Journal of Network and Systems Management最新文献

In recent years, the software-defined networking (SDN) paradigm emerged as an easy way to manage large-scale network infrastructures through programmability brought out and its control plane/data plane decoupling logic. This allows infrastructure and service providers to better deploy, configure and automate their traffic management policies and network equipments. However, congestion control remains a concern due to the evolution of increasingly complex and resource-intensive user requirements [(virtual reality, metaverse, Internet of Things (IoT), Artificial Intelligence (AI), Cloud,...] on network infrastructures. This server state leads to high latency in request processing and data loss. This paper proposes, in such a controller-supervised environment, a congestion management scheme within network service servers to maintain an acceptable quality of service. The strategy relies on work stealing to ensure a better workload balance. Simulations show that the proposed solution can reduce congestion load on the servers by up to 22%, depending on the request grain size, with shorter latency than other works in the literature. Moreover, the proposed solution allows stolen tasks to be completed within a shorter time frame.

The Internet of Body (IoB), a subset of wireless sensor networks, has emerged as a promising technology in the biomedical field. The applications of the IoB, particularly in healthcare and medical applications, have attracted significant attention in recent years. The IoB, also known as a Wireless Body Area Network (WBAN), consists of small sensors placed on the human body, which can collect physiological data and facilitate remote operations such as processing, treatment, assessment and decision-making via the Internet network. This paper presents detailed theoretical and experimental studies on the design of sensors for a 5G-based IoB healthcare monitoring network. The need for efficient and high-performance sensors, in the healthcare industry for enabling continuous monitoring of patient’s health in real-time, is highlighted along this work. In this paper, we propose a novel approach for designing and analyzing the performance of IoB antenna sensors, specifically focusing on channel modeling and power-consumption between wearable wireless sensors. The behavior of the sensors on the human body is studied both theoretically and experimentally for two optimal locations: on the human body waist and on human arm-hand. The results are compared to assess the accuracy of the theoretical model. Despite the complexity of the physiological behavior of the human body, our findings show a good agreement between the theoretical and experimental results. This work provides valuable insights into the design and optimization of IoB/WBANs for real-world medical applications.

The large-scale deployment of cloud composite services distributed over heterogeneous environments poses new challenges in terms of security management. In particular, the migration of their resources is facilitated by recent advances in the area of virtualization techniques. This contributes to increase the dynamics of their configuration, and may induce vulnerabilities that could compromise the security of cloud resources, or even of the whole service. In addition, cloud providers may be reluctant to share precise information regarding the configuration of their infrastructures with cloud tenants that build and deploy cloud composite services. This makes the assessment of vulnerabilities difficult to be performed with only a partial view on the overall configuration. We therefore propose in this article an inter-cloud trusted third-party approach, called C3S-TTP, for supporting secure configurations in cloud composite services, more specifically during the migration of their resources. We describe the considered architecture, its main building blocks and their interactions based on an extended version of the TOSCA orchestration language. The trusted third party is capable to perform a precise and exhaustive vulnerability assessment, without requiring the cloud provider and the cloud tenant to share critical configuration information between each other. After designing and formalizing this third party solution, we perform large series of experiments based on a proof-of-concept prototype in order to quantify its benefits and limits.

The popularity of Internet of Things (IoT) devices in smart homes has raised significant concerns regarding data security and privacy. Traditional machine learning (ML) methods for anomaly detection often require sharing sensitive IoT data with a central server, posing security and efficiency challenges. In response, this paper introduces FedGroup, a novel Federated Learning (FL) method inspired by FedAvg. FedGroup revolutionizes the central model’s learning process by updating it based on the learning patterns of distinct groups of IoT devices. Our experimental results demonstrate that FedGroup consistently achieves comparable or superior accuracy in anomaly detection when compared to both federated and non-federated learning methods. Additionally, Ensemble Learning (EL) collects intelligence from numerous contributing models, leading to enhanced prediction performance. Furthermore, FedGroup significantly improves the detection of attack types and their details, contributing to a more robust security framework for smart homes. Our approach demonstrates exceptional performance, achieving an accuracy rate of 99.64% with a minimal false positive rate (FPR) of 0.02% in attack type detection, and an impressive 99.89% accuracy in attack type detail detection.

Time-slotted Channel Hopping (TSCH) Media Access Control (MAC) was specified to target the Industrial Internet of Things needs. This MAC balances energy, bandwidth, and latency for deterministic communications in unreliable wireless environments. Building a distributed or autonomous TSCH schedule is arduous because the node negotiates cells with its neighbours based on queue occupancy, latency, and consumption metrics. The Minimal TSCH Configuration defined by RFC 8180 was specified for bootstrapping a 6TiSCH network and detailed configurations necessary to be supported. In particular, it adopts Routing Protocol for Low Power and Lossy networks (RPL) Non-Storing mode, which reduces the node’s network awareness. Dealing with unpredicted traffic far from the forwarding node is difficult due to limited network information. Anticipating this unexpected flow from multiple network regions is essential because it can turn the forwarding node into a network bottleneck leading to high latency, packet discard or disconnection rates, forcing RPL to change the topology. To cope with that, this work proposes a new mechanism that implements an RPL control message option for passing forward the node’s cell demand, allowing the node to anticipate the proper cell allocation for supporting the traffic originating by nodes far from the forwarding point embedded in Destination-Oriented Directed Acyclic Graph (DODAG) Information Object (DIO) and Destination Advertisement Object (DAO) RPL control messages. Implementing this mechanism in a distributed TSCH Scheduling developed in Contiki-NG yielded promising results in supporting unforeseen traffic bursts and has the potential to significantly improve the performance and reliability of TSCH schedules in challenging network environments.

