Overview
Published at IEEE ISDFS 2026, this work studies repeated HTTP(S) requests caused by overlapping discovery paths, scanner modules, retries and concurrent workers. Deduplication and memoization are placed at the request dispatch boundary.
The model combines deterministic canonicalization with a policy that decides which request features establish equivalence. Reusing a stored result also requires validity checks, including context binding, time-to-live and safeguards for methods and session-bearing requests. Similar-looking URLs alone are insufficient to establish safe reuse.
The paper describes concurrency controls for limiting simultaneous dispatches for a key, verifies stored results with collision-aware checks and derives bounds for lookup and storage operations. Its guarantees depend on the specified policies and atomicity assumptions; those conditions matter when implementing the mechanism in a real scanner.
Key facts
- Publication
- 2026
- Published in
- IEEE International Symposium on Digital Forensics and Security (ISDFS 2026)
- Topics
- Request deduplication · Memoization · Concurrency · HTTP context