Soft Real-time Systems Research Articles

A novel continual reinforcement learning-based expert system for self-optimization of soft real-time systems

Virtual globes are soft real-time systems, which stream multi-resolution data sets and render world-scale landscapes in real-time. Such systems require an adaptation mechanism to fulfill user and quality of experience (QoE) requirements. Configuration and tuning of parameters to optimize system performance is challenging due to run-time uncertainties. In this work, we propose a continual reinforcement learning-based expert system for self-optimization of a soft real-time system to meet user and quality objectives. The proposed system is capable of learning continuously by utilizing the previously learned knowledge to fulfill adaptation requirements. The proposed managing system interacts with a managed system and evolves a knowledge base of adaptation rules for run-time optimization of the managed system to meet user and quality objectives. We devise a learning classifier system for the evolution, storage, and transfer of adaptation rules using a code fragments-based rich encoding scheme. We transform user goals, system states, and execution environment conditions into observable states. We validated the proposed system for the same and cross-domain knowledge transfer and reuse scenarios for different user goals and execution environments. The results of the proposed system are compared with baseline and state-of-the-art methods where the proposed system outperformed the methods in terms of adaptation accuracy.

Real-Time Synchronization System for Press Machines

A real-time system is one that is put through its paces, meaning that it responds within the allotted time frame or fulfills the deadline. There are two sorts of real-time systems: hard and soft. Both are employed in various situations. When using hard real-time systems, even a delay of a few nanoseconds or microseconds is not permitted. Soft real-time systems offer some leeway in the depiction of time. Modern industries rely heavily and extensively on real-time systems. In order to produce the most precise and accurate results, systems are built to operate in real time. Real-time systems have a significant influence on human health and treatment in the realm of medical science. Real-time systems in medical science have saved countless lives and allowed for the more straightforward treatment of complicated diseases. This research proposal is centred on a "Real-time synchronization system for press machines" to regulate the machines for manufacturing if the number of workpieces exceeds and enable or disable the machines, utilizing sensors, controllers, and more parts. This real-time synchronization system is dependable and reasonably priced; it can be applied to industrial settings or other settings. This system has good sustainability; two press tool machines can use the system.

Development of a Low-Cost Relay Prototype for Real-Time Power Protection Functions

This paper presents a Raspberry Pi 3B+ based low-cost universal relay able to run power protection functions in real-time. The configurations necessary for this single-board computer to be able to provide real-time response are shown, as for latency tests to verify its response time. An experimental circuit was built to send three-phase fault signals from a Relay tester to the low-cost relay, in order to evaluate its response time to clear faults in comparison to a commercial relay. A neural networks algorithm was developed and executed in real-time by the proposed low-cost relay, which is able to differentiate three-phase faults from transient signals created from large load variations. The results show that the low-cost relay is capable of running simple and complex protection functions within a pre-defined runtime and acceptable precision, compared to a commercial protection relay. However, the sampling frequency the low-cost relay is able to handle is limited. The results have also shown that the low-cost relay meet the requirements for a soft real-time system, which is not ideal for practical power protection systems that require hard real-time systems.

Development of Advanced Monitoring System in Real Time Environment

In this paper, the problem of providing a good computerized 
 monitoring system is discussed. As an application, Advanced 
 Monitoring System (AMS) is designed and built for this purpose. 
 AMS is an image processing, a soft real time system, which is 
 concerned with enhancement of the monitoring services with lower 
 cost and a high performance. It seems to be very broad area of 
 research as it relates to many important subjects of computer science. 
 The produced system has accomplished its intended functions and 
 services in trusted fashion.

A Predictable QoS-aware Memory Request Scheduler for Soft Real-time Systems

A memory controller manages the flow of data to and from attached memory devices. The order in which a set of contending memory requests from different tasks are serviced significantly influences the rate of progress and completion times of these tasks. This in turn may affect the Quality-of-Service (QoS) delivered by these tasks. In this article, we focus towards the design of a QoS-aware memory controller targeted towards soft real-time systems. The proposed memory controller tries to generate an urgency-based schedule for the contending memory requests based on the allowable response time latencies associated with each request. The objective is to improve task-level response time predictability while maximizing acquired QoS. Exhaustive experiments carried out using real memory traces and standard simulation tools exhibit the practical efficacy of the proposed memory controller design.

