Georgia Sakellari

A survey of mathematical models, simulation approaches and testbeds used for research in cloud computing

The first hurdle for carrying out research on cloud computing is the development of a suitable research platform. While cloud computing is primarily commercially-driven and commercial clouds are naturally realistic as research platforms, they do not provide to the scientist enough control for dependable experiments. On the other hand, research carried out using simulation, mathematical modelling or small prototypes may not necessarily be applicable in real clouds of larger scale. Previous surveys on cloud performance and energy-efficiency have focused on the technical mechanisms proposed to address these issues. Researchers of various disciplines and expertise can use them to identify areas where they can contribute with innovative technical solutions. This paper is meant to be complementary to these surveys. By providing the landscape of research platforms for cloud systems, our aim is to help researchers identify a suitable approach for modelling, simulation or prototype implementation on which they can develop and evaluate their technical solutions.

The Cognitive Packet Network

Current and future multimedia networks require connections under specific quality of service (QoS) constraints which can no longer be provided by the best-effort Internet. Therefore, ‘smarter’ networks have been proposed in order to cover this need. The cognitive packet network (CPN) is a routing protocol that provides QoS-driven routing and performs self-improvement in a distributed manner, by learning from the experience of special packets, which gather on-line QoS measurements and discover new routes. The CPN was first introduced in 1999 and has been used in several applications since then. Here we provide a comprehensive survey of its variations, applications and experimental performance evaluations.

Admission of QoS aware users in a smart network

Smart networks have grown out of the need for stable, reliable, and predictable networks that will guarantee packet delivery under Quality of Service (QoS) constraints. In this article we present a measurement-based admission control algorithm that helps control traffic congestion and guarantee QoS throughout the lifetime of a connection. When a new user requests to enter the network, probe packets are sent from the source to the destination to estimate the impact that the new connection will have on the QoS of both the new and the existing users. The algorithm uses a novel algebra of QoS metrics, inspired by Warshalls algorithm, to look for a path with acceptable QoS values to accommodate the new flow. We describe the underlying mathematical principles and present experimental results obtained by evaluating the method in a large laboratory test-bed operating the Cognitive Packet Network (CPN) protocol.

Controlling Access to Preserve QoS in a Self-Aware Network

Multimedia traffic and real-time applications created a need for network quality of service(QoS). This demand led to the development of autonomous networks that use adaptive packet routing in order to provide the best possible QoS. Admission control (AC) is a mechanism which takes those networks a step further in guaranteeing packet delivery even under strict QoS constraints. This paper describes a measurement-based admission control algorithm which decides whether a new connection can be served without affecting the existing users of the network, based on the multiple QoS metrics that the users of a self-aware network have specified. Our algorithm promises QoS throughout the lifetime of all accepted connections in the network. The impact that the new call will have, on the QoS of both the new and the existing users, is estimated by sending probe packets and monitoring the networks by exploiting its self-awareness. The decision of whether to accept a new call is made using a novel algebra of QoS metrics, inspired by Warshalls algorithm, which looks for a path with acceptable QoS values that can accommodate the new flow. In this paper we describe the underlying mathematical principles and present experimental results obtained by evaluating the method in a large laboratory test-bed operating the self-aware cognitive packet network (CPN) protocol.

Investigating the Tradeoffs between Power Consumption and Quality of Service in a Backbone Network

Energy saving in networks has traditionally focussed on reducing battery consumption through smart wireless network design. Recently, researchers have turned their attention to the energy cost and carbon emissions of the backbone network that both fixed and mobile communications depend on, proposing primarily mechanisms that turn equipments OFF or put them into deep sleep. This is an effective way of saving energy, provided that the nodes can return to working condition quickly, but it introduces increased delays and packet losses that directly affect the quality of communication experienced by the users. Here we investigate the associated tradeoffs between power consumption and quality of service in backbone networks that employ deep sleep energy savings. We examine these tradeoffs by conducting experiments on a real PC-based network topology, where nodes are put into deep sleep at random times and intervals, resulting in a continuously changing network with reduced total power consumption. The average power consumption, the packet loss and the average delay of this network are examined with respect to the average value of the ON rate and the ON/OFF cycle of the nodes.

