Abstract The explicit consideration of the software architecture supports system evolution and efficient quality assurance. In particular, Architecture-based Performance Prediction (AbPP) assesses the performance for future scenarios (e.g., alternative workload, design, deployment) without expensive measurements for all such alternatives. However, accurate AbPP requires an up-to-date architectural Performance Model (aPM) that is parameterized over factors impacting the performance (e.g., input data characteristics). Especially in agile development, keeping such a parametric aPM consistent with software artifacts is challenging due to frequent evolutionary, adaptive, and usage-related changes. Existing approaches do not address the impact of all aforementioned changes. Moreover, the extraction of a complete aPM after each impacting change causes unnecessary monitoring overhead and may overwrite previous manual adjustments. In this article, we present the Continuous Integration of architectural Performance Model (CIPM) approach, which automatically updates a parametric aPM after each evolutionary, adaptive, or usage change. To reduce the monitoring overhead, CIPM only calibrates the affected performance parameters (e.g., resource demand) using adaptive monitoring. Moreover, a self-validation process in CIPM validates the accuracy, manages the monitoring to reduce overhead, and recalibrates inaccurate parts. Consequently, CIPM will automatically keep the aPM up-to-date throughout the development and operation, which enables AbPP for a proactive identification of upcoming performance problems and for evaluating alternatives at low costs. We evaluate the applicability of CIPM in terms of accuracy, monitoring overhead, and scalability using six cases (four Java-based open source applications and two industrial Lua-based sensor applications). Regarding accuracy, we observed that CIPM correctly keeps an aPM up-to-date and estimates performance parameters well so that it supports accurate performance predictions. Regarding the monitoring overhead in our experiments, CIPM’s adaptive instrumentation demonstrated a significant reduction in the number of required instrumentation probes, ranging from 12.6 % to 83.3 %, depending on the specific cases evaluated. Finally, we found out that CIPM’s execution time is reasonable and scales well with an increasing number of model elements and monitoring data. Graphical Abstract

The increasing adoption of machine learning (ML)-based services has presented challenges in processing sensitive data while ensuring privacy and confidentiality. Homomorphic encryption offers a promising solution by enabling computations on encrypted data. However, applying homomorphic encryption in ML faces challenges regarding efficient structuring, arrangement, and execution of numerical operations. In this paper, we present HEJet: a framework that enables efficient and user-friendly application of neural networks with homomorphic encryption. Our framework maps sequences of numerical computations to an optimized set of instructions that are processed by compilers for homomorphic encryption. Consequently, HEJet provides user-friendly interfaces to utilize advanced neural network structures with homomorphic encryption. Evaluation results on the MNIST dataset highlight its usability and show a significant speedup in inferences between 3% and 48% compared to existing approaches. Additionally, HEJet maintains accuracy levels close to those observed on raw data.

For secure remote processing, homomorphic encryption can be used. It allows operations to be performed on encrypted data. If the processor needs the result’s plaintext, it relies on the data owner to perform the decryption. This poses a vulnerability since a malicious data owner could inject a self–selected value. Moreover, it sees the plaintext result, breaching its confidentiality. In this paper, we propose a solving approach for the CKKS homomorphic encryption scheme. Using a commutative property, the data owner removes its encryption, while a processor’s encryption reinforces integrity and confidentiality. In an evaluation, we demonstrate feasibility regarding computational effort, transmitted data size, and introduced error.

Short development times of software became crucial to stay competitive. However, the quality should not suffer from the faster development processes, which is why increasingly more automation is gaining ground in this context. If models are involved in the development process and used for performance prediction, there are delays due to emerging inconsistencies between different software artifacts. The elimination of these inconsistencies is a time consuming, complex and error prone activity. Currently, there are already approaches for automated consistency preservation of software artifacts. Nevertheless, the limited scope in terms of supported change scenarios is a significant disadvantage.Therefore, we present a comprehensive approach for the maintenance of consistency between the system design and adaptive as well as evolutionary changes. In comparison to existing approaches, the consistency preservation has been significantly extended in our approach to cover a multitude of changes resulting from adaptation and evolution. Ultimately, several validation steps were integrated into the approach, enabling continuous assessment regarding the quality of the consistency preservation. In a case study based evaluation, we measured the accuracy of the updated models and associated performance predictions.

The in-depth consideration of security aspects in modern web infrastructures has become essential to stay competitive. In this context, continuous authentication is a promising approach to prevent the misuse of digital identities. To this end, machine learning (ML) models are well suited to analyze user behavior and to detect anomalies, due to their ability to identify complex patterns and trends that usually cannot be reflected by static rule-based approaches. However, the training of powerful ML models requires large amounts of data, which are often not available within a single organization. Consequently, a federated training of these models by cooperating organizations offers a promising solution, but leads to concerns about coordination, regulations, and quality assurance. To tackle these challenges, we present an approach that combines three research areas: (1) the establishment of continuous user authentication based on (2) a ML model trained by an organized peer-to-peer federation involving different organizations that is underpinned by (3) federated data governance ensuring regulatory compliance and quality of resulting artefacts.

As password-based authentication fails to provide adequate security for online activities such as financial transactions, additional authentication factors are required. Such factors should provide both ease of use and an adequate level of privacy protection, while being easy to implement, operate and maintain, even for applications with thousands of users. As an approach meeting these requirements, we outline Scalable and Privacy- Preserving Continuous Authentication (SPCAuth). SPCAuth determines risk levels for actions that should be authenticated, without requiring explicit user interactions. It analyzes different aspects of user behavior by means of machine learning methods, while preserving the privacy of the affected individuals. SPCAuth trains only a single model per aspect of user behavior being considered, based on observations of all users, ensuring scalability and increasing accuracy for users with infrequent activities. In two experiments, we have confirmed that this key concept enables a scalable and accurate user authentication.

Architecture-based Performance Prediction (AbPP) allows evaluation of the performance of systems and to answer what-if questions without measurements for all alternatives. A difficulty when creating models is that Performance Model Parameters (PMPs, such as resource demands, loop iteration numbers and branch probabilities) depend on various influencing factors like input data, used hardware and the applied workload. To enable a broad range of what-if questions, Performance Models (PMs) need to have predictive power beyond what has been measured to calibrate the models. Thus, PMPs need to be parametrized over the influencing factors that may vary. Existing approaches allow for the estimation of the parametrized PMPs by measuring the complete system. Thus, they are too costly to be applied frequently, up to after each code change. Moreover, they do not keep manual changes to the model when recalibrating. In this work, we present the Continuous Integration of Performance Models (CIPM), which incrementally extracts and calibrates the performance model, including parametric dependencies. CIPM responds to source code changes by updating the PM and adaptively instrumenting the changed parts. To allow AbPP, CIPM estimates the parametrized PMPs using the measurements (generated by performance tests or executing the system in production) and statistical analysis, e.g., regression analysis and decision trees. Additionally, our approach responds to production changes (e.g., load or deployment changes) and calibrates the usage and deployment parts of PMs accordingly. For the evaluation, we used two case studies. Evaluation results show that we were able to calibrate the PM incrementally and accurately.