JP5328829B2 - Sensor network system, sensor data search method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily search real time information from a number of sensors connected to a network. <P>SOLUTION: The system includes: a distributed server DDS which collects data from a plurality of sensors in predetermined periods; a management server DRS which retains locations of data of the sensors; a user terminal UST which requests data to the management server DRS; and a first network NWK-1 which connects the management server, the user terminal, and a plurality of distributed servers. The management server DRS includes a model table MTB which has a preset model name and an information storage destination showing which of the distributed servers the collection destination of data corresponding to this model name is; and a search unit SER which obtains data which has been requested from the user terminal UST from the distributed server DDS corresponding to its information storage destination based on the model table MTB, and returning the data to the user terminal UST. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to a technology that uses information from a large number of sensors connected to a network.

  In recent years, the use of networks such as the Internet is mainly made by accessing stored contents such as documents, images, videos, and sounds from search engines or preset links. In other words, a technique for accessing stored past contents has been completed.

On the other hand, there is a streaming technique for continuously transmitting images from a camera (WEB camera) provided at a predetermined position as a method for transmitting current information. Recently, sensor network technology for acquiring sensing data obtained from a large number of small wireless sensor nodes through a network is being developed (Patent Documents 1 to 5). In recent years, there is an increasing expectation for a sensor network that captures real-world information by a sensor and uses this information at a remote location through the network. Although services on the current Internet are closed to services on the virtual space, the sensor network is essentially different from the current Internet in that it merges with the real space. If integration with real space can be achieved, various situation-dependent services such as time and location can be realized. Traceability can be realized by connecting various objects in real space to the network, and the social needs for “safety” in a broad sense and the “efficiency” needs for inventory management and office work Is possible.
JP 2002-006937 A JP 2003-319550 A JP 2004-280411 A US Patent Application Publication No. 2004/0093239 US Patent Application Publication No. 2004/0103139

  However, although the search engine as shown in the above conventional example can know the position (address) on the network for the data accumulated in the past, real-time information from a huge amount of sensor information connected to the network There is a problem that it is not suitable for efficient search and information change detection.

  An object of the present invention is to realize a sensor network system that can easily retrieve real-time information from a large number of sensors connected to a network.

The present invention comprises a distributed server that stores data transmitted from a plurality of sensor nodes, and a management server connected to the plurality of distributed servers and user terminals via a network, wherein the distributed server includes the plurality of distributed servers. A data processing unit that receives a plurality of types of sensor data transmitted from the sensor node and processes the sensor data by a predetermined processing method, and an accumulation that stores processing data processed by the data processing unit And the management server includes:
A model list storing a plurality of model names set in advance, and an information storage destination for specifying which distributed server is a storage destination of data corresponding to the model name, and data of the sensor node Common to a plurality of model names among elements constituting the model name and a semantic interpretation list in which conversion definitions for converting the information into information understandable by a user using the user terminal are set corresponding to the model name A model binding list that stores definitions associated with information storage destinations of the model list for different types of data, and a search request from the user terminal. A search unit that responds to the user terminal with an information storage destination of data including the model name included in the search request, and data received from the distributed server. A semantic interpretation unit that converts the user terminal into a meaningful information that can be understood by a user using the user terminal and responds to the user terminal on behalf of the distributed server, and the search unit searches from the user terminal. The conversion definition that accepts the request, searches the model binding list for the model name element included in the search request, extracts the corresponding information storage location, and matches the model name element included in the search request Is extracted from the semantic interpretation list, data corresponding to the model name is requested to one of the plurality of distributed servers based on the information storage destination, and one of the plurality of distributed servers is extracted from the search unit. based on the request, to obtain the processed data from the storage unit, and transmits to the management server, the interpretation of the previous SL management server, on the basis of the conversion definition the extracted It converts the data received from one of several distribution server to significant information, and transmits the significant information to the user terminal.

  Therefore, according to the present invention, data collection from the sensors is performed by the distributed server, the location of the sensor data is managed by the management server, the user terminal requests data from the management server, and the data is sent to the distributed server that stores the management server information The management server acquires the data received from the distributed server and responds to the user terminal. As a result, even if the number of sensors becomes enormous, data is not stored in the management server, so it is possible to prevent an excessive load, and while using a large number of sensors, the management server, distributed servers and users It is possible to easily and quickly provide necessary information to the user while suppressing excessive traffic on the network connecting the terminals.

  The user does not need to know the position and function of the sensor, and can request the data from the management server to obtain the desired sensor data as significant information. In a sensor network having a large number of sensors, Can be made easier.

  Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

  FIG. 1 is a configuration diagram of a sensor network system showing a first embodiment of the present invention.

<Outline of system configuration>
The sensor node WSN (wireless sensor node), MSN (wireless mobile sensor node) is installed at a predetermined position or attached to a predetermined object or person, collects information about the environment or information about the attached object, It is a node that transmits information to base stations BST-1 to n. The sensor node is composed of wireless sensor nodes WSN and MSN connected to the base stations BST-1 to n by wireless and wired sensor nodes FSN connected to the network NWK-n by wire.

