国产bbaaaaa片,成年美女黄网站色视频免费,成年黄大片,а天堂中文最新一区二区三区,成人精品视频一区二区三区尤物

首頁> 外文學位 >A simulation-based approach for efficient sensor management of distributed sensor networks.
【24h】

A simulation-based approach for efficient sensor management of distributed sensor networks.

機譯：一種基于模擬的方法，可對分布式傳感器網(wǎng)絡進行有效的傳感器管理。

獲取原文
獲取原文并翻譯 | 示例
  • 摘要
  • 著錄項
  • 相似文獻
  • 相關主題

摘要

Distributed Sensor Networks (DSNs) are sensor fields consisting of intelligent, disparate sensors that are distributed spatially and geographically. DSNs typically incorporate many remote, unattended sensors, creating new capabilities but also introducing new constraints on power, computation, and communications resources. These constraints have created a renewed interest in DSN management strategy, with an increased emphasis on sensor management and efficiency of DSN operations.; Sensor management deals with the inherent tradeoff between DSN performance and resource consumption. In our context, we view it as a sequential activation of sensing assets in such a way as to reach an acceptably certain representation of the search region while minimizing operating costs. In this dissertation, we hypothesize that the use of dynamic control policies based on the underlying conditional probability distribution associated with the current search region representation will lead to more efficient DSN operations, thereby extending the operating lifetime of the DSN.; In order to evaluate the performance of several heuristic dynamic control policies, we developed a simulation model of a military surveillance DSN. We use the simulation model to compare the performance of the dynamic control policies to a Base Policy and a Benchmark Policy. The Base Policy consists of activating all sensors during each sensing iteration. The Benchmark Policy consists of an approximate dynamic programming (ADP) solution to the sensor management problem that uses a simulation-based policy iteration process to minimize DSN operating costs while reaching acceptable levels of certainty about the validity of our search region representation.; This research demonstrates that simple dynamic control policies can be more efficient than the Base Policy and some can even perform comparably with the Benchmark Policy. In addition, the DSN simulation model we developed provides a re search base to (1) investigate the fusion of observations from different sensor types, (2) demonstrate the use of non-imaging sensors to provide adequate situational awareness where “precision” emplacement of more capable sensors is not possible, (3) develop operational concepts to integrate DSN operations with user needs, and (4) experiment to find additional, more-efficient dynamic control policies.
機譯：分布式傳感器網(wǎng)絡（DSN）是由在空間和地理上分布的智能，完全不同的傳感器組成的傳感器領域。 DSN通常包含許多遠程，無人值守的傳感器，不僅創(chuàng)建了新功能，而且還引入了對功率，計算和通信資源的新限制。這些限制使人們對DSN管理策略重新產(chǎn)生了興趣，并越來越重視傳感器管理和DSN操作效率。傳感器管理處理DSN性能和資源消耗之間的固有折衷。在我們的上下文中，我們將其視為感應資產(chǎn)的順序激活，以便在最小化運營成本的同時達到搜索區(qū)域的可接受的確定表示。在本文中，我們假設基于與當前搜索區(qū)域表示相關的潛在條件概率分布的動態(tài)控制策略的使用將導致更有效的DSN操作，從而延長DSN的使用壽命。為了評估幾種啟發(fā)式動態(tài)控制策略的性能，我們開發(fā)了軍事監(jiān)視DSN的仿真模型。我們使用仿真模型將動態(tài)控制策略的性能與基本策略和基準策略進行比較?；静呗园ㄔ诿看胃袘陂g激活所有傳感器。基準策略包括一個針對傳感器管理問題的近似動態(tài)編程（ADP）解決方案，該解決方案使用基于模擬的策略迭代過程來最小化DSN運營成本，同時使我們的搜索區(qū)域表示形式的有效性達到可接受的確定性水平。這項研究表明，簡單的動態(tài)控制策略可能比基本策略更有效，有些甚至可以與基準策略相比。此外，我們開發(fā)的DSN仿真模型提供了一個搜索基礎，以（1）研究來自不同傳感器類型的觀測結(jié)果的融合，（2）演示使用非成像傳感器來提供適當?shù)膽B(tài)勢感知，以實現(xiàn)``精確''定位更強大的傳感器是不可能的;（3）開發(fā)操作概念以將DSN操作與用戶需求集成在一起;（4）實驗以查找其他更有效的動態(tài)控制策略。

著錄項

  • 作者

    Bland, William Steven.;

  • 作者單位

    University of Virginia.;

  • 授予單位 University of Virginia.;
  • 學科 Engineering System Science.
  • 學位 Ph.D.
  • 年度 2003
  • 頁碼 165 p.
  • 總頁數(shù) 165
  • 原文格式 PDF
  • 正文語種 eng
  • 中圖分類 系統(tǒng)科學;
  • 關鍵詞

相似文獻

  • 外文文獻
  • 中文文獻
  • 專利
獲取原文

客服郵箱：kefu@zhangqiaokeyan.com

京公網(wǎng)安備：11010802029741號 ICP備案號：京ICP備15016152號-6 六維聯(lián)合信息科技 (北京) 有限公司?版權所有

  • 客服微信

  • 服務號