Archived project

H2020 SELFNET

Goal: The SELFNET project will design and implement an autonomic network management framework to achieve self-organizing capabilities in managing network infrastructures by automatically detecting and mitigating a range of common network problems that are currently still being manually addressed by network operators, thereby significantly reducing operational costs and improving user experience. SELFNET explores a smart integration of state-of-the-art technologies in Software-Defined Networks (SDN), Network Function Virtualization (NFV), Self-Organizing Networks (SON), Cloud computing, Artificial intelligence, Quality of Experience (QoE) and Nextgeneration networking to provide a novel intelligent network management framework that is capable of assisting network operators in key management tasks: automated network monitoring by the automatic deployment of NFV applications to facilitate system-wide awareness of Health of Network metrics to have more direct and precise knowledge about the real status of the network; autonomic network maintenance by defining high-level tactical measures and enabling autonomic corrective and preventive actions against existing or potential network problems.

For more information: https://selfnet-5g.eu/

SELFNET is supported by the European Commission Horizon 2020 Programme under grant agreement number H2020-ICT-2014-2/671672

Methods: Software-defined Networking, Deep Learning, Network Function Virtualisation, deep packet inspection

Date: 1 June 2015 - 1 July 2018

Updates
0 new
9
Recommendations
0 new
2
Followers
0 new
33
Reads
0 new
648

Project log

Wei Jiang
added a research item
To meet the radical technical requirements specified by ITU-R IMT-2020, the fifth Generation (5G) mobile system will become more complicated and heterogeneous. It imposes a great challenge on today's network managing approaches, which are already costly, vulnerable, time-consuming and therefore inap-plicable to the 5G system. By applying machine learning, a possibility on autonomically self-organizing 5G networks is opened. With minimal human interventions, autonomic management can lower operational expenditure, improve user's experience and shorten time-to-market of new services. In this paper, the concept of intelligence slicing, a flexible and scalable framework for applying machine learning to enable self-organizing 5G networks, is proposed. The life-cycle management of intelligence slices, as well as intelligence domain that is defined as the effective area of a slice, are discussed. Moreover, a proof-of-concept experiment upon a wireless network test-bed is illustrated.
Enrique Chirivella Pérez
added a research item
This paper presents the design and prototype implementation of the SELFNET fifth‐generation (5G) mobile edge infrastructure. In line with the current and emerging 5G architectural principles, visions, and standards, the proposed infrastructure is established primarily based on a mobile edge computing paradigm. It leverages cloud computing, software‐defined networking, and network function virtualization as core enabling technologies. Several technical solutions and options have been analyzed. As a result, a novel portable 5G infrastructure testbed has been prototyped to enable the preliminary testing of the integrated key technologies and to provide a realistic execution platform for further investigating and evaluating software‐defined networking– and network function virtualization–based application scenarios in 5G networks.
Manuel Gil Pérez
added a research item
With the Fifth-Generation (5G) mobile networks set to arrive within the next years, this new generation will transform the industry with a profound impact on its customers as well as on the existing technologies and network architectures. Software-Defined Networking (SDN) and Network Functions Virtualization (NFV) will play key roles for the network operators as they prepare the migration to 5G, allowing them to quickly scale their networks. This paper presents a research work undertaken on this new paradigm of virtualized and programmable networks, aiming to address Self-Organizing Networks (SON) scenarios in a NFV/SDN context, focusing on detection and prediction of potential network and service anomalies. Towards this end, the performance management system performs aggregation, correlation and analysis of data gathered from the virtualized and programmable network elements. In particular, customized catalog-driven tools are developed, and the results show that they are able to successfully address these requirements. Current performance management platforms in production are designed for non-virtualized (non-NFV) and non-programmable (non-SDN) networks, and the knowledge gathered from this research brings some new understanding on how management platforms must evolve in order to be prepared for the upcoming next-generation mobile networks.
Mathias Strufe
added an update
You are interested in our Framework? Find and check-out all the single components in our git repository: https://github.com/selfnet-5g
 
Mathias Strufe
added an update
SELFNET consortium proudly present complete Demo videos of our three use cases Self-Protection, Self-Optimization and Self-Healing. Check out at our YouTube Channel:
 