  For example, the wireless sensor node WSN that is fixedly installed senses the surrounding situation periodically by a mounted sensor, and transmits sensing information to a preset base station BST. The wireless mobile sensor node MSN transmits information to the nearest base station BST on the premise that the mobile mobile sensor node MSN can move such as being carried by a person or mounted on a car. Note that WSN or MSN is used when referring to the entire wireless sensor node (generic name), and subscripts such as WSN-1 to n or MSN-1 to n are used to indicate individual wireless sensor nodes. In the same manner, the other constituent elements are represented without subscripts when indicating generic terms, and with subscripts “−1 to n” when indicating individual components.

  One or a plurality of wireless sensor nodes WSN and MSN are connected to each base station BST-1 to n, and each base station BST-1 to n is connected from each sensor node via a network NWK-2 to n. It is connected to distributed data processing servers DDS-1 to n that collect data. The networks NWK-2 to n are for connecting the base station BST and the distributed data processing server (distributed server) DDS. The number of connections of the distributed data processing server DDS can be varied depending on the size of the system.

  Each distributed data processing server DDS-1 to n includes a disk device DSK for storing data detected by wireless and wired sensor nodes (hereinafter simply referred to as sensor nodes), a CPU and a memory (not shown), and a predetermined program. And collects measurement data from the sensor node as will be described later, and stores data, processes the data, and further performs a directory server (management server) DRS or the network via the network NWK-1 according to a pre-defined condition. Perform actions such as notification to other servers and data transfer. The network NWK-1 is configured with a LAN, the Internet, or the like.

  Here, the data collected from the sensor node is mainly a unique ID for identifying the sensor node and sensed numerical data, and changes according to the time series, but the output of the sensor node is used as it is. It is not in a format that can be easily understood by a user (a user such as a user terminal UST). Therefore, the directory server DRS converts the sensor node output data into a real-world model (human, thing, state, etc.) that can be understood by the user based on a preset definition, and presents it to the user.

  In addition, the object from which the distributed data processing servers DDS-1 to Dn collect data has moved from a sensor node belonging to the base station BST of the network NWK-2 to n to which the distributed data processing server DDS-1 to n is connected or from another base station BST. Data is collected for the data from the wireless sensor node MSN and the above conversion is performed. The wired sensor node FSN may be connected to the distributed data processing servers DDS-1 to DDS-n. Of course, the wired sensor node FSN can be connected to the base station BST, and the base station BST can manage the wired sensor node FSN in the same manner as the wireless sensor node.

  The network NWK-1 includes a directory server DRS that manages a real world model associated with sensing information sent from the distributed data processing server DDS, a user terminal UST that uses the information of the directory server DRS, and a directory server DRS. And a distributed data processing server DDS, a base station BST, and a management terminal ADT for setting and managing sensor nodes are connected. The management terminal may be prepared for a sensor manager who manages sensor nodes and a service manager who manages sensor network services.

  The directory server DRS includes a CPU, a memory, and a storage device (not shown), executes a predetermined program, and manages objects associated with significant information as will be described later.

  That is, when the user requests access to the real world model through the user terminal UST, the directory server DRS accesses the distributed data processing servers DDS-1 to n having the measurement data corresponding to the real world model, and corresponds. Measurement data is acquired, and the sensing data is converted into a form that can be easily understood by the user if necessary, and displayed on the user terminal UST.

  FIG. 2 is a functional block diagram of the sensor network shown in FIG. Here, for the sake of simplicity of explanation, only the configuration of the distributed data processing server DDS-1 among the distributed data processing servers DDS-1 to Dn in FIG. 1 is shown, and connected to the distributed data processing server DDS-1 Of the base stations BST-1 to n, only the configuration of the base station BST-1 is shown. Other distributed data processing servers DDS and base stations BST are configured similarly.

  Details of each part will be described below.

<Base station BST>
A base station BST-1 that collects data from wireless sensor nodes WSN, MSN and wired sensor node FSN (hereinafter referred to as sensor node) includes a command control unit CMC-B, a sensor node management unit SNM, and an event monitoring unit EVM. Is provided. The sensor node transmits measurement data with a preset ID.

  The command control unit CMC-B transmits / receives commands to / from a command control unit CMC-D of the distributed data processing server DDS-1 described later. For example, in accordance with a command from the distributed data processing server DDS-1, the parameter setting of the base station BST-1 is executed or the state of the base station BST-1 is transmitted to the distributed data processing server DDS-1. Alternatively, the parameter setting of the sensor node managed by the base station BST-1 is executed, or the state of the sensor node is transmitted to the distributed data processing server DDS-1.

  The sensor node management unit SNM holds sensor node management information (operating state, remaining power, etc.) managed by the sensor node management unit SNM. When there is an inquiry about the sensor node from the distributed data processing server DDS-1, management information is notified instead of each sensor node. By leaving the processing related to management to the base station BST, the sensor node can reduce its own processing load and suppress excessive power consumption.

  In addition, when the event monitoring unit EVM detects an abnormality, the sensor node management unit SNM updates the management information of the sensor node and notifies the distributed data processing server DDS-1 of the abnormal sensor node. A sensor node abnormality indicates a state in which the function of the sensor node is stopped or stopped, such as when there is no response from the sensor node or when the power of the sensor node falls below a preset threshold value. .

  When the sensor node management unit SNM receives a command (output timing setting) for the sensor node from the command control unit CMC-D, the sensor node management unit SNM transmits the command to the sensor node to perform setting, and completes the setting. After receiving the notification shown from the sensor node, the management information of the sensor node is updated. Note that the output timing of the sensor node indicates, for example, a period when the wireless sensor node WSN periodically transmits data to the base station BST-1.

