Wireless Sensor Networks

Wireless sensor networks (WSNs) exhibit promising development prospects in both economic and technical levels because of their ability to provide a variety of services such as search-and-rescue operations, logistics, vehicle routing and intruder detection. The services provided by WSNs are based on collaboration among small energy-constrained sensor nodes. Due to the large-scale application of WSNs, the demand for organization strategies of sensor nodes are increasingly urgent. Research on WSNs have focused on several research topics and emerged as an important new area in wireless technology. This chapter covers the network characteristics, origin and development, architecture and applications of WSNs.

The routing protocol builds a mechanism for information transmission from the source node to the destination node. It guides transmitting path information of data packets. In this way, a fast and efficient routing protocol has a significant impact on the performance of WSNs. This chapter mainly discusses the fundamental concepts, fundamental structure and several typical routing protocols. On the basis, we propose several improved protocols, On Demand Clustering Improved LEACH Protocol (ODCL), Connectivity-guaranteed and Energy-efficient Clustering Scheme (CECS), Density-Aware Clustering Protocol based on LEACH-C (DACPL) and Hierarchical AODV Protocol for WSN (HAODV). The protocols all belong to clustering algorithms that effectively improve the energy efficiency of traditional routing algorithm.

Network localization is a promising technology for providing high-accuracy positional information in GPS-challenged scenarios. The information is crucial in many location-based applications such as autonomous logistics, building security, as well as search-and-rescue. In this chapter, we discuss some effective localization methods of WSNs. In static sensor networks (SSNs), we present hop-count-based expectation of distance algorithm (HCED). In mobile sensor networks (MSNs), we present range-free Voronoi-based Monte Carlo localization algorithm (VMCL) and an optimal region selection strategy of Voronoi diagram based on VMCL (ORSS-VMCL). In anisotropic SSNs with holes, we present Heuristic Multi-dimensional Scaling (HMDS) and an Extended Kalman Filter Multidimensional Scaling (EKF-MDS) localization method. To solve the problem of Voronoi diagrams based localization scheme (VBLS), an optimal region selection strategy based on VBLS (ORSS-VBLS) is also proposed. Finally, We present a Delaunay triangulation based localization scheme (DBLS) and neighbor constraint assisted distributed localization (NCA-DL), which are effective in refining the distances required for localization.

The coverage control problem is one of the primary issues for Quality of Service (QoS) in WSNs, which effectively improve network connectivity, coverage and self-localization accuracy. In this chapter, we present several effective coverage control algorithms of WSNs. As homogeneity affects coverage rate and algorithm efficiency, we categorize coverage control algorithms into homogeneous algorithms and heterogeneous algorithms. For homogeneous sensor network, we firstly present an autonomous triangular deployment for optimal coverage (ATDOC) algorithm to avoid the nodes stacking problem in traditional virtual force approach. We also present an extended virtual force-based approach (EVFA) that overcome the connectivity maintenance problem in the traditional virtual force algorithm. As the orientation control can eliminate coverage holes and reach reliable connectivity, the Distance and Orientation Control (DOC) algorithm is proposed. For the heterogeneous sensor network, Self-Deployment by Delaunay Triangulation Graph (SDDTG) algorithm is presented and achieves non-redundant coverage. Finally, the Energy Balanced Redeployment Algorithm (EBRA) is proposed to solve the energy-balanced problems in heterogeneous mobile sensor networks.

As a ubiquitous feature of complex networks, community characteristic is common in real world networks especially in wireless sensor network. It is vital to detect community structure in complex networks and to take advantage of extracted information. On the basis, community detection has become a popular theme in the field of complex networks. It focuses on uncovering the affiliation of each node through topology analysis. In this chapter, we have presented the essential knowledge and advanced methods of community detection, including proximate support vector clustering (PSVC), deep auto-encoded extreme learning machine (DA-ELM), deep auto-coded clustering (DAC) and local aggregated differential evolution algorithm (LADE). These methods have been applied in tree-based WSNs and have been proved in terms of robustness and fragility.

The small world phenomena is initially investigated in the sociology that individuals are often linked by a short chain of acquaintances. S. Milgram and his group conduct a series of mail delivery experiments and find that an average of six degrees of separation exists between senders and receivers. Recent researchers have shown that the small worlds are also observed in wireless sensor networks. In this chapter, we discuss the epidemic and immunization process in small world of wireless sensor networks and mainly centre on the tree topologies. We introduce some basic characteristics of small world networks and analyze epidemiogical process in small-world networks. Meanwhile, we present immunization strategies during immunization process.