Encryption Library Research Articles

Better approximation of sigmoid function for privacy-preserving neural networks

Abstract With the development of artificial intelligence technology, privacy protection has become a prominent research direction. Homomorphic encryption is an important and promising technology in the construction of privacy-preserving neural networks. However, computing nonlinear activation functions remains a challenging task in the context of homomorphic encryption. Considering Sigmoid is one of the most commonly used activation functions, this paper proposes a polynomial approximation method for the Sigmoid function and applies it to the homomorphic encryption scheme of CKKS. We design polynomial approximation functions using Improved Taylor Expansion Segmentation (ITES), Least Squares method (LS), and Random Noise Addition method, respectively. These methods are implemented through the CKKS scheme in the homomorphic encryption library of SEAL. The experimental results show that the ITES achieves high fitting accuracy, reducing the average absolute error by 11.09% and the mean squared error by 38.06% compared to previous fitting methods. Therefore, experiments have identified methods with better fitting results in the CKKS scheme, which can meet the requirements of using homomorphic encryption in deep learning technologies. Finally, our analysis points out that different technologies have unique advantages and adapt to different scenarios, e.g., ITES has higher accuracy, LS has higher performance in scenarios with limited computing resources.

Development Paillier's library of fully homomorphic encryption

One of the new areas of cryptography considered-homomorphic cryptography. The article presents the main areas of application of homomorphic encryption. An analysis of existing developments in the field of homomorphic encryption carried out. The analysis showed that existing library implementations only allow processing bits or arrays of bits and do not support division and subtraction operations. However, to solve applied problems, support for performing integer operations are necessary. Because of the analysis, the need to implement the homomorphic division and subtraction operations identified, as well as the relevance of developing our own implementation of a homomorphic encryption library over integers. The ability to perform four operations (addition, difference, multiplication and division) on encrypted data will expand the areas of application of homomorphic encryption. A homomorphic division and subtraction methods proposed that allows the division operation performed on homomorphically encrypted data. An architecture for a library of fully homomorphic operations on integers is proposed. The library supports basic homomorphic operations on integers, as well as homomorphic division method. The article also provides measurements of the time required to perform certain operations on encrypted data and analyzes the efficiency of the developed implementation of the library.

De Bello Homomorphico: Investigation of the extensibility of the OpenFHE library with basic mathematical functions by means of common approaches using the example of the CKKS cryptosystem

Cloud computing has become increasingly popular due to its scalability, cost-effectiveness, and ability to handle large volumes of data. However, entrusting (sensitive) data to a third party raises concerns about data security and privacy. Homomorphic encryption is one solution that allows users to store and process data in a public cloud without the cloud provider having access to it. Currently, homomorphic encryption libraries only support addition and multiplication; other mathematical functions must be implemented by the user. To this end, we discuss and implement the division, exponential, square root, logarithm, minimum, and maximum function, using the CKKS cryptosystem of the OpenFHE library. To demonstrate that complex applications can be realized with this extended function set, we have used it to homomorphically realize the Box–Cox transform, which is used in many real-world applications, e.g., time-series forecasts. Our results show how the number of iterations required to achieve a given accuracy varies depending on the function. In addition, the execution time for each function is independent of the input and is in the range of ten seconds on a reference machine. With this work, we provide users with insights on how to extend the original restricted function set of the CKKS cryptosystem of the OpenFHE library with basic mathematical functions.

Homomorphic Encryption Library, Framework, Toolkit and Accelerator: A Review

Homomorphic Encryption Library, Framework, Toolkit and Accelerator: A Review

A Comparative Study of BFV and CKKs Schemes to Secure IoT Data Using TenSeal and Pyfhel Homomorphic Encryption Libraries

Internet of things (IoT) devices and applications are on the rise, generating large amounts of sensitive and confidential data that need to be processed securely. Due to resource constraints, the data generated is often stored and processed in the cloud. The drawback of data cloud storage and processing is the fact that it can be hacked, leaked, or sold by cloud companies. Fully homomorphic encryption (FHE) allows computation on encrypted data using basic mathematical operations and has recently been successfully implemented using schemes and libraries with better performance. In this paper, the authors propose a mixture of edge-cloud-based security schemes using FHE to secure IoT data. The authors evaluate the performance of two FHE schemes (BFV and CKKS) based on data: encoding speed, encryption speed, arithmetic operations (addition and multiplication) speed, and decryption decoding speed using two Python libraries (TenSEAL and PyFHEl). The encryption and decryption are done at the edge node using a Raspberry Pi 4, while the processing is done at the cloud node using a laptop.