Mathias Strufe
added a research item
In this paper, a wireless test bed for demonstrating intelligent management in virtualized network infrastructure is presented. To instantiate virtual networks on demand including Network Function Virtualization (NFV) OpenStack is applied. In addition, OpenAirInterface (OAI) is integrated into the test bed to support realistic wireless scenarios. Two virtual network environments, that demonstrate self-optimization to improve end-to-end Quality-of-Experience (QoE) and self-protection against distributed denial-of-service (DDoS) attack, are described.
Manuel Gil Pérez
added a research item
This paper reports the design of a complex use case oriented to manage self-organizing capabilities in 5G networks, tailored to be the main scenario during the final execution stage of the H2020 SELFNET project. The use case is devoted to demonstrating that the SELFNET framework is capable of monitoring large volumes of information from multiple and heterogeneous sensors, detecting diverse network problems through Health of Network (HoN) metrics reported by sensors, diagnosing the root causes of problems and deciding the best actions to counter them autonomously, and enforcing the required action plans with the most appropriate sensors and actuators to fulfil the desired operational tasks. Additional intelligence procedures are described to identify new root causes when previously unknown security-related problems come up, as well as the need to provide new cognitive capabilities with the aim of learning how to react next time when those problems arise.
Enrique Chirivella Pérez
added 2 research items
Ultra-High-Definition (UHD) video applications such as streaming are envisioned as a main driver for the emerging Fifth Generation (5G) mobile networks being developed worldwide. This paper focuses on addressing a major technical challenge in meeting UHD users' growing expectation for continuous high-quality video delivery in 5G hotspots where congestion is commonplace to occur. A novel 5G-UHD framework is proposed towards achieving adaptive video streaming in this demanding scenario to pave the way for self-optimisation oriented 5G UHD streaming. The architectural design and the video stream optimisation mechanism are described, and the system is prototyped based on a realistic virtualised 5G testbed. Empirical experiments validate the design of the framework and yield a set of insightful performance evaluation results.
Manuel Gil Pérez
added 2 research items
The next-generation 5G mobile networks are expected to bring a major shift on the management paradigm based on Network Function Virtualization (NFV) and Software-Defined Networking (SDN) compared with its precursor, 4G. Consequently, operators need to significantly change their network architectures, management mechanisms and business models to accommodate and address the 5G challenges. This paper contributes to advancing the operators' management capabilities in the monitoring and analytic domains, looking at the evolution of the existing performance management platform from a leading operator to a new SDN/NFV-enabled platform, in the context of the EU 5G project SELFNET. This work focuses on designing and prototyping the essential functionalities to perform real-time processing over network infrastructure data. This work devises a new Complex Event Processing (CEP) framework, a real-time framework for processing and aggregating big data using a dynamic rule-based approach. This CEP framework has been successfully implemented and deployed, with aggregation rules applied to a Self-Protection use case, which is able to provide in real-time information about detected botnets in the network. From a high-level perspective, this work brings some new understanding about the role of SDN/NFV network management tools for 5G network operators.
Mathias Strufe
added an update
Come and visit us at EuCNC2018 next week in Ljubljana, Slovenia and get first hand information about our project results! Booth #17! Looking forward to talk to you!
 
