Budget-constrained IoT service composition seeks composite services that optimize Quality of Service (QoS), such as response time, while operating within a fixed budget, which is a challenge of practical importance since real-world IoT systems often face strict financial limits. Traditional constraint-handling techniques, such as penalty methods, and random repair operators, often struggle to effectively balance budget feasibility and response time optimization. To address this gap, we propose Cost-Aware Repair-based Genetic Optimization (CARGO-IoT), which combines priority-based selection of infeasible solutions with a two-stage repair strategy. CARGO-IoT focuses on repairing recent solutions with smaller violations and faster response times, using a replace-based method guided by learned probabilities from past solutions and a reduce-based method that eliminates redundant services to meet budget limits. Experiments demonstrate that our approach significantly outperforms state-of-the-art methods in this field, achieving lower response time while maintaining budget feasibility.

Microservice deployment in cloud computing is a challenging combinatorial optimization problem due to the complex dependencies among microservices and the intricate trade-offs among different QoS requirements, e.g., minimizing Energy Consumption (EC) vs. minimizing Communication Overhead (CO). Recently, some hyper-heuristics methods, particularly Genetic Programming Hyper-Heuristics (GPHH), have been proposed to automatically generate heuristics for solving dynamic microservice deployment problems. Meanwhile, Large Language Models (LLMs) are becoming popular for solving various domain-specific problems thanks to their strong ability to learn problem-related knowledge. However, hybridizing GPHH with LLM by combining their abilities in solving complex optimization problems remains unexplored. In this paper, we propose an LLM-enhanced Genetic Programming Hyper-Heuristic (LLM-GPHH) algorithm to evolve heuristics for the dynamic deployment of applications composed of microservices, to jointly optimize EC and CO. Our experiments on real-world datasets demonstrate the effectiveness of the newly proposed LLM-GPHH.

In the domain of Cloud computing, Fog computing is integrated with the Cloud to offer a balanced approach that combines Cloud's scalability with Fog's low latency, enabling efficient software application deployment. However, many current studies overlook the unpredictability of future user requests, such as assuming all requests are known beforehand. User requests often arrive dynamically and may have different quality of service (QoS) preferences. Therefore we need effective methods to handle dynamic application deployment with multiple objectives. This paper tackles this gap by modeling a multi-objective application deployment problem that considers dynamically arriving users' requests on application deployment in a Cloud-Fog environment. We further introduce a multi-objective Genetic Programming Hyper-Heuristic based approach to automatically generate a set of deployment rules that can be chosen according to users' QoS preferences. These rules are generated with different trade-offs of two optimization objectives, i.e., minimizing cost and latency, which can be used for deploying applications dynamically. Our experimental evaluation using real-world data demonstrates that our GPHH approach can generate effective heuristics for deploying applications in an integrated Cloud-Fog environment.

Climate downscaling bridges the gap between coarse-resolution General Circulation Model (GCM) outputs and the fine-resolution data needed for regional assessments. Traditional dynamic and statistical downscaling methods face limitations in computational efficiency and accuracy. Leveraging recent advancements in deep learning, particularly convolutional neural networks (CNNs), we propose an improved method for downscaling coastal sea surface temperature (SST). Our approach introduces a novel composite loss function combining Mean Squared Error with perceptual loss, effectively capturing the high seasonal variations typically in the coastal zones. Additionally, we redesign the YNet CNN to incorporate historical monthly mean SST, enhancing its performance. Comprehensive experiments using real-world SST GCM data and observational monthly SST data show our method significantly outperforms existing techniques in coastal zones up to 50 km offshore. Case studies in New Zealand further demonstrate the reduced errors achieved by our method.

