Research On Node Localization Algorithm In Wireless Sensor Network

Wireless sensor network (WSN) is a new type of informationacquisition and processing mode which is highly combined with the microelectro-mecha nical system, the on-chip system and the wirelesscommunication technology. This network system can be widely used innationa l defense and military, environmental monitoring, trafficma nagement, and med ical and public health fields. In the wireless sensornetwork, the node loca lization algorithm is a critica l issue. First, it is thebasis for a variety of applica tions. In the applica tion of wireless sensornetwork, the node location of the data collection is one of the most basicinformation needed to be known. In the applica tion of environmentaldetection, the target tracking and others, if we don't know the loca tion ofthe point of collection, the collection data will be no significa nce. Second,wireless sensor network have ma ny pivota l technologies need to be used tolocate nodes information. For example, geographical routing protocolneeds the direct using of node loca lization information for datatransmission to avoid the information spreading in the entire network, andthe information can be directed enquiries, and some routing protocol alsoneeds node loca lization information to improve the efficiency of networkrouting topology control mecha nisms, and also need to use nodeloca lization information to construct the topology map, and to assess thedistribution of nodes. The loca lization problem is a technical difficulty. Inthe wireless sensor network, the number of the nodes is often hundreds, themajority of the nodes are randomly distributed, and the applica tionenvironment is often difficult to access for huma n, so it is difficult tola yout the location of each node well in advance. If all the nodes areinsta lled Global Positioning System (GPS), not only increasing costs, butalso consuming the energy soon, so this method is not suitable for practica luse. Therefore it is very important to study the effective loca lizationalgorithm to determine the location of each node. Existing loca lization algorithms can be broadly divided into twocategories: range-free algorithm and ranged algorithm. There are someperforma nce indica tors to evaluate, such as: 1, the positioning precision; 2,node density; 3, anchor node density; 4, power; 5, the cost of; 6, faulttolerant and self-adaptability. This paper will present a broader use of thepositioning algorithm such as convex programming algorithm, MDS-MAPalgorithm, APIT algorithm, DV-hop algorithm, centroid location algorithmto compare. After the comparison, we find that, as the restriction of thesensor node energy, cost, size, and other factors, range-free algorithm has ahigher practicability. Therefore, the current loca lization algorithms havefocused on the studying of range-free algorithm.In the range-free algorithms, the DV-hop algorithm can meet the needsof certain areas both in the node energy and the node loca lization accuracy.But after simulating and the performa nce analyzing by Matlab, we find thatthis algorithm will have a low accuracy when the anchor nodes are far awayfrom the unknown nodes. Then this paper concludes the conception ofanchor influence: in the DV-hop algorithm, the anchor nodes can influencethe location calculating of the unknown nodes, the nearer the dista ncebetween the anchor nodes and the unknown nodes, the higher the accuracyof the location; contrary, lower accuracy. Next, I have concluded theweighted centroid algorithm. The algorithm has the same two phases as theDV-hop algorithm, that is, first of all, the unknown nodes use the dista ncevector excha nge agreement to get hops to the anchor nodes, followed byaccess to the network avera ge jump dista nce. In the third stage, thealgorithm would no longer use the DV-hop algorithm least squares methodto calculate the location of the unknown node, but to use the weightedfactors to carry on the location calculation that the unknown nodes have hadthree or more anchor nodes location information.Weighted centroid algorithm is based on the theory: the weightedfactor reflected that the anchor nodes have the power to decide the centroidcoordinates the using of the weighted factor to indica te the influence of theanchor nodes to the centroid location and reflect the relationship betweenthem. We have simulated both the weighted centroid algorithm and the origina l DV-hop algorithm, after the calculation of the unknown nodecoordinates; we change the number of the anchor nodes and the dista ncebetween the anchor nodes and the unknown nodes, and then calculate theavera ge error value and the ma ximum error va lue. After the comparison, wefind that the weighted centroid algorithm has a higher location accuracytha n the DV-hop algorithm in the circumsta nce of the lower density of theanchor nodes or the farer dista nce between the anchor nodes and theunknown nodes, so I can say that the lower location accuracy for the DVhopalgorithm in this condition has been solved.Energy industry is the lifeblood of economic development, theimportance of the coal industry has become more apparent and isirreplaceable. And the safety of the coal industry should not be optimistic,as mine is operating in the underground movement of large, it is difficult forthe ground persons to control of underground distribution and the workingconditions, when the mine is flood ing, firing and other incidents, the groundstaff unable to understa nd mine timely, it will bring a lot of problems torescue and relief work. Therefore, the fina l paper, according to the need, theuse of wireless sensor networks designed a coal mine personnel positioningsystem pla n. Underground staff Positioning System, is a computer softwareand hardware, information acquisition and processing, wireless sensornetwork communications, multid isciplinary GIS technology as oneintegrated applica tion positioning information technology, the system isthrough the underground staff non-contact information acquisition andprocessing, to achieve the location for the moving or static persons, then wecan implement the auto ma nagement for the aim. The system can be dividedinto control and ma nagement centre and underground network in two parts.The programmer in the system with the following features: 1, disaster relief;2, for the current underground personnel distribution; 3, the investigationpersonnel current location; 4, Underground staff tracking. In this paper, wealso have a network system design pla ns, as well as positioning system map,and we have analyzed and design the location algorithm in this situationIn this paper, I have just done a relatively based research andexperiments, the improved algorithm also have its own shortcomings. For example, the weighted centroid algorithm has a lower accuracy in the higherdensity of the anchor nodes tha n the origina l DV-hop algorithm; thealgorithm complexity is also higher tha n the origina l algorithm.As the wireless sensor network node positioning algorithm research inma ny areas is still in a stage to try and explore, there are ma ny aspects of illconsidered.For example, in this paper, the example of the undergroundmine personnel positioning system pla n, the moving of the unknown nodeshas challenge the existing algorithms. Therefore, the positioning algorithm,we have to be examined further, we should try our best to do more work.