Manuel Gil Pérez
added 2 research items
5G mobile networks are pushing new dynamic and flexible scenarios that demand the automation and optimization of network management processes. In this sense, Self-Organizing Networks (SON) arose to evolve from traditional manual management towards fully autonomic and dynamic processes. Due to the large volumes of data generated in 5G networks, functionalities and capabilities of SON require efficient processes and resource optimization techniques. In particular, self-protection is a critical capability of SON focused on protecting the network resources in a flexible and autonomic way. To achieve self-protection, SON perform different processes ranging from the monitoring of network communications to the analysis, detection, and mitigation of cyber-attacks. In this article, we propose an architecture that combines the Software Defined Networking and Network Functions Virtualization technologies to optimize the usage of network resources for monitoring services. A use case based on botnet detection in 5G networks shows how our architecture ensures the provision of monitoring services in managing self-protection scenarios. Additionally, we describe a set of experiments that confirm the best time calculated by our solution to deploy or reconfigure monitoring and detection services. These experiments consider different aspects like the number of zombies shaping the botnet, their mobility, or network traffic.
Fog and mobile edge computing (MEC) will play a key role in the upcoming fifth generation (5G) mobile networks to support decentralized applications, data analytics and management into the network itself by using a highly distributed compute model. Furthermore, increasing attention is paid to providing user-centric cybersecurity solutions, which particularly require collecting, processing and analyzing significantly large amount of data traffic and huge number of network connections in 5G networks. In this regard, this paper proposes a MEC-oriented solution in 5G mobile networks to detect network anomalies in real-time and in autonomic way. Our proposal uses deep learning techniques to analyze network flows and to detect network anomalies. Moreover, it uses policies in order to provide an efficient and dynamic management system of the computing resources used in the anomaly detection process. The paper presents relevant aspects of the deployment of the proposal and experimental results to show its performance.
Julian Ahrens
added a research item
Efficient network management is one of the key challenges of the constantly growing and increasingly complex wide area networks (WAN). The paradigm shift towards virtualized (NFV) and software defined networks (SDN) in the next generation of mobile networks (5G), as well as the latest scientific insights in the field of Artificial Intelligence (AI) enable the transition from manually managed networks nowadays to fully autonomic and dynamic self-organized networks (SON). This helps to meet the KPIs and reduce at the same time operational costs (OPEX). In this paper, an AI driven concept is presented for the malfunction detection in NFV applications with the help of semi-supervised learning. For this purpose, a profile of the application under test is created. This profile then is used as a reference to detect abnormal behaviour. For example, if there is a bug in the updated version of the app, it is now possible to react autonomously and roll-back the NFV app to a previous version in order to avoid network outages.
Mathias Strufe
added an update
Not yet registered yet? Why then wait? Save your seat now!
WHAT: EFFECTIVE NETWORK MANAGEMENT IN 5G
WHERE: Heidelberg (Germany)
WHEN: 24 May 2018 10:30 – 15:30
(Deadline: 30th of April)
 
Manuel Gil Pérez
added a research item
The upcoming fifth generation (5G) mobile technology, which includes advanced communication features, is posing new challenges on cybersecurity defense systems. Although innovative approaches have evolved in the last few years, 5G will make existing intrusion detection and defense procedures become obsolete in case they are not adapted accordingly. In this sense, this article proposes a novel 5G-oriented cyberdefense architecture to identify cyberthreats in 5G mobile networks efficient and quickly enough. For this, our architecture uses deep learning techniques to analyze network traffic by extracting features from network flows. Moreover, our proposal allows adapting, automatically, the configuration of the cyberdefense architecture in order to manage traffic fluctuation, aiming both to optimize the computing resources needed in each particular moment and to fine tune the behavior and performance of analysis and detection processes. Experiments using a well-known botnet dataset depict how a neural network model reaches a sufficient classification accuracy in our anomaly detection system. Extended experiments using diverse deep learning solutions analyze and determine their suitability and performance for different network traffic loads. The experimental results show how our architecture can selfadapt the anomaly detection system based on the volume of network flows gathered from 5G subscribers’ User Equipments (UE) in real-time and optimizing the resource consumption.
Hans D. Schotten
added 2 research items
In the context of Fifth Generation (5G) mobile networks, the concept of "slice as a service" (SaaS) promotes mobile network operators to flexibly share infrastructures with mobile service providers and stakeholders. However, it also challenges with an emerging demand for efficient online algorithms to optimize the request-and-decision-based inter-slice resource management strategy. Based on genetic algorithms, this paper presents a novel online optimizer that efficiently approaches towards the ideal slicing strategy with maximized long-term network utility. The proposed method requires no a priori knowledge about the traffic/utility models, while providing solid effectiveness, good robustness against non-stationary service scenarios, and high scalability.
The fifth generation (5G) mobile telecommunication network is expected to support multi-access mobile edge computing (MEC), which intends to distribute computation tasks and services from the central cloud to the edge clouds. Towards ultraresponsive, ultra-reliable and ultra-low-latency MEC services, the current mobile network security architecture should enable more decentralized approach for authentication and authorization process. This paper proposes a novel distributed authentication architecture that supports flexible, intelligent and low-cost local authentication with the awareness of network elements, e.g. user equipment, virtual network functions etc., context information.
Manuel Gil Pérez
added a research item
Cybersecurity defense systems have evolved, in fact, new intrusion detection systems (IDS) allows to identify cyber-threats that may have only recently gone unnoticed. However, new protection challenges are raised by the new upcoming fifth generation (5G) mobile technology. The new advanced features of 5G will make that existing detection procedures become obsolete in case they are not adapted accordingly. This paper propose a 5G-oriented architecture to efficiently and quickly conduct the analysis of network flows to identify cyber-threats in 5G mobile networks, by making use of deep learning techniques. Intensive experiments are also included so as to analyze the robustness of the proposed inspection capabilities as well as to determine how many network flows, gathered from 5G subscribers' User Equipments (UE), we can inspect in real-time.
Mathias Strufe
added an update
10th SELFNET plenary meeting started in Aveiro with the feedback from the last review.
 
