Many safety critical real-time systems are considered certified when they meet failure probability requirements with respect to the maximum permitted incidences of failure per hour. In this paper, the mixed-criticalit...
详细信息
Many safety critical real-time systems are considered certified when they meet failure probability requirements with respect to the maximum permitted incidences of failure per hour. In this paper, the mixed-criticality task model with multiple worst case execution time (WCET) estimations is extended to incorporate such system-level certification restrictions. A new parameter is added to each task, characterizing the distribution of the WCET estimations -- the likelihood of all jobs of a task finishing their executions within the less pessimistic WCET estimate. An efficient algorithm named LFF-Clustering is derived for scheduling mixed-criticality systems represented by this model. Experimental analyses show our new model and algorithm out-perform current state-of-the-art mixed-criticality scheduling algorithms.
Cloud computing is a technology which computes the resources and delivered as a service over a network in virtual form. Cloud computing is a bigger and broad concept which allows people to access the services through ...
详细信息
ISBN:
(纸本)9781479966295
Cloud computing is a technology which computes the resources and delivered as a service over a network in virtual form. Cloud computing is a bigger and broad concept which allows people to access the services through the use of internet from various devices. It has rapidly gained the popularity and captures the IT industries, colleges and institutes in recent years. Due to its advanced service benefits, numbers of users are increasing day-by-day. Hence, the need of task scheduling in cloud is increasing. Task scheduling is done to allocate the tasks onto the resources effectively and efficiently. A number of scheduling algorithms have been developed till now. A good scheduling strategy has the capability to adapt to changing environment and type of tasks. In this paper, we proposed a meta-heuristic scheduling algorithm i.e. improved honey bees life scheduling algorithm for a public a cloud (IHBSLA). The new scheduling algorithm is simulated using Cloudsim toolkit. Experimental results showed that our proposed algorithm performs 50% better than honey bees life scheduling algorithm (HBLSA) in terms of cost.
In Wireless Sensor Networks (WSNs) a sensor node periodically produces data which is processed and then transmitted to the base station (BS) for analysis. A key challenge in the WSNs is to schedule the activity of the...
详细信息
ISBN:
(纸本)9781479926084
In Wireless Sensor Networks (WSNs) a sensor node periodically produces data which is processed and then transmitted to the base station (BS) for analysis. A key challenge in the WSNs is to schedule the activity of the node in the cluster based environment for improvement in throughput, energy consumption and delay. In this regards, the paper proposes Cluster-based Myopic and Non-myopic scheduling algorithm (CMNS) which considers the conflict free schedule based on the current and next state. It considers TDMA as the MAC layer protocol and schedules the aggregated packets with consecutive slots. Simulation studies show that, CMNS reduces the number of conflicts, energy consumption and increases the throughput as compared with state-of-the-art solutions.
Leakage power consumption has become a critical limitation in the normally off low-power systems. Power gating provides a promising solution to reduce leakage energy but cannot avoid data loss. Nonvolatile processor h...
详细信息
ISBN:
(纸本)9781467366892
Leakage power consumption has become a critical limitation in the normally off low-power systems. Power gating provides a promising solution to reduce leakage energy but cannot avoid data loss. Nonvolatile processor has paved the way to achieve zero leakage power while maintaining data. However, nonvolatile processor faces severe inrush current problem when all nonvolatile memories are backed up in parallel. A large inrush current will occur and induce IR drop, which deteriorates the stability of the entire system. This paper proposes a distributed backup control architecture for nonvolatile processors to cope with the inrush current problem. Furthermore, we devise corresponding algorithms to accelerate backup operations under given maximum tolerable current constraints. The proposed techniques are evaluated on a simulation platform and a prototype chip. Experimental results demonstrate up to 26.3% reduction in backup time compared with the sequential backup strategy under the same inrush current constraint.
LTE and WiMAX are the most important radio access technologies in the fourth generation mobile networks providing broadband access. The technologies bring strong QoS support, especially for multimedia services such as...
详细信息
LTE and WiMAX are the most important radio access technologies in the fourth generation mobile networks providing broadband access. The technologies bring strong QoS support, especially for multimedia services such as real-time video streaming, requiring bounded packet loss rate and delay. To reach these goals, effective packet scheduling schemes are needed to allocate radio resources. The function is fundamental and very challenging, particularly when considering system level performance and QoS requirements from users. In this paper, we present a low-complexity field-tested QoS downlink scheduler. Besides proportional fairness, the novel mechanism performs unified resource evaluation across traffic flows honoring individual QoS parameters. A resource multiplier is designed to preserve a relative weighting ratio for real-time traffic and prevent low priority flows from starvation. The waste of resources due to block-based allocation is also eliminated. The testing in both lab and field environments shows satisfactory overall throughput gain and most importantly the effective weighted distribution of resources across diverse traffic flows.
