Proof-infused streams: enabling authentication of sliding window queries on streams

  • Authors:
  • Feifei Li;Ke Yi;Marios Hadjieleftheriou;George Kollios

  • Affiliations:
  • Boston University;AT&T Labs Inc.;AT&T Labs Inc.;Boston University

  • Venue:
  • VLDB '07 Proceedings of the 33rd international conference on Very large data bases
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

As computer systems are essential components of many critical commercial services, the need for secure online transactions is now becoming evident. The demand for such applications, as the market grows, exceeds the capacity of individual businesses to provide fast and reliable services, making outsourcing technologies a key player in alleviating issues of scale. Consider a stock broker that needs to provide a real-time stock trading monitoring service to clients. Since the cost of multicasting this information to a large audience might become prohibitive, the broker could outsource the stock feed to third-party providers, who are in turn responsible for forwarding the appropriate sub-feed to clients. Evidently, in critical applications the integrity of the third-party should not be taken for granted. In this work we study a variety of authentication algorithms for selection and aggregation queries over sliding windows. Our algorithms enable the end-users to prove that the results provided by the third-party are correct, i.e., equal to the results that would have been computed by the original provider. Our solutions are based on Merkle hash trees over a forest of space partitioning data structures, and try to leverage key features, like update, query, signing, and authentication costs. We present detailed theoretical analysis for our solutions and empirically evaluate the proposed techniques.