In what follows, we consider that the sensing consumption is negligible, We tackle the
data acquisition approaches related to communication rate reduction. More specifically,
we a r e concerned with the data collection and aggregation techniques that reduce, as
possible, the number of communication as well as t he communicated overhead. In the
next subsection, we observe the impact of the communication on the energy consumption
of sensor nodes.
1.3.3 Communication
Typically, the operational states in a sensor node are: Transmit, Receive, Idle and
Sleep [174, 211]. These states for m a cycle called the duty-cycle. It is a periodic cy-
cle between two states: wake-up state (transmit, receive, idle) and sleep state. Ideally, a
node turns its radio ON only when it is going to transmit or receive data. At all other
times, it should keep its radio off (i. e sleep state). When a radio is on but idle (wa it-
ing f or p ossible traffic), it consumes the same amount of energy as when it is receiving
data [1 83, 214]. The communication cost of a node is composed of the cost of transmis-
sion, reception and listening [155]. The cost of receiving or transmitting a packet is an
order of magnitude greater than the cost of computation. Between the two (tr ansmission
and reception), transmitting a bit consumes twice as much energy as receiving a bit [116].
Tilak et al. show in [230] that the execution of 3000 instructions consumes the same
amount of energy as sending 1 bit over 100 meters by ra dio. Therefore, the main concern
for most WSNs algorithms [28, 146, 168, 256] is to keep the communication cost to a
minimum. In this thesis, we focus on the transmission cost. We aim at reducing the
number of transmissions between nodes (i.e the transmission rate) as well as the number
of transmitted bits (i.e the transmission overhead).
In WSNs, data are transferred (i.e routed) from source sensors to the sink through
multi-hop communication par adigm [7]. For an efficient routing, different schemes have
been adopted, such as chain [137], tree [14] or clusters [264]. We aim at reducing the
transmission cost on one-hop communication between nodes towards the base station.
Description about these schemes is presented in Chapter 2. In this work, we suppose
having a suitable routing protocols that do not interfere with the algorithms proposed in
the following chapters.
1.3.4 Data processing
Nodes in WSNs gather, process and rely information to one or more other nodes. They
collaborate to produce better outcomes during event monitoring. Ideally, data is expected
to be accurate, arrive on time, processed and transmitted with lower energy consumption.
The amount of data exchanged between nodes is expected to be as minimum as possible.
However, data should be significant to respond to applications needs. Fo r these purposes,
collected data need to be processed and a nalyzed in a way to remove any erroneous,
suspicious, redundant or other readings that may cause unnecessary computation and
communication cost. Challenges do not stop here. They continue to grow along with the
trends of producing mobile, smart or a uto no mous objects [235, 272]. Such objects require
intelligent data processing algorithms to avoid bad decisions that affect their energy con-
sumption. For example, an unreliable mobile node produces unpredictable motion and
7
Thèse d'Alia Ghaddar, Lille 1, 2011
© 2011 Tous droits réservés.
http://doc.univ-lille1.fr
剩余113页未读，继续阅读
cpongm
- 粉丝: 5
- 资源: 2万+
我的内容管理
展开
最新资源
- 最优条件下三次B样条小波边缘检测算子研究
- 深入解析：wav文件格式结构
- JIRA系统配置指南：代理与SSL设置
- 入门必备：电阻电容识别全解析
- U盘制作启动盘：详细教程解决无光驱装系统难题
- Eclipse快捷键大全：提升开发效率的必备秘籍
- C++ Primer Plus中文版：深入学习C++编程必备
- Eclipse常用快捷键汇总与操作指南
- JavaScript作用域解析与面向对象基础
- 软通动力Java笔试题解析
- 自定义标签配置与使用指南
- Android Intent深度解析：组件通信与广播机制
- 增强MyEclipse代码提示功能设置教程
- x86下VMware环境中Openwrt编译与LuCI集成指南
- S3C2440A嵌入式终端电源管理系统设计探讨
- Intel DTCP-IP技术在数字家庭中的内容保护
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
