# improved security bounds for generalized feistel networks yaobinshen,chunguo,leiwang 3 table1:...

Post on 27-Jun-2020

0 views

Embed Size (px)

TRANSCRIPT

Improved Security Bounds for Generalized Feistel Networks

Yaobin Shen1, Chun Guo23(�) and Lei Wang1(�)

1 Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China

2 Key Laboratory of Cryptologic Technology and Information Security of Ministry of Education, Shandong University, Qingdao, Shandong, 266237, China

3 School of Cyber Science and Technology, Shandong University, Qingdao, Shandong, China yb_shen@sjtu.edu.cn,chun.guo@sdu.edu.cn,wanglei_hb@sjtu.edu.cn

Abstract. We revisit the security of various generalized Feistel networks. Concretely, for unbalanced, alternating, type-1, type-2, and type-3 Feistel networks built from random functions, we substantially improve the coupling analyzes of Hoang and Rogaway (CRYPTO 2010). For a tweakable blockcipher-based generalized Feistel network proposed by Coron et al. (TCC 2010), we present a coupling analysis and for the first time show that with enough rounds, it achieves 2n-bit security, and this provides highly secure, double-length tweakable blockciphers. Keywords: Block ciphers · Coupling · Tweakable block ciphers · Generalized Feistel networks · Provable security · Mode of operation

1 Introduction 1.1 Feistel Networks Feistel networks consist of several iterative applications of a simple Feistel permutation

ΨFi(A,B) = (B,A⊕ Fi(B)) (1)

for a domain-preserving function Fi : {0, 1}n → {0, 1}n that is typically called its round function. Such networks are not only the high level abstraction of a large number of modern blockciphers including the Data Encryption Standard (DES) [FNS75, Smi71], but also widely used in many other crypto systems (e.g., inverse-free authenticated encryption [Min14]).

A popular approach to analyzing the security of Feistel networks, pioneered by Luby and Rackoff [LR88], is to model the round function Fi as a secret random function. This allows proving its information theoretic indistinguishability, i.e., any distinguisher should not be able to distinguish the Feistel network from a random permutation on 2n-bit strings. With this model, Luby and Rackoff proved the security for 4-round Feistel networks, following which a long series of work has established either better security bounds [Pat90, Mau93, MP03, Vau03, Pat04, HR10a, Pat10] or reduced construction complexity [SP93, Pat93, Nan10, Nan15].

1.2 Generalized Feistel Networks (GFNs) The above classical Feistel networks could be generalized in various manners. Concretely, replacing the domain preserving round function Fi by expanding or contracting ones

Licensed under Creative Commons License CC-BY 4.0. IACR Transactions on Symmetric Cryptology ISSN XXXX-XXXX, Vol. 0, No. 0, pp. 1–33 DOI:XXXXXXXX

mailto:yb_shen@sjtu.edu.cn, chun.guo@sdu.edu.cn, wanglei_hb@sjtu.edu.cn http://creativecommons.org/licenses/by/4.0/ https://doi.org/XXXXXXXX

2 Improved Security Bounds for Generalized Feistel Networks

results in unbalanced Feistel [SK96]; using expanding and contracting round functions in an alternative manner results in alternating Feistel [AB96, Luc96]; finally, partition- ing the inputs into more than two blocks (or branches) results in multi-line generalized Feistel, and the (probably) most popular instances are Type-1, Type-2, and Type-3 Feistel networks [ZMI90], that differ in the relations among the branches. Compared to classical Feistel, the improved flexibility of GFNs significantly widens their applica- tion spectrum, ranging from ultra-lightweight blockciphers [SIH+11], full-domain secure encryption [MRS09], and wide cryptographic permutations [GM16].

Information theoretic security of GFNs could be analyzed in a model similar to classical Feistel, with various “birthday-bound” results showed in [NR99, MRS09, AB96, BR02, BRRS09, Luc96, ZMI90] and “beyond-birthday-bound” results found in [HR10a, Pat10]. Most importantly to this paper, Hoang and Rogaway (henceforth "HR") [HR10a] proved asymptotically optimal security for all the aforementioned types of GFNs via the coupling technique. In detail, with a sufficient number of rounds, all the aforementioned GFNs are CCA-secure up to 2n(1−ε) adversarial queries for any ε > 0. Though appearing nice, it requires a large number of rounds to asymptotically achieve n-bit security.

1.3 Tweakable Blockcipher-based GFN Tweakable permutation (TP) and tweakable blockciphers (TBC) were introduced by Liskov et al. [LRW02]: the former models a family of (efficiently invertible) permutations indexed by a parameter called the tweak, and the latter is a family of keyed TPs. With such primitives, the round function Fi of GFN may be replaced by some other primitives such as a TBC/TP, resulting in more possibilities.

As a concrete instance, Coron et al. [CDMS10] proposed a GFN that turns an n-bit TP with ω-bit tweak (ω > n) into a 2n-bit TP with (ω − n)-bit tweak, i.e., it trades the domain with the tweak space. As tweak extension is generally easier [CDMS10, MI15], this gives rise to a domain extender for TPs/TBCs. In this paper we denote by TGFr[ω, 2n] the r-round variant of Coron et al.’s construction. Coron et al. prove that TGFr[ω, 2n] achieves birthday 2n/2 CCA security when r = 2, and optimal 2n CCA security when r = 3. However, note that the size of the inputs to the underlying TP is actually larger than 2n-bit (i.e., n-bit block plus ω-bit tweak). As recently pointed out by Lee and Lee [LL18], the classical-sense optimal 2n security is actually the birthday-bound for such a TP. Motivated by Lee and Lee’s 24n/3 secure TBC construction, it’s tempting to ask if similar beyond 2n security results could be proved for TGFr[ω, 2n] with r ≥ 4 rounds.

