Network and storage latency attacks to online trading protocols in the cloud

Abstract

Online trading protocols enable participants to trade, barter, or sell goods and services over a private network or the global Net. Due to diversity of network and computational resources at their disposal, participants communicate and carry out their trading with different latencies. This scenario may give rise to latency attacks, where malicious parties (a.k.a. fast traders) exploit lower latency to attack trading protocols’ fairness and increase their income. With the advent of the cloud, the problem of identifying and preventing latency attacks is exacerbated by the fact that cloud providers and privileged users could collude to make latency attacks simpler and more effective. In this paper, we give an overview of network and storage latency attacks in multi-party trading protocols, focusing on cloud peculiarities and providing some empirical recommendations for protocol design.