Making Offline Digital Yen Payments Safe and Affordable

Digital cash must work without an internet connection, but disconnected wallets cannot check whether someone has spent a copied token elsewhere. This design borrows from bacterial plasmids, small pieces of DNA that produce a lasting poison and a short-lived antidote: lose the plasmid, and the antidote runs out. Each digital token pairs a lasting, signed value claim with a proof that expires within minutes. The wallet’s security chip renews that proof using a cryptographic ratchet, which advances through fresh keys and deletes old ones. A token copied without access to the chip loses its valid proof; legitimate recipients bind received money to their own chips and can spend it offline. A fixed set of numbered vouchers limits each wallet’s authorized offline spending. When wallets reconnect, banks check for reused voucher numbers and cover duplicate-payment losses for honest recipients. The design adds expiry, spending limits and explicit fraud coverage to the usual protection of storing money inside a security chip.

Offline digital cash depends on a security chip to stop people copying money. If attackers extract the chip’s secret keys, they can create counterfeits with fresh serial numbers that evade duplicate-number checks. We propose giving each chip a unique hidden fingerprint, much as distributors mark individual video copies to trace leaks. A bank signs a fixed set of fingerprint marks for each chip; every payment includes proof that its mark belongs to that set. Attackers with stolen chip keys therefore have to draw from the provisioned marks when forging payment records. Recipients add their own receipt marks as money changes hands. When devices reconnect, the bank checks for reused marks that an honest chip spends only once, follows receipt trails to locate where counterfeits entered circulation, and statistically compares fingerprints from confirmed counterfeits to identify a compromised source, even if attackers mix several devices’ marks. The marks stay concealed during ordinary payments. Investigators gain evidence beyond repeated serial numbers, and the bank can revoke compromised devices to limit further losses.

After a disaster cuts communications, millions of digital cash wallets can reconnect at once. Replaying every offline payment against the central ledger creates a surge just when online payments are needed again. We borrow the way Amazon’s Dynamo database brings disconnected copies back into agreement. Each wallet records tokens received and tokens spent in two lists that only grow; combining those lists produces the same record regardless of reconnection order. A hash tree, or Merkle tree, summarizes wallet records in one cryptographic fingerprint. At reconnection, the intermediary compares fingerprints, follows only branches that differ, and requests the changed records rather than replaying the entire payment history. Each token also carries a signed transfer trail: spending it twice creates a fork that supplies evidence of the double-spend. Intermediaries spread these comparisons across background sweeps, so reconnection work follows the differences between records instead of forcing every offline payment through one immediate replay queue.