Sensor data collected from climate stations has been used in various scientific applications and environmental monitoring. Maintaining the data quality is essential to guarantee the reliability and accuracy of science outputs, potentially impacting many critical decision making processes. Existing sensor anomaly detection techniques are mostly designed for general purposes, and may not be suitable for climate sensors which require complex handling of seasonality, spatial relationship and sensor interdependency. Current quality control process is deficient in climate sensor drift detection, which is a slow degradation of sensor accuracy over time. Recent development of anomaly detection in climate sensor domain is limited, it’s often constrained to particular sensor types, and not focused on drift detection. In this paper, we present a new drift-aware time series anomaly detection framework which leverages the spatial-temporal correlation of the climate sensor network and significantly improves climate sensor drift detection capability. Moreover, the proposed semi- supervised learning approach helps to generalise the solution for various types of sensors and anomalies. Our experiments using real-world dataset have demonstrated promising and competitive performance in regards to sensitivity, false alarm control, and computational efficiency suitable for real-time or near-real-time applications.

Web service composition composes existing Web services to accommodate users' requests for required functionalities with the best possible quality of services (QoS). Due to the computational complexity of this problem, evolutionary computation (EC) techniques have been employed to efficiently find composite services with near-optimal functional quality (i.e., quality of semantic matchmaking, QoSM for short) or nonfunctional quality (i.e., QoS) for each composition request individually. With a rapid increase in composition requests from a growing number of users, solving one composition request at a time can hardly meet the efficiency target anymore. Driven by the idea that the solutions obtained from solving one request can be highly useful for tackling other related requests, multitasking service composition approaches have been proposed to efficiently deal with multiple composition requests concurrently. However, existing attempts have not been effective in learning and sharing knowledge among solutions for multiple requests. In this article, we model the problem of collectively handling multiple service composition requests as a new multitasking service composition problem and propose a new permutation-based multifactorial evolutionary algorithm based on an estimation of distribution algorithm (EDA), named PMFEA-EDA, to effectively and efficiently solve this problem. In particular, we introduce a novel method for effective knowledge sharing across different service composition requests. For that, we develop a new sampling mechanism to increase the chance of identifying high-quality service compositions in both the single-tasking and multitasking contexts. Our experiment shows that our proposed approach, PMFEA-EDA, takes much less time than existing approaches that process each service request separately, and also outperforms them in terms of both QoSM and QoS.

As a scalable and lightweight infrastructure technology, containers are quickly gaining popularity in cloud data centers. However, dynamic Resource-Allocation in Container-based clouds (RAC) is challenging due to two interdependent allocation sub-problems, allocating dynamic arriving containers to appropriate Virtual Machines (VMs) and allocating VMs to multiple Physical Machines (PMs). Most of existing research works assume homogeneous PMs and rely on simple and manually designed heuristics such as Best Fit and First Fit, which can only capture limited information, affecting their effectiveness of reducing energy consumption in data centers. In this work, we propose a novel hybrid Cooperative Coevolution Genetic Programming (CCGP) hyper-heuristic approach to automatically generate heuristics that are effective in solving the dynamic RAC problem. Different from existing works, our approach hybridizes Best Fit to automatically designed heuristics to coherently solve the two interdependent sub-problems. Moreover, we introduce a new energy model that accurately captures the energy consumption in a more realistic setting than that in the literature, e.g., real-world workload patterns and heterogeneous PMs. The experiment results show that our approach can significantly reduce energy consumption, in comparison to two state-of-the-art methods.

The optimal placement of sensors for environmental monitoring and disaster management is a challenging problem due to its NP-hard nature. Traditional methods for sensor placement involve exact, approximation, or heuristic approaches, with the latter being the most widely used. However, heuristic methods are limited by expert intuition and experience. Deep learning (DL) has emerged as a promising approach for generating heuristic algorithms automatically. In this paper, we introduce a novel sensor placement approach focused on learning improvement heuristics using deep reinforcement learning (RL) methods. Our approach leverages an RL formulation for learning improvement heuristics, driven by an actor-critic algorithm for training the policy network. We compare our method with several state-of-the-art approaches by conducting comprehensive experiments, demonstrating the effectiveness and superiority of our proposed approach in producing high-quality solutions. Our work presents a promising direction for applying advanced DL and RL techniques to challenging climate sensor placement problems.