1.4 Our Contributions For all the GFNs mentioned before, we either improve existing coupling analyzes or present new when non-existing. Concretely, motivated by Lampe and Seurin [LS15] and Nachef et al.’s [NPV17], we improve the coupling analyzes of HR [HR10a, HR10b], and prove the following results:

• For unbalanced Feistel UBFr[m,n], when n ≥ m, we prove 2qt+1 ( 4d nm eq+4q

2n ) t security

bound at (2d nme + 2)t + 2d n me + 1 rounds. The bound is comparable to HR’s

2q t+1 (

(3d nm e+3)q 2n )

t, while the number of rounds is almost halved from HR (4d nme+ 4)t. When n < m, we prove 2qt+1 (

4dmn eq 2n )

t security bound (the same as HR’s bound) at 4t+ 2d nme+ 1 rounds which is much smaller than HR’s (2d

m n e+ 4)t rounds.

• For alternating Feistel ALFr[m,n], we prove 2qt+1 ( 6d nm eq+3q

2n ) t security bound with

(12d nme+ 2)t+ 5 rounds (compared with 2q t+1 (

(6d nm e+3)q 2n )

t with (12d nme+ 8)t rounds of HR). The same improvement holds for numeric alternating Feistel.

Yaobin Shen, Chun Guo, Lei Wang 3

Table 1: Summary of improved CCA bounds in this paper. The rows correspond to the generalized Feistel networks illustrated in Fig. 1 and Fig. 2. Parameters k,m, n, ω,M,N describe the scheme and t determines the number of rounds r. Scheme Previous Bound #rounds Our Bound #rounds UBFr[m,n] n ≥ m 2qt+1

( (3d nm e+3)q

2n

)t (4d nme+ 4)t [HR10a]

2q t+1

( 4d nm eq+4q

2n

)t (2d nme+ 2)t+ 2d

n me+ 1

n < m 2qt+1

( 4dmn eq

2n

)t (2dmn e+ 4)t [HR10a]

2q t+1

( 4d nm eq

2n

)t 4t+ 2d nme+ 1

ALFr[m,n] 2qt+1 (

(6d nm e+3)q 2n

)t (12d nme+ 8)t [HR10a]

2q t+1

( 6d nm eq+3q

2n

)t (12d nme+ 2)t+ 5

NALFr[M,N ] 2qt+1 ( (6dlogM Ne+3)q

N ) t (12dlogM Ne+ 8)t [HR10a]

2q t+1

( 6dlogM Neq+3q

N

)t (12dlogM Ne+ 2)t+ 5

Feistel1r[k, n] 2qt+1 (

2k(k2−k+1)q 2n

)t (2k2 + 2k)t [HR10b] 2qt+1

( 2k(k−1)q

2n

)t (k2 + k − 2)t+ 1

Feistel2r[k, n] 2qt+1 (

2k(k−1)q 2n

)t (2k + 2)t [HR10b] 2qt+1

( 2k(k−1)q

2n

)t 2kt+ 1

Feistel3r[k, n] 2qt+1 (

4(k−1)2q 2n

)t (k + 4)t [HR10b] 2qt+1

( 4(k−1)2q

2n

)t (k + 2)t+ 1

TGFr[ω, 2n] q2n 3 [CDMS10] 2 · (

q t+1

( 30q 22n

)t)1/2 4t+ 2

• For multi-line GFNs Feistel1r[k, n] and Feistel2r[k, n], we prove 2qt+1 (

2k(k−1)q 2n

)t secu-

rity bound with (k2 + k − 2)t+ 1 rounds, and 2qt+1 (

2k(k−1)q 2n

)t with 2kt+ 1 rounds

resp. (compared with 2qt+1 ( 2k(k2−k+1)q

2n ) t with (2k2 + 2k)t rounds, and 2qt+1 (

2k(k−1)q 2n )

t

with (2k + 2)t rounds of HR resp.).

• for type-3 GFN Feistel3r[k, n], we prove 2qt+1 ( 4(k−1)2q

2n ) t security bound with (k+2)t+1

rounds (compared with 2qt+1 ( 4(k−1)2q

2n ) t with (k + 4)t rounds of HR).

For the TBC-based GFN TGFr[ω, 2n], we present the first coupling analysis and prove

2 · (

q t+1

( 30q 22n

)t)1/2 security bound with 4t+ 2 rounds. This for the first time establishes

beyond the birthday bound 2n for TGFr[ω, 2n]. Moreover, it also approaches 22n as the number of rounds t increases. This gives rise to double-length blockciphers with high security: for example, when Deoxys-BC-256 is used, 10 rounds achieve 2 4×1283 ≈ 2170 security. While the efficiency is relatively low, the high security bounds make it suitable in specific application.

1.4.1 Core Ideas for Improvements

Our improvements upon HR [HR10a] are due to more fine-grained analyses of the coupling probabilities. To further illustrate, consider for example the unbalanced Feistel with contracting roun