SDREAM: A Super-Small Distributed REAL-Time Microkernel Dedicated to Wireless Sensors
SDREAM: A Super-Small Distributed REAL-Time Microkernel Dedicated to Wireless Sensors
复制标题
DOI：
10.1108/17427370780000169
复制
发表时间：
2007-01-01
影响因子：
2.6
通讯作者：
De Vaulx, Christophe
中科院分区：
文献类型：
--
作者：
Zhou, Haiying;Hou, Kun;De Vaulx, Christophe
传统的嵌入式操作系统是消耗资源的多任务系统，因此不适合智能无线传感器。本文提出了一种专用于无线传感器的超小型分布式实时微内核（SDREAM）。 SDREAM是一个基于元组的消息驱动的实时内核。它采用元语言：内核建模语言来抽象地定义和描述系统原语。 IPC 和进程同步基于 LINDA 概念：由两个轻原语 (SND: OUT & RCV: IN) 实现的元组模型。在SDREAM中，任务分为两类：周期性任务和优先级任务。周期性任务优先级最高，负责捕捉传感器信号或驱动控制信号；优先级任务有多种优先级，适合时间约束的应用。任务调度采用两级任务调度策略方案，即基于优先级的抢占式调度。 SDREAM 简单高效。它具有灵活的硬件抽象能力，使其能够快速移植到不同的WSN平台和其他微型嵌入式设备中。目前已在多个硬件平台上进行移植和评估。性能结果表明SDREAM需要很少的资源，并且适合硬实时多任务WSN应用并且高效。
Traditional embedded operation systems are resource consuming multitask, thus they are not adapted for smart wireless sensors. This paper presents a super-small distributed real-time microkernel (SDREAM) dedicated to wireless sensors. SDREAM is a tuple-based message-driven real-time kernel. It adopts a meta language: Kernel Modeling Language to define and describe the system primitives in abstract manner. The IPC and processes synchronization are based on the LINDA concept: the tuple model implemented by two light primitives (SND: OUT & RCV: IN). In SDREAM, tasks are classified into two categories: periodic and priority. The periodic task has the highest priority level and is responsible for capturing sensor signals or actuating control signals; the priority task has various priority levels and is suitable for time-constraints applications. A two-level task scheduling policy scheme, named priority-based pre-emptive scheduling, is used for task scheduling. SDREAM is simple and efficient. It has a flexible hardware abstraction capability that enables it to be rapidly ported into different WSN platforms and other tiny embedded devices. Currently, it has been ported and evaluated in several hardware platforms. The performance results show SDREAM requires tiny resource and is suitable and efficient for hard real-time multitask WSN applications.