| With the technology development and constant expansion of application fields inthe past more than ten years, wireless sensors networks (WSN) began to graduallyplay a positive role in the detection, identification, location and tracking, and otherareas of the application, and has become one of the four pillar industries in the field ofhigh and new technology in the21st century. In spite of this, many of the keytechnology of wireless sensor network still need to be addressed or perfect, such asenergy optimization, non-line-of-sight (NLOS) location, coverage scheme, and so on,is one of the hot research topics in the domestic and foreign scholars, and causedwidespread concern.Perception of the application in specific environment, WSN is todetermine or random deployed. Whether network can work normally, it will beaffected by many factors such as the project cost, non line-of-sight error and finiteenergy, and so on. According to the above constraints, the main work and innovationsof this thesis are as follows.(a) In view of the cost constraint, this thesis put forward a kind of coveragescheme of the strip area rhombus deploy(SARC) based on the disk sensing model ofwireless sensor nodes, is used to the effective of node distribution, area coverage andnetwork optimization in monitor area.In the application of WSN, the project needs to consider the cost of the wholesystem. Normally, these costs need to know roughly before deployment. Thiscoverage scheme gives the basic method of establishing mathematical coverage model,and the analytical expressions of nodes number required, can effectively carry outcalculation of the number of nodes and deployment in the case of full coverage andk-connectivity.Firstly, we completed optimization of coverage scheme under the premise of full coverage and k-connectivity in the areas of10001000square meters. Then, wecompared the number of nodes of the different optimal models. The simulation resultsshow that WSN need to make big sacrifices in the number of sensor nodes in order tomaintain network k-connectivity, under the premise of full coverage. At the same time,we has carried on the numerical simulation of the SARC model before and afteroptimization of full coverage and3-connectivity, which focuses on0.3RC RS2situation, can be found that this optimal model can be reduced by more than about athird node, when RC RS0.5. In addition, in order to facilitate area deployment andconsider the node failure due to run out of energy or damage, etc, we also describedRC R S3situation. the simulation experiment results can recommend3-connectivity and full coverage optimal model to the engineering application becauseof the network robustness and sensor nodes overhead, and satisfyRC R S3condition as far as possible, in order to reduce the cost of the systemarchitecture.(b) In view of the NLOS error constraint, this thesis presents a new kind ofanalogous linear programming model which uses a combine of the linear planning andsequential location method, and process the deterioration of error propagationproblem, improve the positioning precision of non-line-of-sight environments.Node communication will be seriously influenced by NLOS, especially in theharsh environments such as the dense forest, canyon, tunnels of twists and turns, etc.Scholars try to solve it by the many different ways, one is to use the nodes with theGPS module which provides a high-precision localization information, but the cost ishigher and location accuracy is very lower or the nodes can not locate; second, themethod is to increase the number of nodes, as far as possible to circumvent the effectsof non-line-of-sight error, but also increase the cost of the whole system, and result inthe serious phenomenon of the data congestion and packet loss; the three method isSoft processing, studies the algorithm of multiple nodes cooperative localizationmethod under a non line-of-sight environment. Our works is mainly based on the third,the corresponding mathematical model is established, and a comparative experimentwas carried out.To establish a linear objective function, linear programming method requiresline-of-sight distance estimation of at least three anchor nodes. However, WSN isoften not enough distribution of anchor nodes. So basically, this condition is difficult to satisfy. In order to get more information and enlarge the range of localizationcoverage, this thesis put forward a new kind of analogous linear programmingframework which combines with linear programming method and the sequentiallocalization method. First of all, we need to establish artificial non-line-of-sightlocation estimation, and these estimates should be brought into the linearprogramming method, to facilitate completion of blind node localization; Second, weuse the thought of virtual anchor node, when a blind node location once get, willimmediately updated to a new virtual anchor node, thus assist to complete other blindnode localization.We did simulation experiments of sequential least squares estimate and the newframework about the two technical indexes of coverage and positioning accuracy, andalso investigated the affect of the number of anchor nodes and blind nodes to the twotechnical indexes. First, we analyzed coverage and the number of localization usingthe cumulative distribution function, and found that the new framework has greatadvantages in2-hop and3-hop, thanks to inhibition of cooperative localization tonon-line-of-sight error. Second, we have made the performance comparison of theaverage coverage and the localization error, experiments show that the newframework has better performance than sequential least squares estimate in3-hop.Atthe same time, we also found that the average coverage effects are not obvious by theincrease of the hop count, and increasing of the number of blind node will havenegative influence on localization accuracy and computational complexity. So, weneed to compromise between the two aspects according to the practical application.(c) In view of the energy constraint, this thesis design a simple energy efficientrouting protocol which used in the evaluation of WSN node energy consumptiontrends, go a step further, we put forward a kind of non-uniform Progressivedeployment algorithm (NUPDA) which realizing the equalization of energyconsumption of nodes to prolong the network lifetime.Network node need to consume large energy in carrying out detection,processing and communication tasks. Due to the energy is limited, limited energyoften leads to the monitoring blind hole, network partition and even complete lossperception ability. Scholars mainly carry out research of the network lifetimeextension technology from two aspects, one is the development of energy scavengingtechnology, by capturing solar energy, wind energy and other forms of external energyprovide the energy for network. Second is to carry on research on the premise oflimited energy, through development of low energy design and energy balance technology of network node to achieve prolong network lifetime.First of all, according to the extensibility, energy efficient, simple and practicalprinciple, we design a simple energy efficient routing protocol based on the wirelesssensor node energy consumption model, and present the operation method and thebasic flow of this protocol. Then, we mainly evaluated the performance of theprotocol about time consume at which the first node fails due to depleting its energysource, and at which the sink becomes unreachable, due to its last neighbor failing, atwhich network node survival rate is lower than the corresponding threshold, and aboutthe comparison of average residual energy of nodes. At the same time, we carry outexperiment of the information acquisition. The conclusion provides an importantreference for our understanding of the trends of WSN node energy consumption.Based on the above work, we focus on2-D deployment model. First of all, wedefine lifetime of a single node and the network. Then, we divide the energyconsumption level of sense field according to the energy consumption trend, andintervene different level lifetime by increasing the number of nodes, and present anetwork energy equalization algorithm---NUPDA; Then, on the premise of the samenumber of nodes, node number of non-uniform distribution of different levels and thecorresponding lifetime was given. Finally, we contrasted the uniform and non-uniformdisplay. The results show that the proposed algorithm has greatly improved in thenetwork lifetime. |
PDF Full Text Request