Manuel Gil Pérez
added 2 research items
Preserving the privacy of the users' information should be an essential requirement in information management systems. The mobility provided by context-aware services has increased the complexity of ensuring this challenge by allowing users to obtain, share, and provide information at anytime and anywhere. Addressing this challenge requires automatic mechanisms that allow users to control their information at real-time and on demand. In order to achieve these requirements, we proposed several context-aware solutions that allow users to manage the privacy of their information through policies. The privacy-policies managed by our solutions let users decide at real-time what, where, when, how, to whom, and at which level of precision they want to reveal their information.
Can we trust ICT (Information & Communication Technology) systems? Every single day a handful of previously unknown security vulnerabilities on these environments are published, dangerously feeding the lack of trust feeling that many end users already exhibit with respect to ICT. In order to disrupt and even invert such a perilous tendency (hindering the wide adoption of ICT and all its associated benefits), a number of research challenges in the field of cyber security need to be addressed. This paper presents some of these key challenges, offering initial thoughts on how to tackle each of them.
Manuel Gil Pérez
added a research item
Managing network resources in charge of monitoring network flows generated by potential attackers is a complex process that should be carried out in an automatic way. Self-Organizing Networks should be able to guarantee the provision of flow-based monitoring services to later detect cyber-attacks and react against them automatically. In this sense, we present a policy-based system oriented to the SDN paradigm in charge of managing automatically the network resources to ensure the provision of flow-based monitoring services. Using our proposal, network administrators define policies to switch-on/off, balance, or create/dismantle physical and virtual network resources considering the users' mobility, the network statistics, and the location of the infrastructure.
Manuel Gil Pérez
added a research item
Botnets are one of the most powerful cyberthreats affecting continuity and delivery of existing network services. Detecting and mitigating attacks promoted by botnets become a greater challenge with the advent of 5G networks, as the number of connected devices with high mobility capabilities, the volume of exchange data, and the transmission rates increase significantly. Here, a 5G-oriented solution is proposed for proactively detecting and mitigating botnets in a highly dynamic 5G network. 5G subscribers’ mobility requires dynamic network reconfiguration, which is handled by combining software-defined network and network function virtualization techniques.
Mathias Strufe
added an update
Last preperation meeting before the 2nd review tomorrow in Herzliya/Israel started right now.
 
