A Usable and Robust Continuous Authentication Framework Using Wearables

摘要: One-time login process in conventional authentication systems does not guarantee that the identified user is actual throughout session. However, it necessary to re-verify identity periodically a session, which lacking existing one-time systems. Continuous , re-verifies without breaking continuity of can address this issue. methods for Authentication are either reliable or usable. In paper, we introduce usable and Wearable-Assisted (WACA), relies on sensor-based keystroke dynamics data acquired through built-in sensors wearable (e.g., smartwatch) while typing. The transparently compared with registered profile initially logged-in one-way classifiers. With this, WACA continuously ensures current who initially. We implemented framework evaluated its performance extensively real devices users. empirical evaluation reveals feasible, error rate as low 1 percent 30 seconds processing time 2-3 20 seconds. computational overhead minimal. Furthermore, capable identifying insider threats very high accuracy (99.2 percent) also robust against powerful adversaries such imitation statistical attackers. believe work has practical far-reaching implications future field.