With the development and continuous progress of society, wireless sensor networks have been widely used. However, due to the limited energy of wireless sensor nodes, the application environment of wireless sensor networks is relatively fragile. Here, I search and sort out the fault diagnosis technology of wireless sensor networks for everyone. Welcome to read, I hope it will help you! For more information, please continue to pay attention to our fresh graduates training network!
Wireless sensor networks are composed of a large number of sensor nodes. Because sensor nodes are cheap and miniature, and the utilization rate of wireless sensor networks is high, especially in the monitoring field of wireless sensor networks, the monitoring task of wireless sensors is completed in the form of self-organization and cooperation, so its application prospect is very broad, but the working ability of sensor nodes is limited, and system failures will inevitably occur.
1 wireless sensor network fault evaluation index
The performance evaluation index of wireless sensor network fault diagnosis is based on the network characteristics and network application of wireless sensors, and its standards are mainly reflected in four aspects: diagnostic accuracy, diagnostic accuracy in special environment, energy efficiency and diagnostic time.
Diagnostic accuracy. The fault diagnosis accuracy of wireless sensor is the most direct evaluation method of fault diagnosis mechanism, especially in the environment with high network security, if the fault diagnosis accuracy cannot be guaranteed, it will lead to security loopholes in the sensor network system, that is to say, the fault diagnosis accuracy will fail. The accuracy of diagnosis is mainly based on the first process of fault diagnosis, and the degree of coincidence between the diagnosed node state and the actual node state is analyzed. Fault false alarm rate and fault recognition rate are two indexes to evaluate faults.
Diagnostic accuracy in special environment. The application of wireless sensor network in special environment has specific diagnostic accuracy, such as natural disasters, man-made destruction and other special environmental factors. Due to the uneven distribution of fault nodes in the network, there may be excessive evacuation or excessive concentration of nodes in the fault area, and the common diagnostic accuracy is not suitable, so the fault can only be evaluated with the diagnostic accuracy of special environment.
Energy efficiency. Affected by the energy supply of wireless sensor networks, energy efficiency becomes the first problem to be considered in fault diagnosis and evaluation mechanism. An energy-efficient fault diagnosis mechanism can promote the service life of the network, thus ensuring the continuous supply of energy in the monitoring and calculation of the sensor network. Factors directly related to energy efficiency include data communication, processing and acquisition.
Diagnostic time. After the wireless sensor network is put into use, if fault diagnosis is needed, the relationship between nodes in the sensor needs to be judged cooperatively, mainly because the number of active nodes is relatively large, and if the nodes have contact faults, it will definitely cause huge energy consumption pressure on the wireless sensor network, so the time for node fault diagnosis cannot be too long.
2 wireless sensor network fault diagnosis classification
The faults of wireless sensor networks mainly come from the nodes of sensors, which are mainly manifested in four modules, namely, energy battery power supply module, wireless network communication module, sensor processing module and sensor module. Based on the operation and use of wireless sensor networks, there will be various problems in its components and components, such as communication interference, line aging, power consumption, loose wiring and so on, which will lead to the failure of wireless sensor networks.
2. 1 node level fault
Node-level faults mainly occur in nodes of sensor networks. Most faults are mainly caused by the problems of sensor nodes themselves, which are divided into soft faults and hard faults. Soft fault refers to the node fault that can affect communication instantly without affecting the operation of wireless sensor networks. Hard fault refers to the direct damage to the node itself and the sensor network, such as the damage to the node itself, unreasonable power supply arrangement or insufficient power supply energy, which will cause the failure of the wireless sensor network.
2.2 Network level fault
Network-level fault refers to that the nodes of wireless sensor are normal, but there are restrictive problems in transmission and cooperation between nodes, which leads to abnormal network connection, blocked communication, information loss, IP deviation, illegal intrusion and so on. This kind of fault directly affects the network, which is very obvious, and the fault appears very quickly and has a very wide range of influence, which belongs to the more sensitive fault in wireless sensor networks.
2.3 functional level fault
The failure of functional level of wireless sensor network will affect the whole network. For example, a failure at the functional level will cause the convergence point in the network to fail to receive and collect all the information running in the network normally. The main causes of functional failures are the restart, death and failure of sensor nodes, link failures and routing equipment failures.
2.4 Data-level fault
Data-level fault refers to the normal behavior of sensor nodes, but the wrong data information is transmitted, which leads to the formation of wrong data perception in the network. Data-level faults are relatively hidden. Only after careful detection can we find that the sensor node has transmitted the wrong sensing data, because even if the node senses the data transmission error, there is no problem in its own form, which invisibly reduces the running performance of the wireless sensor network and will lead the network administrator to check and repair it wrongly.
