Ad Hoc Prevention of Double-spending in Offline Payments

Full abstract

Read the full abstract

Digital payments are vital for daily activities, and they need to be available even when communication with payment service providers is impossible. The major problem faced in offline payments is that the same money can be spent multiple times, known as double-spending. Most solutions address double-spending either by deterrence or by using secure hardware. This paper introduces an offline payment method to prevent double-spending with resource-constrained mobile devices. This solution is based on an arbitrary digital payment system enabling users to prevent double-spending by sharing information over ad hoc contacts. To reduce the overhead of sharing this data, we recommend using Bloom filter to store the spent money identifiers and suggest an algorithm to control the false positive rate. To further reduce the likelihood of double-spending, the payment protocol is designed to ensure that the funds used in each transaction are not under the control of the payer. We evaluated the proposed solution using several real-world mobility datasets by considering users’ contacts as payment and information-sharing opportunities. The findings indicate that our proposed solution effectively prevents double-spending, akin to ideal conditions where users have unlimited storage and bandwidth. The evaluation results show that with 240 bytes to store and share information about spent funds, the system achieves an accuracy of over 85% in detecting double-spending attempts.