Table 9 Resilience readiness and big data cyber security aspects

Responsibility

HMI configuration to meet the big data cybersecurity needs. Tailoring of the big data analytics results considering the cyber security concerns. F.e. malicious attacker manipulating the HMI to cause an incorrect action [26], injecting false data or invalid commands. The actors shall have sufficient level of data analytics knowhow to distinguish the false data on UI. The possible commands, command names, action names, updates and event names could be derived via data analytics from the system data and the UI/HMI could be having validity checks. The timing/duration of the action/update in HMI/UI could be compared with the update timing/duration/effort statistics/historical traces/logs of the system that have been performing with real data (duration of querying, UI update etc.).

Mutual Impacts

Threat model [27], common system view, standardization? Security view as an architectural view? Design patterns defined for security? Security as a service?

Situational Intelligence

Security as a lifecycle issue. How to coordinate security practices in requirements, design, code analysis and test stages? Are there common high level security requirement sets for big data systems? Common code fallacies causing vulnerabilities or common test approaches to detect them?

Operation Resources

Attacks targeting the recovery and replication management, that are specific for big data systems. The secure strategy for data replication and its effects on performance. Attacks that cause data replication or unexpected recovery. Attacks against system configuration that injects error to the configuration managed by the operator, validation of the configuration (shall be automated or manual?). The detections of noise injection at data fusion, what are the data analytics methodologies for this? Are there any software libraries that verifies and validates the data analytics process against attacks, f.e. at the time of fusion?

Mutability

The encryption management strategy would be application architecture specific. What are the encryption management strategies applied for the security of the big data system? Are there any communication route or structure adaptation of the system to meet the cybersecurity requirements?

What are the attack detection strategies, i.e. checking the mean time between failures? Is the system compatible with the contemporary cybersecurity tools, i.e. ease of modification, integration or monitoring?

Modularity

Is the system compatible with the contemporary cybersecurity tools, i.e. ease of modification, integration or monitoring? Could the component availability be measured as an attack parameter? What are the security criteria considered while applying the data and information refinement?

Event Mechanisms

What are the cybersecurity qualifications considered while adopting a driver driver, kernel function? Which are the secure functions and how to assess the maturity of the function from the cybersecurity aspect? How is the security ensured for the manual modes, i.e. training of the individual or system adaptations such as command verification or peer review?