FlowFence: Practical Data Protection for Emerging IoT Application Frameworks
summary
FlowFence requires apps to split sensitive processing into Quarantined Modules that run in sandboxes and return opaque handles instead of raw data, with taint labels and a trusted sink enforcing publisher and consumer policies. Implementation uses a KV store and event channels for secure data sharing, needs common OS features only, and ports three real IoT apps with source size rising from 232 to 332 lines on average. Evaluation shows practical costs — a face‑recognition door controller adds a 4.9% latency overhead.
AI-written summary · about
chapters (6)
- 0:00Start
- 3:01Instruction-Level Flow Analysis Techniques
- 7:29Face Recognition App Example
- 12:36Publisher and Consumer Flow Policies
- 19:20Evaluation Overview
- 23:16Summary
related talks
50mUsing FOSS to Protect Embedded Systems Data in Transit
14mBeyond the Device: Why IoT Data Security is the Real Challenge
12mTransparent Microsegmentation in Smart Home IoT Networks
14mCloudFlow: Identifying Security-sensitive Data Flows in Serverless
23mKeynote: Zephyr & IoT Security
55mFog Computing and IoT Solution Architecture
33mIoT Security: Past, Present and Future
59mIntro to Flow