Tasks at EdgeCampus have a sensitivity label: public (e.g., map rendering), internal (course analytics), personal (face recognition on campus video, health data). Resources have a trust level: student phones (low), building edge servers (high, campus-managed, secure boot), public cloud (medium: provider-managed; high only with confidential-computing VMs at 30 % extra cost and 10 % slower).
- Define placement rules (which sensitivity may run where) and justify them with the threats of this chapter.
- Extend a scheduler from this course (the deadline placer of Lab OS13.4 or HEFT from Chapter 6) with these rules. Describe the change precisely (pseudo-code) and its effect on the feasibility of deadlines.
- Numerical example: design 6 tasks (2 of each sensitivity) with deadlines and sizes, and 3 resources; show the placement and the makespan/cost with and without security constraints. What is the "cost of security"?
- Discuss how encryption in transit (e.g., +2 ms per MB) and attestation checks affect the model, and one open research question.