Cybersecurity of Digital Service Chains

Data is the key driver for the digital economy. Besides its clear business meaning, this statement accounts for most of the innovations and new paradigms introduced by the software industry in the last decade. As a matter of fact, the most remunerative business today is not the software per se, but the possibility to create value-added services for specific domains: industry, smart city, smart grid, e-Health, multimedia, etc. The real competitive advantage in this scenario is given by the agility to implement ever new digital value chains that emerge, evolve, and dissolve much faster than ever. New computing models and software architecture have been progressively introduced to bring more agility in the creation and management of new digital services and products. The recurring buzzword that conveniently represents this attitude is "orchestration", meaning the capability to implement (semi-)autonomous systems that are able to evolve with self-properties (self-configuration, self-management, self-healing, self-protection, etc.). Concrete achievements in this respect consist of a number of management frameworks and interfaces for cyberphysical systems and telecommunication infrastructures, including TOSCA, ETSI NFV, and FIWARE. They actually allow us to compose digital resources from multiple domains (cloud, IoT, networks, data) into high-value services in a seamless way, without caring about technical details concerning hardware and software provisioning. The downside of this evolution is represented by cybersecurity aspects, which have not yet been addressed in a satisfactory way. Despite the effort in making software- defined systems ever more smart and autonomous, cybersecurity processes still largely depend on human skill and expertise. Relying on individuals' ability for hardening, verification of security properties, attack detection, and threat identification is no longer practical, and it is clearly an unacceptable practice, especially when critical infrastructures and large chains are involved. Motivated by this substantial imbalance between software management paradigms and cybersecurity models, the GUARD project has advocated the transition towards more agile security and privacy processes, which could follow the dynamics of modern digital infrastructures and services. The scope has extended to service integrity and data sovereignty, including, therefore, attack detection and data tracking aspects. The main objective is the introduction of similar models to those already used for software management, namely ones with the ability to orchestrate security capabilities in order to build advanced and agile detection and analytic processes. This book provides an overall review of the main concepts, architectures, technologies, and results from the GUARD project, covering both technical and non-technical aspects, i.e., legal and ethical issues.