MPC/FHE-based implementation of MTCS

Develop practical implementations of the Multilateral Trade Credit Set-off (MTCS) solver using multi-party computation and/or homomorphic encryption to operate on encrypted obligation graphs without exposing cleartext data, providing performance and integrity comparable to trusted execution environments.

Background

Cycles currently proposes a privacy architecture that uses trusted execution environments (TEEs) for private computation and zero-knowledge proofs (ZKPs) for integrity of outputs. While performant, TEEs introduce a hardware-based trust assumption and potential vulnerabilities.

The authors explicitly defer MPC/FHE implementations of MTCS, which would reduce reliance on TEEs by enabling computation over encrypted data with distributed trust or fully homomorphic techniques. Practical realizations must address performance, security, and verifiability at scale.