  The base station BST manages the wireless sensor nodes WSN, MSN and the wired sensor node FSN under control set in advance, and transmits data measured by each sensor node to the distributed data processing server DDS.

<Distributed data processing server DDS>
The distributed data processing server DDS-1 includes a disk device DSK that stores a database DB, a command control unit CMC-D, an event action control unit EAC, and a database control unit DBC as described later.

  The command control unit CMC-D communicates with the base station BST and a directory server DRS described later, and transmits / receives commands and the like.

  Each time the event action control unit EAC receives measurement data from the sensor node from the base station BST, the event action control unit EAC obtains an ID corresponding to the measurement data, that is, a data ID, and stores data from a table (event table ETB in FIG. 10) described later. An event occurrence rule corresponding to the ID is read to determine whether an event has occurred according to the value of the measurement data. Further, the event action control unit EAC executes an action corresponding to the occurrence of the event corresponding to the data ID. If only one sensor is provided in the sensor node, the sensor node ID for identifying the sensor node may be used as the data ID.

  Then, as the content of the action execution, the measurement data is converted into the processing data based on the rules for each data ID set in advance by the user or the like, or the measurement data and the processing data are stored in the database DB through the database control unit DBC. Or processing such as notifying the directory server DRS.

  In this embodiment, as shown in FIG. 1, by arranging a plurality of distributed data processing servers DDS that aggregate some of them locally (or locally) with respect to a plurality of base stations BST, Information from a large number of sensor nodes can be distributed and processed. For example, a distributed data processing server DDS may be provided for each floor in an office or the like, and a distributed data processing server DDS may be provided for each building in a factory or the like.

  The disk device DSK of the distributed data processing server DDS-1 includes the sensor node WSN, MSN, and FSN measurement data received from the base station BST, the processed data obtained by processing these measurement data, the base station BST, and the wireless sensor node WSN. , Device data related to the MSN and the wired sensor node FSN are stored as a database DB.

  Then, the database control unit DBC of the data processing server DDS-1 stores the measurement data that is the output of the sensor node sent from the event action control unit EAC in the database DB. If necessary, the processing data obtained by numerically processing the measurement data or merging with other data is stored in the database DB. The device data is updated as needed in response to a request from the administrator terminal ADT.

<Directory server DRS>
The directory server DRS that manages the plurality of distributed data processing servers DDS includes a session control unit SES that controls communication from the user terminal UST and the administrator terminal ADT connected via the network NWK-1, and, as will be described later. The model management unit MMG, model table MTB, device management unit NMG, action control unit ACC, and search engine SER are included.

  The model management unit MMG is a real-world model table MTB in which the correspondence between the real-world model (object) that is easy for the user to understand and the measurement data or processing data collected from the sensor nodes by the distributed data processing server DDS is set. It is managed by the world model list MDL.

  The directory server DRS also manages position information (links such as URLs) of locations where measurement data or processing data corresponding to the real world model exists. That is, the user can directly access measurement information of the sensor node that changes every moment by designating the real world model. While the measurement data and processing data from the sensor node increase with the passage of time, the size of the real world model information does not change with the passage of time, only the contents thereof. Details of this real world model will be described later.

  The real world model table MTB is stored in a predetermined storage device (not shown) of the directory server DRS.

  The action control unit ACC of the directory server DRS communicates with the event action control unit EAC and the command control unit CMC-D of the distributed data processing server DDS, and makes a request for setting an event action from the user terminal UST or the administrator terminal ADT. Accept. Then, the contents of the received event or action are analyzed, and function sharing between the directory server DRS and the distributed data processing servers DDS-1 to n according to the analysis result is set. One action or event may involve not only one distributed data processing server DDS but also a plurality of distributed data processing servers DDS-1 to n.

  The search engine SER performs a search on the database DB of the distributed data processing server DDS with reference to the real world model table MTB based on the search request for the object received by the session control unit SES.

  If the search request is a query, the database DB is matched according to the contents of the query and the query is subjected to SQL (Structured Query Language) conversion to execute the search. Note that the database DB to be searched may span a plurality of distributed data processing servers DDS.

  This query corresponds to “latest data acquisition (snapshot / stream)”. “Latest data acquisition (stream)” corresponds to the action setting of the action control unit ACC. That is, an action setting that always transfers the corresponding data to the user terminal may be set in the event action control unit EAC of the corresponding distributed data processing server DDS.

  Next, the device management unit NMG includes a distributed data processing server DDS connected to the network NWK-1 to form a sensor network, a base station BST connected to the distributed data processing server DDS, and a sensor connected to the base station BST. Nodes are managed in an integrated manner.

  The device management unit NMG provides an interface related to registration and search of the distributed data processing server DDS, base station BST, and sensor node to the administrator terminal ADT, and manages the state of each device and the state of the sensor node.

  The device management unit NMG can issue commands to the distributed data processing server DDS, the base station BST, and the sensor nodes, and manages the resources of the sensor network using these commands. The sensor node receives a command from the device management unit NMG via the command control unit CMC-B of the upper base station BST, and the base station BST passes through the command control unit CMC-D of the upper distributed data processing server DDS. Receives a command from the device management unit NMG.