A distributed admission control mechanism for multi-criteria QoS

Admission control based on multiple criteria has always been a desirable feature for computer networks. In the past, we presented such a system, but it was based on a centralised decision mechanism, which by itself constitutes a significant limitation. Here, we describe a decentralised version of the initial admission control algorithm, while still keeping the multiple-criteria aspect, with each user being able to specify their Quality of Service (QoS) metrics at the level of the individual. Our scheme decides whether a new call should be allowed to enter the network based on measurements of the QoS metrics on each link of the network before and after the transmission of probe packets. The decision is based on a novel algebra of QoS metrics, inspired by Warshalls algorithm that searches whether there is a feasible path to accommodate the new flow without affecting the existing users. The decision is made at each source node individually, based on either only personal information or also on information exchange among the nodes that are involved. The performance of the algorithm is evaluated in terms of QoS throughout the lifetime of all connections. The experiments presented in this paper were conducted in an actual laboratory test-bed of realistic topology and under highly congested circumstances. The results are particularly encouraging.

Cloud-Based Cyber-Physical Intrusion Detection for Vehicles Using Deep Learning

Detection of cyber attacks against vehicles is of growing interest. As vehicles typically afford limited processing resources, proposed solutions are rule-based or lightweight machine learning techniques. We argue that this limitation can be lifted with computational offloading commonly used for resource-constrained mobile devices. The increased processing resources available in this manner allow access to more advanced techniques. Using as case study a small four-wheel robotic land vehicle, we demonstrate the practicality and benefits of offloading the continuous task of intrusion detection that is based on deep learning. This approach achieves high accuracy much more consistently than with standard machine learning techniques and is not limited to a single type of attack or the in-vehicle CAN bus as previous work. As input, it uses data captured in real-time that relate to both cyber and physical processes, which it feeds as time series data to a neural network architecture. We use both a deep multilayer perceptron and recurrent neural network architecture, with the latter benefitting from a long-short term memory hidden layer, which proves very useful for learning the temporal context of different attacks. We employ denial of service, command injection and malware as examples of cyber attacks that are meaningful for a robotic vehicle. The practicality of computation offloading depends on the resources afforded onboard and remotely, and the reliability of the communication means between them. Using detection latency as the criterion, we have developed a mathematical model to determine when computation offloading is beneficial given parameters related to the operation of the network and the processing demands of the deep learning model. The more reliable the network and the greater the processing demands, the greater the reduction in detection latency achieved through offloading.

Computation offloading of a vehicle’s continuous intrusion detection workload for energy efficiency and performance

Computation offloading has been used and studied extensively in relation to mobile devices. That is because their relatively limited processing power and reliance on a battery render the concept of offloading any processing/energy-hungry tasks to a remote server, cloudlet or cloud infrastructure particularly attractive. However, the mobile device’s tasks that are typically offloaded are not time-critical and tend to be one-off. We argue that the concept can be practical also for continuous tasks run on more powerful cyber-physical systems where timeliness is a priority. As case study, we use the process of real-time intrusion detection on a robotic vehicle. Typically, such detection would employ lightweight statistical learning techniques that can run onboard the vehicle without severely affecting its energy consumption. We show that by offloading this task to a remote server, we can utilse approaches of much greater complexity and detection strength based on deep learning. We show both mathematically and experimentally that this allows not only greater detection accuracy, but also significant energy savings, which improve the operational autonomy of the vehicle. In addition, the overall detection latency is reduced in most of our experiments. This can be very important for vehicles and other cyber-physical systems where cyber attacks can directly affect physical safety. In fact, in some cases, the reduction in detection latency thanks to offloading is not only beneficial but necessary. An example is when detection latency onboard the vehicle would be higher than the detection period, and as a result a detection run cannot complete before the next one is scheduled, increasingly delaying consecutive detection decisions. Offloading to a remote server is an effective and energy-efficient solution to this problem too.

Energy-Aware Admission Control for Wired Networks

Overprovisioning and redundancy has contributed towards better survivability and performance in networks, but has led to inefficient use of energy. Proposals for energy aware networks of the near future aim to reduce the energy consumption by switching off or putting to sleep individual network devices. Here we propose a mechanism that is taking this concept once step further through the use of admission control. Admission control has been traditionally used in wired networks to control traffic congestion and guarantee quality of service. We propose a two-fold approach. First, an admission control mechanism delays the users that are projected to be the most energy demanding, and whose acceptance would require the turning on of devices. At the same time, an auto-hibernation mechanism regulates the rate at which machines are turned off due to inactivity. Collectively, the two mechanisms contribute towards energy saving by monitoring both at the level of entry in the network and at the level of active operation.

A Decentralised, Measurement-based Admission Control Mechanism for Self-Aware Networks

This paper presents a decentralised Admission Control (AC) algorithm, based on the centralised proposed in [1{4]. Our algorithm is a multiple criteria AC algorithm, where each user can specify the QoS metrics that interest him/her, and decides whether a new call should be allowed to enter the network based on measurements of the QoS metrics on each link of the network before and after the transmission of probe packets. Our algorithm will be brie°y described and we will present experimental results, conducted in a large laboratory test-bed, under highly congested circumstances.