Efficient Distributed Linear Programming with Limited Disclosure

Journal of Information Systems

Vaidya

Wang

2013

Self Cite

In the contemporary information era, the ubiquitous collection of data from different parties frequently accommodates significant mutual benefits to the involved participants. However, data is a double-bladed sword. Inappropriate access or use of data by the recipients may pose serious privacy issues that explicitly harm the data owners. In the past decade, swiftly increasing privacy concerns arise in many business processes such as supply chain management. How to protect the private information of different participants in the supply chain has become a key multidisciplinary research problem in information systems, production and operations management, computer science, and mathematics. Specifically, in the real world, manufacturers, distributors, and retailers commonly collaborate with each other to cater to the demands of supplying and marketing. In their traditional cooperation, all the parties completely share their proprietary information so as to jointly optimize their operations (e.g., maximize their profit or minimize their cost). Now, they realize that completely sharing such information would bring considerable negative impact to themselves. For overcoming this, some recent research results begin to make the following ideal occasion possible—all the participants collaboratively solve a realistic problem without revealing any private proprietary information to each other. In this paper, we primarily review the literature on the applications of privacy-preserving techniques to supply chain collaboration among multiple parties. We first identify various private proprietary information required in the supply chain collaboration, and discuss several potential privacy-preserving techniques. Then, we review the relevant research results from theory to applications. Since intensive collaboration in modern supply chains opens even more opportunities in both academia and industry, we finally outline the future research trend and the potential challenges in this promising area.

Section: Future Research Trend and Challenge Discussionmentioning

confidence: 99%

Section: Linear Programmingmentioning

confidence: 99%

A Survey of Privacy-Aware Supply Chain Collaboration: From Theory to Applications

Journal of Information Systems

Vaidya

Wang

2013

Self Cite

“…Specifically, Li and Atallah and Vaidya proposed privacy‐preserving techniques for solving two‐party linear programming (LP) problems. Hong and Vaidya and Mangasarian protected the privacy while solving different forms of multiparty LP problems and their corresponding applications . Besides LP, some other classic optimization problems can be also securely solved with limited disclosure in literature (e.g., ).…”

Section: Related Workmentioning

confidence: 99%

An efficient and privacy-preserving scheme for P2P energy exchange among smart microgrids

Goel

Liu

2015

Int. J. Energy Res.

Self Cite

SUMMARYTo date, increasing number of entities on the smart grid begin to establish their local energy generator for ensuring reliability and resilience of power supply. These 'microgrids' can either connect to the power grid or isolate themselves from the grid by consuming their locally generated or stored energy. In reality, some microgrids may have excessive energy while the others may have to request extra energy from the main grid. To better balance the demand and supply of the distributed smart microgrids, it is desired to develop peer-to-peer (P2P) energy exchange models that enable microgrids to interactively exchange their local energy instead of consuming energy delivered from the main grid. However, in this scenario, all the microgrids have to disclose their private information (e.g., demand load and energy storage amount) to each other in the exchange. To tackle these issues, in this paper, we first formulate several novel energy exchange optimization problems that minimize the global energy loss during the exchange in different scenarios, and then develop an efficient and privacy-preserving scheme to solve the energy exchange optimization problems without private information disclosure. We also extend the privacy-preserving scheme to a collusion-resistant scheme in which all the microgrids cannot learn any additional information through colluding with each other. The performance of our proposed approaches is experimentally validated on real microgrid data.

“…Du [12] and Vaidya [13] transformed the linear programming problem by multiplying a monomial matrix to both the constraint matrix and the objective function, assuming that one party holds the objective function while the other party holds the constraints. Bednarz et al [14,15] pointed out a potential attack to the above transformation approach, which has been resolved in [3,16]. In addition, Mangasarian presented two transformation approaches for horizontally partitioned linear programs [17] and vertically partitioned linear programs [18] respectively.…”

Section: Related Workmentioning

confidence: 99%

Collaboratively Solving the Traveling Salesman Problem with Limited Disclosure

Lecture Notes in Computer Science

Vaidya

et al. 2014

Self Cite

Abstract. With increasing resource constraints, optimization is necessary to make the best use of scarce resources. Given the ubiquitous connectivity and availability of information, collaborative optimization problems can be formulated by different parties to jointly optimize their operations. However, this cannot usually be done without restraint since privacy/security concerns often inhibit the complete sharing of proprietary information. The field of privacy-preserving optimization studies how collaborative optimization can be performed with limited disclosure. In this paper, we develop privacy-preserving solutions for collaboratively solving the traveling salesman problem (TSP), a fundamental combinatorial optimization problem with applications in diverse fields such as planning, logistics and production. We propose a secure and efficient protocol for multiple participants to formulate and solve such a problem without sharing any private information. We formally prove the protocol security under the rigorous definition of secure multiparty computation (SMC), and demonstrate its effectiveness with experimental results using real data.