Fine-grained parallelism framework with predictable work-stealing for real-time multiprocessor systems

Lately, parallel task models have received much attention in the development of real-time multiprocessor systems, as they allow highly compute-intensive tasks to have shorter deadlines which is very much required in modern reactive systems. However, missing modularity and portability can make parallel programming a cumbersome endeavor. As a consequence, compute-intensive sectors in the desktop and server segment have relied on parallelism frameworks such as Intel Threading Building Blocks, Cilk and OpenMP. These parallelism frameworks, however, are optimized for decent average case performance and consequently, do not meet the strict requirements imposed by real-time systems.In this paper, we present a proof-of-concept parallelism framework which was implemented in particular for soft real-time systems and having tight timing and safety requirements of such critical systems in mind. The proposed runtime system implements static memory allocation in a work-stealing environment that conforms to the strict space and tight probabilistic time bounds of work-stealing schedulers. Furthermore, we evaluate the performance of this framework by conducting multiprogrammed benchmarks on a real-time embedded multicore architecture.

Fault Tolerance in the Joint EDF-RMS Algorithm: A Comparative Simulation Study

Failure is a systemic error that affects overall system performance and may eventually crash across the entire configuration. In Real-Time Systems (RTS), deadline is the key to successful completion of the program. If tasks effectively meet the deadline, it means the system is working in pristine order. However, missing the deadline means a systemic fault due to which the system can crash (hard RTS) or degrade inclusive performance (soft RTS). To fine-tune the RTS, tolerance is the critical issue and must be handled with extreme care. This article explains the context of fault tolerance with improvised Joint EDF-RMS algorithm in RTS. The backup method has been derived to prevent the system from being recursively migrating the same task. If any task migrates three times, this migrated task will get shifted to the backup queue. This backup queue assigns the task to a backup processor and is destined for final execution. For performance evaluation purposes, a relative graph between fault and failure rates, failure and total processor utilization along with other averages have been evaluated. Furthermore, these archived results are compared with fault-tolerant Earliest Deadline First (EDF) and Rate Monotonic Scheduling (RMS) algorithms independently in relatively similar conditions. These comparisons show better performance against overloading conditions.

Tardiness Bounds for Sporadic Gang Tasks Under Preemptive Global EDF Scheduling

Following the trend of increasing autonomy in cyber-physical systems, parallel embedded architectures have enabled devices to better handle the large streams of data and intensive computation required by such autonomous systems. However, while the explosion of highly-parallel platforms has seen a proportional growth in the number of applications/devices that utilize these platforms, the embedded systems community's understanding of how to build time-predictable, safety-critical systems with parallel platforms has not kept pace. As a well-motivated but challenging parallel scheduling model, gang scheduling requires all parallel threads of each parallel task to simultaneously execute in unison, which is in contrast to traditional, multi-threaded parallel scheduling, where a parallel task may spawn multiple threads, and each thread will be scheduled independently of other threads of the same task. While increasing research efforts on hard real-time (HRT) gang scheduling have recently been seen, the problem of gang scheduling in the context of soft real-time (SRT) systems, where provably bounded deadline tardiness can be tolerated, has hardly been studied yet. In this article, we derive and prove the first tardiness bounds for sporadic gang task systems under preemptive GEDF scheduling. A total utilization bound for SRT-schedulability is required for ensuring such tardiness bounds but it is shown to be tight with respect to the platform capacity and maximum parallelism-induced idleness. Furthermore, we also empirically evaluate the effects of different degrees of task parallelism upon the SRT-schedulability.

Towards soft real-time fault diagnosis for edge devices in industrial IoT using deep domain adaptation training strategy

Artificial intelligence and industrial internet of things (IIoT) have been rejuvenating the fault diagnosis systems in Industry 4.0 for avoiding major financial losses caused by faults in rotating machines. Meanwhile, the diagnostic systems are provided with a number of sensory inputs that introduce variations in input space which causes difficulty for the algorithms in edge devices. This issue is generally dealt with bi-view cross-domain learning approach. We propose a soft real-time fault diagnosis system for edge devices using domain adaptation training strategy. The investigation is carried out using deep learning models that can learn representations irrespective of input dimensions. A comparative analysis is performed on a publicly available dataset to evaluate the efficacy of the proposed approach which achieved accuracy of 88.08%. The experimental results show that our method using long short-term memory network achieves the best results for the bearing fault detection in an IIoT environmental setting.

