Incremental search provides real-time suggestions as users type their queries. However, recent studies demonstrate that its encrypted search traffic can disclose privacy-sensitive data through side channels. Specifically, attackers can derive information about user keystrokes from observable traffic features, like packet sizes, timings, and directions, thereby inferring the victim’s entered search query. This vulnerability is known as a remote keystroke inference attack. While various attacks leveraging different traffic features have been developed, accompanied by obfuscation-based countermeasures, there is still a lack of overall and in-depth understanding regarding these attacks and defenses. To fill this gap, we conduct the first comprehensive evaluation of existing remote keystroke inference attacks and defenses. We carry out extensive experiments on five well-known incremental search websites, all listed in Alexa’s top 50, to evaluate and compare their real-world performance. The results demonstrate that attacks utilizing multidimensional request features pose the greatest risk to user privacy, and random padding is currently considered the optimal defense balancing both efficacy and resource demands. Our work sheds light on the real-world implications of remote keystroke inference attacks and provides developers with guidelines to enhance privacy protection strategies.

Typical Internet of Things (IoT) systems are event-driven platforms, in which smart sensing devices sense or subscribe to events (device state changes), and react according to the preconfigured trigger-action logic, as known as, automation rules. "Events" are essential elements to perform automatic control in an IoT system. However, events are not always trustworthy. Sensing fake event notifications injected by attackers (called event spoofing attack) can trigger sensitive actions through automation rules without involving authorized users. Existing solutions verify events via "event fingerprints" extracted by surrounding sensors. However, if a system has homogeneous sensors that have strong correlations among them, traditional threshold-based methods may cause information redundancy and noise amplification, consequently, decreasing the checking accuracy. Aiming at this, in this paper, we propose "EScope" , an effective event validation approach to check the authenticity of system events based on device state correlation. EScope selects informative and representative sensors using an Neural-Network-based (NN-based) sensor selection component and extracts a verification sensor set for event validation. We evaluate our approach using an existing dataset provided by Peeves. The experiment results demonstrate that EScope achieves an average 67% sensor amount reduction on 22 events compared with the existing work, and increases the event spoofing detection accuracy.