  Note that examples of commands issued by the device management unit NMG via the command control unit CMC-D include reset, parameter setting, data erasure, data transfer, and routine event / action setting.

<Example of sensor node>
Next, an example of a sensor node is shown in FIGS.

  FIG. 3 is a block diagram illustrating an example of the wireless sensor node WSN. The wireless sensor node WSN includes a sensor SSR that measures a state quantity (temperature, pressure, position, etc.) or a change in state quantity (low temperature / high temperature, low pressure / high pressure, etc.), a controller CNT that controls the sensor SSR, A radio processing unit WPR that communicates with the base station BST, a power supply POW that supplies power to each block SSR, CNT, and WPR, and an antenna ANT that performs transmission and reception are configured.

  The controller CNT reads the measurement data of the sensor SSR at a preset period or irregularly, adds the preset ID of the sensor node to the measurement data, and transfers it to the wireless processing unit WPR. In some cases, the measurement data is provided with time information of sensing as a time stamp. The wireless processing unit WPR transmits the data sent from the controller CNT to the base station BST.

  The wireless processing unit WPR transmits a command received from the base station BST to the controller CNT, and the controller CNT analyzes the received command and performs a predetermined process (for example, setting change). Further, the controller CNT transmits information on the remaining power (or charge amount) of the power supply POW to the base station BST via the wireless processing unit WPR. The controller CNT itself monitors the remaining power (or amount of charge) of the power supply POW, and when the remaining power falls below a preset threshold, an alarm indicating that the power is lost is transmitted to the base station BST. You can leave it.

  Since the wireless processing unit WPR performs long-time measurement with limited power, it operates intermittently and reduces power consumption as shown in FIG. In the figure, in the sleep state SLP, the controller CNT stops driving the sensor SSR, switches from the sleep state to the operation state at a predetermined timing, drives the sensor SSR, and transmits measurement data.

  Although FIG. 3 shows an example in which one wireless sensor node WSN is provided with one sensor SSR, a plurality of sensors SSR may be arranged. Alternatively, instead of the sensor SSR, a memory storing a unique identifier ID may be provided, and the wireless sensor node WSN may be used as a tag.

  3 and 4, the power source POW may be configured to use a battery or include an autonomous power generation mechanism such as a solar battery or vibration power generation. Also, the wireless mobile sensor node MSN may be configured in the same manner as in FIGS.

  FIG. 5 is a detailed view showing an example of a sensor node connected to the distributed data processing server DDS-1 shown in FIGS.

  In the present embodiment, an example in which sensor nodes are provided in the office, the A meeting room, and B is shown.

  A base station BST-0 is installed in the office, and a wireless sensor node WSN-0 having a pressure-sensitive (pressure) switch is arranged as a sensor SSR in an office chair. The wireless sensor node WSN-0 is set to communicate with the base station BST-0.

  A base station BST-1 is installed in the A conference room, and wireless sensor nodes WSN-3 to 10 having pressure-sensitive switches are arranged as sensors SSR in the chair of the A conference room. In the conference room A, a wired sensor node FSN-1 having a temperature sensor is installed and connected to the base station BST-1. Each of the wireless sensor nodes WSN-3 to 10 and the wired sensor node FSN-1 are set to communicate with the base station BST-1.

  Similarly, the base station BST-2 is installed in the conference room B, the chairs of the conference room B have wireless sensor nodes WSN-11 to 17 having pressure-sensitive switches as sensors SSR, and wired sensors having temperature sensors. Node FSN-2 is installed and connected to base station BST-2.

  Employees who use these offices, the A meeting room, and the B meeting room are equipped with a wireless sensor node MSN-1 that also serves as a name tag. The wireless sensor node MSN-1 is a name tag including a temperature sensor SSR-1 that measures the body temperature (or ambient temperature) of the employee and a tag TG-1 that stores an identifier (employee ID) unique to the employee. Composed. The wireless sensor node MSN-1 can transmit the employee ID and the temperature as measurement data to the base station BST-0, 1 or 2.

<Operational concept of sensor network>
Next, an outline of the operation of the sensor network shown in FIGS. 1 to 5 will be described with reference to FIGS.

  FIG. 6 is a block diagram showing the relationship between the measurement data of the object and the sensor node, which is a specific form of the real world model, showing the start time of the measurement, and FIG. Indicates the state.

  In FIG. 6, the directory server DRS generates the following objects in advance as real world models and defines them in the real world model list MDL of the real world model table MTB. Here, the case of Mr. Suzuki as an employee who uses the office or the A meeting room and the B meeting room of FIG. 5 is shown, and this person is wearing the wireless sensor node MSN-1 shown in FIG. To do.

  As shown in the sensor information table STB of FIG. 12, the sensor information table is stored so that the measurement data (ex. Temperature) and position information of each sensor node MSN are stored in the distributed data processing server DDS designated as the data storage destination. Is defined. Here, the position information of the sensor node MSN can be obtained as ID information of the base station BST that detects the sensor node MSN.

  In the real world model list MDL of the real world model table MTB, it is defined that the object (OBJ-1) of Mr. Suzuki's position has a data entity at the storage location of measurement data 1 (LINK-1). The correspondence between the real world model and the actual data position is managed.

