Statistical privacy-preserving message broadcast for peer-to-peer networks.

Area of Science:

• Computer Science
• Cryptography
• Network Security

Background:

• Financial blockchains face public scrutiny regarding transaction originator privacy.
• Existing adaptive diffusion methods are unsuitable for current internet peer-to-peer networks due to unrealistic models.
• Securing transaction origins is crucial for decentralized financial systems.

Purpose of the Study:

• To develop a statistical privacy-preserving broadcast protocol for financial blockchains.
• To adapt the adaptive diffusion approach for realistic peer-to-peer network conditions.
• To enhance the privacy of transaction originators in decentralized networks.

Main Methods:

• Modeled unstructured peer-to-peer networks as organically growing graphs.
• Applied normal distribution to model shortest path distributions.
• Developed statistical estimators for distribution parameters using multivariate models.
• Analyzed model behavior concerning network size (nodes) and connectivity (edges).

Main Results:

• The proposed protocol enhances privacy for transaction originators on financial blockchains.
• Shortest path distributions in modeled networks follow a normal distribution.
• The model's performance scales logarithmically with network size and inversely exponentially with edge addition.
• Optimal forwarding probabilities for maximum privacy were determined.

Conclusions:

• The transformed adaptive diffusion protocol offers improved privacy for financial blockchain transactions.
• The network and statistical models provide a foundation for privacy-preserving broadcast protocols.
• This research contributes to securing decentralized financial infrastructure against privacy attacks.