ORide: A Privacy-Preserving yet Accountable Ride-Hailing Service
summary
ORide implements somewhat-homomorphic encryption with ciphertext packing and number-theoretic transforms to let the service compute squared Euclidean distances over encrypted coordinates and return packed distances for riders to decrypt. Optimized flow drops rider bandwidth to ~200 KB, driver uplink needs stay below 2 Mbps, service-provider work falls from ~100,000 ms to ~200 ms, and decryption/key ops are under 10 ms. Evaluations show anonymity set >1000 in many Manhattan zones and extra driver detour <0.5 km in 75% of rides.
AI-written summary · about
chapters (15)
- 0:00Intro
- 1:08Privacy Problems
- 2:30Adversarial Assumptions
- 3:12Attacks Considered
- 3:52Desired Usability and Accountability Features
- 4:29ORide Overview
- 5:58Oblivious Ride Matching - Intuition
- 11:08Accountability Mechanisms
- 13:32Payment and Reputation Rating
- 14:10Protocol Analysis
- 15:40Evaluation
- 17:02Implementation
- 17:37Per-Ride Overhead
- 18:46Riders' Anonymity Set vs. Drivers' Bandwidth
- 21:21Conclusion
related talks
16mOblivious Plug&Charge: A Privacy-Preserving EV Charging Scheme based on ORAM
18mDecentralization is the New Encryption
23mOblivP2P: An Oblivious Peer-to-Peer Content Sharing System
22mXRD: Scalable Messaging System with Cryptographic Privacy
21mDriving connected mobility forward
11mHow Ory Network prevents fraud 
36mSecuring Every Bit
17mEncrypted Access Logging for Online Accounts: Device Attributions without...