  That is, in the real world model list MDL, the object of Mr. Suzuki's position (OBJ-1) is associated with the storage position of the distributed data processing server DDS corresponding to the measurement data 1 (LINK-1). 6 and 7, the position information from wireless sensor node MSN-1 indicating the position of Mr. Suzuki (for example, defined as “where to be connected to base station BST”) is distributed from distributed data processing server DDS. It is stored in the disk device DSK1.

  When viewed from the user terminal UST, the value of Mr. Suzuki's position (OBJ-1) appears to exist in the real world model table MTB of the directory server DRS, but the actual data is not the directory server DRS but a preset distribution. It is stored in the disk device DSK1 of the data processing server DDS.

  Mr. Suzuki's seating (OBJ-2) is a real world model so that the seating information obtained from the pressure sensitive switch (WSN-0) mounted on the office chair is stored in the measurement data 2 (LINK-2). Defined in table MTB. Furthermore, the distributed data processing server DDS corresponding to the measurement data 2 and the storage location are defined. 6 and 7, the seating information from the MSN-1 and the wireless sensor node WSN is stored in the disk device DSK2 of the distributed data processing server DDS.

  The object of Mr. Suzuki's temperature (OBJ-3) is defined in the real world model table MTB so that the temperature measured by the temperature sensor SSR-1 of the wireless sensor node MSN-1 is stored in the measurement data 3 (LINK-3). Is done. Furthermore, the distributed data processing server DDS corresponding to the measurement data 3 and the storage location are defined. 6 and 7, the temperature from MSN-1 is stored in the disk device DSK3 of the distributed data processing server DDS.

  In the object A conference room member (OBJ-4), the employee name obtained from the information of the wireless sensor node MSN connected to the base station BST-1 of the A conference room is stored in the measurement data 4 (LINK-4). As defined in the real world model table MTB. When the pressure sensitive switch (WSN-3 to 10) is not used, the number of the A meeting rooms can be obtained from the number of wireless sensor nodes MSN detected by the base station BST-1 in the A meeting room within a certain unit time. good. Furthermore, the distributed data processing server DDS corresponding to the measurement data 4 and the storage location are defined. 6 and 7, the personal information of each employee from the wireless sensor node MSN is stored in the disk device DSK4 of the distributed data processing server DDS.

  The object A meeting room number (OBJ-5) is stored in the measurement data 5 (LINK-5) so that the number obtained from the pressure sensitive switch (WSN-3 to 10) of the A meeting room is stored in the measurement data 5 (LINK-5). Defined in MTB. Furthermore, the distributed data processing server DDS corresponding to the measurement data 5 and the storage location are defined. 6 and 7, the seating information of the wireless sensor nodes WSN3 to 10 is stored in the disk device DSK5 of the distributed data processing server DDS.

  The object A meeting room temperature (OBJ-6) is defined in the real world model table MTB so that the temperature measured by the wired sensor node FSN-1 of the meeting room A is stored in the measurement data 6 (LINK-6). The Further, the distributed data processing server DDS corresponding to the measurement data 6 and the storage location are defined. 6 and 7, the temperature from FSN-1 is stored in the disk device DSK6 of the distributed data processing server DDS.

  That is, each object OBJ defined in the real world model table MTB stores a storage location (LINK) corresponding to the measurement data, and the target data seems to exist in the directory server DRS from the user terminal UST. However, the actual data is stored in the distributed data processing server DDS.

  In the information storage destination LINK, a storage location of data usable by the user, such as measurement data measured by the sensor node or processed data obtained by converting the measurement data into a form understandable to the user, is set. Collection of measurement data from the sensor node is performed by each distributed data processing server DDS. Further, if an event action is set as will be described later, processing is added to the measurement data, and predetermined processing data is processed. The data is stored in the distributed data processing server DDS.

  Data collection and processing from actual sensor nodes are performed by the distributed data processing server DDS, and the directory server DRS manages the real world model, information storage destination, sensor node definitions, and the like.

  Thereby, the user of the user terminal UST does not need to know the location of the sensor node, and can retrieve desired data corresponding to the measured value (or processed data) of the sensor node by searching the object OBJ. is there.

  The directory server DRS manages the storage destination (link destination) for each object OBJ, and the actual data is stored and processed by the distributed data processing server DDS. Even if the number of sensor nodes becomes enormous, It is possible to prevent an excessive load on the data processing server DDS. That is, it is possible to suppress the traffic of the network NWK-1 connecting the directory server DRS, the distributed data processing server DDS, and the user terminal UST from being excessive while using a large number of sensor nodes.

  In FIG. 7 after a predetermined time has elapsed from the state of FIG. 6, the actual measurement data from the sensor node is written in the disk devices DSK1 to 6 of the distributed data processing server DDS, and the amount of data increases with the passage of time.

  On the other hand, the storage destinations LINK-1 to LINK-6 corresponding to the objects OBJ-1 to -6 set in the model list MDL of the real world model table MTB of the directory server DRS do not change the amount of information even if time passes. The contents of the information indicated by the storage destinations LINK-1 to LINK-6 change.

  That is, the relationship between the amount of information of the objects OBJ-1 to 6 managed by the directory server DRS and the amount of data of the measurement data 1 to 6 managed by the distributed data processing server DDS and the time are as shown in FIG. Whereas the amount is constant, the measurement data increases over time.

