11.07.2014 Views

Identity-Based Encryption Protocols Using Bilinear Pairing

Identity-Based Encryption Protocols Using Bilinear Pairing

Identity-Based Encryption Protocols Using Bilinear Pairing

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Encrypt: Let v = (v 1 , . . . , v j ) be the identity under which a message M ∈ G 2 is to be<br />

encrypted. Choose s to be a random element of Z p . The ciphertext is<br />

(C 0 = M × e(P 1 , P 2 ) s , C 1 = sP, B 1 = sV 1 (v 1 ), . . . , B j = sV j (v j )).<br />

Decrypt: Let C = (C 0 , C 1 , B 1 , . . . , B j ) be a ciphertext and the corresponding identity<br />

v = (v 1 , . . . , v j ). Let (d 0 , d 1 , . . . , d j ) be the decryption key corresponding to the identity v.<br />

The decryption steps are as follows.<br />

Verify whether C 0 is in G 2 , C 1 and the B i ’s are in G 1 . If any of these verifications fail,<br />

then return bad, else proceed with further decryption as follows. Return<br />

∏ j<br />

k=1<br />

C 0 ×<br />

e(B i, d i )<br />

.<br />

e(d 0 , C 1 )<br />

It is standard to verify the consistency of decryption.<br />

6.3 Security<br />

We first state the result on security and discuss its implications.<br />

Theorem 6.3.1. The protocol HIBE-spp described in Section 6.2 is (ɛ hibe , t, q)-IND-ID-CPA<br />

secure assuming that the (t ′ , ɛ dbdh )-DBDH assumption holds in 〈G 1 , G 2 , e〉, where ɛ hibe<br />

2ɛ dbdh /λ; t ′ = t + χ(ɛ hibe ) and<br />

χ(ɛ) = O(τq + O(ɛ −2 ln(ɛ −1 )λ −1 ln(λ −1 ));<br />

τ is the time required for one scalar multiplication in G 1 ;<br />

λ = 1/(2(2σ(µ l + 1)) h ) with µ l = l(N 1/l − 1), N = 2 n and σ = max(2q, 2 n/l ).<br />

We further assume 2σ(1 + µ l ) < p.<br />

Before proceeding to the proof, we discuss the above result. The main point of the theorem<br />

is the bound on ɛ hibe . This is given in terms of λ and in turn in terms of µ l . We simplify<br />

this bound.<br />

Since l ≥ 1, we have 1 + µ l = 1 + l(N 1/l − 1) ≤ lN 1/l = l2 n/l . Consequently,<br />

ɛ hibe ≤ 2ɛ dbdh<br />

= 4(2σ(µ l + 1)) h ɛ dbdh<br />

λ<br />

≤ 4(2σl2 n/l ) h ɛ dbdh<br />

= 4(2l2 n/l ) h σ h ɛ dbdh (6.3.2)<br />

The reduction is not tight; security degrades by a factor of 4(2l2 n/l ) h σ h . We now consider<br />

several cases. The actual value of degradation depends on the value of q, the number of key<br />

extraction queries made by the adversary. A value of q used in the previous chapter and also<br />

in earlier analysis is q = 2 30 [47]. We will use this value of q in the subsequent analysis.<br />

65<br />

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!