Showing papers by "Jarrod Trevathan published in 2006"

Undesirable and fraudulent behaviour in online auctions

Abstract: Online auctions are a popular means for exchanging items over the Internet. However, are many inherent security and fairness concerns. Participants can behave in an undesirable and fraudulent manner in an attempt to gain an advantage at the expense of rivals. For example, a bidder might seek to suppress the price by bid sniping, or the seller could introduce fake bids to inflate the price. In addition, an outsider or rival seller can lure away bidders by directly offering them better deals, or a malicious seller can auction mis-represented or non-existent items. This conduct is a problem as it results in market failure, thereby inhibiting the usefulness of online auctions as an exchange medium. While cryptography has been used to provide security in terms of bid authentication and privacy, there is no documented means to prevent many of the aforementioned problems. This paper investigates undesirable and fraudulent behaviour in online auctions. We examine the following practices: bid shielding, shill bidding, bid sniping, siphoning and selling non-existent or misrepresented items. We describe the characteristics of such behaviour and how to identify it in an auction. We also provide recommendations for recourse against undesirable and fraudulent participants.

An Anonymous and Secure Continuous Double Auction Scheme

Abstract: A Continuous Double Auction (CDA) allows many buyers and sellers to continuously submit bids for the purchase and sale of a commodity (e.g., online share trading). Protocols protecting privacy in this type of powerful market mechanism are essential. However, until recently the security of CDAs has been given limited coverage. This paper describes a new scheme for conducting an anonymous and secure CDA. We show that any existing secure group signature scheme can be used to implement a CDA which has the following characteristics: unforgeability, anonymity, unlinkability, exculpability, coalitionresistance, verifiability, robustness and traceability. Furthermore, bidders can be added to and removed from the auction without affecting the process of the auction. Our scheme is more flexible than the only existing secure CDA scheme, which in contrast provides only a limited subset of these characteristics.

RAS: a system for supporting research in online auctions

Abstract: Online auctioning is unparalleled as rhe fastest growing exchange medium to emerge from electronic commerce technology. Buyers and sellers located around the world now auction various items from the latest DVD to rare collectibles. eBay and uBid are among the most successful and popular of the commercial online auctioneers. They use an auctioning process based on a type of auction referred to as an English auction. In an English auction. bidders outbid each other for an item. The winner is the bidder with the highest bid.

Secure Online English Auctions.

Abstract: Security and privacy in online auctions is a major concern as auction participants have many opportunities to cheat (e.g., repudiate bids, not deliver items, etc.). Online auctions such as those used by eBay are based on a type of auction referred to as an English auction. Dispite the English auction being the most popular type of auction, it has received less security coverage than other types of auctions (e.g., sealed-bid auctions). An existing proposal for a “secure” English auction prevents the Auctioneer from closing the auction early and from blocking bids, but does not protect a bidder’s anonymity. Another proposal provides anonymity, but does not stop an Auctioneer from skewing its clock or blocking bids. This paper proposes a new scheme for conducting secure and anonymous online English auctions using a modified type of group signature. Trust is divided among three servers owned by separate companies to ensure anonymity and fairness. Our scheme solves the problems of the existing English auction schemes and has following characteristics: unforgeability, anonymity, unlinkability, exculpability, coalition-resistance, verifiability, robustness, traceability, revocation, one-off registration, unskewability and unblockability. Our scheme has comparable efficiency to the existing schemes for the enhanced security and privacy it provides.

Secure Online English Auctions

Abstract: Security and privacy in online auctions is a major concern as auction participants have many opportunities to cheat (e.g., repudiate bids, not deliver items, etc.). Online auctions such as those used by eBay are based on a type of auction referred to as an English auction. Dispite the English auction being the most popular type of auction, it has received less security coverage than other types of auctions (e.g., sealed-bid auctions). An existing proposal for a “secure” English auction prevents the Auctioneer from closing the auction early and from blocking bids, but does not protect a bidder’s anonymity. Another proposal provides anonymity, but does not stop an Auctioneer from skewing its clock or blocking bids. This paper proposes a new scheme for conducting secure and anonymous online English auctions using a modified type of group signature. Trust is divided among three servers owned by separate companies to ensure anonymity and fairness. Our scheme solves the problems of the existing English auction schemes and has following characteristics: unforgeability, anonymity, unlinkability, exculpability, coalition-resistance, verifiability, robustness, traceability, revocation, one-off registration, unskewability and unblockability. Our scheme has comparable efficiency to the existing schemes for the enhanced security and privacy it provides.

Variable Quantity Market Clearing Algorithms

Abstract: Market clearing is the process of matching buy and sell bids in securities markets The allocative efficiency of such algorithms is important, as the Auctioneer is typically paid a commission on the number of bids matched and the volume of quantity traded Previous algorithms have concentrated on price issues This paper presents several market clearing algorithms that focus solely on allocating quantity among matching buy and sell bids The goal is to maximise the number of bids matched, while at the same time minimise the amount of unmatched quantity The algorithms attempt to avoid situations resulting in unmarketable quantities (ie, quantities too small to sell) Algorithmic performance is tested using simulated data designed to emulate the Australian Stock Exchange (ASX) and other world stock markets Our results show that it is difficult to avoid partial matchings as the complexity of doing so is NP-complete The optimal offline algorithm for partial quantity matching is used as a benchmark to compare online matching strategies We present three algorithms that outperform the ASX’s strategy by increasing the number of bids matched, the amount of quantity matched, and the number of bids fully matched

Randomised dynamic traitor tracing

Abstract: Dynamic traitor tracing schemes are used to trace the source of piracy in broadcast environments such as cable TV. Dynamic schemes divide content into a series of watermarked segments that are then broadcast. The broadcast provider can adapt the watermarks according to the pirate's response and eventually trace him/her. As dynamic algorithms are deterministic, for a given set of inputs, the tracing algorithm will execute exactly the same way each time. An adversary can use this knowledge to ensure that the tracing algorithm is forced into executing at its worst case bound. In this paper we review dynamic traitor tracing schemes and describe why determinism is a problem. We amend several existing dynamic tracing algorithms by incorporating randomised decisions. This eliminates any advantage an adversary has in terms of the aforementioned attack, as he/she no longer knows exactly how the tracing algorithm will execute. Simulations show that the randomising modifications influence each dynamic algorithm to run at its average case complexity in terms of tracing time. We provide an efficiency analysis of the amended algorithms and give some recommendations for reducing overhead.

Randomised Dynamic Traitor Tracing

Abstract: Dynamic traitor tracing schemes are used to trace the source of piracy in broadcast environments such as cable TV. Dynamic schemes divide content into a series of watermarked segments that are then broadcast. The broadcast provider can adapt the watermarks according to the pirate's response and eventually trace him/her. As dynamic algorithms are deterministic, for a given set of inputs, the tracing algorithm will execute exactly the same way each time. An adversary can use this knowledge to ensure that the tracing algorithm is forced into executing at its worst case bound. In this paper we review dynamic traitor tracing schemes and describe why determinism is a problem. We ammend several existing dynamic tracing algorithms by incorporating randomised decisions. This eliminates any advantage an adversary has in terms of the aforementioned attack, as he/she no longers knows exactly how the tracing algorithm will execute. Simulations show that the randomising modifications influence each dynamic algorithm to run at its average case complexity in terms of tracing time. We provide an efficiency analysis of the amended algorithms and give some recommendations for reducing overhead.