  For example, several hundred sensor nodes are connected to one base station BST, several base stations BST are connected to one distributed data processing server DDS, and several tens of distributed data processing servers DDS are connected to one directory server DRS. When connected, the total number of sensor nodes is several thousand to several tens of thousands. Assuming that each sensor node sequentially sends data every minute, about 100 to 1,000 measurement data per second is sent to the distributed data processing server DDS to determine the presence or absence of an event. In the case of a predetermined action, processing such as notification and data processing occurs. If these processes are realized by one or a small number of servers, the load on the server itself and the load on the network for connecting to the server become extremely large. Furthermore, since the collected data and the processed data are accessed from the user terminal UST, the load on the server and the network is further increased in order to provide the data to the user terminal UST.

  Therefore, the access is accepted from the user terminal UST, the directory server DRS that manages the storage destination of the sensor node information, the plurality of base stations BST, the data is collected from the sensor nodes assigned to the base station BST, The distributed data processing server DDS to be processed is divided.

  Information from sensor nodes is collected and distributed by a plurality of distributed data processing servers DDS, and data is stored and processed by each distributed data processing server DDS, thereby collecting and processing data of a large number of sensor nodes. It is possible to distribute and prevent the load from being concentrated on a specific server.

  On the other hand, the directory server DRS centrally manages the storage destination of the information obtained from the measurement data of the sensor node, and provides the correspondence between the object and the storage destination LINK to the user terminal UST. Even if the user does not know the physical position of the sensor node or the like, significant information can be obtained from the data storage position by making an inquiry to the directory server DRS about the target object. That is, by centrally managing the storage destination of the information with the directory server DRS, regardless of the location of the sensor node, the user terminal UST obtains measurement data or processed data related to the target sensor node when accessing the directory server DRS. Can do it.

Then, the directory server DRS converts the data obtained from the distributed data processing server DDS into information understandable by the user based on the attribute-specific semantic interpretation list ATL, and provides it to the user terminal UST.
The objects stored in the directory server DRS are set / changed according to the system configuration to be constructed, and do not change with time unlike the measurement data detected by the sensor node. The management part is not affected by the load fluctuation of the measurement data over time. Therefore, since the data of the sensor node is directly exchanged with the distributed data processing server DDS, the load on the network NWK-1 connected to the directory server DRS can be prevented from becoming excessive.

  6 and 7 show the case where the disk devices DSK are connected to the respective distributed data processing servers DDS, but one distributed data processing server DDS is provided as shown in FIG. In addition, a plurality of disk devices DSK may be provided, and the disk devices DSK may be connected to a group of a plurality of distributed data processing servers DDS.

<Relationship between measurement data and events>
Next, FIG. 9 and FIG. 10 show the relationship between the measurement data collected by the distributed data processing server DDS and the event action based on the measurement data.

  In FIG. 9, the event action control unit EAC of the distributed data processing server DDS includes an event table ETB that associates measurement data collected from the base station BST with events.

  As shown in FIG. 10, the event table ETB includes a data ID (corresponding to the data IDs in FIGS. 12 and 14) assigned to each sensor node and given to the measurement data, and an event occurrence determination condition regarding the measurement data. One record is composed of a certain EVT and a data storage DHL for determining whether or not measurement data is stored in the database DB.

  For example, in the figure, when the measurement data with the data ID “XXX” has a value greater than A1, the occurrence of an event is notified to the directory server DRS. The measurement data with the data ID “XXX” is set to be written to the disk device DSK every time the data arrives.

  In the distributed data processing server DDS, the measurement data received from the base station BST is first received by the sensing data ID extraction unit IDE, and a data ID that is an ID given to the measurement data is extracted. Further, the sensing data ID extraction unit IDE sends the measurement data to the latest data memory LDM.

  The extracted data ID is sent to the event search unit EVS, and the event table ETB is searched. If there is a record with the matching data ID, the event content EVT and measurement data of the record are sent to the event occurrence determination unit EVM.

  The event occurrence determination unit EVM compares the value of the measurement data with the event content EVT, and notifies the occurrence of the event to the directory server DRS through the directory server interface DSI if the condition is satisfied. In addition, the event occurrence determination unit EVM transmits a request for data storage DHL to the latest data memory.

  The DB control unit DBC receives data from the latest data memory LDM and writes it to the disk device DSK for data for which the data storage DHL of the event table ETB is YES.

  When the directory server interface DSI receives a measurement data reference request from the directory server DRS, the distributed data processing server DDS sends the access request to the data access reception unit DAR.

  If the access request is the latest data, the data access receiving unit DAR reads the measurement data corresponding to the data ID included in the access request from the latest data memory LDM and returns it to the directory server interface DSI. Alternatively, if the access request is past data, the measurement data corresponding to the data ID included in the access request is read from the disk device DSK and returned to the directory server interface DSI.

  As described above, in the distributed data processing server DDS, the latest data among the sensor node data collected from the base station BST is held in the latest data memory LDM, and only data that is expected to be referred to later is stored. Record in the disk device DSK. In addition, it is possible to set only data when an event occurs to record data in the disk device DSK. In this case, it is possible to prevent an increase in disk usage due to data collected periodically (observation intervals). With the above method, it is possible to manage a plurality of base stations BST (that is, a large number of sensor nodes) with one distributed data processing server DDS.

<Details of Device Manager NMG and Model Manager MMG>
<Device Manager NMG>
FIG. 11 shows details of the device management unit NMG, the model management unit MMG, and the real world model table MTB of the directory server DRS shown in FIG.

