Most of today's real-time embedded systems consist Of a heterogeneous mix of fully-programmable processors, fixed-function components or hardware accelerators, and partially-programmable engines. Hence, system des...
详细信息
ISBN:
(纸本)9780769528007
Most of today's real-time embedded systems consist Of a heterogeneous mix of fully-programmable processors, fixed-function components or hardware accelerators, and partially-programmable engines. Hence, system designers are faced with an array of implementation possibilities for an application at hand. Such possibilities typically come with different tradeoffs involving cost, power consumption and packaging constraints. As a result, a designer is no longer interested in one implementation that meets the specified real-time constraints (i.e. is schedulable), but would rather like to identify all schedulable implementations that expose the different possible performance tradeoffs. In this paper we formally define this multicriteria schedulability analysis problem and derive a polynomial-time approximation algorithm for solving it. This result is interesting because the problem of optimally computing even one schedulable solution in our setup (and in most common setups) is computationally intractable (NP-hard). Further our algorithm is reasonably easy to implement, returns good quality (approximate) solutions, and offers significant speedups over optimally computing all schedulable tradeoffs.
In order to speed up the efficiency in designing PISA, a series of resource constraints are imposed. The resource scheduling situation under data dependency, control dependency, and various resource dependency constra...
详细信息
In order to speed up the efficiency in designing PISA, a series of resource constraints are imposed. The resource scheduling situation under data dependency, control dependency, and various resource dependency constraints is to be explored. In this regard, a mixed integer programming model is developed and the optimization objective is determined to keep the number of occupied pipeline stages as short as possible. On the one hand, the data pre-processing work can be carried out, and the data can be transformed into a matrix of 0-1 variables to facilitate the control of the basic information in each module, and the processed data are judged to determine whether the decision variables can satisfy the constraints; on the other hand, the constraints limiting the objective function already in the problem are determined, and the corresponding equations are listed for the three different data dependencies of read-after-write, write-after-read, and write-after-write. At the same time to find the constraints implied by the problem itself, according to the linking relationship of the basic blocks in each board to adjust the order of the boards, due to the directed basic block linking information determines between adjacent boards, any basic block of the latter board must be linked with a certain basic block of the previous board, that is, the remaining hidden constraints can be found by the Big M method. After determining the complete decision variables, constraints and objective functions, a mixed integer programming model can be established to optimize the minimum number of occupied pipeline levels as the objective, and the results are calculated by using the lingo solver, and finally the results are optimized and verified by the heuristic algorithm - particle swarm algorithm to obtain the pipeline of problem one The number of basic blocks is the most when the number of levels is 4, and the corresponding number is 57, and the total number of pipeline levels is 53 at this ti
We consider the problem of scheduling task graphs on two types of unrelated resources, which arises in the context of task-based runtime systems on modern platforms containing CPUs and GPUs. In this paper, we focus on...
详细信息
ISBN:
(数字)9781728168760
ISBN:
(纸本)9781728168777
We consider the problem of scheduling task graphs on two types of unrelated resources, which arises in the context of task-based runtime systems on modern platforms containing CPUs and GPUs. In this paper, we focus on an algorithm named HeteroPrio, which was originally introduced as an efficient heuristic for a particular application. HeteroPrio is an adaptation of the well known list scheduling algorithm, in which the tasks are picked by the resources in the order of their acceleration factor. This algorithm is augmented with a spoliation mechanism: a task assigned by the list algorithm can later on be reassigned to a different resource if it allows to finish this task *** propose here the first theoretical analysis of the HeteroPrio algorithm in the presence of dependencies. More specifically, if the platform contains m and n processors of each type, we show that the worst-case approximation ratio of HeteroPrio is between 1 + max (m/n, n/m) and 2 + (m/n, n/m). Our proof structure allows to precisely identify the necessary conditions on the spoliation strategy to obtain such a guarantee. We also present an in-depth experimental analysis, comparing several such spoliation strategies, and comparing HeteroPrio with other algorithms from the literature. Although the worst case analysis shows the possibility of pathological behavior, HeteroPrio is able to produce, in very reasonable time, schedules of significantly better quality.
Hyperledger Fabric (Fabric for short), a typical consortium blockchain platform, has released many versions that support different consensus protocols to ensure that the system can withstand faults (such as crashes, n...
详细信息
Hyperledger Fabric (Fabric for short), a typical consortium blockchain platform, has released many versions that support different consensus protocols to ensure that the system can withstand faults (such as crashes, network partitions, or network shutdowns). However, if there is a malicious node in the system, Fabric cannot guarantee any behavior of the malicious node. Therefore, when a certain number of nodes are attacked simultaneously, it directly affects the security and efficiency of the entire system. In the face of multi-channel Fabric network architecture, the parallel processing of transactions by node groups linked on different channels significantly improves the throughput of the Fabric network. However, it may also cause system congestion and safety problems. To solve this problem, this paper designs a Lyapunov-based scheduling algorithm for the performance optimization of Fabric. When considering the system's security probability and queue accumulation, we give the maximum consensus verification rate under the minimum security probability guarantee. Finally, we installed the latest stable version v2.0 of Fabric on the cloud server to verify the algorithm's effectiveness.