A Hybrid Multiprocessor Scheduler for Soft Real-Time System with Processor Affinity Concept

Now a days numerous real-time OS like QNX, VxWorks, LynxOS and real-time extension of Linux practicing the processor affinity concept to schedule the real-time tasks and it offers more flexible approach instead of traditional approaches defined in the literature. Limiting the migration may reduce the context switch overhead and improve the cache performance but it largely influence on other parameters which are most important for the any real-time operating system. It degrades overall performance by decreasing the schedulability of tasks, increase the deadline miss ratio. This paper presents the problem in existing processor affinity based approach which confines the schedulability of the tasks and designed a novel processor affinity based algorithm to enhance the schedulability of the tasks and decreases the average deadline miss ratio by providing the flexible migration policy with the priority reassignment mechanism.

Optimized scheduling algorithm for soft Real-Time System using particle swarm optimization technique

Scheduling of tasks in Real-Time Systems is based on static or dynamic priority like earliest deadline first (EDF) and rate monotonic, respectively. The static scheduler does not give assurance of scheduling all tasks during the underload scenario, whereas dynamic scheduler performs poorly during an overload scenario. This paper has proposed a swarm intelligence-based scheduling algorithm that can overcome both the situations. This paper has used particle swarm optimization (PSO) based swarm technique to design the new scheduling approach. It considers each task as a particle and applied modified PSO technique to identify the most critical task to execute. The efficiency of the newly proposed method has been compared with existing EDF and ACO based scheduling algorithms considering two significant parameters, the success ratio and the effective CPU utilization. All three algorithms have been tested on the simulator with a Soft Real-time periodic task set on 500 timelines. It has been observed that during the underload scenario, the proposed algorithm performs equally to EDF and ACO based algorithms. During overload and highly overload situations, the proposed algorithm performs batter compared to EDF and ACO based algorithms.

Applied machine learning for a zero defect tolerance system in the automated assembly of pharmaceutical devices

Creating reliable and robust quality control systems that identify process errors while having a low number of false rejects is a considerable challenge in the automated manufacturing industry. Especially in the pharmaceutical industry, where a product's quality has to be ensured at all costs, a large amount of false rejects is acceptable to guarantee the integrity of all released products. As standard quality control systems mainly perform a binary classification, most of them do not provide insights about the reason behind rejections. As a result, the underlying reason for the rejects, such as degradation in equipment or wrong settings in process parameters, often goes unnoticed. Yet, these systems are based on conservative approaches that incorporate the uncertainties related to the measurement system and process variation such as batch-to-batch variations and assembly tolerances. In this contribution, a new data-driven quality control system is suggested. The system is based on well-established machine learning methods that differentiate multiple types of errors in the assembly processes of medical products. Trained on process data, the system's functionality is demonstrated in a pre-study and two real industrial use cases. Moreover, application-specific differences are discussed. It is shown that for the two use cases and a limited number of batches the system not only detects 100% of all defective products but also limits the number of false rejects to an acceptable amount. In all of the application examples, the system has the potential to be executed as a soft real-time system that allows integration into industrial processes. Moreover, it is shown that the algorithm can present the extracted knowledge in various forms understandable for humans, allowing for more informed decision making.

Statically optimal dynamic soft real-time semi-partitioned scheduling

Semi-partitioned scheduling is an approach to multiprocessor real-time scheduling where most tasks are fixed to processors, while a small subset of tasks is allowed to migrate. This approach offers reduced overhead compared to global scheduling, and can reduce processor capacity loss compared to partitioned scheduling. Prior work has resulted in a number of semi-partitioned scheduling algorithms, but their correctness typically hinges on a complex intertwining of offline task assignment and online execution. This brittleness has resulted in few proposed semi-partitioned scheduling algorithms that support dynamic task systems, where tasks may join or leave the system at runtime, and few that are optimal in any sense. This paper introduces EDF-sc, the first semi-partitioned scheduling algorithm that is optimal for scheduling (static) soft real-time (SRT) sporadic task systems and allows tasks to dynamically join and leave. The SRT notion of optimality provided by EDF-sc requires deadline tardiness to be bounded for any task system that does not cause over-utilization. In the event that all tasks can be assigned as fixed, EDF-sc behaves exactly as partitioned EDF. Heuristics are provided that give EDF-sc the novel ability to stabilize the workload to approach the partitioned case as tasks join and leave the system.

