Valid

JR/T 0196-2020Multi-party secure computing financial application technical specifications (English PDF)

Preview PDF

This is a limited preview

Buy now to download the full PDF (12 pages)

Need quotation for your management?

Get a formal quotation for JR/T 0196-2020 in 30 minutes.

Issued by

People's Bank of China

Level / Type

National · Recommended

Issue date

November 24, 2020

Implementation date

November 24, 2020

Scope

JR/T 0196-2020 (Multi-party secure computing financial application technical specifications) is available as an English-translated PDF.

JR/T 0196-2020 — This document specifies the basic requirements, security requirements, performance requirements, etc. of the financial application of multi-party secure computing technology.

This document is applicable to the product design and software development of financial institutions that carry out multi-party secure computing financial applications.

Found what you need?

Document preview — JR/T 0196-2020

National Standard of the People's Republic of China

Classification: A 11

Issued by: People's Bank of China

Foreword
1 Scope
2 Normative references
3 Terms and definitions
4 Abbreviations
5 Overview
5.1 MPC participants and work sequence
5.2 Application goals
5.3 General requirements
6 Basic requirements
6.1 Data input

Foreword

This document starts from the provisions of GB/T 1.1-2020 "Guidelines for Standardization Work Part 1: Structure and Drafting Rules of Standardization Documents" grass.

1 Scope

This document specifies the basic requirements, security requirements, performance requirements, etc. of the financial application of multi-party secure computing technology.

This document is applicable to the product design and software development of financial institutions that carry out multi-party secure computing financial applications.

2 Normative references

The contents of the following documents constitute the indispensable clauses of this document through normative references in the text. Among them, dated reference documents, Only the version corresponding to that date is applicable to this document; for undated references, the latest version (including all amendments) is applicable to this document file.

GB/T 18391.1-2009 Information Technology Metadata Registration System (MDR) Part 1: Framework GB/T 32400-2015 Information Technology Cloud Computing Overview and Vocabulary

3 Terms and definitions

The following terms and definitions apply to this document.

Secure multi-party computation; MPC A method based on multi-party data collaboration to complete the calculation goal, and achieve no leakage of the privacy data of each party except for the calculation result and the information that can be deduced According to the cryptographic technology.

Note. The technologies often used in multi-party secure computing include obfuscated circuits, inadvertent transmission, secret sharing, homomorphic encryption, etc.

Party

A natural or legal person involved in multi-party secure computing.

[Source. GB/T 32400-2015, 3.1.6, with modification] 3.3 Calculation factor Data generated based on multi-party safe calculation of input data.

Note. Including input factor, output factor and intermediate factor. input factor means that the data provider can perform subsequent calculations after the data provider performs the data input process the output factor refers to the data returned to the result user to restore the final calculation result after the calculation party performs the calculation; the intermediate factor refers to the data in the calculation party Data generated during the calculation process.

Data input

Secret sharing, obfuscation circuit, homomorphic encryption and other technologies are used to process the private data provided by the data provider to transform the data into the process of entering factors.

Data output

The process of using secret sharing, obfuscating circuits, homomorphic encryption and other technologies to process the output factors to obtain the calculation results.

Computation node

The software, computer, virtual computer or cluster that the computing party executes the multi-party secure computing protocol or algorithm logic.

Note. A computing party corresponds to a computing node and management domain, and provides an external interactive interface, such as IP address, port, etc.

Security model

Assumptions made about the participant's behavioral patterns.

Note. Different MPC protocols can be based on different security models. The security models can be divided into two types. semi-honest model and malicious attack model. The semi-honest model is participation when contacting and processing the private data of other parties, the party shall, on the basis of strictly abiding by the agreement and specifications, try its best to mine from the contacted and processed data Provide effective information; the malicious attack model is that the participants may perform any behavior and do their best to obtain effective information about private data, such as deviation from the agreement.

Disputes, collusion with others, etc., such a participant is also called a dishonest participant.

