1. The Ekert Protocol
In 1991, Artur Ekert proposed a quantum key distribution protocol based on entangled particles and Bell correlations. Chris Bernhardt presents a simplified version using three measurement bases.
2. Entangled Pairs
Alice and Bob each receive one qubit from the Bell state |Φ+⟩ = (|00⟩ + |11⟩)/√2. Measuring both qubits in the same basis produces matching random bits.
3. Why One Fixed Basis Is Insecure
If Alice and Bob always use the standard basis, Eve can intercept Bob’s qubit, measure it in that basis, and send a replacement prepared in the measured state. Alice and Bob can still obtain matching bits while Eve learns their values.
4. Randomly Choosing Three Bases
Alice and Bob independently choose among three bases, each with probability 1/3. There are nine equally likely basis pairs, and three use the same basis. Therefore, the expected fraction of same-basis measurements is 1/3 and the different-basis fraction is 2/3.
5. Generating and Testing the Key
After measurement, Alice and Bob publicly compare their basis choices without revealing their secret bit values. Same-basis results supply potential key bits; different-basis results are used to test correlations. For 900 pairs, approximately 300 are expected to use the same basis and 600 different bases.
6. Detecting Eve
In Bernhardt’s simplified example, the expected agreement rate for different-basis measurements is 1/4 (25%) with the original entanglement. A particular intercept-and-measure attack changes it to 3/8 (37.5%). This change can reveal interference statistically.
7. Why Eve May Go Undetected in One Measurement
If Eve, Alice, and Bob all use the same basis, Eve can obtain the bit and send Bob a matching replacement. Their results can still agree. Detection relies on the pattern across many measurements and different bases, not on every intercepted bit causing a mismatch.
8. Two Different 25% Probabilities
When Alice and Bob choose independently among four equally likely bases, the probability of choosing the same basis is 1/4. This is different from the 1/4 agreement probability for different-basis measurement outcomes in Bernhardt’s three-basis Bell example.
9. Reliability and Security
Experimental correlations have statistical uncertainty and may be disturbed by noise as well as eavesdropping. Practical quantum key distribution also requires authenticated classical communication, error correction, privacy amplification, and finite-key security analysis. These are qualifications beyond the simplified book presentation.
Key Takeaway
Perfectly matching bits alone do not prove secrecy. Random measurement bases and quantum correlation tests help Alice and Bob detect interference before accepting a shared secret key.
Note metadata
- Note type: learning-note
- Subject: learning-notes
- Source: Quantum Computing for Everyone