Domain Generation Algorithms (DGAs) are often used for generating huge amounts of domain names to maintain command and control between the infected computer and the bot master. By establishing as needed a great number of domain names, attackers may mask their C2 servers and escape detection. Many malware families have switched to a stealthier contact approach. Therefore, the traditional methods become ineffective. Over the past decades, many researches have started to use artificial intelligence to create systems able to detect DGA in traffic, but these works do not use the same data to evaluate their models. This article proposes a comparative methodology to compare machine learning models based on unsupervised clustering and then applied this methodology to study the best models belonging to neural network methods and traditional machine learning methods to detect DGAs. We extracted 21 linguistic features based on the analysis of alphanumeric and n-gram, we studied the correlation between these features in order to reduce their number. We examine in detail those Machine learning algorithms and we discuss the drawbacks and strengths of each method with specific classes of DGA to propose a new switch case model that could be always reliable to detect DGAs.

Wireless Sensor Networks (WSNs) play a crucial role in diverse applications, encompassing environmental monitoring, healthcare, and industrial automation. However, these networks are susceptible to various security threats, underscoring the need for robust attack detection systems. In this paper, we propose a Stochastic Machine Learning-Based Attack Detection System for WSNs that leverages the synergy of Hidden Markov Models (HMMs) and Gaussian Mixture Models (GMMs). The proposed system employs Principal Component Analysis for dimensionality reduction in the WSN dataset, thereby retaining essential routing features while mitigating the number of variables. Additionally, iterative machine learning Expectation-Maximization is employed to train the HMMs and GMMs, empowering the system to accurately detect and classify malicious activities and erroneous routing data. To evaluate the system’s efficacy, a series of experiments were conducted, entailing variations in the parameters of both HMMs and GMMs. Notably, the findings underscore that the configuration comprising 3 HMMs and 4 GMMs surpasses other combinations, achieving an exceptional accuracy level of 94.55%. Furthermore, a comprehensive comparison is drawn between the proposed system and common machine learning classifiers. This analysis unequivocally highlights the system’s superiority in terms of accuracy and overall performance. Notable is the system’s exceptional performance in cross-validation, consistently achieving accuracies within the range of 0.96 to 0.98. The proposed Stochastic Machine Learning-Based Attack Detection System introduces a highly promising approach to fortify the security of WSNs. The amalgamation of rigorous experimentation, comparative analysis, and impressive results underscores its potential as an effective security enhancement tool.

We propose a theoretical analysis of a novel source routing scheme called XSR. XSR uses linear encoding operation to both (1) build the path labels of unicast and multicast data transfers; (2) perform fast computational efficient routing decisions compared to standard table lookup procedure without any packet modification all along the path. XSR specifically focuses on decreasing the computational complexity of forwarding operations. This allows packet switches (e.g, link-layer switch or router) to perform only simple linear operations over a binary vector label that embeds the path. We provide analytical proofs demonstrating that XSRs efficiently compute a valid unicast or multicast path label over any finite fields ({mathbb {F}}_{2^w}). Furthermore, we show that this path label can be used for both the forward and return unicast paths, unlike other source routing algorithms that require recomputing a label for the return path. Compared to recent approaches based on modular arithmetic, XSR computes the smallest label possible and presents strong scalable properties, allowing it to be deployed over any kind of core vendor or datacenter networks.

Task allocation is the mechanism which enables the allotment of sensing tasks to participating users in a mobile crowdsensing (MCS) environment. Task allocation plays a vital role in the management of resources in crowdsensed networks which deploy mobile participants or devices. While conventional task allocation techniques focus on maximizing profit for either the platform or the user, our proposed task allocation scheme, called Load Balanced Task Allocation (LBTA) is geared towards user-oriented task allocation in order to mainly address altruistic MCS campaigns in which participants voluntarily contribute towards a common goal such as in citizen science-based projects. This paper deals with the problem of task allocation using a load balanced approach while trying to maximize the allocation of tasks at the same time. For this, we propose and formulate the LBTA algorithm, which is an extension of a greedy algorithm. The proposed LBTA algorithm has been compared with a known algorithm and their relative performances have been analysed. Simulation results demonstrate that the proposed algorithm performs better than the baseline algorithm for time-dependent MCS systems that operate without a budget constraint, and comparatively better up to a certain budget for those systems with budgeting limitations.