3 wireless sensor network fault diagnosis technology
The fault diagnosis of wireless sensor network is mainly aimed at the period when it is put into use. It judges whether there is a fault in wireless sensor network by analyzing the information transmitted by the network, and detects the basic root cause of the fault according to the state of the fault. Fault diagnosis of wireless sensor networks is a complex system engineering, and its environment and operating characteristics determine the difficulty of fault diagnosis. In order to reduce the difficulty of diagnosis, in general, fault diagnosis should be based on the daily measurement data of each sensor node, supplemented by additional information transmitted by node data, so as to promote the efficiency of fault diagnosis.
The indicators of fault diagnosis in wireless sensor networks are high-quality service and effective protection of energy, and the measurement indicators of fault diagnosis strategy mainly include false alarm rate and detection rate, in which the false alarm rate feeds back the proportion of invalid alarms in the diagnosis report to the total alarms, and a low false alarm rate can indicate that the diagnosis results are highly reliable; The detection rate feeds back the proportion of detected faults in the total network faults, which is opposite to the error reporting rate. The higher the detection rate, the more effective the diagnosis strategy is. At present, the research on fault diagnosis technology of wireless sensor networks mainly focuses on sensor faults and scene types, discusses the functions and reading faults of sensor nodes, and analyzes the fault diagnosis technology of wireless sensor networks.
3. 1 sensor node reading fault diagnosis technology
Node reading fault diagnosis technology is mainly aimed at the wrong measurement data in wireless sensor networks. The situation of generating erroneous data mainly includes network security attacks and node component damage caused by external environment interference. Aiming at the node reading fault, the following diagnostic techniques are proposed. Diagnostic technology of (1)WMFDS. This technology mainly measures the spatial correlation of data between sensor nodes, and the closer the nodes are, the greater the similarity of measurement results. Therefore, based on the spatial relationship of normal readings, a WMFDS diagnosis method is proposed, which mainly analyzes the failure rate and distribution density between two nodes to determine whether there is a problem in the nodes. This method can also be used to weight adjacent nodes, but this method can only be used for readings of nodes with spatial correlation.
(2) Find diagnostic techniques. This technology uses the characteristics of continuous monitoring of wireless sensor nodes in the monitoring area to perceive emergencies in the network. The data reading of this node can feed back the distance from the incident point to the node, and the signal strength of the sensor node is inversely proportional to the distance, that is, the greater the relative distance, the weaker the signal strength of the node, and the change of the signal strength of the node is called monotonic change characteristic, so the monotonic characteristic of the node is the criterion for judging the reading failure of the feedback node. For example, a faulty node will exhibit the opposite phenomenon to the monotony of relative distance.
(3)CSN diagnosis technology. This diagnostic technique has some limitations. It mainly takes mobile devices as the detection object, and uses accelerometers to obtain the seismic motion of nodes. There will be a threshold value for the readings of faulty nodes, which is quite different from the actual history. Analyze the proportion of nodes by computer. If the threshold is high, it means that there is something wrong with this node.
3.2 sensor node network fault diagnosis technology
The network failure of sensor nodes is mainly manifested in the phenomenon that the link is affected by environmental factors, which leads to the decline of network reliability. Aiming at the network fault of sensor nodes, three main diagnosis technologies are proposed:
(1) network software debugging method. The debugging agent is used in the sensor nodes, and the network state at the nodes is analyzed by the debugging command of the software, and the network data of the nodes is collected to determine the source of the node network fault.
(2) Specific model reasoning method. Specific model reasoning methods mainly include distributed and centralized methods. Distributed diagnosis technology is aimed at all nodes in the network and adopts a decision-making method from local to whole. The representative methods of distributed diagnosis technology are LD2 and TinyD2, and finally the diagnosis report is obtained through the integration of node networks. Centralized diagnosis technology is to implant small detectors in the nodes of the network and classify and group the application data passing through the nodes. However, the detector's ability to analyze the obtained information is very limited, which requires the participation of the sensor system, and based on this, the node network fault is diagnosed in detail.
(3) Silent fault diagnosis technology. This diagnostic technique has certain particularity among the three techniques. It can effectively diagnose inexperienced faults, such as AD diagnosis technology, which is a typical representative. By changing the correlation diagram between various diagnostic information of nodes, we can find hidden faults in the network, that is, silent faults. This technology can improve the accuracy of fault diagnosis and reduce the frequency of fault occurrence.
To sum up, using wireless sensor fault diagnosis technology to diagnose the problems existing in wireless sensor networks, and timely and effective treatment, on the one hand, can improve the working efficiency of wireless sensor networks, on the other hand, improve the utilization rate of wireless sensor networks, so the normal operation of wireless sensor networks has promoted the development and improvement of China's economic and social benefits to a certain extent.
To sum up, the attention of wireless sensor networks in the world is relatively high, which permeates many scientific technologies, such as wireless communication technology, sensor technology and information processing technology. The research of wireless sensor has important economic and social benefits. On the one hand, effective fault diagnosis technology of wireless sensor networks can improve the utilization efficiency of wireless sensors, on the other hand, it has certain practical value for energy saving.
;