Research On Transmission Strategy Of Self - Organizing Wireless Sensor Network Based On Mobile Data Collection

In recent years, Wireless Sensor Networks (WSN) has been widely used inmany fields, among most of which batteries are used to supply power for sensors,thus the energy of the sensor nodes have become the bottle neck of WSN. Energyefficiency is an important index of the property of the sensor networks. One typicalproblem is the low energy node, which results from the fact that sensor nodes lose toSink use up energy quickly because it takes more communication load and whichleads to the early ending of functioning cycle of the network. And this problemremains in spite of the fact that multi-level radius delivering and uneven distributionof sensor nodes methods have been used to balance the consumption of networkenergy.In recent years, some researchers employed mobile nodes method to solve theabove problems in WSN. As MDC (mobile data collector), mobile nodes collect datawithin communication ranges and process the data it collects or tranmit them to Sink.This kind of network is called WSN-MDC. With this network model, topologystructure of the network, positioning of the nodes, routing selection of the nodes androuting plannning and time extension of MDC and the engergy balanced datatransmission strategy will be studied.This article provides specific analysis on the characteristics of different stagesfrom the WSN-MDC network activation phase, node locations discovering phase,routing discovery phase to the stable transmission phase of information. And studiesare carried on the performance of the network topology, the positioning of the nodes,the establishment and updates of routing paths, and the planning of MDC routingand algorithm of the data transmission. Through the integration of the algorithm ofvarious stages, the QOS of the network and network life cycle have been improved.The following innovations are carried out in the course of the research:1. A WSN-MDC network model with dual band is designed. TakingWSN-MDC as the research object, a dual band network model with a mobile datacollector is designed(DBWSN-MDC); Based on this model, the system and structure,the structure of the network level, coverage, connectivity and performanceparameters of network is studied; the function of the sensor node is defined; Thefunction index of the selected software protocols and hardware of node is clarified;In this model, the anchor point was not set and the sensor nodes were randomly deployed in appointed region. Besides, node had no location capability and the initialstate was the same. When the node channel conflict occurred, MAC protocol is usedto eliminate the conflict.2. A novel DV-Hop algorithm based on improved artificial bee colony algorithm(DCABCDV-Hop) is proposed. Firstly, the bionic intelligent computing theory isstudied and the artificial bee colony algorithm(ABC) is analyzed; and based on theoriginal ABC, the opposite learning method is introduced to construct the initialsolution set, the artificial bee colony algorithm based on divide and conquer strategy(DCABC) is put forward. The experimental results showed that the algorithm hadgood effect on convergence acceleration, and the ability of global search andefficiency is improved; then, the DV-Hop positioning algorithm is studied and thefitness function of the algorithm was gotten by deducing, which is applied in theDCABC, then, the optimal location information of node was obtained. Simulationexperiment results showed that DCABCDV-Hop algorithm can effectively improvethe positioning accuracy, and the ideal effect is achieved.3. In view of the network model (DBWSN-MDC) which was proposed in thesecond chapter, a RSSI localization algorithm based on mobile data collection(RSSI-MDC) is designed. The network is divided into a plurality of square grids, andthe grid nodes of adjacent region used two kinds of band to communicate. Duringthe execution of the RSSI-MDC, stops of MDC is preseted, and the set of stay pointis constructed, and then the mobile path of MDC based on stay point set is designed;MDC moved from the initial position along the path established, and when it reachedthe stay point, MDC stopped and communicated with the nodes within the RSSIrange, until the traversal of all the stay point is completed. After the MDC got thedata, then the coordinates of nodes is calculated according to the position coordinatesof stay point and trilateration. Through simulation experiments, it is proved thatRSSI-MDC has better effectiveness of positioning, and is suitable for the practicalneed of Greenhouse Wireless Sensor network.4. In view of DBWSN-MDC, a mobile data collection strategies based ondouble Layer (DLDC) is proposed. First of all, the MDC set the stay point accordingto the nodes and grid layout of the network model; on this basis, the collection pathof MDC is designed; then, inside each grid, the temporary sink within each gridinternal called Rendezvous Point (RP) was selected according to the position of staypoint; the multi-object optimization routing algorithm (MOR) is designed. Amulti-objective optimization function is designed by taking the factors of residual energy, RP node distance and link quality into account, which is taken as a measureof the selection of node forwarding data; finally, the data collection work of MDC iscompleted through the movement of the pre-designed path for each RP. The strategyachieved the two consideration between delay and energy consumption, andprovided theoretical guidance for application of delay tolerate network (e.g.Greenhouse Wireless Sensor Network Application).This article designed the dual band network model with a mobile data collector,and focused on studying three algorithms from the two aspects of node location anddata collection. The common goal of the three algorithms was to reduce networkenergy consumption, to prolong the network life cycle. The first node localizationalgorithm was a kind of range-free localization algorithms, and had stronguniversality. Through the positioning process, node was activated, and knew theposition coordinates of their own and neighbor node set; the second kinds oflocalization algorithm is an improved RSSI algorithm based on DBWSN-MDC. Theinput is non-activated node while the output is these nodes which are activated andthe coordinate set of nodes within the communication range are known; the thirdalgorithm is implemented based on the second location algorithm; the input wasnodes within the grid which were known and activated, and the output was theultimate collection of information which was collected by MDC.To sum up, based on the WSN-MDC network, the paper designed the specificnetwork model. All stages of data transmission were deeply analyzed, and from thenetwork, location, routing to collect data, the difficulties of reducing energyconsumption, prolonging the network life cycle were analyzed respectivelyaccording to the characteristics of different stages; besides, different efficientalgorithm is designed and the overall performance of WSN-MDC is enhancedcomprehensively, and through the means of theoretical proof and simulationexperiment, the work in this article is proved to be effective and advanced.