In this paper we perform an analysis of the main software methods for consumption reduction in real time systems (RTS). The study covers both scheduling methods and energy consumption models for RTS, where we focus on...
详细信息
In this paper we perform an analysis of the main software methods for consumption reduction in real time systems (RTS). The study covers both scheduling methods and energy consumption models for RTS, where we focus on sensor nodes. Further on, we propose a new consumption model for a large number of node configurations, which suites most of the main scheduling algorithms. We also introduce a software environment which integrates the new model and provides a set of reports and graphical results to compare the power consumption efficiency of the scheduling algorithms.
We consider the maximum throughput scheduling problem in a millimeter-wave wireless personal area network in which users can use adaptive modulation and coding schemes to change their data rates. The scheduling proble...
详细信息
ISBN:
(纸本)9781479935123
We consider the maximum throughput scheduling problem in a millimeter-wave wireless personal area network in which users can use adaptive modulation and coding schemes to change their data rates. The scheduling problem is to map transmissions to time slots so that the total throughput is maximized. Due to the ultra-wide bandwidth of the mm Wave band, bad scheduling tends to waste significant channel resource. It is worth the effort to consider a more sophisticated scheduling scheme than the simple serial TDMA scheme. The challenge is that the achieved data rate of one flow is limited by the interference from other transmissions in the same slot, which is unknown until the scheduling decision is known. We propose two scheduling algorithms for variable data rate transmissions. The first algorithm is a greedy algorithm, which always chooses the best option at the moment;the second one uses sorting to decide the order that flows are included in a slot. Both algorithms are interference-aware. The algorithms can be applied to transmissions with omnidirectional antennas as well as directional antennas. The simulation results show that the proposed algorithms achieve higher throughput than previous work for adaptive-rate scheduling.
The FlexRay bus is a modern standard used in the automotive industry. It offers deterministic message transmission with zero jitter while using time-triggered scheduling in the static segment. When several vehicle var...
详细信息
ISBN:
(纸本)9781479932351
The FlexRay bus is a modern standard used in the automotive industry. It offers deterministic message transmission with zero jitter while using time-triggered scheduling in the static segment. When several vehicle variants (i.e. different models and their versions) share the same signal, the car manufacturers require to schedule such signal at the same time in all vehicle variants. This requirement simplifies the signal traceability and diagnostics in different vehicle variants using the same platform and simplifies reuse of components and tools. In this paper, we propose a first fit based heuristic algorithm which creates the schedules for several vehicle variants at once, while transmitting a given signal at the same time in all the schedules. The scheduling algorithm also takes the time constraints as release dates and deadlines into account. Finally, different algorithm versions are compared on benchmark sets and low computational time demands are validated on large instances.
WiMAX, IEEE 802.16 is a wireless telecommunication technology, which supports broadband products to offer high speed internet access to the users with always on connectivity, covering more range and mobility features ...
详细信息
ISBN:
(纸本)9781479960545
WiMAX, IEEE 802.16 is a wireless telecommunication technology, which supports broadband products to offer high speed internet access to the users with always on connectivity, covering more range and mobility features as compared to other broadband technologies like Wi-Fi. It provides high quality services by making the classes of services as real time and non-real time and attached a scheduling flow with each class of application to satisfy their QOS (Quality of service) parameters like latency, jitter, throughput, etc. The scheduling algorithms for supporting this QOS in literature, works in high to low priority order which may cause users with low priority to suffer large waiting time and are sometimes dropped from the queue due to crossing latency limits. The paper focus on this perspective and proposed a downlink BS scheduler for WiMAX heterogeneous data in which, it modified the existing Greedy Latency algorithm, used distance deduction scheme and hybrid with SJF scheduler. The simulation results carried out in MATLAB shows that the proposed technique has less processing time or latency, low packet loss rate and high throughput in comparison to existing scheduling algorithms.
Complexity of cloud services acts as a barrier towards adopting cloud to some of the Cloud Service Users. Cloud Service Middleware plays an important role to get rid of such problems. The middleware manages and optimi...
详细信息
Complexity of cloud services acts as a barrier towards adopting cloud to some of the Cloud Service Users. Cloud Service Middleware plays an important role to get rid of such problems. The middleware manages and optimizes the cloud resources to execute various jobs submitted by the users. A middleware can be enhanced to utilize the idle time of the reserved resources in cloud environment by scheduling these resources free of cost to jobs submitted by the same Cloud Service User (CSU) or a different CSU. This enhancement not only makes it possible to utilize the resources to their fullest extent, but also reduces the usage cost of the CSU who reserved the resources (or a different CSU in certain cases). However, finding the mapping between the jobs and available pool of resources is a key challenge to the design of a middleware. This paper proposes some scheduling algorithms to find such mappings that minimizes the job execution cost within public cloud.
暂无评论