In epidemic management, the unpredictable dynamics of outbreaks, the constraints of available resources, and the complexity of variable interactions pose a significant challenge in designing effective management rules. While traditional mathematical models offer insights into the spread of the outbreaks, they are not effective due to the lack of available information on the outbreaks. Moreover, they are valuable for the evaluation of different management rules but do not provide management rules. In this paper, we propose a novel approach that leverages Reinforcement Learning (RL) to automatically design epidemic management rules. By using RL, we formulate the problem as determining the optimal daily inspection frequency during an outbreak with the goal of minimizing both the inspection cost and the epidemic size. A management rule is trained using a deep neuroevolution algorithm called Evolutionary Strategy (ES). An epidemic simulator is developed to provide a realistic and dynamic environment for evaluating our proposed approach's effectiveness. Extensive experiments have shown that our RL-based approach outperformed conventional methods and can better balance the trade-off between the inspection cost and epidemic size. The findings from this work provide not only a framework for autonomous learning of epidemic management rules, but also potential implications for real-world epidemic control and policy-making.

Test costs for large industrial designs increase rapidly in recent years. On-chip test compression hardware has become a pragmatic technology to cut down the overall test costs by reducing the test data volume. Determining the input and output channel counts of test compression hardware that results in minimum test data volume is thus a critical issue. In this article, efficient methods to estimate test pattern counts for an extensive range of input/output counts are developed. These methods require only a small number of ATPG runs. The estimation results can then be utilized to determine the test data volume for each input/output configuration. The configuration with the estimated lowest test data volume thus can be determined. The pattern count results of each configuration for a design can also be used to determine the best suitable configuration when the design is to be embedded in an SoC system.

Test cost has become a critical issue for large industrial integrated circuits. Various test compression techniques have been adopted in the industry to reduce test cost. However, appropriate input and output channel counts must be selected to utilize the test compression technology best. This paper presents an efficient and effective method to estimate the test pattern counts under different compression configurations for the Embedded Deterministic Test (EDT) compression technique. In searching for the accurate estimation method, we build mathematical models that reveal the internal relationship among different compression configurations. The models are established based on novel theoretical analysis as well as actual experimental data. Accurate estimation of test pattern counts for a wide range of compression configurations can be obtained based on the results of only two ATPG runs. Experimental results on nine industrial circuits show that the average error rate of pattern count estimation is about 5%, with very few outliers. With the proposed method, a test compression designer can easily pick the best input and output channel configuration to fit the design needs.

Service-oriented architecture (SOA) is a widely adopted software engineering paradigm that encourages modular and reusable applications. One popular application of SOA is web service composition, which aims to loosely couple web services to accommodate complex goals not achievable through any individual web service. Many approaches have been proposed to construct composite services with optimized Quality of Service (QoS), assuming that QoS of web services never changes. However, the constructed composite services may not perform well and may not be executable later due to its component services' failure. Therefore, it is important to build composite services that are robust to stochastic service failures. Two challenges of building robust composite services are to efficiently generate service composition with near-optimal quality in a large search space of available services and to accurately measure the robustness of composite services considering all possible failure scenarios. This article proposes a novel two-stage GA-based approach to robust web service composition with an adaptive evolutionary control and an efficient robustness measurement. This approach can generate robust composite service at the design phase, which can cope with stochastic service failures and maintain high quality at the time of execution. We have conducted experiments with benchmark datasets to evaluate the performance of our proposed approach. Our experiments show that our method can produce highly robust composite services, achieving outstanding performance consistently in the event of stochastic service failures, on service repositories with varying sizes.

