[Paper Review] Strong Anonymity for Mesh Messaging

Neil Perry, Bruce Spang+2 more|arXiv (Cornell University)|Jul 8, 2022

Opportunistic and Delay-Tolerant Networks4 citations

TL;DR

This paper proposes a secure, anonymous mesh messaging system for protesters in internet-blackout scenarios, using broadcast-based encryption and cover traffic to hide all metadata—including sender, receiver, timing, and message patterns—while scaling to support over 3,000 users under realistic conditions. The system achieves strong anonymity against nation-state adversaries through a novel threat model and protocol design that prevents traffic correlation and metadata leakage.

ABSTRACT

Messaging systems built on mesh networks consisting of smartphones communicating over Bluetooth have been used by protesters around the world after governments have disrupted Internet connectivity. Unfortunately, existing systems have been shown to be insecure; most concerningly by not adequately hiding metadata. This is further complicated by the fact that wireless communication such as Bluetooth is inherently a broadcasting medium. In this paper, we present a new threat model that captures the security requirements of protesters in this setting. We then provide a solution that satisfies the required security properties, hides all relevant metadata, scales to moderately sized protests, and supports group messaging. This is achieved by broadcasting all messages in a way that limits the overhead of duplicate messages, ensuring that ciphertexts do not leak metadata, and limiting what can be learned by observing user behavior. We also build a model of our system and numerically evaluate it to support our claims and analyze how many users it supports. Finally, we discuss further extensions that remove potential bottlenecks in scaling and support substantially more users.

Motivation & Objective

• To define a formal security model for anonymous mesh messaging in adversarial, internet-disrupted environments, particularly for protesters facing nation-state surveillance.
• To address the critical lack of metadata protection in existing mesh messaging apps like Bridgefy and FireChat, which have been shown vulnerable to traffic analysis and deanonymization.
• To design and implement a system that ensures end-to-end confidentiality and strong anonymity, including hiding sender, receiver, timing, and communication patterns.
• To evaluate system scalability through modeling and numerical analysis under realistic protest conditions, including crowd density and message frequency.
• To explore extensions for supporting larger protests while maintaining strong privacy guarantees.

Proposed method

• Design a novel threat model that captures the capabilities of nation-state adversaries in internet-blackout protest scenarios, including global network visibility and control over some users.
• Use broadcast-based message dissemination to prevent metadata leakage, ensuring that ciphertexts do not reveal sender, receiver, or timing information.
• Implement cover traffic generation to mask real message patterns and prevent traffic correlation attacks by adversaries observing user behavior.
• Apply a modified broadcast algorithm based on prior work to ensure all devices receive messages while minimizing duplicate message overhead.
• Use symmetric-key encryption and a key management scheme that ensures forward secrecy and prevents long-term tracking.
• Build a system model to simulate message delivery latency and throughput under varying user counts, crowd densities, and message rates.

Experimental results

Research questions

• RQ1 Can a mesh messaging system be designed to provide strong anonymity against a nation-state adversary in internet-disrupted environments?
• RQ2 How can metadata such as sender, receiver, timing, and message frequency be hidden in a Bluetooth-based mesh network with no centralized infrastructure?
• RQ3 What is the maximum number of users a mesh messaging system can support while maintaining low latency and strong anonymity under realistic protest conditions?
• RQ4 How can cover traffic and broadcast mechanisms be combined to prevent traffic correlation attacks without excessive network overhead?
• RQ5 What architectural extensions can be used to scale the system to support tens of thousands of users while preserving privacy guarantees?

Key findings

• The system supports up to 200 users sending messages every 30 seconds under pessimistic crowd density assumptions, with message delivery within several seconds.
• Under more realistic conditions, the system supports over 3,000 users sending messages every minute, with delivery times under two minutes.
• The system achieves strong anonymity by hiding all metadata, including sender, receiver, message timing, and communication patterns, through broadcast-based encryption and cover traffic.
• Numerical modeling confirms that the system scales effectively for moderately sized protests, with low message loss and predictable delivery latency.
• The protocol is resilient to passive adversaries with global network visibility, as traffic correlation is prevented by cover traffic and broadcast obfuscation.
• Extensions such as clique-based broadcasting could further scale the system to support tens of thousands of users, though with potential trade-offs in anonymity or user friendship limits.