SoK: New Insights into Fully Homomorphic Encryption Libraries via Standardized Benchmarks

Fully homomorphic encryption (FHE) enables arbitrary computation on encrypted data, allowing users to upload ciphertexts to cloud servers for computation while mitigating privacy risks. Many cryptographic schemes fall under the umbrella of FHE, and each scheme has several open-source implementations with its own strengths and weaknesses. Nevertheless, developers have no straightforward way to choose which FHE scheme and implementation is best suited for their application needs, especially considering that each scheme offers different security, performance, and usability guarantees. To allow programmers to effectively utilize the power of FHE, we employ a series of benchmarks called the Terminator 2 Benchmark Suite and present new insights gained from running these algorithms with a variety of FHE back - ends. Contrary to generic benchmarks that do not take into consideration the inherent challenges of encrypted computation, our methodology is tailored to the secure computational primitives of each target FHE implementation. To ensure fair comparisons, we developed a versatile compiler(called T2) that converts arbitrary benchmarks written in a domain - specific language into identical encrypted programs running on different popular FHE libraries as a backend.Our analysis exposes for the first time the advantages and disadvantages of each FHE library as well as the types of applications most suited for each computational domain(i.e., binary, integer, and floating - point).

On the security of fully homomorphic encryption for data privacy in Internet of Things

SummaryTo achieve data privacy in Internet of Things (IoT), fully homomorphic encryption (FHE) technique is used to encrypt the data while allowing others to compute on the encrypted data. However, there are many well‐known problems with FHE such as chosen‐ciphertext attack security and circuit privacy problem. In this article, we demonstrate that a famous FHE application named Brakerski/Fan–Vercauteren scheme, a circuit privacy application based on fast private set intersection, and an encoding application that encodes integer or floating point numbers based on Microsoft Simple Encryption Arithmetic Library homomorphic encryption library, are insecure against chosen ciphertext attacks due to insecurity of the underlying fully homomorphic schemes. These results show that using cryptographic primitives even with security proofs causes serious security vulnerabilities on the applications themselves. The results also give evidences that the security of adopted cryptographic primitives in IoT should be proved in appropriate formal security models as well as proof of the scheme itself.

A Novel Homomorphic Approach for Preserving Privacy of Patient Data in Telemedicine.

Globally, the surge in disease and urgency in maintaining social distancing has reawakened the use of telemedicine/telehealth. Amid the global health crisis, the world adopted the culture of online consultancy. Thus, there is a need to revamp the conventional model of the telemedicine system as per the current challenges and requirements. Security and privacy of data are main aspects to be considered in this era. Data-driven organizations also require compliance with regulatory bodies, such as HIPAA, PHI, and GDPR. These regulatory compliance bodies must ensure user data privacy by implementing necessary security measures. Patients and doctors are now connected to the cloud to access medical records, e.g., voice recordings of clinical sessions. Voice data reside in the cloud and can be compromised. While searching voice data, a patient’s critical data can be leaked, exposed to cloud service providers, and spoofed by hackers. Secure, searchable encryption is a requirement for telemedicine systems for secure voice and phoneme searching. This research proposes the secure searching of phonemes from audio recordings using fully homomorphic encryption over the cloud. It utilizes IBM’s homomorphic encryption library (HElib) and achieves indistinguishability. Testing and implementation were done on audio datasets of different sizes while varying the security parameters. The analysis includes a thorough security analysis along with leakage profiling. The proposed scheme achieved higher levels of security and privacy, especially when the security parameters increased. However, in use cases where higher levels of security were not desirous, one may rely on a reduction in the security parameters.

Implementing Linear Regression with Homomorphic Encryption

Benefiting from the computing power and storage power of cloud computing, machine learning tasks usually choose to upload data and models to a third-party server. However, third-party servers may face data privacy leaks in the process of data acquisition, storage, and use, especially in the fields of finance, medical treatment, and biometrics. Homomorphic encryption technology supports ciphertext calculation without decryption, and can be used for machine learning model training and prediction in the ciphertext domain. We use the homomorphic encryption library SEAL to reconstruct the linear regression protocol, use six data sets to conduct experiments, and analyze its performance and security.