Security parameter

A set of parameters used to measure the security strength or cracking difficulty of a multi-party secure computing protocol.

Note. MPC security parameters mainly include dishonesty thresholds, statistical security parameters, and calculated security parameters. The dishonesty threshold is that the multi-party secure computing protocol allows collusion the maximum value of dishonest participants, when the value is less than half of the number of participants, the agreement is said to be the honest majority, otherwise the agreement is the dishonest majority the statistical security parameter is an integer l, the probability distribution of the calculation factor generated according to the input data, and the calculation of random simulation without knowing the input data Calculate the probability distribution of the factor, the two are statistically indistinguishable (the statistical distance is not higher than 2-l); the calculated safety parameter is an integer k, which represents a polynomial the computational complexity for an attacker to crack a multi-party secure computing protocol is O(2k).

Dataset

One or more data providers participate in the data collection of multi-party secure calculations.

3.10

Metadata

Data that defines and describes other data.

[Source. GB/T 18391.1-2009, 3.2.16] 3.11 Calculation engine A group of computing nodes that perform multi-party secure computing formed by each computing party connected through a network.

3.12

Private data

The data entered by the data provider, the data obtained by the result user, and the data in the algorithm parameters and model parameters that need to be protected.

3.13

Effective number of digits enob for values that have no decimal places and end with several zeros, the number of digits from the leftmost digit of the non-zero digit to the right is subtracted from invalid zeros (that is, only is the number of zeros used for positioning. For other numerical values, the number of digits obtained by counting from the leftmost digit of a non-zero number to the right.

3.14

MPC accuracy

Used to measure the accuracy of multi-party security calculation results.

Note. Compared with the plaintext calculation result of the same data, the more consecutive identical effective digits, the higher the accuracy. For the case where there are multiple values in the calculation result, you can in practice, the accuracy of each value is measured or the accuracy is calculated after fitting multiple values into one value.

3.15

Correctness

It is used to measure the consistency of multi-party security calculations with the same data plaintext calculation results within a certain MPC accuracy range.

4 Abbreviations

The following abbreviations apply to this document.

CA. Certification Authority RA. Registration Authority (Registration Authority) TPS. The number of transactions per second (Transactions Per Second)

5 Overview

5.1 MPC participants and work sequence

5.1.1 MPC participants

The MPC participants explained as follows.

a) Task initiator. Trigger MPC tasks, complete task resource configuration before task execution, and verify the availability of resources.

b) Scheduler. Configure computing tasks, manage and coordinate other participants to perform tasks.

c) Algorithm provider. Provide calculation logic and algorithm parameters for MPC. When algorithm parameters have protection requirements, the algorithm parameters should be considered for private data, the algorithm provider is regarded as the data provider.

d) Data provider. Provide MPC with the required privacy data, and convert the privacy data into input factors through MPC data input processing.

Give it to the computing party. The number of data providers in an MPC calculation task is greater than or equal to 2.

e) Computing party. Provide computing power support for MPC. The computing party receives the input factors of the data provider and performs calculations.

The output factor is sent to the consumer of the result. The number of computing parties in an MPC computing task should be greater than or equal to 2, and cannot be determined by the same entity.

The entity assumes the role of multiple computing parties.

f) Result consumer. receiving the MPC calculation result, there can be one or more result consumers of an MPC calculation task.

5.1.2 Working sequence

The MPC task sequence includes steps such as task creation, task allocation, data input, task calculation, and result analysis, as shown in Figure 1.MPC the task sequence is detailed as follows.

a) Task creation.

1) The task initiator configures and verifies the resources required for the MPC task calculation, and initiates the calculation task.

2) The data provider authorizes all data usage, except when the task initiator and the data provider are the same entity.

The data provider can entrust the dispatcher to authorize the use of the data, or pre-authorize the data before the task is created. number According to the use authorization and subsequent task allocation phases can be combined and executed.

b) Assignment of tasks.

