Feedback Control Scheduling Research Articles

Towards QoS-Based Embedded Machine Learning

Due to various breakthroughs and advancements in machine learning and computer architectures, machine learning models are beginning to proliferate through embedded platforms. Some of these machine learning models cover a range of applications including computer vision, speech recognition, healthcare efficiency, industrial IoT, robotics and many more. However, there is a critical limitation in implementing ML algorithms efficiently on embedded platforms: the computational and memory expense of many machine learning models can make them unsuitable in resource-constrained environments. Therefore, to efficiently implement these memory-intensive and computationally expensive algorithms in an embedded computing environment, innovative resource management techniques are required at the hardware, software and system levels. To this end, we present a novel quality-of-service based resource allocation scheme that uses feedback control to adjust compute resources dynamically to cope with the varying and unpredictable workloads of ML applications while still maintaining an acceptable level of service to the user. To evaluate the feasibility of our approach we implemented a feedback control scheduling simulator that was used to analyze our framework under various simulated workloads. We also implemented our framework as a Linux kernel module running on a virtual machine as well as a Raspberry Pi 4 single board computer. Results illustrate that our approach was able to maintain a sufficient level of service without overloading the processor as well as providing an energy savings of almost 20% as compared to the native resource management in Linux.

Improving the quality of service of real-time database systems through a semantics-based scheduling strategy

A real-time database management system (RTDBMS) aims to manage applications with a large number of data being accessed by update and user transactions having to meet some time constraints. We are interested in improving the quality of service (QoS) in RTDBMSs by optimising the execution of transactions to improve meeting their deadlines. A new scheduling strategy is developed and evaluated into a QoS management architecture called feedback control scheduling (FCS) proposed for RTDBMSs. The proposed strategy is based on the new advanced earliest deadline first based on transactions aggregation and data semantic links (AEDF-TAL-DSL) protocol we developed. It introduces new parameters relating to the aggregation links (defined according to the type of operations composing the transactions and the data they access) existing between transactions as well as the semantic links appearing between the users' queries. We also show the contributions provided by our approach through simulation results.

An Optimization of Manufacturing Systems using a Feedback Control Scheduling Model

In complex production system that involves multiple process, unplanned disruption often turn to make the entire production system vulnerable to a number of problems which leads to customer’s dissatisfaction. However, this problem has been an ongoing problem that requires a research and methods to streamline the entire process or develop a model that will address it, in contrast to this, we have developed a feedback scheduling model that can minimize some of this problem and after a number of experiment, it shows that some of this problems can be eliminated if the correct remedial actions are implemented on time.

QoS management in real-time spatial big data using a feedback control scheduling

Heterogeneous real-time spatial data management is a very active research domain nowadays. We are talking about a real-time spatial big data that process a large amount of heterogeneous data accessed simultaneously by two types of transactions update transactions and user transactions. In these applications, it is desirable to execute transactions within their deadlines using a real-time spatial data. But the real-time spatial big data can be overloaded and many transactions may miss their deadlines, or real-time spatial data can be violated. To address these problems, we proposed, as a first contribution, a new architecture called feedback control scheduling architecture for real-time spatial big data (FCSA-RTSBD) (Hamdi et al., 2015). Then, we propose, as a second contribution, two-shadow speculative concurrency control (SCC-2S) with priority and imprecise computation (SCC-2S-P-IC). Finally, a simulation study is shown to prove that our contributions can achieve a significant performance improvement using the TPC-DS (TPC, 2014) benchmark.

