Resumen de A model of a secure intelligent trade agent
In this paper, we propose a model of a secure intelligent trade agent. In building this agent, we adapt our agent to use with a secure and efficient electronic cash (executable digital-cash, x-cash) [1] to enable agents to carry funds and make payments on-site without running the risk of “pick-pocketing”. It is a means of binding an offer to accompanying goods or payment, enabling the processes of searching and paying to be unified. The result is a mechanism by which electronic trades can occur in a highly distributed setting with strong security guarantees. This model provides an interface by which users obtain the certificates from the bank and authorizing them to make payments. The user can login to the agent repository (AR) for issuing offer programs w. The program w along with the certificate constitutes a piece of x-cash. An agent together with the executable digital cash (x-cash) which represents a buyer will be sent across the Internet via “asymmetric proxy reencryption protocol [2,9], executing methods (w) on a number of merchant’s server (on remote server systems). It takes as input some item, and outputs the amount which the buyer is willing to pay for that item.
If the seller wishes to sell an item (Q), he/she can take the x-cash and the item Q to the buyer’s bank. By running the program (w) on Q, The buyer’s bank can determine how much to pay the seller. The buyer’s bank may then hold the item for buyer or otherwise arrange to send it to him. The trade is thus completed in a secure fashion without any direct contacts between seller and buyer.
Intelligent trade agents must be able to buy goods and pay for them without any human intervention. In this paper, we introduce a model of agent based system, describe some variants and sketch proofs of its security properties
