The Resilience Model Supporting IIoT System Trustworthiness
(6). This makes the value of P true. This is
how these techniques and approaches help
to achieve the resilience goals considered
earlier in this article. environment and constraints. These
approaches and techniques are considered
active because they influence the factors
determining resilience according to (6).
The approaches listed in Table E-1 of the
Draft NIST Special Publication 800-160
VOLUME 2 may be implemented either at
design phase or at runtime. Not all approaches can be implemented for
every given system. Depending on the initial
state and functional constraints the
stakeholders may consider the options of
how to increase system resilience. The
classification shown in the Table 1 helps to
clarify these options.
Approaches implemented at design phase
provide the foundation for building
resilience capacity. These approaches and
techniques are mostly passive. They set up
the types and appropriate ranges for the
factors of the generalized predicate P in (6).
Approaches used at runtime help in dynamic
realignment of algorithms, resources and
data according to dynamically changing
Table 1 – Classification of resilience approaches and techniques according to the proposed model
Factors Active resilience techniques and Passive resilience techniques and
approaches implemented at approaches implemented at design
runtime
phase
ST Non-Persistent Services / NP
Consistency Analysis / CD
Algorithmic
structure of
the functions
determining
the control
process
behavior Non-Persistent Connectivity / NP
Orchestration / CD
Dynamic
Segmentation
Isolation / Se
and Architectural Diversity / Di
Design Diversity / Di
Temporal Unpredictability / Up
Contextual Unpredictability / Up
Synthetic Diversity / Di
Supply Chain Diversity / Di
Distributed Functionality / DP
Restriction / Ra
Replacement / Ra
Specialization / Ra
Predefined Segmentation / Se
September 2018
- 126 -