The ever-increasing complexity of communication networks demands ever better solutions for Network Management Automation (NMA). Over the last decade, teams at Nokia Bell Labs have developed multiple solutions for NMA. The recent solutions focus on the application of machine learning for those challenges, to evolve the networks to what may be termed Cognitive Autonomous Networks (CAN). The ideas underlying this evolution have been compiled into a book (with the same title) that will be published by Wiley in the summer of this year 2020 (see ).
This tutorial will summarize the hypothesis of the book as well as the major ideas presented therein. We begin the discussion with a management view of communication networks and their evolution, from which we identify the existing and the evolving management challenges and the need for advanced concepts in Network Management automation. We then review the state of the art in network management automation by summarizing the 3GPP standardized SON solutions and the concepts behind their implementation. This provides a basis for the new ideas, where we motivate the possibility of applying advanced cognitive concepts towards network management. Therein, we attempt to differentiate cognition from self-organization, highlighting the (quasi-)orthogonal components in cognition, i.e. we describe the features which an entity must possess for it to be considered cognitive. We then discuss the different means through which cognitive capabilities can be added to networks. Next, we present some work from researchers at Nokia Bell Labs that demonstrates examples of the application of cognitive techniques towards NM. Finally, we discuss the open questions and challenges that need to be addressed over the next few years in order to realize the fully Cognitive Autonomous Networks.
 Stephen S. Mwanje, Christian Mannweiler, “Towards Cognitive Autonomous Networks: Network Management Automation for 5G and Beyond”, Wiley 2020, to appear.
Stephen S. Mwanje is a Senior Research Engineer in the “Network Automation” team at Nokia Bell Labs in Munich, Germany. He is working on the application of machine learning for network automation and the corresponding system challenges of dealing with non-deterministic multi-agent systems that result from deploying multiple learning agents. He leads the team’s research activities on development of network intelligence with a special focus on its use in Radio Access Networks. Stephen earned his PhD/Dr.-Ing. degree (with summa cum laude) at the Ilmenau University of Technology, in Germany. His thesis on multi agent coordination of cognitive self-organizing network functions proposed the first approach towards the operation of multiple cognitive functions in cellular networks. He holds a BSc. Electrical Engineering from Makerere University in Kampala, Uganda and MS Electrical Engineering from the University of Rochester in Rochester, NY. He has 8 years experience in mobile network operations, where he managed various projects on network planning, deployment, and optimization; microwave radio engineering and spectrum/interference management as well as fiber-optic network planning, deployment and operations. He intermittently taught multiple undergraduate courses in electrical engineering and informatics between 2005 and 2011. He is (co-) author of numerous articles and papers on automation in mobile network management and lead co-editor and author of the book titled “Towards Cognitive Autonomous Networks - Network Management Automation for 5G and beyond”.
Compounding operational complexity and cost, diverging service requirements and exploding degrees of freedom in hybrid terrestrial and aerial architecture being conceived for 6G, hinge the technical and financial viability of future mobile networks on achieving zero touch automation. However, despite success in other domains, in mobile networks attempts to leverage AI for enabling automation remain hampered by two fundamental challenges: 1) Sparsity of the training data; 2) Hyper parameterization. These challenges remain largely uncovered by the existing tutorials/literature on AI for mobile networks. The goal of this tutorial is to first introduce the zero-touch deep automation framework and then provide case study based in-depth analysis of the multi-faceted implications of the data sparsity and hyper-parameterization on the practical performance of AI based solutions for mobile networks. The tutorial then delves into selected promising approaches for addressing the two problems. Some of the approaches to be discussed to address the sparsity challenge include, one shot learning, inductive transfer learning, transductive transfer learning, unsupervised transfer learning, leveraging different types of network geometries, generative adversarial networks and novel realistic synthetic data generation methods. To address the hyper-parameterization problem state of the art techniques such as AutoML, Neuro-Evolutionary Algorithms, Deep Reinforcement Learning, Federated Learning and Bayesian optimization will be discussed through case studies. The tutorial will conclude with identification of new practical research problems and potential opportunities therein to trigger the much-needed research effort to enable zero touch automation in emerging mobile networks.
