Side Channels Flashcards
What is a side-channel attack?
An attack based on information from the physical implementation of a system, rather than its theoretical properties.
Why are side-channel attacks significant in cryptosystems?
Cryptosystems often leak information through timing, power consumption, or electromagnetic emissions, revealing secret data.
What are timing attacks?
Exploiting variations in operation times to infer sensitive information, such as cryptographic keys.
What is power analysis in side-channel attacks?
Observing power consumption patterns to deduce information about computations or secret keys.
What is differential power analysis?
Using statistical analysis of power consumption traces to extract cryptographic keys.
How do cache timing attacks work?
Manipulating cache states and observing timing differences to infer victim activity and data access patterns.
What vulnerability does Spectre exploit?
Speculative execution and cache timing to access private information within the same process.
How does the “Flush + Reload” technique work?
Fluch cache lines, wait for victim activity, then reload and time accesses to see which lines the victim accessed.
What are the steps in a Spectre attack?
- Train branch predictor.
- Exploit mis-speculation to access restricted memory.
- Analyze cache impacts to extract secrets.
What are laser-based audio injection attacks?
Using lasers to simulate audio signals in voice-controlled systems to manipulate them remotely.
How can motion sensors be exploited?
Inferring speech or keystrokes from subtle vibrations detected by accelerometers and gyroscopes.
How can keyboards be compromised?
Monitoring electromagnetic emissions or acoustics to infer typed characters.
How can reflections compromise privacy?
Analyzing reflections from screens or surfaces to recover displayed or typed information.
How can sound be extracted from videos?
Using high-frame-rate video analysis to infer sound vibrations (e.g., SIGGRAPH “Visual Microphone”).
What is website fingerprinting?
Identifying accessed webpages based on traffic patterns or power consumption profiles.
