A Skill-Key Encrypted Auction Scheme with Ring Signatures for Privacy-Preserving Heterogeneous Task Allocation
Privacy-preservation in multi-robot task allocation is critical for applications where robot capabilities, task requirements, or schedules must remain confidential. Additionally, while multi-robot task allocation has been extensively studied, privacy-preserving approaches for heterogeneous systems remain largely unexplored. This work addresses privacy-concerned heterogeneous task allocation in a multi-robot system where tasks require diverse skills and robots possess complementary capabilities, necessitating coalition formation. This work proposes a novel communication scheme applied to auction-based task allocation using skill-key symmetric encryption of messages to protect task requirements and individual group schedules and ring signatures to ensure bidder anonymity and validation during schedule formation. Shannon entropy is employed to quantify privacy levels and compare our approach to other efforts in the field. Traditional metrics are also applied in the experimental results to demonstrate and analyze the relationships between privacy-preservation, scheduling optimality, and computational overhead.
Advisor: Professor Carlo Pinciroli (RBE)
Committee: Professor Robert Walls (Cyber Security) and Professor Kevin Leahy (RBE)