An Energy-Efficient Periodic Resource Model for Bounded Delay-Tolerant Real-Time Systems

In the past few years, the energy consumption of real-time systems has become an essential topic of research, especially with the rise of the internet of things. In hard real-time systems, sensitivity to timing is the focal feature of system behaviors. Real-time systems must consider time constraints (mainly deadlines) to deliver the proper results. Therefore, typical real-time systems are inefficient in resource reservation, where resource supply is always higher than the workload demand and thus consumes more power than required. In firm and soft real-time systems, some deadline violations can be tolerated in a specific time interval. The worst-case execution time can be delayed to abound range. Therefore, instead of over-provisioning resources when considering the demand cannot be higher than supply, we propose a new algorithm that can efficiently estimate the resource reservation and reduce power consumption. The proposed algorithm is theoretically studied, and results show that while still satisfying the delay requirements of control systems, significant energy savings (up to 78.5%) could be achieved.

Reachability and reward checking for stochastic timed automata

Stochastic timed automata are an expressive formal model for hard and soft real-time systems. They support choices and delays that can be deterministic, nondeterministic or stochastic. Stochastic choices and delays can be based on arbitrary discrete and continuous distributions. In this paper, we present an analysis approach for stochastic timed automata based on abstraction and probabilistic model checking. It delivers upper/lower bounds on maximum/minimum reachability probabilities and expected cumulative reward values. Based on theory originally developed for stochastic hybrid systems, it is the first fully automated model checking technique for stochastic timed automata. Using an implementation as part of the Modest Toolset and four varied examples, we show that the approach works in practice and present a detailed evaluation of its applicability, its efficiency, and current limitations.

Execution of Partial State Machine Models

The iterative and incremental nature of software development using models typically makes a model of a system incomplete (i.e., partial) until a more advanced and complete stage of development is reached. Existing model execution approaches (interpretation of models or code generation) do not support the execution of partial models. Supporting the execution of partial models at the early stages of software development allows early detection of defects, which can be fixed more easily and at a lower cost. This paper proposes a conceptual framework for the execution of partial models, which consists of three steps: static analysis, automatic refinement, and input-driven execution. First, a static analysis that respects the execution semantics of models is applied to detect problematic elements of models that cause problems for the execution. Second, using model transformation techniques, the models are refined automatically, mainly by adding decision points where missing information can be supplied. Third, refined models are executed, and when the execution reaches the decision points, it uses inputs obtained either interactively or by a script that captures how to deal with partial elements. We created an execution engine called PMExec for the execution of partial models of UML-RT (i.e., a modeling language for the development of soft real-time systems) that embodies our proposed framework. We evaluated PMExec based on several use-cases that show that the static analysis, refinement, and application of user input can be carried out with reasonable performance and that the overhead of approach, which is mostly due to the refinement and the increase in model complexity it causes, is manageable. We also discuss the properties of the refinement formally and show how the refinement preserves the original behaviors of the model.

Online Resource Management for Improving Reliability of Real-Time Systems on “Big–Little” Type MPSoCs

Heterogeneous multiprocessor systems on a chips (MPSoCs) consisting of cores with different performance/power characteristics are widely used in many real-time embedded systems, where both soft-error reliability and lifetime reliability are key concerns. Although existing efforts have investigated related problems, they either focus on one of the two reliability concerns or propose time-consuming scheduling algorithms that cannot adequately address runtime workload and environmental variations. This paper introduces an online framework which is adaptive to runtime variations and maximizes soft-error reliability while satisfying the lifetime reliability constraint for soft real-time systems executing on MPSoCs that are composed of high-performance cores and low-power (LP) cores. Based on each core’s executing frequency and utilization, the framework performs workload migration between high-performance cores and LP cores to reduce power consumption and improve soft-error reliability. Experimental results based on different hardware platforms show that the proposed approach reduces the probability of failures due to soft errors by at least 17% and 50% on average compared to a number of representative existing approaches that satisfy the same lifetime reliability constraints.

