black box linear bias

src/tea3/tea3linearbias.py

@@ -0,0 +1,66 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from random import Random
+
+from .tea3 import Tea3
+
+def rand_key(rng: Random) -> list[int]:
+    return [rng.getrandbits(8) for _ in range(10)]
+
+@dataclass
+class BiasResult:
+    samples: int
+    matches: int
+    p_match: float
+    bias: float
+    estimated_needed_samples: float
+
+def estimate_black_box_output_bias(
+    *,
+    frame_number: int,
+    output_byte_index: int = 0,
+    output_bit: int = 0,
+    samples: int = 10000,
+    seed: int = 1,
+) -> BiasResult:
+    rng = Random(seed)
+    matches = 0
+
+    for _ in range(samples):
+        key = rand_key(rng)
+        tea = Tea3(frame_number, key)
+
+        out_byte = 0
+        for _ in range(output_byte_index + 1):
+            tea.next_byte()
+        out_byte = tea.iv_view() >> 56
+
+        bit = (out_byte >> output_bit) & 1
+        if bit == 0:
+            matches += 1
+
+    p_match = matches / samples
+    bias = abs(p_match - 0.5)
+    needed = float("inf") if bias == 0 else 1.0 / (bias * bias)
+
+    return BiasResult(
+        samples=samples,
+        matches=matches,
+        p_match=p_match,
+        bias=bias,
+        estimated_needed_samples=needed,
+    )
+
+def main():
+    frame = 0x12345678
+
+    res = estimate_black_box_output_bias(
+        frame_number=frame,
+        output_byte_index=0,
+        output_bit=0,
+        samples=20000,
+        seed=123,
+    )
+    print("black-box output bias")
+    print(res)