Earliest Deadline First Policy Research Articles

Fluid limits for earliest-deadline-first networks

This paper analyzes fluid scale asymptotics of two models of generalized Jackson networks employing the earliest deadline first (EDF) policy. One applies the ‘soft’ EDF policy, where deadlines are used to determine priority but jobs do not renege, and the other implements ‘hard’ EDF, where jobs renege when deadlines expire, and deadlines are postponed with each migration to a new station. The arrival rates, deadline distribution and service capacity are allowed to fluctuate over time at the fluid scale. Earlier work on EDF network fluid limits, used as a tool to obtain stability of these networks, addressed only the soft version of the policy, and moreover did not contain a full fluid limit result. In this paper, tools that extend the notion of the measure-valued Skorokhod map are developed and used to establish for the first time fluid limits for both the soft and hard EDF network models.

First steps towards an IEEE 802.1AS clock for EDF scheduling in distributed real-time systems

The EDF (Earliest Deadline First) scheduling policy can take advantage of the existence of a synchronization clock mechanism to reach higher utilizations in distributed real-time systems, as indicated by theoretical research. In this work, we give the first steps to enable the usage of EDF and a global clock in a distributed system. To this end, our proposal relies on Linux to provide the global clock and on a Real-Time Operating System (RTOS) to provide the global EDF policy. The objective is the evaluation of global EDF to verify to what extent the theoretical results are met.

Edge minimality of EDF resource sharing networks

We consider a general real-time network with soft customer deadlines, in which users require service from several shared resources simultaneously. We call the service protocol for such a network edge minimal (locally edge minimal) if it minimizes globally (resp., locally) in time, in a suitable sense, the system resource idleness with respect to customers with lead times not greater than any given threshold value on all the routes of the network. We show that the preemptive Earliest-Deadline-First (EDF) service discipline is edge minimal. Moreover, we characterize the preemptive EDF policy as a protocol making the underlying network locally edge minimal. Our arguments are pathwise, independent on the network topology and requiring very mild assumptions, or even no assumptions, on the model stochastic primitives. Application of our characterization to fluid model analysis for EDF resource sharing networks is also discussed.

IASA: an energy-efficient scheduling algorithm for real-time tasks with lock-free objects

In this paper, we are interested in the energy-efficient scheduling of real-time tasks which may access lock-free objects in a non-ideal DVS processor. Compared to conventional lock-based real-time synchronisation protocols, lock-free objects have the advantage of the avoidance of priority inversion and deadlock. However, because lock-free objects permit concurrent operations to interfere with each other, the access to a lock-free object might take an arbitrarily long time to complete owing to repeated interferences. In this paper, we propose an algorithm, called interference-aware speed assignment (IASA), for scheduling of real-time tasks with lock-free objects. The IASA assigns proper execution speeds to tasks scheduled by the earliest deadline first (EDF) scheduling policy. The execution speeds are calculated based on the schedulability analysis of the EDF policy with the cost of the interferences of lock-free objects. As a result, the energy consumption can be reduced while meeting the timing constraints of real-time tasks.

Performance evaluation of the earliest deadline first policy over ad hoc networks

To satisfy the Quality of Service (QoS) requirements of delay-sensitive flows, we propose in this paper to support the Earliest Deadline First (EDF) policy over wireless ad hoc networks. Hence, we enhance the IEEE 802.11 protocol to support the real-time EDF scheduling policy. Therefore, we develop a Markov-chain-based analysis modelling the backoff process of the EDF policy. Then, we propose to evaluate the performance of EDF in terms of saturation throughput and service time delay bounds. Service time delay bounds are obtained by inverting service time Z-transforms. Analytical results are validated and extended by simulation using the NS-2 network simulator. Finally, we evaluate the performance of the EDF policy in a multi-hop environment in the presence of ad hoc routing protocols.

Heavy traffic analysis for EDF queues with reneging

This paper presents a heavy-traffic analysis of the behavior of a single-server queue under an Earliest-Deadline-First (EDF) scheduling policy in which customers have deadlines and are served only until their deadlines elapse. The performance of the system is measured by the fraction of reneged work (the residual work lost due to elapsed deadlines) which is shown to be minimized by the EDF policy. The evolution of the lead time distribution of customers in queue is described by a measure-valued process. The heavy traffic limit of this (properly scaled) process is shown to be a deterministic function of the limit of the scaled workload process which, in turn, is identified to be a doubly reflected Brownian motion. This paper complements previous work by Doytchinov, Lehoczky and Shreve on the EDF discipline in which customers are served to completion even after their deadlines elapse. The fraction of reneged work in a heavily loaded system and the fraction of late work in the corresponding system without reneging are compared using explicit formulas based on the heavy traffic approximations. The formulas are validated by simulation results.