Search

Major challenges of assisting passengers to safely and quickly escape from ships when an emergency occurs, include complex realistic features such as human behavior uncertainty, dynamic human traversal times, and the computation and communication delays in the systems that offer advice to users during an emergency. In this paper, we present simulations that examine the influence of these key features on evacuation performance in terms of evacuation time. The approach is based on our previously proposed lookup table-based ship passenger evacuation method, i.e., ANT. The simulation results we present show that delays in the users' reception of instructions significantly impair the effectiveness of the evacuation service. In contrast, behavior uncertainty has a weaker influence on the performance of the navigation method. In addition, these effects also vary with the extent of the behavior uncertainty, the dynamics of the traversal time distributions, and the delay in receiving directions. These findings demonstrate the importance of carefully designing evacuation systems for passenger ships in a way that takes into account all realistic features of the ship's indoor evacuation environment, including the crucial role of information technology., Comment: 6 pages, 9 figures, WF-IoT conference

This paper proposes a novel Self-Supervised Intrusion Detection (SSID) framework, which enables a fully online Deep Learning (DL) based Intrusion Detection System (IDS) that requires no human intervention or prior off-line learning. The proposed framework analyzes and labels incoming traffic packets based only on the decisions of the IDS itself using an Auto-Associative Deep Random Neural Network, and on an online estimate of its statistically measured trustworthiness. The SSID framework enables IDS to adapt rapidly to time-varying characteristics of the network traffic, and eliminates the need for offline data collection. This approach avoids human errors in data labeling, and human labor and computational costs of model training and data collection. The approach is experimentally evaluated on public datasets and compared with well-known {machine learning and deep learning} models, showing that this SSID framework is very useful and advantageous as an accurate and online learning DL-based IDS for IoT systems.

Cyberattacks are increasingly threatening networked systems, often with the emergence of new types of unknown (zero-day) attacks and the rise of vulnerable devices. While Machine Learning (ML)-based Intrusion Detection Systems (IDSs) have been shown to be extremely promising in detecting these attacks, the need to learn large amounts of labelled data often limits the applicability of ML-based IDSs to cybersystems that only have access to private local data. To address this issue, this paper proposes a novel Decentralized and Online Federated Learning Intrusion Detection (DOF-ID) architecture. DOF-ID is a collaborative learning system that allows each IDS used for a cybersystem to learn from experience gained in other cybersystems in addition to its own local data without violating the data privacy of other systems. As the performance evaluation results using public Kitsune and Bot-IoT datasets show, DOF-ID significantly improves the intrusion detection performance in all collaborating nodes simultaneously with acceptable computation time for online learning.

IoT Servers that receive and process packets from IoT devices should meet the QoS needs of incoming packets, and support Attack Detection software that analyzes the incoming traffic to identify and discard packets that may be part of a Cyberattack. Since UDP Flood Attacks can overwhelm IoT Servers by creating congestion that paralyzes their operation and limits their ability to conduct timely Attack Detection, this paper proposes and evaluates a simple architecture to protect a Server that is connected to a Local Area Network, using a Quasi Deterministic Transmission Policy Forwarder (SQF) at its input port. This Forwarder shapes the incoming traffic, sends it to the Server in a manner which does not modify the overall delay of the packets, and avoids congestion inside the Server. The relevant theoretical background is briefly reviewed, and measurements during a UDP Flood Attack are provided to compare the Server performance, with and without the Forwarder. It is seen that during a UDP Flood Attack, the Forwarder protects the Server from congestion allowing it to effectively identify Attack Packets. On the other hand, the resulting Forwarder congestion can also be eliminated at the Forwarder with "drop" commands generated by the Forwarder itself, or sent by the Server to the Forwarder., Comment: 8 pages, 13 figures

The IoT is vulnerable to network attacks, and Intrusion Detection Systems (IDS) can provide high attack detection accuracy and are easily installed in IoT Servers. However, IDS are seldom evaluated in operational conditions which are seriously impaired by attack overload. Thus a Local Area Network testbed is used to evaluate the impact of UDP Flood Attacks on an IoT Server, whose first line of defence is an accurate IDS. We show that attacks overload the multi-core Server and paralyze its IDS. Thus a mitigation scheme that detects attacks rapidly, and drops packets within milli-seconds after the attack begins, is proposed and experimentally evaluated.

