Wireless Sensor Networks And Applied Research

Wireless sensor networks (WSN) is a wireless self-organized network, which iscomposed by numerous sensors deployed densely and randomly in a geographicalscenario, and is used to sense, collect and process sensed information collaboratively.Research on WSN is very significant in promoting the application of informationtechnology in national defense, meeting the requirements of some special applicationfield, boosting the transfer of the core technology achievements and finallypopularizing WSN to social life as a new point for economic growth.Five topics referred to in this dissertation are basic theories, QoS model, topologyalgorithms, routing algorithms and protosystem. The main work and innovations isas follows.1. Methodology and QoS Mechanism based on Task of WSNIn this section, based on the analyses of characters, this dissertation discusses theconformation features and research connotations of WSN, presents a task-basedclassification method and the modalities of different network are analyzed. After that,this dissertation makes the extraction and definitions of the basic networkarchitectures, gives a important principle named Node Coherence, points out thedifference between traditional networks and WSN, proves the connectivity ofadjacent clusters of WSN. Besides, aiming at the problems of routing and softwaredevelopment, this dissertation gives the basic design principles, and raises a soft busbased modular architecture of node software. At the end of this section, thisdissertation analyzes the relations between QoS and three essential workingproblems of WSN, presents the concept of Active QoS, the design scheme,infrastructure, procedure of working, state evaluation are discussed, its feasibility isvalidated.2. Active QoS Model of WSNIn this section, this dissertation presents an Active QoS Model based on service andmiddleware of WSN, named QISM (QoS Infrastructure based on Service and Middleware), and discuss the design scheme, structure, important modules andworking processes. QISM is a software layer between protocol stack and applicationlayer of WSN, and can be classified into application dynamic control process andtask active accommodation process according to different executive entities, servicepublishing and subscription mechanism is the foundation of it. The importantcharacteristics of QISM include the active regulation mechanism based on feedbackand negotiation between application and network, the optimizing scheme based onapplications and statues and roles, the architecture based on middleware and servicepublishing and subscription, the expression method based on Topic and FunctionRegion. Through the active regulation to applications and network, QISM make thenetwork support applications and make the applications adapt the network furthest.Proved by the simulation based on the active regulation model, the quality ofapplications accomplishment is guaranteed better, and the lifetime of networks isprolonged at the same time.3. Dynamic Cluster Organization Algorithm of WSNIn this section, this dissertation presents a simple distributed algorithm for clusterorganization of WSN, named DSCO (Distributed Simple Cluster OrganizationAlgorithm), discusses the design scheme, algorithm realization, and defines someevaluation indexes. Considering the process ability and limited energy of nodes,unreliable channels due to complicated work environment, DSCO is a localalgorithm using local knowledge and without routing tables, so the nodes areindependent each other. Simple logic, connectionless information exchange andasynchronism are important features of the algorithm. Through the network globalreorganization and cluster heads alternation of DSCO, the network is recoverable,while the energy consumption of nodes is averaged and the lifetime of network isprolonged. Through the analysis of the algorithm simulation, DSCO has fastclustering speed, proportional distribution of cluster nodes, shows good performanceon clustering rate and topology control, and provides high adaptability in large-scaleWSN.4. ACS Algorithm based on Altitude Information of WSNIn this section, this dissertation presents an improved ACS algorithm based onAltitude Information (AI) and Ants Withdrawal (AW), named ACSA (an improved ACS algorithm with AI), brings forward the concept of AI, and discusses the designscheme, algorithm realization, and performance evaluations of ACSA. ACS (AntColony System) is an important evolvement and component of ACO (Ant ColonyOptimization), has specific predominance on solving combinatorial optimizationproblem. But ACS can't be used directly in WSN, due to the long calculation timeand slow convergence speed of ACS, and the resource constraint of WSN. AI isglobal heuristic information which correlative with source and destination node,represents the possibility of ants traveling from current node to destination, andreflects the influence of the distance between current and destination node, theincluded angle of the lines drawn from source node and current node respectively todestination node, on path optimization. By using AI, the algorithm can get fasterconvergence speed; by using AW, the algorithm can increase the probability of antstraveling to destination node, and avoid local optimization. The algorithm simulationillustrates that ACSA can get more stable and robust result, faster calculation andconvergence speed than ACS. The study on ACSA brings forward a solution to useACO in WSN.5. Protosystem based on Cluster of WSNIn this section, this dissertation presents a design scheme of WSN protosystem,brings forward the node hardware and Bluetooth gateway infrastructures, anddiscusses the essential design ideas of CMR (Cluster-based Multihop Routing)component, the infrastructure and the message structure of CMR in detail. At the endof this section, an application demonstration of this protosystem is given for thepurpose of validation. The protosystem consists of 30 micro sensor nodes, has fastcalculation capability and wide bandwidth, supports I~2C, UART, and many kinds ofsensors, such as light, temperature, magnetic field and acceleration sensors, etc. Theprotosystem is compatible with Crossbow Mica2, the Bluetooth-based Sink gatewayand cluster-based multihop routing protocol is important innovation. Theprotosystem satisfies the needs of test bed in the research of WSN, many importanttheories and technologies about system research and development of WSN istestified through the design and development process of protosystem. We do not findany similar system in domestic researches as best as we can.