  First, the device management unit NMG of the directory server DRS searches a sensor information table STB for managing sensor nodes, a registration interface for registering / changing sensor nodes in the sensor information table STB, and the contents of the sensor information table STB. It has a search interface. Here, the sensor information table STB is managed by the sensor manager terminal ADT-A.

  As shown in FIG. 12, the sensor information table STB includes a data ID assigned in advance for each sensor node, a sensor type indicating the type of the sensor node, a meaning indicating the measurement target of the sensor node, and a measurement measured by the sensor node. The contents of the value, the installation location indicating the position (or target) where the sensor node is installed, the observation interval indicating the period at which the sensor node detects the measurement value from the measurement target, and the storage location of the measured data (distributed data processing server) A data storage destination indicating storage positions on DDS-1 to DDS) is composed of one record, and is managed using an ID for identifying a sensor node as an index.

  For example, the tag TG-1 of the wireless sensor node MSN-1 configured as the name tag shown in FIG. 5 is assigned the data ID 01, and the measurement target is the location (position) of the wireless sensor node MSN-1. The measurement cycle is every 30 seconds, indicating that the measurement data is stored in the distributed data processing server DDS-1. Similarly, the sensor SSR-1 arranged in the wireless sensor node MSN-1 configured as the name tag is assigned the data ID 02, the measurement target is the ambient temperature, and the measurement cycle is every 60 seconds. The measurement data is stored in the distributed data processing server DDS-2.

  The sensor information table STB is set from the sensor administrator terminal ADT-A, and the sensor administrator and the service administrator refer to the sensor information table STB, so that the function, position, and measurement data of the sensor node are obtained. You can know where to store.

  If the sensor node measures the data at a different period, the sensor interval is set to “event” as in the seating sensor with sensor node ID = 03 in FIG. Only when a specific state is detected, this state is notified to the distributed data processing server DDS.

<Model Manager MMG>
Next, the model management unit MMG and the real world model table MTB shown in FIG. 11 will be described.

  The real world model table MTB managed by the model management unit MMG includes an attribute-specific semantic interpretation list ATL for interpreting the meaning of the measurement data of the sensor node and the objects OBJ-1 to OBJ-1 shown in FIG. The real-world model list MDL indicating the correspondence between the model name of n and the actual information storage position, and the model binding list MBL indicating the correlation between the objects OBJ-1 to n.

  Then, the model management unit MMG manages each list of the real world model table MTB, so that the attribute-specific semantic interpretation list management unit ATM that manages the attribute-specific semantic interpretation list ATL and the real-world model list MDL that manages the real-world model list MDL. A world model list management unit MDM and a model binding list management unit MBM for managing the model binding list MBL are provided. Each management unit is used to register / change a list, and to search each list. Each has a search interface.

  Here, it is assumed that the real world model table MTB is managed by a service manager using the service manager terminal ADT-B. In FIG. 11, the sensor manager terminal and the service manager terminal may use one management terminal ADT as shown in FIG.

  In addition, the user terminal UST using the sensor network searches the object OBJ from a desired list via the search interface of each list.

  First, the attribute-specific semantic interpretation list ATL managed by the attribute-specific semantic interpretation list management unit ATM includes the return values (measured values) from the sensor nodes WSN, MSN, and FSN and the processed data converted by the distributed data processing server DDS as they are. Since a user of the user terminal UST (hereinafter simply referred to as a user) cannot understand the value of, a table for converting the output value of the sensor node into significant information is provided as shown in FIG. FIG. 13 is preset corresponding to the objects OBJ-1 to n.

  In FIG. 13, the name table ATL-m is associated with Mr. Suzuki's position OBJ-1 shown in FIG. 6, and as shown in FIG. 12, the sensor type MSN-1 whose sensor type is the name tag. The personal name corresponding to the return value (measured value) from the identifier set in the set tag TG is associated with the meaning column.

  In FIG. 13, a place table ATL-p is a table showing the position of the employee wearing the name tag, and the place name corresponding to the return value (for example, the ID of the base station to which the sensor node is connected) is in the column of meaning. It is associated. For example, a return value of 01 means that the location is an office.

  Moreover, the seating table ATL-s of FIG. 13 shows the seating state of the chair in the office shown in FIG. 5 or the conference room A, and is provided for each chair (each wireless sensor node WSN-3 to 10). The seated state (present or absent) corresponding to the return value (measured value) of the wireless sensor nodes WSN-3 to 10 is stored. For example, when the return value is 00, it indicates that the user is present (sitting), and when the return value is 01, it indicates that the user is absent.

  Similarly, the temperature table ATL-t in FIG. 13 is a table showing values indicated by the temperature sensors (MSR-1, SSR-1, FSN-1, 2) shown in FIG. A function f (x) for converting measured data) to temperature y is stored in the meaning column.

  In FIG. 13, the number of people table ATL-n is a table showing the number of employees in the A meeting room, and the return value (the number of seats in the A meeting room chair sensor or the number of mobile nodes MSN in the A meeting room) is shown. The corresponding number of people is associated with the meaning column.

  As described above, the attribute-specific semantic interpretation list ATL is a list that defines the meaning of the measurement data, and each table is set according to the generated object.