1) The dispatcher verifies the legitimacy of the task request information, including the legitimacy of identity verification and data authorization.

2) After the verification is passed, the task configuration information is generated and sent to the data provider, the calculation party and the result user.

3) The data provider, the calculation party and the result user will verify after receiving the task configuration information.

4) Each participant saves the task configuration information sent and received.

c) Data input.

1) The data provider reads data from the data source and generates input factors, which are sent to the designated computing party through a secure channel.

2) The data provider saves the task configuration information and records the sent input factors.

d) Task calculation.

1) The computing node receives the input factors of each data provider, and performs collaborative calculations to generate output factors in accordance with the MPC protocol.

2) Send the output factor to the user of the result.

e) Result analysis.

1) The user of the result analyzes the output factor to obtain the calculation result.

2) Document the results.

Figure 1 MPC working sequence diagram

5.2 Application goals

MPC should meet the requirements of data privacy, data legality, correctness of calculation results, and acceptability of calculation performance. The specific requirements are as follows.

a) The private data of each participant should not be obtained or inferred by other participants. The user of the result can divide the information derived from the calculation result outer.

b) The original private data of other participants should not appear in the calculation process.

c) Each participant shall participate in the MPC calculation according to the agreed role of the calculation task.

d) The private data used in computing tasks should be authorized by the corresponding data provider in advance.

e) The calculation result should meet the correctness requirements and be obtained only by the user of the result.

f) Computing performance should meet specific application requirements.

5.3 General requirements

The overall requirements of MPC financial applications include three parts. basic requirements, safety requirements and performance requirements, as shown in Figure 2.

The basic requirements include data input, algorithm input, collaborative computing, result output and scheduling management, etc., which are mainly for data extraction.

Supplier, algorithm provider, computing party, result user, dispatcher.

Security requirements include requirements for protocol security, privacy data security, authentication and authorization, password security, communication security, evidence storage and logs.

Performance requirements put forward performance index requirements for MPC financial applications such as calculation delay, throughput, and calculation accuracy.

Figure 2 Overall requirements for MPC financial applications

6 Basic requirements

6.1 Data input

The specific data input requirements of the data provider are as follows.

a) The data provider shall convert the private data into input factors, provide them to the designated computing node, and ensure that they are under the set security model the input data cannot be derived from the input factor.

b) The data provider shall conduct unified management of data sources, data sets, metadata, etc..

1) Data source management.

--Should support different types of data source access, including but not limited to databases and files, database types such as relational data Database, column database, data warehouse, etc., file types such as txt, csv, xml, key-value, etc.;

--Extensible to support new data types.

2) Data set management.

--Should support the addition and deletion of data sets;

--Should support the user, purpose and amount of the specified data set;

--Should support the data set access status query function, and display the status of all data set access tasks;

--Should support the function of monitoring the data set's participation in the calculation status, such as being involved in the calculation, finished use, etc.

3) Metadata management.

--Should support the use of metadata to describe data sets;

-Should support metadata query function, including name, mark, description, size, sample, type and other information;

--Should support the provision of metadata information of the data set to the data demander.

c) Data should be available...

......
This preview omits tables, figures, formulas and parts of the technical clauses. The complete document — 12 pages — is available in the English PDF.

Referenced standards

Normative references

GB/T 18391.1-2009Information technology - Metadata registries (MDR) - Part 1: Framework

GB/T 32400-2015

How to Download JR/T 0196-2020

1
Add to cart
Click "Download PDF" on this page and choose "Buy this standard". You can add more standards before checkout.
2
Checkout
Enter your email and billing details. Payment is processed securely by Stripe (cards or bank transfer). Prefer bank details? Request a quotation (min. order $300).
3
Instant delivery (0–9 sec)
Delivery is automatic: within seconds of payment you'll receive an email with a secure download link. The link stays valid for 72 hours.
4
Invoice included
A tax invoice is attached to the confirmation email. Need a custom invoice? Contact us.