Hybrid Scheduler (S_LST) for Soft Real-Time System based on Static and Dynamic Algorithm

In the Soft Real-Time System scheduling process with the processor is a critical task. The system schedules the processes on a processor in a time interval, and hence the processes get chance to executes on the processor. Priority-driven scheduling algorithms are sub-categorized into mainly two categories called Static Priority and Dynamic Priority Scheduler. Critical Analysis of more static and dynamic priority scheduling algorithms have been discussed in this paper. This paper has covered the static priority algorithms like Rate Monotonic (RM) and Shortest Job First (SJF) and the dynamic priority algorithms like Earliest Deadline First (EDF) and Least Slack Time First (LST). These all algorithms have been analyzed with preemptive process set and this paper has considered all the process set are periodic. This paper has also proposed a hybrid approach for efficient scheduling. In a critical analysis, it has been observed that while scheduling in underload situation dynamic priority algorithms perform well and even EDF also make sure that all process will meet their deadline. However, in an overload situation, the performance of dynamic priority algorithms reduce quickly, and most of the task will miss its deadline, whereas static priority scheduling algorithms miss a few deadlines, even it is possible to schedule all processes in underload situation, whereas in an overload situation, the static algorithms perform well compared to the dynamic scheduler. This paper is proposing one Hybrid algorithm call S_LST which uses the concept of LST and SJF scheduling algorithm. This algorithm has been applied to the periodic task set, and observations are registered. We have observed the Success Ratio (SR) & Effective CPU Utilization (ECU) and compared all algorithms in the same conditions. It is noted that instead of using LST and SJF as an independent algorithm, Hybrid algorithm S_LST performs well in underload and overload scenario. Practical investigations have been led on a huge dataset. Data Set consists of the 7000+ process set, and each process set has one to nine processes and load varies between 0.5 to 5. It has been tried on 500-time unit to approve the rightness everything being equal.

ACO Based Scheduling Method for Soft RTOS with Simulation and Mathematical Proofs

The Ant Colony Optimization (ACO) algorithm is a mathematical model enlivened by the system searching conduct of ants. By taking a gander at the qualities of ACO, it is most suitable for scheduling of tasks in soft real-time systems. In this paper, the ACO based scheduling method for the soft real-time operating system (RTOS) has been profound with mathematical and practical proof. In Mathematical proof, three different Propositions and two Theorems have been given, which prove the correctness of the proposed algorithm. Practical experiments also support mathematical proofs. During the investigation, observations are gathered with different periodic task set. Algorithms have been observed regarding Success Ratio (SR) and Effective CPU utilization (ECU). ACO based scheduling algorithm has been compared with the Earliest Deadline First (EDF) algorithm with parameter SR and ECU. The EDF is dynamic scheduling algorithm and it is most suitable in RTOS when task set is preemptable. It is noted that the new algorithm is equally efficient during under loaded conditions when CPU load is less than one. ACO based scheduling algorithm performs superior during the overloaded conditions when CPU load is more than one where as EDF algorithm performance degraded in overload condition. Empirical study has been executed with a hefty Dataset consist of more than 7500 task set, and a set contains different one to nine processes where CPU load is dynamic for each process set and differ from 0.5 to 5. Algorithms have been executed on five-hundred-time unit for each process set to authenticate the accuracy of both algorithms.

Improving Reliability of Soft Real-Time Embedded Systems on Integrated CPU and GPU Platforms

Multiprocessor systems on a chip consisting of integrated CPUs and GPUs are suitable platforms for real-time embedded applications requiring massively parallel processing. For such applications, lifetime reliability due to permanent faults and soft-error reliability due to transient faults are major concerns. Detailed execution profiling has revealed that a CUDA task’s CPU execution time significantly increases if the task executes on a different core than the operating system (OS). Based on this observation, an extended task model is introduced to consider the execution time dependencies among tasks and the OS. A hybrid framework is proposed to improve soft-error reliability while satisfying a lifetime reliability constraint for soft real-time systems executing on integrated CPU and GPU platforms. This framework: 1) reduces the total utilization of cores and improves soft-error reliability via off-line task mapping; 2) achieves a higher lifetime reliability through task migration at run time; and 3) improves soft-error reliability by dynamically scaling frequencies of CPU and GPU cores. The experimental results show that the proposed framework leads to a system that can execute without soft errors for at least 4 days (4 times) and 6 days (6 times) longer, on average, than existing approaches.