Automatically finding execution scenarios to deploy security-sensitive workflows

Abstract

We introduce a new class of analysis problems, called Scenario Finding Problems (SFPs), for security-sensitive business processes that-besides execution constraints on tasks-define access control policies (constraining which users can execute which tasks) and authorization constraints (such as Separation of Duty). The solutions to SFPs are concrete execution scenarios that assist customers in the reuse and deployment of security-sensitive workflows. We study the relationship of SFPs to well-known properties of security-sensitive processes such as Workflow Satisfiability and Resiliency together with their complexity. Finally, we present a symbolic approach to solving SFPs and describe our experience with a prototype implementation on real-world business process models taken from an on-line library.

Pattern-Based Approach to the Workflow Satisfiability Problem with User-Independent Constraints

“…The key idea behind the PBT algorithm is that it is not necessary to search the space of plans; it is sufficient to search the space of patterns checking, for each eligible complete pattern, if it is authorised. (This idea was used in the preliminary version [37] of this paper; a similar idea was also exploited by dos Santos et al [24,25] who however use different algorithmic techniques such as Petri nets and Datalog. )…”

“…We have considered the instances appearing in practice from [6,24,25], namely, TRW (Trip Request Workflow), ITIL (IT Financial Reporting), and ISO (Budgeting for Quality Management) having 5, 7, 9 steps (tasks) and 5, 2, 3 not-equals constraints, respectively. More details and description of user-step authorisations for 3 satisfiable and 3 unsatisfiable versions of these instances can be found in Section 5 of [25]. Our new solver PBT correctly solved each of the six instances in less than 0.001 sec.…”

“…(Note the term "workflow" originally arose from the flow of the steps between users, however, in this context, the time ordering is not relevant -the challenge is to make a feasible assignment for all the steps.) The WSP has important applications and has been extensively studied in the security research community [4,5,6,14,15,17,18,25,50].The WSP is NP-complete, and it has been difficult to solve, even for some moderately-sized instances [12,47]. Work in WSP has attempted to render solving of the WSP practical by finding a subclass of problems that admit fixed parameter tractable (FPT) algorithms; informally speaking, this means that there is a small parameter k such that the problem is exponential in k but polynomial in the size of the problem.…”

“…(Note the term "workflow" originally arose from the flow of the steps between users, however, in this context, the time ordering is not relevant -the challenge is to make a feasible assignment for all the steps.) The WSP has important applications and has been extensively studied in the security research community [4,5,6,14,15,17,18,25,50].…”

Last man standing: Static, decremental and dynamic resiliency via controller synthesis

“…Some works have addressed workflow resiliency probabilistically, e.g., [10,11], whereas other works addressed it by modifying the constraints, e.g., [12,13]. Several approaches consider static resiliency only (e.g., [14,15,16]), one of them also considers decremental resiliency [16], whereas, to the best of our knowledge, an exact approach to dynamic resiliency (i.e., an approach that returns a dynamic execution plan if and only if a workflow is resilient without modifying the problem nor returning an execution plan that may fail) remains unexplored; see also [17] for a very recent survey on workflow satisfiability and resiliency.…”