Alberto Huertas
added 2 research items
5G mobile networks are pushing new dynamic and flexible scenarios that require the automation of the management processes performed by network administrators. To this end, Self-Organizing Networks (SON) arose with the goal of moving from traditional manual management processes towards an automatic and dynamic perspective. The orchestration of the monitoring services is an essential task to conduct self-configuration, self-healing, and self-optimization processes required by SONs. In this context, we propose a solution that efficiently orchestrates the monitoring services by managing the network resources automatically. In particular, we propose a 5G-oriented architecture that integrates the Software Defined Networking (SDN) and Network Functions Virtualization (NFV) technologies to monitor and orchestrate the whole life-cycle of monitoring services considering information of the network control plane.
The Big Data age is characterized by the explosive increase of data managed by electronic systems. Healthcare Information Management systems are aware of this situation having to adapt services and procedures. This, along with the fact that the proliferation of mobile devices and communications has also promoted the use of context-aware services ubiquitously accessible, means that protecting the privacy of the patients’ information is an even greater challenge. To address this issue, a mechanism that allows patients to manage and control their private information is required. We propose the preservation of patients’ privacy in a health scenario through a multicontext-aware system called h-MAS (health-related multicontext-aware system). h-MAS is a privacy-preserving and context-aware solution for health scenarios with the aim of managing the privacy of the users' information in both intra- and inter-context scenarios. In a health scenario, h-MAS suggests a pool of privacy policies to users, who are aware of the health context in which they are located. Users can update the policies according to their interests. These policies protect the privacy of the users' health records, locations, as well as context-aware information being accessed by third parties without their consent. The information on patients and the health context is managed through semantic web techniques, which provide a common infrastructure that makes it possible to represent, process, and share information between independent systems more easily.
Manuel Gil Pérez
added a research item
This is the first white paper of the 5G PPP Security Working Group. Launched in early April 2016 and led by 5G-ENSURE, this WG encompasses all Phase 1 projects either active and/or interested in 5G security. The largest contributions are from the two projects most active in security (5G-ENSURE and CHARISMA), but most of the 5G PPP Phase 1 Projects have joined the 5G PPP Security WG and provided inputs to this white paper (namely SELFNET, VirtuWind, COGNET, 5G-NORMA, Speed-5G, 5GEx, SONATA). This white paper describes the 5G PPP Security Landscape of Phase 1 projects, covering the scope in 5G PPP Phase 1 Projects with specific reference to 5G Security. The objective of this white paper is thus twofold: first get the reader acquainted with 5G Security the way it has been addressed through Phase 1 in terms of the "What" and "Why" but also, and probably most importantly, pave the way for Phase 2 Projects so they can leverage the achievements resulting from this first phase. While this white paper has been produced in the context of the 5G PPP Phase 1, it is hoped that it can serve as a reference document in other contexts.
Mattia Zago
added a research item
Despite the important effort dedicated to maximise the security of network infrastructures and the services provided on top of them some major cyberthreats, such as botnets, persist. If classic botnets are nowadays difficult to track due to the amount of devices that may potentially be subjugated, at the dawn of the 5G era, with an even increased number of devices powered by ultra-wide broadband, this problem seems to be even more relevant. In this context, the paper at hand describes a system designed considering the integration of multiple machine-learning algorithms with some other technologies like Software-Defined Networks (SDN) and Network Virtual Functions (NFV) that are usually being considered as enablers in 5G. The intention of this system is to analyse, detect and react against existing and new coming cyber threats, including botnets.
Marian Ulbricht
added a research item
As line-speeds and packet losses are sufficient well for most applications, reduction of latency and jitter are gaining in importance. We introduce and discuss the architecture of a novel networking device providing low-latency switching and routing. It integrates an up-to-date FPGA with a standard x86-64 processor and is targeted to Time-Sensitive Networking (TSN) and machine-to-machine communication (M2M). First results show a cut-through latency of 2 - 2.5 µs for its 12 Gigabit Ethernet ports and full line rate packet processing. It features frequency synchronization across networks and is easily extendable, enabling researchers to build experiments in areas like industrial, automotive, and 5G mobile access networks, with highest precision, repeatability, and ease.
Mathias Strufe
added an update
Just finished our 8th plenary meeting. During the three days with status updates on the three use cases and their implementation progress, the consortium defined a good roadmap for the 2nd review in september this year. Furthermore in parallel sessions partners could define the details needed for the overall integration of all SELFNET modules in one testbed.
 