Dr. Ali Imran is founding director of AI4Networks Research Centre (www.ai4networks.com) at the University of Oklahoma. AI4networks is the first academic centre in to be exclusively created for research on zero touch deep network automation. The centre is host to TurboRAN (http://bsonlab.com/TurboRAN/ )-a purpose-built cellular testbed for enabling experimental research on zero touch automation–and numerous multinational R&D projects on AI for wireless networks. Dr. Imran is also co-founder of a start-up AISON (www.aison.co ) that has launched world’s first deep AI enabled RAN automation and performance optimization solution currently being evaluated by several operators around the world for its game changing gains over current SON paradigm. Dr Imran’s research on network automation has played pioneering role in this area and has been supported by over $4M in nationally and internationally competitive research grants. On this topic, he has published over 100 refereed journal and conference papers and has several patents granted and pending. His work includes some of the most influential publications in the area of mobile network automation. The impact of his work on network automation has been recognized by several prestigious awards such as VPR Outstanding International Impact Award at the University of Oklahoma, 2018, IEEE Green ICT YP International award 2017, and best paper award IEEE CAMAD 2013. In 2019 he has been named William H. Barkow Presidential Professor at the University of Oklahoma for his contributions to this field. Dr. Imran is routinely invited to serves as an advisor to key stakeholder in cellular network eco-system and as a speaker and a panellist on international industry fora and academic conferences on this topic. He is an Associate Fellow of Higher Education Academy (AFHEA), UK; president of ComSoc Tulsa Chapter; Senior Member IEEE, Member of Advisory Board for Special Technical Community on Big Data at IEEE Computer Society, and board member of ITERA. For more detailed bio of Dr. Imran see: www.ali-imran.org.
Dr. Muhammad Imran received his M.Sc. (Distinction) and Ph.D. degrees from Imperial College London, UK, in 2002 and 2007, respectively. He is currently a Chair Professor at the University of Glasgow and visiting Professor at 5GIC Surrey. He is the founding director of the Communications Sensing and Imaging Research labs and a regular invited speaker on several 5G related talks, industrial panels and policy events for wireless communication technology. He has a global collaborative research network spanning both academia and key industrial players in the field of wireless communications. He has led a number of multimillion pounds international research projects encompassing the areas of Internet of Things (IoT), energy efficiency, fundamental performance limits, sensor networks and self-organizing cellular networks. He led the physical layer research for 5G innovation centre at Surrey (an outdoor cellular testbed developed at University of Surrey with a grant of above $50m). He has supervised 40+ successful Ph.D. graduates and published over 400 peer-reviewed research papers including more than 30 IEEE transactions. He is associate editor of IEEE Transactions on Communications, IEEE Access, IEEE Communications Letters and guest editor of several special issues in IEEE journals. He is a chair for several tracks in highly reputed international conferences and workshops including forthcoming IEEE ICC 2020 (co-chair for Next Generation Networking Symposium). He has been awarded IEEE Comsoc’s Fred Ellersick award 2014 and FEPS Learning and Teaching award 2014. He has also been shortlisted for Wharton-QS Stars Reimagine Education Awards 2014. He is a senior member of IEEE, Fellow of IET and a Senior Fellow of Higher Education Academy (SFHEA), UK.
Smart contracts enable automatic enforcement and execution of contract terms. When used together with blockchains, certain common criteria on immutability, and secure access by the stakeholders are satisfied by the infrastructure. Such smart contracts can be covering various domains, from traditional business-to-business agreements to controlling processes for the Industrial Internet of Things (IIoT). There are interesting, business-driven target areas within the Industrial IoT domain, including sectors such as supply chain (including manufacturing, transportation and logistics), maintenance, energy trading, grids, and even healthcare. When compared to consumer IoT, these systems have special requirements: certain level of real-time, security, engineering complexity, multi-stakeholder visibility, fast transaction and asset traceability. While the Distributed Ledger Technology (DLT) already addresses some areas of these (such as multi-stakeholder visibility or asset traceability), Blockchain Technology (BCT) provides additional value for security, building trust, and reducing cost while accelerating transactions of service agreements.