Botnet attacks are a major threat to networked systems because of their ability to turn the network nodes that they compromise into additional attackers, leading to the spread of high volume attacks over long periods. The detection of such Botnets is complicated by the fact that multiple network IP addresses will be simultaneously compromised, so that Collective Classification of compromised nodes, in addition to the already available traditional methods that focus on individual nodes, can be useful. Thus this work introduces a collective Botnet attack classification technique that operates on traffic from an n-node IP network with a novel Associated Random Neural Network (ARNN) that identifies the nodes which are compromised. The ARNN is a recurrent architecture that incorporates two mutually associated, interconnected and architecturally identical n-neuron random neural networks, that act simultneously as mutual critics to reach the decision regarding which of n nodes have been compromised. A novel gradient learning descent algorithm is presented for the ARNN, and is shown to operate effectively both with conventional off-line training from prior data, and with on-line incremental training without prior off-line learning. Real data from a 107 node packet network is used with over 700,000 packets to evaluate the ARNN, showing that it provides accurate predictions. Comparisons with other well-known state of the art methods using the same learning and testing datasets, show that the ARNN offers significantly better performance.

This paper presents several novel algorithms for real-time cyberattack detection using the Auto-Associative Deep Random Neural Network, which were developed in the HORIZON 2020 IoTAC Project. Some of these algorithms require offline learning, while others require the algorithm to learn during its normal operation while it is also testing the flow of incoming traffic to detect possible attacks. Most of the methods we present are designed to be used at a single node, while one specific method collects data from multiple network ports to detect and monitor the spread of a Botnet. The evaluation of the accuracy of all the methods is carried out with real attack traces. These novel methods are also compared with other state-of-the-art approaches, showing that they offer better or equal performance, at lower computational learning and shorter detection times as compared to the existing approaches.

Classical Random Neural Networks (RNNs) have demonstrated effective applications in decision making, signal processing, and image recognition tasks. However, their implementation has been limited to deterministic digital systems that output probability distributions in lieu of stochastic behaviors of random spiking signals. We introduce the novel class of supervised Random Quantum Neural Networks (RQNNs) with a robust training strategy to better exploit the random nature of the spiking RNN. The proposed RQNN employs hybrid classical-quantum algorithms with superposition state and amplitude encoding features, inspired by quantum information theory and the brain's spatial-temporal stochastic spiking property of neuron information encoding. We have extensively validated our proposed RQNN model, relying on hybrid classical-quantum algorithms via the PennyLane Quantum simulator with a limited number of \emph{qubits}. Experiments on the MNIST, FashionMNIST, and KMNIST datasets demonstrate that the proposed RQNN model achieves an average classification accuracy of $94.9\%$. Additionally, the experimental findings illustrate the proposed RQNN's effectiveness and resilience in noisy settings, with enhanced image classification accuracy when compared to the classical counterparts (RNNs), classical Spiking Neural Networks (SNNs), and the classical convolutional neural network (AlexNet). Furthermore, the RQNN can deal with noise, which is useful for various applications, including computer vision in NISQ devices. The PyTorch code (https://github.com/darthsimpus/RQN) is made available on GitHub to reproduce the results reported in this manuscript.

Owing to the increasing frequency and destruction of natural and manmade disasters to modern highly-populated societies, emergency management, which provides solutions to prevent or address disasters, have drawn considerable research over the last few decades and become a multidisciplinary area. Because of its open and inclusive nature, new technologies always tend to influence, change or even revolutionise this research area. Hence, it is imperative to consolidate the state-of-the-art studies and knowledge to meet the research needs and identify the future research directions. The paper presents a comprehensive and systemic review of the existing research in the field of emergency management from both the system design aspect and algorithm engineering aspect. We begin with the history and evolution of the emergency management research. Then the two main research topics of this area, "emergency navigation" and "emergency search and rescue planning", are introduced and discussed. Finally, we suggest the emerging challenges and opportunities from system optimisation, evacuee behaviour modelling and optimisation, computing patterns, data analysis, energy and cyber security aspects., Comment: 33 pages, 3 figures

Artificial Neural Network (ANN) based techniques have dominated state-of-the-art results in most problems related to computer vision, audio recognition, and natural language processing in the past few years, resulting in strong industrial adoption from all leading technology companies worldwide. One of the major obstacles that have historically delayed large scale adoption of ANNs is the huge computational and power costs associated with training and testing (deploying) them. In the mean-time, Neuromorphic Computing platforms have recently achieved remarkable performance running more bio-realistic Spiking Neural Networks at high throughput and very low power consumption making them a natural alternative to ANNs. Here, we propose using the Random Neural Network (RNN), a spiking neural network with both theoretical and practical appealing properties, as a general purpose classifier that can match the classification power of ANNs on a number of tasks while enjoying all the features of a spiking neural network. This is demonstrated on a number of real-world classification datasets.

