My course on WSN and VANET Security
This course is included in the Doctoral Program in Information and Communication Technology at the University of L'Aquila. It deals with the security management principles - according to ISO 31000 Risk Management - and cyber risk quantification techniques applicable to a special class of ad hoc networks, namely Wireless Sensor Networks (WSN) and Vehicular Ad hoc NETworks
(VANET). VANETs can be futher splitted into inter-vehicular and intra-vehicular ad hoc networks.
Ad hoc networks can be considered one of the underlying technologies to enable Operational Technologies (OT) and currently there is a process of management convergence between ICT and OT where typical cyber threats and risks in ICT domain are passing into OT domain. This leads to the application of the security management frameworks as well as NIS and NIS2 security requirements also to OT domain.
Hence course objective is introducing ad hoc networks - WSN and VANET - security issues, challenges and solutions through a security management process according to the current frameworks and standards principles, first of all ISO 31000 "Risk Management": this course aims to be a concrete guideline to manage and perform a Cyber Security Management Process compliant to current best practises.
According to ISO 31000, we can refer to a "security management process" instance as a cyclic 4-steps process (Deming cycle) which starts from (1) cyber risk identification, (2) risk evaluation and quantification, (3) risk mitigation to ensure that the residual risk approaches the "acceptable risk" defined by the organization and (4) continuous monitoring to ensure that (3) remains true along time otherwise another security management process instance would be started. The "acceptable risk" identifies the "Minimum Security Requirement" or the "Required Security Level" (RSL) that should be regarded as an input given by the organization to the security designer. Evaluation and quantification methods for cyber risk (CRQ) are challenging issues: pure quantitative techniques are hardly applicable due to the difficulty in representing all involved parameters in a closed form: however different well-defined semi-quantitative techniques for CRQ have been defined and can be successfully employed. On the other side, the security performance indicators - basically in terms of deterrence delay and false positives / negatives rates - offered by the security functions can be quantitatively estimated with good approximation. Values for the security performance indicators determine the "Offered Security Level" (OSL). The best security design will result when OSL meets RSL: in such a case the applied security functions makes "acceptable" the residual risk of the organization.
ISO 31000 defines active and passive cybersecurity countermeasures according to the availability of feedbacks about the behaviour of the organization along time (in case of a system, behaviour corresponds to the state sequence along time) or not: passive countermeasures results only in deterrence. From a technical point of view, passive countermeasures include cryptographic techniques to preserve data confidentiality, sender authentication (signature) and message authentication (content), while active countermeasures include estimation and classification techniques for system behavior to detect and classify the anomalies induced in the system by cyber attacks.
ISO 31000 principles inspire methods and procedures for security management in specific engineering domains: e.g. ISO 27000 family for the ICT domain, ISO 21434 for the automotive domain.
Raffinate mathematical methods from the "information theoretic security", i.e. information theory applied to security introduced by C. E. Shannon in 1949 with its masterwork "Communication Theory of Secrecy Systems", are introduced to compute the security performance of a crypto system and the performance upper bound in a behavior estimation problem.
Standard passive security functions for WSNs and VANETs are still based on the ordinary cryptographic mechanisms (symmetric, asymmetric, hybrid schemes) but the resource constraints in terms of energy and memory savings of microprocessors embedded into WSN and VANET push to innovative and raffinate techniques such as elliptic curve cryptography (ECC), identity-based cryptography (IBC) which allows direct access to public keys and, in the future, post quantum cryptography (PQC).
Active security functions are based on behavior estimators and classifiers derived from the theory of Discrete Event Dynamic Systems (DEDS) and Maximum Likelihood Estimation (MLE) techniques with the aid of Machine Learning / Artificial Intelligence algorithms.
The course deals with the most relevant passive and active security functions applied to ad hoc networks from literature and those that have been designed and developed at Univaq labs in recent industrial projects as SEAMLESS, EMERGE, SHINE-ON, CYBORG: In particular TAKS (Topology Authenticated Key Scheme) and its ECC-based version EC-TAKS as well as the intrusion detection system WIDS (WPM-based Intrusion Detection System) and MVET (Mean-Variance Evaluation Technique) are introduced.
