# US 2024/0187206 A1 · Method and system for protecting cryptographic operations against side-channel attacks

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0187206 A1
- Document ID: US20240187206A1
- Status: Published application
- Kind: A1
- Application number: 18/212,372
- Filed: 2023-06-21
- Published: 2024-06-06
- Earliest claimed: 2022-06-22
- Assignees: Secure-Ic SAS
- Applicants: Secure-Ic SAS
- Inventors: Sofiane TAKARABT; Sylvain GUILLEY
- Web page: https://patentgrep.com/patent/US20240187206A1

## CPC

- H04L 9/002: {Countermeasures against attacks on cryptographic mechanisms (network architectures or network communication protocols for protection against malicious traffic H04L63/1441)}
- H04L 9/003: {for power analysis, e.g. differential power analysis [DPA] or simple power analysis [SPA]}
- H04L 9/0618: {Block ciphers, i.e. encrypting groups of characters of a plain text message using fixed encryption transformation}
- H04L 9/085: {Secret sharing or secret splitting, e.g. threshold schemes}
- H04L 9/3093: {involving Lattices or polynomial equations, e.g. NTRU scheme}
- H04L 2209/046: of operations, operands or results of the operations
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A cryptographic system for executing operations of a cryptographic scheme applied to encrypt a data block is provided. The cryptographic system comprises a first sharing processing unit configured to execute an arithmetic sharing function applied to the data block and providing as an output a first and a second arithmetic share component. The cryptographic system further comprises a second sharing processing unit and a computation processing unit. The second sharing processing unit is configured to determine a random component and to execute a Boolean sharing function applied to the first arithmetic share component and to the random component and providing as an output an intermediate Boolean share component. The computation processing unit is configured to execute a recursive carry computation function configured to compute a first, a second and a third Boolean share component from the second arithmetic share component, the random component and the intermediate Boolean share component.

## Claims

### Claim 1

A cryptographic system for executing operations of a cryptographic scheme, said operations being applied to a data block u to generate encrypted data corresponding to a protected data block X, wherein said cryptographic system comprises a first sharing processing unit configured to execute an arithmetic sharing function applied to said data block u using at least one secret key, and providing as an output a first arithmetic share component A and a second arithmetic share component M, wherein the cryptographic system further comprises a second sharing processing unit and a computation processing unit, said second sharing processing unit being configured to determine a random component A(0), said second sharing processing unit being further configured to execute a Boolean sharing function applied to said first arithmetic share component A and to said random component A(0) and providing as an output an intermediate Boolean share component A(1), said computation processing unit being configured to execute a recursive carry computation function configured to compute a first Boolean share component X(0), a second Boolean share component X(1) and a third Boolean share component X(2) from said second arithmetic share component M, said random component A(0) and said intermediate Boolean share component A(1), and
wherein said protected data block X being generated from said first Boolean share component X(0), said second Boolean share component X(1) and said third Boolean share component X(2).

### Claim 2 (depends on 1)

The cryptographic system of claim 1, wherein said intermediate Boolean share component A(1) and said second arithmetic share component M are associated with said k bit-precision, said computation processing unit being configured to initialize three carry values C0(0), C0(1), and C0(2) to class zero,
wherein the computation processing unit is configured to execute the recursive carry computation function by iteratively determining three carry components Ci+1(0), Ci+1(1), and Ci+1(2) for each index i, initially set to 0, until i=k−2, the index i being incremented for each next iteration, the three carry components Ci+1(0), Ci+1(1), and Ci+1(2) at a current iteration corresponding to an index i being determined according to:
Ci+1(0)=Mi⊗(Ai(1)⊕Ci(1)⊕Ci(2))⊕Ai(1)⊗(Ci(1)⊕Ci(2));
Ci+1(1)=Mi⊗(Ai(0)⊕Ci(0))⊕Ai(0)⊗(Ci(0)⊕Ci(1));
Ci+1(2)=Ai(1)⊗Ci(0)⊕Ai(0)⊗Ci(2);
and wherein said computation processing unit is configured to determine the 3-sharing Boolean components Xi(0), Xi(1) and Xi(2) corresponding to each index i comprised between 0 and k according to:
Xi(0)=Ci(0)⊕Mi;
Xi(1)=Ci(1)⊕Ai(0);
Xi(2)=Ci(2)⊕Ai(1).

### Claim 3 (depends on 1)

The cryptographic system of claim 1, wherein said Boolean sharing function is executed for each ith bit corresponding to an index i ∈[0, k−1] according to the following equation: Ai(1)=Ai(0)⊕Ai.

### Claim 4 (depends on 1)

The cryptographic system of claim 1, wherein said cryptographic scheme is a lattice-based cryptography scheme.