QoS management in real-time spatial big data using a feedback control scheduling

QoS management in real-time spatial big data using a feedback control scheduling

Improving QoS in real-time data warehouses by using feedback control scheduling

Nowadays the update frequency for traditional data warehouses cannot meet the objectives of real-time data analysis relying on data freshness. To alleviate this problem, the real-time data warehouse (RTDW) technology has emerged. A RTDW allows decision makers to access and analyse fresh data as fast as possible in order to support real-time decision processes. The RTDW must often deal with transient usage charges, due to the unpredictability of access to data. The purpose of this paper is two-fold: to maintain the behaviour of the RTDW at a stable state, as well as the reduction of the number of transactions responsible for not meeting their deadline. Moreover, we focus on optimisation techniques to speed up query processing; in particular, a query response time optimisation and storage space optimisation. This paper proposes our FCSA-RTDW architecture (feedback control scheduling architecture for RTDW) which deals with quality of service management by optimising the resources used and reducing significantly the RTDW overloads.

QOS MANAGEMENT IN REAL-TIME SPATIAL BIG DATA USING FEEDBACK CONTROL SCHEDULING

Abstract. Geographic Information System (GIS) is a computer system designed to capture, store, manipulate, analyze, manage, and present all types of spatial data. Spatial data, whether captured through remote sensors or large scale simulations has always been big and heterogenous. The issue of real-time and heterogeneity have been extremely important for taking effective decision. Thus, heterogeneous real-time spatial data management has become a very active research domain. Existing research has principally focused on querying of real-time spatial data and their updates. But the unpredictability of access to data maintain the behavior of the real-time GIS unstable. In this paper, we propose the use of the real-time Spatial Big Data and we define a new architecture called FCSA-RTSBD (Feedback Control Scheduling Architecture for Real-Time Spatial Big Data). The main objectives of this architecture are the following: take in account the heterogeneity of data, guarantee the data freshness, enhance the deadline miss ratio even in the presence of conflicts and unpredictable workloads and finally satisfy the requirements of users by the improving of the quality of service (QoS).

具有通信受限的网络控制系统量化反馈控制

This paper is concerned with the joint design of quantized feedback control and dynamic scheduling for networked control systems with constrained communication capacity. A new discrete-time switched system is proposed to the networked control system subject to both communication constraints and input quantization. By using Lyapunov stability theory, sufficient conditions of mode- and quantization-dependent robust stabilization and a quantized feedback controller are derived for networked control system. Based on the model of switched system, a dynamic scheduling policy is designed to guarantee the asymptotic stability of close-loop system. An illustrative example is finally given to show the effectiveness of the proposed method.

Energy Efficient Scheduling For Real-Time Embedded Systems With Precedence And Resource Constraints

Energy consumption is a critical design issue in real-time systems, especially in battery- operated systems. Maintaining high performance, while extending the battery life between charges is an interesting challenge for system designers. Dynamic Voltage Scaling and Dynamic Frequency Scaling allow us to adjust supply voltage and processor frequency to adapt to the workload demand for better energy management. Usually, higher processor voltage and frequency leads to higher system throughput while energy reduction can be obtained using lower voltage and frequency. Many real-time scheduling algorithms have been developed recently to reduce energy consumption in the portable devices that use voltage scalable processors. For a real-time application, comprising a set of real-time tasks with precedence and resource constraints executing on a distributed system, we propose a dynamic energy efficient scheduling algorithm with weighted First Come First Served (WFCFS) scheme. This also considers the run-time behaviour of tasks, to further explore the idle periods of processors for energy saving. Our algorithm is compared with the existing Modified Feedback Control Scheduling (MFCS), First Come First Served (FCFS), and Weighted scheduling (WS) algorithms that uses Service-Rate-Proportionate (SRP) Slack Distribution Technique. Our proposed algorithm achieves about 5 to 6 percent more energy savings and increased reliability over the existing ones.