At the end of the course, students should be able to carry on a simplified but complete security management process instance applied to a case of ad hoc network application / service, to identify a subset of most hazardous risks, to assess system vulnerabilities, to quantify actual risks (or AS IS risks) and, according to a predefined "small acceptable risk", identify "the required security levels" to be met by the "offered security levels" through an early conceptual security design using the security functions available from the standard as block elements.
Slides:
The course. Risk Based Thinking. Elements of a wireless network, Wireless Network Taxonomy: Wireless Sensor Network (WSN) vs. Mobile Ad hoc Network (MANET) vs. Vehicular Ad hoc Network (VANET).
Part I: Security analysis applied to WSN and VANET
I.1 The framework of Security Management. From Risk to Security Management: Security Management Process, Approaches for Risk Evaluation, Techniques for Risk Evaluation, P-I Matrix and isorisk curves, FTA – CVSS, NIST SP 800-30 Guide for Conducting a Risk Assessment. Security management in the automotive domain: ISO / SAE 21434, Threat Analysis and Risk Assessment (TARA), Cybersecurity Risk Quantification technique EVITA, Guide line for TARA execution using EVITA. Reference Cyber Security functions: Security metrics, Timing Constraints, Cyber Risk Mitigation.
I.2 The case of WSN. Definition of WSN. Applications, design issues, reference WSN architecture. IEEE 802.15.4.
I.3 The case of VANET. Definition of VANET. VANET
vs. MANET. VANET applications. Inter-Vehicular communications systems.
Intra-Vehicular communications systems.
I.4 Threats and Attacks against WSN and VANET: Classification of Cyber attackers, Classification of attacks, Cyber attacks against WSN, Cyber attacks against VANET, Cyber attacks against Intra-Vehicle Communications, Classification of the Security Functions.
Part II: Mitigation Measures: Security Techniques for WSN and VANET
II.1 Passive Security Functions. Mathematical background: Kerckhoffs' principle. The Shannon's lessons. Modular Arithmetic, Generating Prime Numbers, Generating Pseudo-random Numbers, Elliptic Curve (EC) Algebra, Discrete Logarithm Problem and its EC version, Zero Knowledge Proof. Techniques: Ciphering, Hash functions, Message authentication codes, Digital signatures. Key Establishment Protocols (KEP): Symmetric KEP, Asymmetric KEP, Hybrid KEP, Authentication of public key. Key Management Protocols (KMP). IEEE 802.15.4 security, IEEE 1609.2 security, secure routing.
II.2 Active Security Functions. Mathematical background: dynamic systems, discrete events dynamical systems (DEDS), the Intrusion Detection Problem: DEDS modeling using Petri Nets, mapping PN into a finite automaton (FA), identification of observables and hidden states, state sequence estimation. Behavior Classifier. Information Theoretic Model of an Intrusion Detection System. Techniques: Machine Learning, Anomaly Detection System, audit data, Representation Model: Rules Based Techniques, Statistics Based Techniques. Classification Model.
II.3 Security Techniques made in Univaq. TAKS/ECTAKS: TAKSx driving ideas & main features, TAKS definition, TAK equations, geometric interpretation of TAK, authenticated network topology. ECC-based TAKS (ECTAKS), ECTAKS vs. ECDHE. ECTAKS schemes for encryption / decryption and sender signature. Intro to ECTAKS Security Proof. Selective Secure Clusterwise Communications. TAKSx release chronology. WIDS/MVET: WIDS driving ideas & main features, reference architecture, technique, security analysis. WIDS for IEEE 802.15.4 systems. MVET driving ideas & main features, reference architecture, technique, performance analysis.
II.4 VANET Security and Privacy. V2X communications security: architecture, analysis, privacy preserving solutions. Intra-Vehicle communications security: vulnerabilities, countermeasures.