A Privacy-preserving Online Double Auction For Spectrum Allocation

The rapid development of wireless communication industry is rapidly depleting the limited spectrum resources.Researchers proposed to realize the spectrum reallocation through auction and improve the overall utilization of spectrum resources.However,most of the existing works only consider the economic robustness while neglecting the privacy leakage of users.The few existing secure spectrum auction mechanisms fail to provide adequate security,lack protection for users’ geographic location information and request of time-slots,and are not suitable for scenarios where spectrum requests arrive online.To solve these problems,we propose a privacy-preserving and truthful online double auction for spectrum allocation,namely PS-TODA.Compared with the existing works,our scheme provides a comprehensive and strong protection for users’ privacy information,including users’ bid value,geographic location and request of time-slots.To be specific,the innovation and main contents of this paper include the following three points:First,we adopt homomorphic encryption technology to encrypt users’ sensitive information,and design a series of security building blocks,including ciphertext addition,subtraction,multiplication,comparison and selection protocols,by utilizing cryptography technologies such as garbled circuits and oblivious transfer protocol.These security building blocks can support various arithmetics over encrypted real numbers by non-secret key parties,and the security building blocks are well-applicable to other secure spectrum auction models.Second,we design modularized user privacy protection mechanisms for three important stages of spectrum auction.In this paper,we improve the existing dynamic spectrum allocation model by designing a novel privacy preserving buyer grouping protocol to realize spectrum reuse.Besides,we utilize secure multi-party computation technology to ensure users’ privacy,and the economic characteristics.The scheme of this paper is extensible for other secure spectrum auction models,each module can be modified,updated and extended conveniently.Third,we propose a series of system optimization strategies to reduce system overhead,ensuring the real-time performance of the system.Thus,our system is more suitable for practical business applications.Finally,theoretical analysis of the security and performance of this scheme proves that our scheme is secure in the semi-honest adversaries model,and the performance of our scheme is evaluated through experiments.Experimental results show that our proposed scheme has good spectrum allocation efficiency and reasonable computation andcommunication costs,and our scheme is suitable for practical commercial applications.