The Dynamic Multi-Workflow Scheduling (DMWS) problem aims to allocate highly complex tasks modeled as workflows to cloud resources while optimizing workflow brokers' interests. A workflow broker offers workflow execution services to end-users with agreed Service Level Agreements (SLA) while reducing its total VM rental fees in the meantime. Most existing DMWS-related research works focus on minimizing the workflow makespan by using either heuristics or hyper-heuristics techniques. However, these techniques were either designed for static workflow scheduling based on prior workflow information and/or the simplified cloud environment. In this paper, the DMWS problem is formulated to collectively minimize VM rental fees and SLA violation penalties. Moreover, we introduce a novel priority-based deep neural network scheduling policy that can flexibly adapt to a changing number of VMs and workflows. To train the new policy, a new Evolutionary Strategy based Reinforcement Learning (ES-RL) is developed and implemented. Different from gradient-based deep reinforcement learning algorithms, ES-RL has its advances in effectively training population-based and generally applicable policies in parallel as well as robustness to hyper-parameter settings. Our experiments with real-world datasets show that ES-RL can effectively train scheduling policies that can significantly reduce the costs by more than 90% compared to the state-of-the-art scheduling policies.

As the complexity of industrial integrated circuits continue to increase rapidly, test data compression has now become a de facto technology for large designs to reduce the overall test cost. During the design for test (DFT) planning, it is critical to understand the impact of using different numbers of input/output test channels on test coverage, test cycles, and test data volume. In this paper, two approaches to predict the test pattern counts and test data volumes with different input channel counts are presented, one with the compression tool able to generate channel-scaling patterns and the other without this capability. The results can be used to determine the scan test configuration that results in the smallest or near smallest test data volume. Experiments on industrial circuits show that the average error rates of pattern count prediction for most circuits are less than 10% for both approaches. The error rates of the predicted smallest data volumes are all less than 3.5%. The total ATPG run time can be reduced by a factor of more than 10X compared to the currently used trial-and-error approach.

Timing exceptions are commonly used to indicate that the timing of certain paths have been relaxed so as to enable the design to meet timing closure. Generating scan-based test patterns without considering timing exceptions can lead to invalid test responses, resulting in unpredictable test quality impact. The existing simulation-based solution masks out unreliable signals after a test pattern is generated. If the signals required for detecting the target fault are unreliable and masked out, the generated test pattern fails to detect the target fault, and it is discarded. To achieve an acceptable test coverage, several iterations of test generation with a randomized decision-making process are typically required where different tests are generated for target faults. In this paper, an innovative deterministic ATPG algorithm called TEA (Timing Exception ATPG) is proposed to prevent the generated test patterns from being impacted by timing exceptions. The deterministic algorithm is compatible with the existing simulation-based approach. In this simulation environment, TEA is complete such that for a target fault, the test pattern generated is guaranteed to detect it. If a test pattern cannot be generated using TEA, the target fault is untestable given the timing exception paths in the design and the existing simulation environment. Compared to the existing simulation-based approach, using TEA can generate a more effective test set, improving test coverage, test pattern count, and the total ATPG run time significantly.

Web service composition aims to loosely couple web services to accommodate complex goals, which can not be accomplished by any existing web service. Many researchers have been working on such service composition problems with the aim to find composite services with optimized Quality of Service (QoS) and/or Quality of Semantic Matchmaking (QoSM). Due to the huge search space of this NP-hard problem, Evolutionary Computation techniques have been popularly utilized to search for solutions with near-optimal QoS and QoSM. A majority of these works share a common assumption that QoS of web services seldom or never changes. However, the execution of composite services obtained from the design stage may fail due to unexpected service failures at the execution stage. In this paper, we introduce a robust service composition approach with the goal to build robust composite services that serve as the blueprint/baseline for service execution. These baseline composite services can cope with unexpected interruptions in a robust manner, by applying local search to resume their feasibility while maintaining high quality at the time of execution. Our experiments show that our new approach can significantly outperform a state-of-the-art service composition method (without explicitly considering the robustness) in terms of both effectiveness and efficiency in the event of unexpected service failures.