100 Gb/s dynamically programmable SDN-enabled hardware encryptor for optical networks

We have successfully demonstrated an on-demand programmable hardware encryptor for optical networks enabled by the software-defined network (SDN) and tested its application in a quantum key distribution (QKD) use case. We believe our proposed encryptor is unique in the way that it combines on-demand programmable encryption algorithms with a 100G Ethernet network interface to provide a high-transmission capacity with flexible security. We have successfully formed an on-demand encryption library consisting of AES-256, AES-192, AES-128, Camellia-256, XOR, and no-encryption configurations. Our system has shown a network throughput of 91.3 Gb/s for AES variations and 90.2 Gb/s for Camellia-256, while theoretical encryption throughputs were 160 Gb/s. In addition, the goodputs of all the encryption schemes were measured as at least 90.4 Gb/s. For a faster reconfiguration of the field-programmable gate array (FPGA), partial reconfiguration technology has been used, and reconfiguration times of 2.6 s for encryption and 2 s for decryption have been achieved. The FPGA configuration rate was 3.35 MB/s. When our proposed design was tested in a QKD use case with 256-bit keys, the highest achieved key consumption rate was 27 keys/s, which corresponded to the minimum granularity of 474 MB per key. Thus, with a 256-bit key, up to a 6912 b/s key consumption rate was achieved. In addition, the encryptor’s end-to-end network latency has been tested using 300k standard Internet Control Message Protocol (ICMP) pings. For all encryption schemes, the latency was measured as 0.093 m s ± 0.028 m s on average, but the encryption/decryption processes did not have a meaningful latency impact with microsecond precision.

Embedding and classifying test execution traces using neural networks

Classifying test executions automatically as pass or fail remains a key challenge in software testing and is referred to as the test oracle problem. It is being attempted to solve this problem with supervised learning over test execution traces. A programme is instrumented to gather execution traces as sequences of method invocations. A small fraction of the programme's execution traces is labelled with pass or fail verdicts. Execution traces are then embedded as fixed length vectors and a neural network (NN) component that uses the line-by-line information to classify traces as pass or fail is designed. The classification accuracy of this approach is evaluated using subject programs from different application domains—1. Module from Ethereum Blockchain, 2. Module from PyTorch deep learning framework, 3. Microsoft SEAL encryption library components, 4. Sed stream editor, 5. Nine network protocols from Linux packet identifier, L7-Filter and 6. Utilities library, commons-lang for Java. For all subject programs, it was found that test execution classification had high precision, recall and specificity, averaging to 93%, 94% and 96%, respectively, while only training with an average 14% of the total traces. Experiments show that the proposed NN-based approach is promising in classifying test executions from different application domains.

SEAL-Embedded: A Homomorphic Encryption Library for the Internet of Things

The growth of the Internet of Things (IoT) has led to concerns over the lack of security and privacy guarantees afforded by IoT systems. Homomorphic encryption (HE) is a promising privacy-preserving solution to allow devices to securely share data with a cloud backend; however, its high memory consumption and computational overhead have limited its use on resource-constrained embedded devices. To address this problem, we present SEAL-Embedded, the first HE library targeted for embedded devices, featuring the CKKS approximate homomorphic encryption scheme. SEAL-Embedded employs several computational and algorithmic optimizations along with a detailed memory re-use scheme to achieve memory efficient, high performance CKKS encoding and encryption on embedded devices without any sacrifice of security. We additionally provide an “adapter” server module to convert data encrypted by SEAL-Embedded to be compatible with the Microsoft SEAL library for homomorphic encryption, enabling an end-to-end solution for building privacy-preserving applications. For a polynomial ring degree of 4096, using RNS primes of 30 or fewer bits, our library can be configured to use between 64–137 KB of RAM and 1–264 KB of flash data, depending on developer-selected configurations and tradeoffs. Using these parameters, we evaluate SEAL-Embedded on two different IoT platforms with high performance, memory efficient, and balanced configurations of the library for asymmetric and symmetric encryption. With 136 KB of RAM, SEAL-Embedded can perform asymmetric encryption of 2048 single-precision numbers in 77 ms on the Azure Sphere Cortex-A7 and 737 ms on the Nordic nRF52840 Cortex-M4.

