| Wireless sensor networks contain numerous sensors which are interrelated, process each other and transmit some information through wireless communicate. From the end of the last century, field bus technology has begun to be used in sensor network, which was used to build intelligent sensor networks, and the large multifunctional sensors are used there. Then wireless sensor network is beginning to take shape. Wireless sensor networks now being used in military, medical, environmental testing, and other aspects. Existing wireless sensor networks were generally consistent with the low power, scalable, adaptive and simple feature. At present there are many researches in the area. The research can be divided into two parts: node level and network level. The content of node level includes sensor technologies, low power chip technology, wireless communications technology; The content of network level include low power routing technology, low power MAC protocol and co-location technology, clock synchronization, data fusion technology. This paper is a study of the location technology.In the location technology of wireless sensor network, the nodes are divided into beacon nodes and unknown nodes according to whether the node is aware of its own position. The beacon nodes occupy a small proportion of the whole network; we can obtain their precise position through the GPS. The beacon nodes are the referenced nodes of the unknown nodes. Besides the beacon nodes, the left are all unknown nodes, they find their position according to the position of the beacon nodes.Generally, we divide location algorithm into range-based location algorithm and range-free location algorithm. The former required the measurement of the absolute distance or position between nodes, and used the actual distance between nodes to calculate the position of unknown nodes; the latter did not measured the absolute distance or position between nodes, but used the estimated distance between nodes to calculate the position of the node.Range-based location is implemented by measuring the absolute distance or position between nodes. Specific process is usually divided into three stages: The first is the stage of measuring distance, unknown nodes measure the distance or angle to the neighbor nodes, and calculate distance or angle to the nearby beacon nodes. When calculating the distance from unknown nodes to the nearby beacon nodes, we can only calculate straight-line distance or use the hop distance to be approximate straight-line distance. The second is the stage of locating. After calculating the distance or angle to three or more beacon nodes, unknown nodes use trilateration, triangulation or maximum likelihood estimation to confirm its own position. The third is the stage of correcting. We should improve the precision of the result and reduce the error. In the range-based location, there are four kinds of method when measuring the distance or position between nodes: time of arrival based location mechanism (TOA), time of difference arrival based location mechanism (TDOA), angle of arrival based location mechanism (AOA) and received signal strength indicator based location mechanism (RSSI). This paper mainly studies the RSSI location mechanism.We must pay attention to some characteristic of RSSI. When some conditions are not satisfied, RSSI is trustless. There are some factors affecting RSSI: antenna length, antenna orientation, node height and some weather environment, such as temperature and humidity, also indoor and outdoor are different. We can not control the temperature and humidity, so we suppose them changeless. Under these conditions, let us find out how the other factors affect RSSI. Over some experiment, at a certain scope, the longer the antenna is, the more obvious the law of the RSSI is. The closer to 2πthe angle between the nodes is, the more obvious the law of the RSSI is. When the height of the node is equal or bigger to 2m, the law between the length and RSSI is very good, the effect of the other condition are tiny. Then, I did some tests to get some data and calculated the relation between the length and RSSI: length = 0.4425×RSSI2 + 10.2174×RSSI +78.6547. So, the location system studied in this paper must satisfy some conditions: outdoor, changeless temperature and humidity, the height of the node is about 2m ( ?+1 0cm). Without these condition, maybe, the result is error or trustless.The whole location system is based on TinyOS and nesC language, and the nodes are composed by MSP430 and CC2420. The system can be divided into three parts: beacon nodes, unknown nodes and listening nodes. Beacon nodes periodic send broadcast message which contains the node's ID and position. After unknown nodes have received these messages, they obtain and record the RSSI till finishing 10, then we calculate the average of the RSSI. After finishing the 3 RSSI averages, we could calculate the lengths of the unknown node to the 3 beacon nodes according to the relation between the length and RSSI. Finally the unknown node calculates the position according to the trilateration and sends it with the node's ID. The listening nodes can receive the message, so we can see the position of the unknown nodes. Every location system has some errors; the end of this paper does some work about error. There are about four sources of the error: error from the node's own calculating, error from the relation between RSSI and the length, error from the trilateration, error from the outside environment. The system enlarges the multiple in order to reduce or eliminate the errors about float when calculating; the errors from environment factor are ineluctable; there are only three beacon nodes, so RSSI value has a certain degree of randomness. The result of locating maybe trustless if some message is affected and antenna length or antenna orientation both can change the RSSI value tinily. We should do some work, such as get the average of RSSI or the weighted RSSI, to reduce the RSSI randomicity. Now, we make some experiment at the same environmental conditions: selecting the same nodes and locating the node with the single RSSI mode, average RSSI mode and weighted RSSI mode per 5m, maximum is 30m. According to the result, error from the average RSSI mode is the lowest, error from the single RSSI is topmost and error from the weighted RSSI is in the middle. The average RSSI mode is selected in the location system of this paper.It is believed that with the maturity of the WSN technology, the development of the topology frame and the improvement of the network carrying capacity, its application scope will be wider. It has more precise localization and more mature networks to improve the management and control of the vehicle on the highway efficiency. All these can create abundance of higher economical and social returns for the benefit of human being. |
PDF Full Text Request