times faster than a baseline approach in the three studied data sets.

The presence of connected devices in homes introduces numerous threats to privacy via the analysis of the encrypted traffic these devices generate. Prior works have shown that traffic attributes such as packet size combined with machine learning techniques enable the inference of private information from Internet of Things users. One of the commonly used techniques to mitigate those privacy threats is traffic obfuscation, such as packet padding. Most padding mechanisms that were previously proposed statically select the number of bytes inserted in the packets, which incurs high overhead and ineffective privacy improvement. These static mechanisms are particularly unsuitable for networks whose traffic patterns are significantly dynamic, such as smart homes. This article proposes an adaptive packet padding approach based on software-defined networking (SDN) that adjusts the number of bytes inserted into packets in response to variations in the home network utilization. The proposed technique monitors the network to instruct a padding mechanism through a representational state transfer (REST) interface proposed in this article. This mechanism ensures that the length of packets generated by connected devices is modified. The evaluation includes four supervised learning mechanisms, random forest (RF), support vector machine (SVM), decision tree, and k-nearest neighbors (KNNs), to measure privacy improvement through the metrics accuracy, recall, and F1-score. Goodput, jitter, and packet loss induced by the proposal are also evaluated. Our proposal is shown to overcome the state-of-the-art solutions in privacy preservation with a significantly lower overhead. For instance, the accuracy of RF on identifying devices decreases from 96% to 4.96%.

In this paper we propose a circuit reallocation strategy for elastic optical networks aware of the effects of the physical layer. The proposed strategy aims to reorganize active circuits so that new circuit requests that are pending blockage are established. The strategy was applied with the algorithms CS, KSPC, KSP-RQoTO and MD-PC. The strategy was evaluate using the bandwidth blocking probability metric and components of the circuit blocking probability. The results were obtained using the real EON and NSFNet topologies. In terms of the probability of bandwidth blocking for the EON topology, for example, the reduction at the last load point was 81.01%, 89.2%, 80.93% and 78.27% for the CS, KS-PC, KSP-RQoTO and MD-PC algorithms, respectively. The blocking probability reduction for the NSFNet topology was 87%, 95.55%, 92.19% and 76.66%, for CS, KS-PC, KSP-RQoTO and MD-PC, also considering the last load point. The reductions in the probability of bandwidth blocking were due to the reallocations of circuits carried out by the strategy proposed in this work.

Traffic analysis of side-channel information represents a major concern in consumer IoT privacy, since it can lead to a leakage of user data and behavior. In this paper, we propose a lightweight padding-based mechanism to obfuscate one type of side-channel information, the packet length, in order to enhance user privacy. A prototype of the mechanism is implemented in a virtual machine along with five different machine learning classifiers, which are used to identify IoT devices by means of packet length patterns. Results show that the packet padding mechanism is able to reduce the accuracy of the classifiers in at least 75%, thus enhancing the privacy of the IoT communication. Additionally, the delay, jitter, throughput and packet loss are used to quantify the impact of the mechanism on the communication performance of the IoT devices. It is shown that the communication overhead generated the by proposed mechanism was kept to a minimum.

In this paper, cryptographic schemes are applied to Ethernet-based layer-2 communication to provide authenticated encryption to safety- critical automotive control data. Confidentiality, integrity and authenticity are provided by combining AES with HMAC. Experimental results using low-cost hardware show that, despite the introduced cryptographic overhead, latency requirements are comfortably met for this type of communication.

Energy efficiency has become a salient element in the design and operation of modern telecommunications networks. To support energy efficiency in passive optical network (PON) systems, the ITU-T has recently standardized a new power management technique, namely the Watchful Sleep mode, which unifies the two previously standardized techniques (i.e., Doze and Cyclic Sleep modes). The new mode can emulate each of the Doze and Cyclic modes, and it outperforms them in terms of energy efficiency and reduced network signaling. In this paper, we present the first analytical model for numerically measure the performance of the Watchful Sleep mode. Specifically, the new model estimates the ONU's energy efficiency under different network scenarios given a set of quality-of-service constraints. To validate our model, we perform comparisons with simulation results. These comparisons demonstrate that the proposed model can accurately capture the energy efficiency of PON systems.