We study stability regions of multi-rate Gaussian multiple access (MAC) and broadcast (BC) networks with centralized scheduling algorithms. Techniques are presented to characterize stability regions of BC and MAC netw...
详细信息
ISBN:
(纸本)9781424422562
We study stability regions of multi-rate Gaussian multiple access (MAC) and broadcast (BC) networks with centralized scheduling algorithms. Techniques are presented to characterize stability regions of BC and MAC networks with peak power constraints and average power constraints. The duality property that relates the MAC and BC information theoretic capacity regions is found to extend to their stability regions as well, in the average power constraint case.
This paper introduces a high-level synthesis method based on an adaptive evolutionary algorithm that utilizes as soon as possible (ASAP) scheduling algorithm to reduce switching power of CMOS circuits. Experimental ve...
详细信息
ISBN:
(纸本)0780375963
This paper introduces a high-level synthesis method based on an adaptive evolutionary algorithm that utilizes as soon as possible (ASAP) scheduling algorithm to reduce switching power of CMOS circuits. Experimental verification of the algorithm has been performed for a set of test circuits chosen from ISCAS'85 and ISCAS'89 benchmarks, proving its efficiency. The obtained reduction of power consumption varies for different benchmarks from 3 to 42 percent, without deteriorating the throughput of the whole system. Allowing deterioration of the throughput by 10 to 50 percent, the power reduction has been even larger (up to 52 percent).
Motivated by recent work on scheduling with predicted job sizes, we consider the performance of scheduling algorithms with minimal advice, namely a single bit. The analysis of such schemes, besides demonstrating the p...
详细信息
ISBN:
(纸本)9781611976830
Motivated by recent work on scheduling with predicted job sizes, we consider the performance of scheduling algorithms with minimal advice, namely a single bit. The analysis of such schemes, besides demonstrating the power of very limited advice, is quite natural. In the prediction setting, one bit of advice can be used to model a simple prediction as to whether a job is "large" or "small";that is, whether a job is above or below a given threshold. Further, one-bit advice schemes can correspond to mechanisms that tell whether to put a job at the front or the back for the queue, a limitation which may be useful in many implementation settings. Finally, queues with a single bit of advice have a simple enough state that they can be analyzed in the limiting mean-field analysis framework for the power of two choices. Our work follows in the path of recent work by showing that even small amounts of even possibly inaccurate information can greatly improve scheduling performance.
The duration of time for which each application locks each shared resource is critically important in composing multiple independently-developed applications upon a shared "open" platform. The concept of res...
详细信息
ISBN:
(纸本)9780769528007
The duration of time for which each application locks each shared resource is critically important in composing multiple independently-developed applications upon a shared "open" platform. The concept of resource hold time (RHT) - the largest length of time that may elapse between the instant that an application system locks a resource and the instant that it subsequently releases the resource - is formally defined and studied in this paper An algorithm is presented for computing resource hold times for every resource in an application that is scheduled using Earliest Deadline First scheduling, with resource access arbitrated using the Stack Resource Policy. An algorithm is presented for decreasing these RHT'S without changing the semantics of the application or compromising application feasibility.
This paper considers packet scheduling over a broadcast channel with packet erasures to multiple receivers with different messages (multiple uni-cast) each with possibly different hard deadline constraints. A novel me...
详细信息
ISBN:
(纸本)9781538643280
This paper considers packet scheduling over a broadcast channel with packet erasures to multiple receivers with different messages (multiple uni-cast) each with possibly different hard deadline constraints. A novel metric is proposed and evaluated: the global deadline outage probability, which gives the probability that the hard communication deadline is not met for at least one of the receivers. The cut-set upper bound is derived and a scheduling policy is proposed to determine which receiver's packets should be sent in each time slot. This policy is shown to be optimal among all scheduling policies, i.e., it achieves all boundary points of cut-set upper bounds when the transmitter knows the erasure patterns for all the receivers ahead of making the scheduling decision. An expression for the global deadline outage probability is obtained for two receivers and is plotted and interpreted for various system parameters. These plots are not Monte-Carlo simulations, and hence the obtained expression may be used in the design of future downlink broadcast networks. Future extensions to per-user deadline outage probabilities as well as to scenarios with causal knowledge of the channel states are briefly discussed.
Storage system is the infrastructure of big data. Performance analysis of hard disk drive (HDD) plays a fundamental role to improve the efficiency of storage system. State-dependent M/G/1/K queue is used to model HDD,...
详细信息
ISBN:
(纸本)9781509067817
Storage system is the infrastructure of big data. Performance analysis of hard disk drive (HDD) plays a fundamental role to improve the efficiency of storage system. State-dependent M/G/1/K queue is used to model HDD, but it does not have a closed-form solution in the literature. In this paper, we use an M/G/1/K with state-dependent service time to formulate the dynamics of disk random access, where the service time depends on the queue length (batch size of requests determined by the queue length). A numerical computation approach is then proposed to compute the steady state distribution of this queuing model. By utilizing the block structure of transition probability matrix, we further develop an approach to speed up the computation, which can reduce the model complexity from O(K-6) to O(K-3). Finally, we apply this approach to a case study of hard disks of Western Digital Corp. It demonstrates the efficiency of our approach and gains useful insights for the optimization of storage system.
暂无评论