dsac (digital signature aggregation and chaining)

27
DSAC (Digital Signature Aggregation and Chaining) Digital Signature Aggregation & Chaining An approach to ensure integrity of outsourced databases

Upload: heinz

Post on 01-Feb-2016

34 views

Category:

Documents


0 download

DESCRIPTION

DSAC (Digital Signature Aggregation and Chaining). Digital Signature Aggregation & Chaining An approach to ensure integrity of outsourced databases. Contents. Signature Aggregation Mechanisms Chaining Mechanism Comparison of the results with previous work. ODB. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: DSAC (Digital Signature Aggregation and Chaining)

DSAC(Digital Signature Aggregation and Chaining)

Digital Signature Aggregation & Chaining

An approach to ensure integrity of

outsourced databases

Page 2: DSAC (Digital Signature Aggregation and Chaining)

Contents

Signature Aggregation Mechanisms

Chaining Mechanism

Comparison of the results with previous work

Page 3: DSAC (Digital Signature Aggregation and Chaining)

ODB

Outsourced Data Base(ODB) model : Client stores its data at an external data base service provider.

Concern: Ensure the database security & integrity.

Page 4: DSAC (Digital Signature Aggregation and Chaining)

Authenticity: The tuples in the result set have not been tampered i.e correctness.

Integrity: No valid tuples have been omitted from the result set i.e completeness

Page 5: DSAC (Digital Signature Aggregation and Chaining)

Result set & Size of a result set

Result result includes all the tuples matching the query predicates.

Size : 0-n, or 2^n subsets,

where, n is total number of tuples in the database.

Page 6: DSAC (Digital Signature Aggregation and Chaining)

Merkle Hash Tree

Use to prove existence of an element in a set. For eg. prove x1 exists in the set y={x2, x6, x1, x9}

Constructed as binary tree where leaves are hash value of corresponding element.

Non leaf & Leaf nodesRoot of the MHT is digitally signed using

public key signature scheme (RSA/ DSA)

Page 7: DSAC (Digital Signature Aggregation and Chaining)

MHT example…

Page 8: DSAC (Digital Signature Aggregation and Chaining)

Auth DS (Authenticated Data Structures)

Approach to prove correctness

Uses MHT to prove correctness of the result set.

Limitation : Need to pre-compute and store a potentially large number of authenticated data structures to answer queries.

Completeness issue not answered

Page 9: DSAC (Digital Signature Aggregation and Chaining)

VB Tree Approach

Uses a modified MHT

Not only root of MHT is signed but all nodes as well

Limitation: Consumes large storage space and increased verification time.

Provides proof of correctness

Completeness issue not answered !

Page 10: DSAC (Digital Signature Aggregation and Chaining)

Drawbacks…Overheads associated with building,

storing and updating data structures in AuthDS and VB tree.

Signs each individual tuple before storing.Server stores tuples along with its

corresponding signature.In response to a query, server sends both

tuple and its signature.

Page 11: DSAC (Digital Signature Aggregation and Chaining)

Drawbacks(contd.)

Query reply set consists of thousands of

tuples.

Sending/ receiving and verifying signature of each tuple.

Expensive for the querier.

Page 12: DSAC (Digital Signature Aggregation and Chaining)

DSAC: Correctness

Combines multiple individual signatures in the result set into a unified/ aggregated signature.

Verifying a unified signature is same as verifying signatures of each individual tuple in the result set.

Page 13: DSAC (Digital Signature Aggregation and Chaining)

DSAC: Completeness

Includes the boundary tuples as well to ensure all the tuples matching the query is returned.

Link the tuple level signatures to form a signature chain.

Page 14: DSAC (Digital Signature Aggregation and Chaining)

Constructing signature chains

If h() is a hash function such as SHA, || denotes concatenation, IPRi denotes immediate predecessor tuple

along dimension ‘i’ , l being number of searchable dimensions, SK is private signing key of the data owner

Page 15: DSAC (Digital Signature Aggregation and Chaining)

then the signature of a tuple ‘r’ can be computed as follows

Page 16: DSAC (Digital Signature Aggregation and Chaining)

Computing IPR of a tuple

Sort tuple in increasing order of the attribute value for each dimension.

IPR of a given tuple in a given dimension is a tuple with highest value of the attribute that is less than the value of that tuple.

Each tuple has as many IPRs as the number of searchable dimensions.

Page 17: DSAC (Digital Signature Aggregation and Chaining)

Example of signature chainingConsider tuple R5

Page 18: DSAC (Digital Signature Aggregation and Chaining)

Completeness (contd.)

In this way, server answers range queries by releasing all matching tuples, boundary tuples as well as aggregated signature.

Signature chain proves querier that server has returned all tuples in the query range proving completeness.

Page 19: DSAC (Digital Signature Aggregation and Chaining)

Compleness(contd.)

Querier on receiving the result set:

Verifies the values in boundary tuples are just beyond the query range.

Page 20: DSAC (Digital Signature Aggregation and Chaining)

Building a result set

Compute the tuple set Ts={Ra…Rz} Compute Tn consisting of immediate

predecessor and successor nodes

Tn= {R(a-1), R(b+1)} Obtain corresponding signature of each tuple Calculate the aggregate the signature

Page 21: DSAC (Digital Signature Aggregation and Chaining)

(Contd)

Chain the signature of all tuples along with its corresponding IPR

Now, the result consists of {Ts, Tn, Sign(r), ∑}

Page 22: DSAC (Digital Signature Aggregation and Chaining)

Analysis of DSAC scheme

We compare the DSAC scheme with other prominent correctness/ completeness guarantee schemes such as AuthDS and VB tree.

Page 23: DSAC (Digital Signature Aggregation and Chaining)

Query Verification Time (Naïve approach vs DSAC)

Page 24: DSAC (Digital Signature Aggregation and Chaining)

VO Size (Naïve approach vs DSAC approach)

Page 25: DSAC (Digital Signature Aggregation and Chaining)

Freshness

Freshness : The result set in response to a query should be the recent snapshot of the database.

Prevents the server from replaying the old signature chains, hence freshness is part of data integrity concerns.

Page 26: DSAC (Digital Signature Aggregation and Chaining)

Further scope

How to reduce the size of the verification object.

{Ts, Tn, Sign(r), ∑}

Freshness Issues

Page 27: DSAC (Digital Signature Aggregation and Chaining)

Reference

DSAC : An approach to ensure integrity of outsourced databases using signature aggregation and chaining

- Authors : Maithili Narasimha & Gene Tsudik

Computer Science Department

University of California, Irvine