...
首页> 外文期刊>Computer networks >Improving the accuracy of timing in scalable WSN simulations with communication software execution models
【24h】

Improving the accuracy of timing in scalable WSN simulations with communication software execution models

机译:通过通信软件执行模型提高可扩展WSN模拟中的定时精度

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       

摘要

Emerging infrastructure-less network architectures such as WSNs consist of devices that perform packet processing in software. General-purpose network simulators do currently not possess models to simulate the intra-node delay of such devices. For example, a TelosB mote with TinyOS spends seven ms on processing packets with a size of 36 bytes and fifteen ms on packets of 124 bytes. The core problem addressed in this work is that simulation does not include such delays, and therefore, the results are inaccurate. To overcome this problem, we create a communication software execution model of TelosB that accounts for its temporal behavior to enable more accurate WSN simulations in the ns-3 simulator. A challenge is to create a tracing framework for TinyOS that can be used to accurately and reliably trace the behavior of a very resource-constrained system. By analyzing the software execution of TelosB running TinyOS in the emulator Cooja/MSPSim and on a real device, we discover discrepancies in the temporal behavior. The evaluation of our model shows that it is scalable and accurate; the simulated intra-OS delay deviates at most 5% from the intra-OS delay in the real mote. When we include the model in simulations, the forwarding capacity of a mote is decreased by 36%. The WSN community can use this model for more realistic simulations, and future WSN mote models will be easier to make with it as a foundation.
机译:较少的新兴的基础架构网络架构(如WSN)包括在软件中执行数据包处理的设备。通用网络模拟器目前不具有模拟此类设备的节点内延迟的模型。例如,TelosB微尘与TinyOS的花费上的大小为36个字节,在124个字节的数据包15毫秒处理分组7毫秒。在这项工作中解决的核心问题是模拟不包括这样的延迟,因此,结果不准确。为了克服这个问题,我们创建了一个Telosb的通信软件执行模型,其考虑其时间行为,以在NS-3模拟器中启用更准确的WSN模拟。挑战是为TineOS创建一个追踪框架,可以用于准确可靠地跟踪一个非常资源受限系统的行为。通过分析仿真器Cooja / MSPSIM中的TineOS的TinoS的软件执行,我们在真实设备上发现了在时间行为中的差异。我们模型的评估表明它是可扩展和准确的;模拟的OTRA-OUT延迟偏离了真实的操作系统内的ORA延迟的5%。当我们在仿真中包含模型时,MOTE的转发能力下降36%。 WSN社区可以使用此模型进行更现实的模拟,未来的WSN MOTE模型将更容易地作为基础。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号