The Cloud computing paradigm has revolutionised the computer science horizon during the past decade and has enabled the emergence of computing as the fifth utility. It has captured significant attention of academia, industries, and government bodies. Now, it has emerged as the backbone of modern economy by offering subscription-based services anytime, anywhere following a pay-as-you-go model. This has instigated (1) shorter establishment times for start-ups, (2) creation of scalable global enterprise applications, (3) better cost-to-value associativity for scientific and high performance computing applications, and (4) different invocation/execution models for pervasive and ubiquitous applications. The recent technological developments and paradigms such as serverless computing, software-defined networking, Internet of Things, and processing at network edge are creating new opportunities for Cloud computing. However, they are also posing several new challenges and creating the need for new approaches and research strategies, as well as the re-evaluation of the models that were developed to address issues such as scalability, elasticity, reliability, security, sustainability, and application models. The proposed manifesto addresses them by identifying the major open challenges in Cloud computing, emerging trends, and impact areas. It then offers research directions for the next decade, thus helping in the realisation of Future Generation Cloud Computing., Comment: 51 pages, 3 figures

This paper proposes new nonnegative (shallow and multi-layer) autoencoders by combining the spiking Random Neural Network (RNN) model, the network architecture typical used in deep-learning area and the training technique inspired from nonnegative matrix factorization (NMF). The shallow autoencoder is a simplified RNN model, which is then stacked into a multi-layer architecture. The learning algorithm is based on the weight update rules in NMF, subject to the nonnegative probability constraints of the RNN. The autoencoders equipped with this learning algorithm are tested on typical image datasets including the MNIST, Yale face and CIFAR-10 datasets, and also using 16 real-world datasets from different areas. The results obtained through these tests yield the desired high learning and recognition accuracy. Also, numerical simulations of the stochastic spiking behavior of this RNN auto encoder, show that it can be implemented in a highly-distributed manner., Comment: 10 pages (a small edit to the abstract)

We assume that, within the dense clusters of neurons that can be found in nuclei, cells may interconnect via soma-to-soma interactions, in addition to conventional synaptic connections. We illustrate this idea with a multi-layer architecture (MLA) composed of multiple clusters of recurrent sub-networks of spiking Random Neural Networks (RNN) with dense soma-to-soma interactions, and use this RNN-MLA architecture for deep learning. The inputs to the clusters are first normalised by adjusting the external arrival rates of spikes to each cluster. Then we apply this architecture to learning from multi-channel datasets. Numerical results based on both images and sensor based data, show the value of this novel architecture for deep learning., Comment: 10 pages (a small edit to the abstract)

Most Software Defined Networks (SDN) traffic engineering applications use excessive and frequent global monitoring in order to find the optimal Quality-of-Service (QoS) paths for the current state of the network. In this work, we present the motivations, architecture and initial evaluation of a SDN application called Cognitive Routing Engine (CRE) which is able to find near-optimal paths for a user-specified QoS while using a very small monitoring overhead compared to global monitoring which is required to guarantee that optimal paths are found. Smaller monitoring overheads bring the advantage of smaller response time for the SDN controllers and switches. The initial evaluation of CRE on a SDN representation of the GEANT academic network shows that it is possible to find near-optimal paths with a small optimality gap of 1.65% while using 9.5 times less monitoring., Comment: This is a non-final version of the paper submitted to IEEE ICC 2016