This tutorial aims to reveal the opportunities and challenges for smart service contracts as well as presenting real-life examples. First it provides an overview and definitions smart contracts executed with BCT. Next, it describes some special requirements of the Industrial IoT domain together with ideas of utilizing BCT to cover these needs. While discussing benefits, the tutorial reveals some drawbacks as well. The tutorial provides insights on various use-cases of employing BCT and smart service contracts in healthcare, electricity trading, production, asset tracking or proactive maintenance. Regarding implementations, the tutorial helps participants entering in Ethereum- and Hyperledger Fabric-based solutions.
Pal Varga, PhD, holds an associate professor position at BME, where he teaches various subjects both in English and in Hungarian. He is the main lecturer for “Infocommunications”, and “IoT frameworks and industrial applications” for BSc and MSc students respectively, and "5G Networks, Services and their Synergies with Industrial IoT" for PfD students. On one side, his research interest includes network- and service management, network monitoring (mobile core, as well as high-speed networks), bottleneck detection, anomaly detection, QoS, QoE issues (especially metrics for multimedia), Distributed Denial of Service attack detection and mitigation, hardware accelerated intrusion detection and prevention systems. On the other side, he is very active in the Industrial IoT community, where his research covers IoT frameworks, interoperability and integrability issues, heterogeneous IoTsystems, protocol translation, service oriented architectures, service discovery, end-point management, energy trading systems, Industrie4.0 use-cases (mainly production plant management and asset tracking), IoT security, IoT lifecycle management, smart service contracts, and Blockchains for IIoT.
Ferenc Nandor Janky currently is a PhD student at Budapest University of Technology and Economics (BME) where his thesis research topic is around smart contracts, as well as process and life-cycle modelling in Industrial IoT frameworks. He graduated with a Master's in Electrical Engineering from BME with a specialization in Incofommuncation Systems in 2013. He has several years industrial experience gained at various telecommunications companies like Vodafone, AITIA International Inc., Ericsson. Beside of the PhD studies he is currently working in the financial industry developing low-latency trading applications.
IPFS resembles past and present efforts to build and deploy Information-Centric Networking approaches to content storage, resolution, distribution and delivery. IPFS and libp2p , which is the modular network stack of IPFS, are based on name-resolution based routing. The resolution system is based on Kademlia DHT and content is addressed by flat hash-based names. IPFS sees significant real-world usage, with over 250,000 daily active network nodes, millions of end users and wide adoption by several other projects in the Decentralised Web space, but not only. An adjacent project to IPFS, which was also masterminded and is also being developed withinProtocol Labs (the umbrella company of IPFS and libp2p) is Filecoin. Filecoin is a cryptocurrency that supports a decentralised storage and delivery network. Storage and retrieval miners are rewarded according to their contribution to the network and the mechanics of filecoin secure the network against malicious activity.
Dr. Yiannis Psaras is a Research Scientist at Protocol Labs. He is part of the Resilient Networks Lab where, together with David, he is working on identifying and addressing future challenges that the IPFS and libp2p protocols will face. He is particularly interested in content routing design and optimization, the performance of libp2p’s pubsub protocol, content naming and generally architectural extensions to support the resilience and growth of the IPFS network. Before joining Protocol Labs he has spent 10+ years in academia, the majority of them as a Lecturer at University College London, where he investigated resource management techniques for current and future networking architectures with particular focus on “function-centric networks” to realise distributed and decentralised edge computing. He held a prestigious EPSRC Early Career Fellowship in the area of “content-oriented and service-centric edge-computing architectures”. He has also been heavily involved in the effort to shift the Internet towards an Information-Centric Networking environment.