Oblivious transfer: using noise to protect privacy

Oblivious transfer (OT) is a privacy-preserving communication primitive allowing someone to choose one piece of information from a set, without revealing which one they chose and without gaining access to the rest. In simple terms, it enables private selection: the receiver gets only the requested information, while the sender does not learn what was selected. This makes oblivious transfer an important building block for privacy-preserving digital services.

A key challenge is that oblivious transfer cannot be created from ordinary message exchange alone. If two parties only exchange regular messages, without any additional trusted setup, physical resource, or special assumption, then this level of privacy is impossible to guarantee. The same limitation also applies in standard quantum settings: quantum communication alone does not automatically solve the problem.

However, this does not mean oblivious transfer is impossible in practice. Communication channels are often noisy, especially in wireless and shared network environments. That noise can create useful uncertainty: different users may receive slightly different versions of the same transmission. Under the right conditions, this uncertainty can be turned into a privacy resource, enabling oblivious transfer with strong information-theoretic guarantees.

This idea is especially relevant for future communication networks, where many users share the same wireless medium. Instead of treating the physical layer only as a transport mechanism, researchers are exploring how the communication channel itself can help provide privacy. Current research studies oblivious transfer over shared noisy channels, the impact of colluding users, and practical implementations using coding techniques such as polar codes over AWGN channels.

Potential use cases include:

• Private database access and information retrieval: A user can retrieve one item from a database without revealing which item was requested.
• Privacy-preserving authentication: A system can verify access to selected information without exposing unnecessary personal data.
• Secure multi-party computation: OT is a core building block for protocols where several parties jointly compute a result while keeping their inputs private.
• Private digital services: Users can access content, recommendations, or credentials without revealing their exact choice or intent.
• Future 6G and multi-user networks: Network noise and different channel observations may support privacy-aware services at the infrastructure level.

The broader message is simple: noise is not always just a problem to overcome. In privacy-preserving communication, it can also become a useful resource. By understanding how noisy channels support oblivious transfer, researchers are working toward future network architectures where privacy is built not only into software, but also into the communication infrastructure itself.

In QSTARS, we develop and simulate new oblivious-transfer protocols for secure data exchange between mutually untrusted parties.