Service-oriented computing is a computing paradigm that creates reusable modules over the Internet, often known as Web services. Web service composition aims to accomplish more complex functions by loosely coupling web services. Researchers have been proposing evolutionary computation (EC) techniques for efficiently building up composite services with optimized non-functional quality (i.e., QoS). Some of these techniques employ multi-objective EC algorithms to handle conflict qualities in QoS for fully automated service composition. One recent state-of-art work hybridizes NSGA-II and MOEA/D, which allows the multi-objective service composition problem to be decomposed into many scalar optimization subproblems, where a simple form of local search can be easily applied. However, their local search is considered to be less effective and efficient because it is randomly applied to a predefined large number of subproblems without focusing on the most suitable candidate solutions. In this paper, we propose a memetic NSGA-II with probabilistic model-based local search based on Estimation of Distribution Algorithm (EDA). In particular, a clustering technique is employed to select suitable Pareto solutions for local search. Each selected solution and its belonged cluster members are used to learn a distribution model that samples new solutions for local improvements. Besides that, a more challenging service composition problem that optimizes both functional and non-functional quality is considered. Experiments have shown that our method can effectively and efficiently produce better Pareto optimal solutions compared to other state-of-art methods in the literature.

Web service composition aims to provide added values by loosely coupling web services to accommodate users' complex requirements. Evolutionary computation techniques have been used to efficiently find near-optimal composite services to satisfy users' requirements reasonably well. Often, the quality of a composite service is measured by two important quality criteria that are related to the non-functional quality (i.e., Quality of service, QoS for short) and function quality (i.e., Quality of semantic matchmaking, QoSM for short). One recent work [2] proposed a Hybrid method that combines NSGA-II and MOEA/D with swap-based local search to enhance the performance of NSGA-II. This Hybrid method handles two quality criteria in QoS as two trade-off objectives. However, the local search of this method is randomly applied to a predefined large number of subproblems without focusing on the most suitable candidate solutions. In this paper, we propose a memetic NSGA-II with EDA-based local search. Particular, EDA performs the local improvements of a few well-selected composite services in different regions of the Pareto front. We also aim to handle two practical trade-off objectives with respect to QoS and QoSM. Our experiments have shown that our proposed method outperforms the recent state-of-the-art algorithms and the baseline NSGA-II method with respect to effectiveness and efficiency.

Cloud computing enables enormous computational resources to be scheduled as parallel workflow applications. Most traditional heuristics can only solve one particular scheduling problem. For example, Heterogeneous Earliest Finish Time (HEFT) and Greedy algorithms allocate resources to given ordered list of tasks using a specific single heuristic, which only caters for a specific scheduling problem, e.g. the fixed number of tasks in a workflow and available resources. Many researchers considered the heterogeneous workflows and cloud resources in scheduling in order to minimize the cost and makespan, but the solutions provided are only for specific workflow pattern. In this paper, we demonstrate a workflow scheduling problem which considers the combination of heterogeneous workflows as well as heterogeneous computing resources. We proposed Flexible Scheduling using Genetic Programming (FSGP) approach to minimize the total cost and makespan of heterogeneous workflows in the cloud. The performance of our proposed FSGP is regardless of the number of tasks in the workflow, available resources and workflow patterns. We evaluated our proposed approach using a benchmark dataset. Performance evaluation of some well-known algorithms such as HEFT and greedy algorithms exhibit that our FSGP approach perform better than other competing algorithms.

This article presents a next-generation test data compression scheme. It builds on the isometric compression paradigm, but makes it more flexible and elevates encoding efficiency to values unachievable through state-of-the-art sequential compression schemes. Furthermore, its programmable selection of full-toggle scan chains ensures high test coverage and virtually eliminates compression aborts. The presented approach follows from a fundamental observation that among test cube care bits, only a very few have a status of necessary assignments (their locations cannot be changed), whereas the remaining ones have alternative sites. These test cubes are used to form circular test templates which synergistically control a decompressor and guide back ATPG to find assignments yielding highly compressible test patterns. A redesigned low-silicon-area decompressor is also capable of reducing switching rates in scan chains with a new test power control scheme. The experimental results obtained for large industrial designs and other benchmark circuits confirm the superiority of the proposed scheme over existing techniques and are reported herein.