Quasi-decentralized Scheduled Output Feedback Control of Process Systems Using Wireless Sensor Networks

This paper presents a quasi-decentralized output feedback control structure for multi-unit plants with limited state measurements and distributed control systems that exchange information over a resource-constrained wireless sensor network (WSN). The networked control structure brings together model-based feedback control, state estimation and sensor scheduling to enforce closed-loop stability while simultaneously minimizing the rate of communication over the WSN. Initially, an observer-based output feedback controller is designed for each unit. To conserve the resources of the wireless devices, communication between the local control systems is suspended periodically for extended time periods during which each control system relies on models of the plant units to generate the necessary control action. Communication is then re-established at discrete time instances according to a certain schedule that determines the order and times at which the wireless sensor suites transmit the state estimates needed to update the states of the models embedded in the target units. By analyzing the combined discrete-continuous behavior of the scheduled closed-loop plant, we explicitly characterize the stability of the networked closed-loop system in terms of the communication rate, the sensor transmission schedule, the accuracy of the models, as well as the controller and observer design parameters. The results are illustrated using a chemical plant example where it is shown that by judicious management of the interplays between the control, communication and scheduling design parameters, it is possible to stabilize the plant while simultaneously enhancing the savings in WSN resources beyond what is possible with concurrent transmission configurations.

Feedback-Controlled Server for Scheduling Aperiodic Tasks

This paper proposes a scheduling scheme using feedback control to reduce the response time of aperiodic tasks with soft real-time constraints. We design an algorithm based on the proposed scheduling scheme and Total Bandwidth Server (TBS) that is a conventional server technique for scheduling aperiodic tasks. We then describe the feedback controller of the algorithm and give the control parameter tuning methods. The simulation study demonstrates that the algorithm can reduce the mean response time up to 26% compared to TBS in exchange for slight deadline misses. Keywords—Real-Time Systems, Aperiodic Task Scheduling, Feedback-Control Scheduling, Total Bandwidth Server

Experimental evaluation of linear time-invariant models for feedback performance control in real-time systems

In recent years a new class of soft real-time applications operating in unpredictable environments has emerged. Typical for these applications is that neither the resource requirements nor the arrival rates of service requests are known or available a priori. It has been shown that feedback control is very effective to support the specified performance of dynamic systems that are both resource insufficient and exhibit unpredictable workloads. To efficiently use feedback control scheduling it is necessary to have a model that adequately describes the behavior of the system. In this paper we experimentally evaluate the accuracy of four linear time-invariant models used in the design of feedback controllers. We introduce a model (DYN) that captures additional system dynamics, which a previously published model (STA) fails to include. The accuracy of the models are evaluated by validating the models with regard to measured data from the controlled system and through a set of experiments where we evaluate the performance of a set of feedback control schedulers tuned using these models. From our evaluations we conclude that second order models (e.g., DYN) are more accurate than first order models (e.g. STA). Further we show that controllers tuned using second order models perform better than controllers tuned using first order models.

Specification and management of QoS in real-time databases supporting imprecise computations

Real-time applications such as e-commerce, flight control, chemical and nuclear control, and telecommunication are becoming increasingly sophisticated in their data needs, resulting in greater demands for real-time data services that are provided by real-time databases. Since the workload of real-time databases cannot be precisely predicted, they can become overloaded and thereby cause temporal violations, resulting in damage or even a catastrophe. Imprecise computation techniques address this problem and allow graceful degradation during overloads. In this paper, we present a framework for QoS specification and management consisting of a model for expressing QoS requirements, an architecture based on feedback control scheduling, and a set of algorithms implementing different policies and behaviors. Our approach gives a robust and controlled behavior of real-time databases, even for transient overloads and with inaccurate runtime estimates of the transactions. Further, performance experiments show that the proposed algorithms outperform a set of baseline algorithms that uses feedback control.

Semi-automatic control system for hydraulic shovel

A semi-automatic control system for a hydraulic shovel has been developed. Using this system, unskilled operators can operate a hydraulic shovel easily and accurately. A mathematical control model of a hydraulic shovel with a controller was constructed and a control algorithm was developed by simulation. This algorithm was applied to a hydraulic shovel and its effectiveness was evaluated. High control accuracy and high-stability performance were achieved by feedback plus feedforward control, nonlinear compensation, state feedback and gain scheduling according to the attitude.