Feature Fusion Methods for Indexing and Retrieval of Biometric Data: Application to Face Recognition With Privacy Protection

Computationally efficient, accurate, and privacy-preserving data storage and retrieval are among the key challenges faced by practical deployments of biometric identification systems worldwide. In this work, a method of protected indexing of biometric data is presented. By utilising feature-level fusion of intelligently paired templates, a multi-stage search structure is created. During retrieval, the list of potential candidate identities is successively pre-filtered, thereby reducing the number of template comparisons necessary for a biometric identification transaction. Protection of the biometric probe templates, as well as the stored reference templates and the created index is carried out using homomorphic encryption. The proposed method is extensively evaluated in closed-set and open-set identification scenarios on publicly available databases using two state-of-the-art open-source face recognition systems. With respect to a typical baseline algorithm utilising an exhaustive search-based retrieval algorithm, the proposed method enables a reduction of the computational workload associated with a biometric identification transaction by 90%, while simultaneously suffering no degradation of the biometric performance. Furthermore, by facilitating a seamless integration of template protection with open-source homomorphic encryption libraries, the proposed method guarantees unlinkability, irreversibility, and renewability of the protected biometric data.

완전 동형 암호 라이브러리의 성능 분석

완전 동형 암호 라이브러리의 성능 분석

Private queries on encrypted genomic data

BackgroundOne of the tasks in the iDASH Secure Genome Analysis Competition in 2016 was to demonstrate the feasibility of privacy-preserving queries on homomorphically encrypted genomic data. More precisely, given a list of up to 100,000 mutations, the task was to encrypt the data using homomorphic encryption in a way that allows it to be stored securely in the cloud, and enables the data owner to query the dataset for the presence of specific mutations, without revealing any information about the dataset or the queries to the cloud.MethodsWe devise a novel string matching protocol to enable privacy-preserving queries on homomorphically encrypted data. Our protocol combines state-of-the-art techniques from homomorphic encryption and private set intersection protocols to minimize the computational and communication cost.ResultsWe implemented our protocol using the homomorphic encryption library SEAL v2.1, and applied it to obtain an efficient solution to the iDASH competition task. For example, using 8 threads, our protocol achieves a running time of only 4 s, and a communication cost of 2 MB, when querying for the presence of 5 mutations from an encrypted dataset of 100,000 mutations.ConclusionsWe demonstrate that homomorphic encryption can be used to enable an efficient privacy-preserving mechanism for querying the presence of particular mutations in realistic size datasets. Beyond its applications to genomics, our protocol can just as well be applied to any kind of data, and is therefore of independent interest to the homomorphic encryption community.

Cryptoleq: A Heterogeneous Abstract Machine for Encrypted and Unencrypted Computation

The rapid expansion and increased popularity of cloud computing comes with no shortage of privacy concerns about outsourcing computation to semi-trusted parties. Leveraging the power of encryption, in this paper, we introduce Cryptoleq: an abstract machine based on the concept of one instruction set computer, capable of performing general-purpose computation on encrypted programs. The program operands are protected using the Paillier partially homomorphic cryptosystem, which supports addition on the encrypted domain. Full homomorphism over addition and multiplication, which is necessary for enabling general-purpose computation, is achieved by inventing a heuristically obfuscated software re-encryption module written using Cryptoleq instructions and blended into the executing program. Cryptoleq is heterogeneous, allowing mixing encrypted and unencrypted instruction operands in the same program memory space. Programming with Cryptoleq is facilitated using an enhanced assembly language that allows the development of any advanced algorithm on encrypted data sets. In our evaluation, we compare Cryptoleq's performance against a popular fully homomorphic encryption library, and demonstrate correctness using a typical private information retrieval problem.

SHIELD: Scalable Homomorphic Implementation of Encrypted Data-Classifiers

Homomorphic encryption (HE) systems enable computations on encrypted data, without decrypting and without knowledge of the secret key. In this work, we describe an optimized Ring Learning With Errors (RLWE) based implementation of a variant of the HE system recently proposed by Gentry, Sahai and Waters (GSW). Although this system was widely believed to be less efficient than its contemporaries, we demonstrate quite the opposite behavior for a large class of applications. We first highlight and carefully exploit the algebraic features of the system to achieve significant speedup over the state-of-the-art HE implementation, namely the IBM homomorphic encryption library (HElib). We introduce several optimizations on top of our HE implementation, and use the resulting scheme to construct a homomorphic Bayesian spam filter, secure multiple keyword search, and a homomorphic evaluator for binary decision trees. Our results show a factor of $10\times$ improvement in performance (under the same security settings and CPU platforms) compared to IBM HElib for these applications. Our system is built to be easily portable to GPUs (unlike IBM HElib) which results in an additional speedup of up to a factor of $103.5\times$ to offer an overall speedup of $1{,}035\times$ .