Media streaming services are responsible for the significant increase of Internet traffic. This fact generates the need for better computing resources management in order to maintain such services always available. Cloud computing offers an elastic infrastructure that provides computing resources on demand in order to maintain such services always available. Moreover, emerging network architectures, such as Software Defined Networks, provide mechanisms that simplify the implementation of advanced network functions. In this context, this paper proposes an elastic approach to manage a video streaming service. The solution includes an OpenFlow-based load balancer, and an elasticity management mechanism that increases or decreases the amount of active streaming servers, considering customer's demand. A prototype was developed using a private cloud infrastructure and experiments were performed to show the viability and effectiveness of our solution, in terms of load balancing and elasticity functionality.

Large scale networks, such as those deployed by Service Providers (SPs), employ robust architectures, capable of supporting large volumes of traffic with very different characteristics. Their network equipment has significant processing load, being responsible for building both a routing logic and the routing of traffic itself. By having the network control implemented in a distributed manner and being built with a limited number of vendors, these networks have limitations of control and traffic engineering, hindering the differentiation among SPs. Additionally, the network intelligence is hidden in the network equipment, making innovations very slow and conditioned to the vendors interests. As an alternative option, this work proposes a Software Defined Networking (SDN)-OpenFlow network architecture that attempts to improve the previously mentioned problems, and, at the same time, to solve the arising difficulties related to the SDN network centralized feature. With the proposed architecture, a robust SP SDN-OpenFlow network is created to support high controller response times and controller outages, without additional delays in the creation of flows and with significant reduction of the controller load. A prototype has been built using Open vSwitch as a virtualization software for OpenFlow clients, Mininet for the topology construction and Ryu as the controller, all with OpenFlow 1.3 support. The obtained results are general and can be extended to other types of networks.

In the past few years, Ethernet Passive Optical Networks (EPONs) have rapidly gained large popularity in broadband access networks. However the current standard has no management protocols aiming at reducing power consumption. In this paper, we propose an Energy Management Mechanism (EMM) within the IEEE 802.3ah control scheme. The main idea is to switch Optical Network Units (ONUs) to sleep mode and determine a suitable wakeup time schedule at the Optical Line Terminal (OLT). In the proposed EMM there is a trade-off between maximizing the power saving and guaranteeing the network performance. We compare two types of downstream scheduling schemes, Upstream Centric Scheduling (UCS) and Downstream Centric Scheduling (DCS), which are different in the way they assign 'active' and 'sleep' states to ONUs. Simulation results in terms of energy consumption and queuing delay are shown for the EMM based EPON system.

This paper presents a novel hierarchical and grid based reconfigurable optical-wireless network (GROW-Net) architecture. GROW-Net supports scalable and flexible integration of a large scale wireless mesh network in a municipal environment. Under the GROW-Net architecture, a joint evolution strategy is proposed to allow graceful upgrades in both optical backbone and wireless mesh network. To alleviate the known throughput bottleneck in mesh networks, a capacity enhancing technique based on flexible cell-splitting is proposed. The technique allocates wavelength resources to unutilized dark fibers using the unique structure of GROW-Net's reconfigurable and colorless access gateways. To analyze the effectiveness of the approach, throughput performance of the wireless mesh network is analyzed using a high fidelity simulator. The wireless mesh network in the simulation employs a distributed and cooperative medium access control protocol within a time division multiple access and time division duplex framework. Through the high fidelity simulation, results show the degree of throughput enhancement by using cell splitting method. The evolution strategy further enables the backbone to scale its bandwidth and adapt to future high throughput and very high throughput wireless technologies. The performance of the proposed backbone is demonstrated experimentally over the optical testbed.

In this paper we present a method for calculating the exact blocking probability in WDM rings with full, ubiquitous wavelength conversion under Poissonian, homogeneous traffic with any spatial profile. A new scalable procedure is presented to derive the normalization constant of the Erlang's classical model for ring networks. In the same analytical framework, formulas are derived for the utilization rate and blocking probability of the ring.

The paper introduces a new analytical approach for estimating blocking probabilities in linear all-optical networks. The assumptions of the classical Lee approximation are discussed and their effects are evaluated. The paper proposes a better substitute for the independent link assumption. The new assumption takes all active paths, as well as all free links, as independent objects on the network topology. The new model is shown to generate estimates that exactly fit the blocking probabilities obtained through simulations on linear topologies using Poissonian, single-wavelength spatially homogeneous traffic. Finally, the independent wavelength assumption is also evaluated.

The performance improvements imparted by first-fit wavelength assignment algorithms on blocking WDM rings without wavelength conversions are investigated. Computationally inexpensive algorithms are derived for the minimization of: (a) the blocking probability immediately after each assignment; and (b) the increment of the channel blocking probability at the assigned wavelength. Their performance is evaluated by simulation and compared with the basic random and priority algorithms.