Timing Attack Threat Analysis

Version: 1.0
Last Updated: 2025-01-02
Security Classification: PUBLIC

Overview

Timing attacks exploit variations in execution time to extract sensitive information. These attacks are particularly dangerous because they can be mounted remotely and leave no trace in system logs.

Attack Vectors

Cache Timing Attacks

Flush+Reload: Monitor cache lines for access patterns
Prime+Probe: Fill cache and measure evictions
Evict+Time: Evict data and measure reload time

Branch Prediction Attacks

Spectre-style: Exploit speculative execution
BTB attacks: Branch Target Buffer manipulation

Algorithm-Specific Timing

Table lookups: AES S-box timing
Modular operations: RSA/ECC scalar multiplication
Comparison operations: Password/MAC verification

Vulnerable Operations

Operation	Risk Level	Common In	Mitigation
Table lookups	High	AES, DES	Bitsliced implementation
Conditional branches	High	All algorithms	Constant-time selection
Division/Modulo	High	RSA, DSA	Montgomery multiplication
Memory access	Medium	All algorithms	Cache-oblivious algorithms
Floating-point	Medium	Falcon, NTRU	Fixed-point arithmetic

Mitigation Strategies

Constant-Time Implementation

// Bad: Timing depends on secret
if (secret_bit) {
    operation_a();
} else {
    operation_b();
}

// Good: Constant-time selection
result = ct_select(secret_bit, value_a, value_b);

Cache-Line Alignment

Align sensitive data to cache boundaries
Use full cache lines for lookups
Implement scatter-gather techniques

Blinding Techniques

Add random delays (limited effectiveness)
Randomize execution order
Use masking for intermediate values

Platform-Specific Considerations

x86/x64

Use AES-NI for constant-time AES
Avoid variable-time instructions (DIV, IDIV)
Consider TSX for atomic operations

ARM

Use NEON for constant-time operations
Beware of early termination in multiplication
Utilize cryptographic extensions when available

WebAssembly

Limited timing precision (mitigation)
No direct cache control
Rely on runtime protections

Detection Methods

Statistical Analysis: Measure timing distributions
Differential Testing: Compare implementations
Automated Tools: dudect, ctgrind, MemorySanitizer
Manual Review: Code inspection for branches

Real-World Examples

Lucky Thirteen (2013): TLS CBC padding oracle
CacheBleed (2016): OpenSSL RSA key extraction
Minerva (2019): ECDSA timing attack
Hertzbleed (2022): Frequency scaling side-channel