Pirates: Anonymous Group Calls Over Fully Untrusted Infrastructure

TL;DR

Pirates addresses the challenge of enabling real-time anonymous group voice calls over fully untrusted infrastructure. It leverages computational private information retrieval (PIR) with multi-bucket parallelization and a hierarchical server architecture, coupled with LPCNet-based low-bandwidth encoding, to hide communication patterns while delivering sub-$400$ ms mouth-to-ear latency for small groups. The work provides formal proofs of communication unobservability and delivers an in-depth evaluation comparing PIR schemes and scalability, showing that adding servers improves group size and latency performance. Overall, Pirates demonstrates the feasibility of anonymous group calls today and identifies clear directions for supporting larger and dynamic groups as PIR techniques advance.

Abstract

Anonymous metadata-private voice call protocols suffer from high delays and so far cannot provide group call functionality. Anonymization inherently yields delay penalties, and scaling signalling and communication to groups of users exacerbates this situation. Our protocol Pirates employs PIR, improves parallelization and signalling, and is the first group voice call protocol that guarantees the strong anonymity notion of communication unobservability. Implementing and measuring a prototype, we show that Pirates with a single server can support group calls with three group members from an 11 concurrent users with mouth-to-ear latency below 365ms, meeting minimum ITU requirements as the first anonymous voice call system. Increasing the number of servers enables bigger group sizes and more participants.

Figures (6)

• Figure 1: Simplified interaction between a client $A$, and Pirates's servers (coordinator $\mathcal{C}$, relay $\mathcal{R}$, and worker $\mathcal{W}$)
• Figure 2: Pirates operates in epochs that are split into three phases: Mapping Generation, Dialing, and Communication. The communication phase consists of multiple subrounds. Not to scale, size of phase does not correspond to duration.
• Figure 3: Mouth-to-ear latency measurements in Pirates for varying numbers of clients (a) and varying group size (b). Ratio refers to the ratio between worker processing time and snippet length.
• Figure 4: Correlation between the number of workers and the total time it takes to distribute voice data and compute PIR answers in Pirates and Addra.
• Figure 5: Results of interrupted playback survey.

Theorems & Definitions (14)

• definition thmcounterdefinition: Query Privacy
• lemma thmcounterlemma
• theorem thmcountertheorem
• proof
• lemma thmcounterlemma
• proof
• lemma thmcounterlemma
• proof
• lemma thmcounterlemma
• proof