Mathias Strufe
added a research item
The maintenance and management for the current Fourth Generation (4G) networks are still in a manual and semi-automatic manner, which are costly and time-consuming. This imposes a great challenge on the network management of heterogeneous, software-defined and virtualized Fifth Generation (5G) systems. With the advent of network intelligence, a possibility on intelligent management is opened for the 5G system. Without interventions of network administrators, the novel approach can autonomically deal with network failures, cyber-attacks and inefficient resource utilization, which in turn can lower operational expenditure, improve user's experience and reduce time-to-market of new services. In this paper, the reference architecture, functionality, closed-loop control, enabling algorithms of the network intelligence are presented. An intelligent 5G test-bed is set up and the experimental results verify the feasibility and effectiveness.
Mathias Strufe
added a research item
The Fifth Generation (5G) system is envisioned to become more complicated, which imposes a great challenge on network management. But today’s manual and semi-automatic managing approaches are already costly and time-consuming.Thanks to new technologies, such as software-defined networking and network function virtualization, a possibility of autonomicmanagement is opened for the 5G system. In this context, a novel management framework in software-defined and virtualized network is proposed by the SELFNET project so as to lower operational expenditure, improve user’s experience and reduce time-to-market of services. As a key part of this framework, an Autonomic Manager (AM) is designed to provide the network intelligence by means of machine learning techniques. In this paper, the functionality and mechanism of the AM, as well as an intelligence control loop, are presented. A 5G test-bed established to demonstrate the autonomic management, along with some results on collecting and selecting network metrics, are illustrated.
Manuel Gil Pérez
added a research item
5G networks are envisioned to support substantially more users than the current 4G does as a direct consequence of the anticipated large diffusion of Machine-2-Machine (M2M) and Internet of Things (IoT) interconnected devices, often with significantly higher committed data rates than general bandwidth currently available into Long Term Evolution (LTE) and broadband networks. The expected large number of 5G subscribers will offer new opportunities to compromise devices and user services, which will allow attackers to trigger much larger and effective cyber-attacks. Significant advances in network management automation are therefore needed to manage 5G networks and services in an efficient, scalable, and effective way while protecting users and infrastructures from a wide plethora of advanced security threats. This paper presents a novel self-organized network management approach for 5G mobile networks where autonomic capabilities are tightly combined with Network Functions Virtualization (NFV) and Software-Defined Networking (SDN) technologies so as to provide an effective detection and mitigation of cyber-attacks.
Manuel Gil Pérez
added a research item
Cyber-attacks can affect both continuity and delivery of services. Such attacks threaten to directly affect subscribers of emerging 5G services. To protect them, 5G PPP project SELFNET has developed solutions particularly aimed at botnet attacks.
Mathias Strufe
added 21 project references
Mathias Strufe
added a research item
Traditional networks are characterized by wasting considerable amount of energy that could be reduced drastically. The challenge of energy saving should be managed efficiently, where the mobility of users and services are nominated to play a significant role as well as the use of the Software Defined Networking (SDN) paradigm. Besides the network management supported by the SDN paradigm, we highlight the management of the network infrastructure at run-time, considering aspects like the energy efficiency. In this paper, we present an energy-aware and policy-based system oriented to the SDN paradigm, which allows managing the network infrastructure dynamically at run-time and on demand through policies. With these policies, any network using our solution will be able to reduce energy consumption by switching on/off its resources when they are inefficient, and creating virtualized network resources like proxies to reduce the network traffic. The experiments conducted demonstrate how the energy consumption is reduced when enforcing the proposed policies, considering aspects such as the number of base stations, their cell sizes, and the number of active devices in a given time, among other.
Mathias Strufe
added a project goal
The SELFNET project will design and implement an autonomic network management framework to achieve self-organizing capabilities in managing network infrastructures by automatically detecting and mitigating a range of common network problems that are currently still being manually addressed by network operators, thereby significantly reducing operational costs and improving user experience. SELFNET explores a smart integration of state-of-the-art technologies in Software-Defined Networks (SDN), Network Function Virtualization (NFV), Self-Organizing Networks (SON), Cloud computing, Artificial intelligence, Quality of Experience (QoE) and Nextgeneration networking to provide a novel intelligent network management framework that is capable of assisting network operators in key management tasks: automated network monitoring by the automatic deployment of NFV applications to facilitate system-wide awareness of Health of Network metrics to have more direct and precise knowledge about the real status of the network; autonomic network maintenance by defining high-level tactical measures and enabling autonomic corrective and preventive actions against existing or potential network problems.
For more information: https://selfnet-5g.eu/
SELFNET is supported by the European Commission Horizon 2020 Programme under grant agreement number H2020-ICT-2014-2/671672