This paper addresses the use of Big Data and machine learning based analytics to the real-time management of Internet scale Quality-of-Service Route Optimisation with the help of an overlay network. Based on the collection of large amounts of data sampled each $2$ minutes over a large number of source-destinations pairs, we show that intercontinental Internet Protocol (IP) paths are far from optimal with respect to Quality of Service (QoS) metrics such as end-to-end round-trip delay. We therefore develop a machine learning based scheme that exploits large scale data collected from communicating node pairs in a multi-hop overlay network that uses IP between the overlay nodes themselves, to select paths that provide substantially better QoS than IP. The approach inspired from Cognitive Packet Network protocol, uses Random Neural Networks with Reinforcement Learning based on the massive data that is collected, to select intermediate overlay hops resulting in significantly better QoS than IP itself. The routing scheme is illustrated on a $20$-node intercontinental overlay network that collects close to $2\times 10^6$ measurements per week, and makes scalable distributed routing decisions. Experimental results show that this approach improves QoS significantly and efficiently in a scalable manner., Comment: 9 pages

The increasingly wide application of Cloud Computing enables the consolidation of tens of thousands of applications in shared infrastructures. Thus, meeting the quality of service requirements of so many diverse applications in such shared resource environments has become a real challenge, especially since the characteristics and workload of applications differ widely and may change over time. This paper presents an experimental system that can exploit a variety of online quality of service aware adaptive task allocation schemes, and three such schemes are designed and compared. These are a measurement driven algorithm that uses reinforcement learning, secondly a "sensible" allocation algorithm that assigns jobs to sub-systems that are observed to provide a lower response time, and then an algorithm that splits the job arrival stream into sub-streams at rates computed from the hosts' processing capabilities. All of these schemes are compared via measurements among themselves and with a simple round-robin scheduler, on two experimental test-beds with homogeneous and heterogeneous hosts having different processing capacities., Comment: 10 pages, 9 figures

State-of-the-art emergency navigation approaches are designed to evacuate civilians during a disaster based on real-time decisions using a pre-defined algorithm and live sensory data. Hence, casualties caused by the poor decisions and guidance are only apparent at the end of the evacuation process and cannot then be remedied. Previous research shows that the performance of routing algorithms for evacuation purposes are sensitive to the initial distribution of evacuees, the occupancy levels, the type of disaster and its as well its locations. Thus an algorithm that performs well in one scenario may achieve bad results in another scenario. This problem is especially serious in heuristic-based routing algorithms for evacuees where results are affected by the choice of certain parameters. Therefore, this paper proposes a simulation-based evacuee routing algorithm that optimises evacuation by making use of the high computational power of cloud servers. Rather than guiding evacuees with a predetermined routing algorithm, a robust Cognitive Packet Network based algorithm is first evaluated via a cloud-based simulator in a faster-than-real-time manner, and any "simulated casualties" are then re-routed using a variant of Dijkstra's algorithm to obtain new safe paths for them to exits. This approach can be iterated as long as corrective action is still possible., Comment: Submitted to PerNEM'15 for review

Mobile networks are vulnerable to signalling attacks and storms that are caused by traffic patterns that overload the control plane, and differ from distributed denial of service (DDoS) attacks in the Internet since they directly attack the control plane, and also reserve wireless bandwidth without actually using it. Such attacks can result from malware and mobile botnets, as well as from poorly designed applications, and can cause service outages in 3G and 4G networks which have been experienced by mobile operators. Since the radio resource control (RRC) protocol in 3G and 4G networks is particularly susceptible to such attacks, we analyze their effect with a mathematical model that helps to predict the congestion that is caused by an attack. A detailed simulation model of a mobile network is used to better understand the temporal dynamics of user behavior and signalling in the network and to show how RRC based signalling attacks and storms cause significant problems in the control plane and the user plane of the network. Our analysis also serves to identify how storms can be detected, and to propose how system parameters can be chosen to mitigate their effect., Comment: Submitted to the IEEE TETC special issue on "Emerging topics in Cyber Security"

Voice over IP (VOIP) has strict Quality of Service (QoS) constraints and requires real-time packet delivery, which poses a major challenge to IP networks. The Cognitive Packet Network (CPN) has been designed as a QoS-driven protocol that addresses user-oriented QoS demands by adaptively routing packets based on online sensing and measurement. This paper presents our design, implementation and evaluation of a "Voice over CPN" system where an extension of the CPN routing algorithm has been developed to support the needs of voice packet delivery in the presence of other background traffic flows with the same or different QoS requirements., Comment: 8 pages, 4 figures