### Claim 5 (depends on 4)

The cryptographic system of claim 4, wherein said random component A(0) being determined with a bit-precision value k and wherein said cryptographic system further comprises an additional computation processing unit configured to determine an optimum coefficient value Qk−1 from an initial coefficient value Q chosen from prime numbers, and to provide the bit-precision value k.

### Claim 6 (depends on 5)

The cryptographic system of claim 5, wherein said additional computation processing unit is configured to initially set an initial value km of said bit-precision value k to [log2 Q] and determine a critical value s as
s
=
⌊
Q
2
⌋
,
and wherein said additional computation processing unit is configured to determine said optimum coefficient value Qk−1 by iteratively computing, for each value k, k being initially set to km and being incremented for each next iteration:
a coefficient value Qk−1 corresponding to the current value of k according as [2k/Q],
two intermediate comparison values t1=s×Qk−1 and t2=(s+1)×Qk−1, the iterations being terminated if t1<2k−1 and t2≥2k−1,
the coefficient value Qk−1 and the bit-precision value k corresponding to the last iteration being returned by the additional computation processing unit.

### Claim 7 (depends on 5)

The cryptographic system of claim 5, wherein said cryptographic system further comprises two converting processing units configured to multiply respectively each of the two arithmetic share components A and M by said optimum coefficient value Qk−1.

### Claim 8 (depends on 7)

The cryptographic system of claim 7, wherein said converting processing unit is further configured to apply a [q/4] offset to the first arithmetic share component A.

### Claim 9 (depends on 4)

The cryptographic system of claim 4, wherein said cryptographic system further comprises a first evaluation processing unit and a second evaluation processing unit, the first evaluation processing unit being configured to determine a first final boolean share component B comprising a set of elements Bi, i ∈[0, k−1], each component Bi being determined as Bi=Xi(0)⊕Xi(1), and wherein the second evaluation processing unit is configured to determine a second final boolean share component N comprising a set of elements Ni, i∈[0, k], each component Ni being determined equal to said third Boolean share component Xi(2),
and wherein said first and second evaluation processing units are configured to return the kth elements Bk−1 and Nk−1 of the first final boolean share component and of the first final boolean share component.

### Claim 10 (depends on 1)

The cryptographic system of claim 1, wherein said second sharing processing unit and said computation processing unit comprise electronic circuits comprising electronic logic gate.

### Claim 11

A method of executing operations of a cryptographic scheme, the method comprising the step of executing an arithmetic sharing function applied to a data block u and providing as an output a first arithmetic share component A and a second arithmetic share component M, wherein the method comprises the steps of:
determining a random component A(0) with a bit-precision value k, and executing a Boolean sharing function to said first arithmetic share component A and to said random component A(0) and providing as an output an intermediate Boolean share component A(1),
executing a recursive carry computation configured to compute a first Boolean share component X(0), a second Boolean share component X(1) and a third Boolean share component X(2) from said second arithmetic share component M, said random component A(0) and said intermediate Boolean share component A(1).

### Claim 12 (depends on 11)

A method of claim 11, wherein said intermediate Boolean share A(1) and said second arithmetic share component M are associated to said k bit-precision, the method comprising the step of initializing to class zero three carry values C0(0), C0(1), and C0(2),
wherein said recursive carry computation comprises the step of iteratively determining three carry components Ci+1(0), Ci+1(1), and Ci+1(2) for each index i, initially set to 0, until i=k−2, the index i being incremented for each next iteration, the three carry components Ci+1(0), Ci+1(1), and Ci+1(2) at a current iteration corresponding to an index i being determined according to:
Ci+1(0)=Mi⊗(Ai(1)⊕Ci(1)⊕Ci(2))⊕Ai(1)⊗(Ci(1)⊕Ci(2));
Ci+1(1)=Mi⊗(Ai(0)⊕Ci(0))⊕Ai(0)⊗(Ci(0)⊕Ci(1));
Ci+1(2)=Ai(1)⊗Ci(0)⊕Ai(0)⊗Ci(2);
and wherein the method comprises the step of determining the 3-sharing boolean components Xi(0), Xi(1) and Xi(2) corresponding to each index i comprised between 0 and k according to:
Xi(0)=Ci(0)⊕Mi;
Xi(1)=Ci(1)⊕Ai(0);
Xi(2)=Ci(2)⊕Ai(1);

### Claim 13 (depends on 11)

A method of claim 11, wherein said Boolean sharing function comprise the step of executing for each ith bit corresponding to an index i ∈[0, k−1] according to the following equation Ai(1)=Ai(0)⊕Ai.
