Clustering-Based Low Power Adaptive Wireless Sensor Network Protocol Architecture

無線網路在近幾年有著非常快速的發展，大多的無線裝置或手持式系統是由電池來供應電源。因此，電源消耗的多寡大大影響了整個系統的生命週期。在這篇論文中，我們提出一個低功率無線感測網路架構，利用每一個單一感測網路節點有數個不同發送功\\\率的特性，搭配階層以及叢集的概念，使得網路拓撲不再一定像以往一樣需要是均勻分布，而卻能夠因應該區域節點實際分布情況，分成數個涵蓋大小不一的叢集，且每個叢集之間彼此相互連通。此外，每個叢集會依據節點的移動速度，剩餘電量，以及叢集間傳送訊息所需要消耗的功\\\率，選出適當的叢集閘道，並決定以其為首的叢集涵蓋範圍之所需要的最小發射功\\\率，來負責建構出整個以叢集為基礎之網路骨架。而非叢集閘道之節點可以依據本身與叢集閘道之間的距離，選擇其傳送資料的最小傳送功率，並在累計一定閒置時間之後而進入休眠模式，來節省不必要的電量的消耗。To extend the lifetime of the wireless sensor network, the technique of clustering is an effective way to achieve the power conservation, network scalability and load balance. In this paper, we propose the architecture of the wireless sensor network with multiple levels of transmission power and take advantage the concept of hierarchy and clustering to have different cluster ranges among the chosen clusterheads. A network with clusters of different ranges of the covered area can have better topology control for reducing the unnecessary interference and save large amount of energy. Moreover, it can also improve the network load balance for the condition of non-homogeneous dispersion of sensor nodes. Meanwhile, each cluster will choose the clusterhead in accordance with the moving speed, residual energy, intra-cluster communication cost and decide the minimum power of the cluster range for the clusterhead to cover the entire cluster area of which it takes charge. On the other hand, the non-clusterheads can choose its own minimum transmission power level according to the distance between itself and the clusterhead. In addition, the non-clusterheads will enter sleeping state after a period of idle time to save the unnecessarily wasted power.In addition, we also investigate the tradeoff between energy consumption and throughput for the cluster-based hierarchical model with one-hop model and multi-hop model inside a cluster. Also we provide an adaptive hand-over mechanism for the re-election of the clusterhead which can prolong the network lifetime.Chapter 1 1 Introduction 1 1.1 Wireless Sensor Network 2 1.2 The Origin of Sensor Network 2 1.3 Characteristics of Sensor Network 4 1.4 Motivation 6 1.5 Thesis Organization 8 Chapter 2 9 Background and Related Work 9 2.1 Low Power Designs at Different Layers 9 2.1.1 Power Saving at the Data-Link Layer 9 2.1.2 Power Saving at the Network Layer 12 2.1.3 Power Saving at the Transport Layer 12 2.2 Location Based Energy Aware Protocols 13 2.2.1 Geographic Adaptive Fidelity (GAF) 13 2.2.2 Geographic and Energy Aware Routing (GEAR) 15 2.3 Clustering Based Energy Aware Protocols 18 2.3.1 The Concept of Clustering Networks 18 2.3.2 Topology Control 19 2.3.3 Cluster Formation 21 2.3.4 Low-Energy Adaptive Clustering Hierarchy (LEACH) 22 2.3.5 Threshold-sensitive Energy Efficient Sensor Network Protocol 23 2.3.6 Hybrid Energy-Efficient Distributed Clustering (HEED) 25 Chapter 3 27 Proposed Low Power Adaptive Wireless Sensor Network Architecture 27 3.1 Dynamic Cluster Range Selection Phase 28 3.2 Clusterhead Cost Computation Phase 31 3.3 Cluster Formation Phase 32 3.4 Node State Transition and Re-election Mechanism 34 3.5 Routing Model 39 3.6 Mobility 40 Chapter 4 42 Experimental Results 42 4.1 Radio and Energy Models 42 4.1.1 Radio Propagation Model 42 4.1.2 Energy Consumption Model 44 4.2 Simulation Environment and Parameters 47 4.2.1 Simulator 47 4.2.2 Simulation Parameters 48 4.3 Cluster Formation Results Analysis 49 4.4 Power Consumption and Network Lifetime 53 4.5 Network Performance 57 Chapter 5 60 Conclusion 60 Bibliography 6