  Next, the real world model list MDL is set in advance by a service administrator or the like. As shown in FIG. 14, the position of information corresponding to the model name set for each object is stored in the information link destination. The That is, the model name, the information link destination, and the data ID that are paired constitute the real world model list MDL.

  The directory server DRS manages only significant information understandable by the user using the model list MDL, and the location of this significant information is one of the distributed data processing servers DDS-1 to n. For this reason, in the object OBJ defined in the model list MDL, the location of significant information entity is set in the information link destination. This information link destination is set in advance by a service administrator or the like. Similarly, the data ID is a value corresponding to sensor data (data obtained directly from a sensor node or data obtained by processing) that is a source of an object value.

  In FIG. 14, for example, an information link destination LINK-1 is stored for Mr. Suzuki's position OBJ-1, and a URL, a path, and the like are stored in this information link destination. When an object is searched, significant information (object entity) can be obtained from the information link destination.

  For example, when a keyword or the like is transmitted from the user terminal UST to the search engine SER of the directory server DRS, a list of model names including the keyword from the model names of the model list MDL is returned from the search engine SER. When the user operating the user terminal UST selects a desired model name, first, the directory server DRS acquires data corresponding to the information link destination from the distributed data processing server DDS set to the information link destination LINK.

  The directory server DRS converts the acquired data into information understandable by the user based on the attribute-specific semantic interpretation list ATL, and transmits the information to the user terminal UST.

  Therefore, the user can acquire necessary information as information that can be recognized without knowing the knowledge and location of each sensor node.

  In the distributed data processing server DDS, since it is not necessary to convert the data collected from the sensor nodes into a form that can be understood by the user each time the data is collected, the distributed data processing server that collects and manages data of a large number of sensor nodes. DDS load can be greatly reduced. This data conversion process is performed by the directory server DRS as required based on a request from the user, so that unnecessary conversion processes can be suppressed and the sensor network resources can be functioned without waste. Become.

  Next, the model binding list MBL indicating the correlation between the objects OBJ-1 to n is a collection of information related to elements common to the objects OBJ of the real world model list MDL.

  As an example of the model binding list MBL, as shown in FIG. 15, among the objects OBJ of the real world model list MDL, the common element is related to “person name” (“Mr. Suzuki” in the figure) and “Meeting room A”. It is what extracted what to do. For example, as the object OBJ related to the personal name “Mr. Suzuki” registered in the meaning column of the name table ATL-m of the attribute-based semantic interpretation list ATL in FIG. 13, the position OBJ-1 and the seated state OBJ- 2. There is a temperature OBJ-3, and the link destination of the object related to the name of Mr. Suzuki is “Position” LINK-1, “Sitting state” LINK-2, “Temperature” LINK-3 and a tree shape as shown in the figure This is set as a model binding list MBL-P related to the name of a person.

  Similarly, when the real world model list MDL is viewed from the element A meeting room, there are objects OBJ-4 to 6 of “members”, “number of persons”, and “temperature”, and information on objects related to the place of the meeting room A Link destinations LINK-4 to 6 are set in a tree shape with “members”, “number of people”, and “temperature” as shown in the figure, and this is used as a model binding list MBL-R for the conference room A.

  As described above, the model binding list MBL is information obtained by associating different information of common elements among the elements of the objects of the real world model list MDL. The association of the model bind list MBL is set in advance by a service administrator or the like.

<Operation of Model Manager MMG>
Next, the operation of the sensor network system will be described below.

<Register sensor node>
First, a sensor node registration procedure will be described with reference to FIGS. 16 and 17. The sensor administrator installs the sensor node in a predetermined place or person, and then registers the sensor node in the directory server DRS according to the time chart of FIG.

  In FIG. 16, first, the sensor administrator connects to the directory server DRS from the sensor administrator terminal ADT-A, and calls the registration interface of the device management unit NMG. Then, according to the data format shown in FIG. 17 from the sensor administrator terminal ADT-A, a newly added data ID, sensor type, attribute, measurement value, installation location, observation interval, data storage destination are set, and the directory server DRS A registration request is transmitted to the device management unit NMG (RG-1). Here, prior to registration, it is assumed that a data storage destination is secured and an attribute is specified for the distributed data server DDS that receives sensor node data.

  Upon receiving this registration request, the device management unit NMG of the directory server DRS adds the information of the sensor node that has made the registration request to the sensor information table STB shown in FIG. Then, the device management unit NMG assigns an identifier ID to the newly added sensor node. The sensor node ID may be assigned from the sensor administrator terminal ADT-A.

  The device management unit NMG assigns the storage location of the measurement data of the sensor node requested to register to the distributed data processing server DDS designated as the data storage location, and then assigns one record of the sensor information table STB. Finalize.

  Then, the device management unit NMG returns a completion notification (ACK) indicating that a new record has been added to the sensor administrator terminal ADT-A, and ends the registration process.

  Although not shown in the figure, the distributed data processing server DDS that has received the registration notification of the sensor node from the directory server DRS has a predetermined ID for the sensor node having the ID corresponding to the base station BST corresponding to the “installation location” in FIG. Command to detect the measured value from the sensor node at the observation interval. In the sensor management unit SNM of the base station BST, the commanded data ID and the observation interval are registered.

  As a result, the new sensor node can communicate with the base station BST to which it belongs and transmit measurement data to the distributed data processing server DDS to which this sensor node belongs.

<Object definition>
Next, processing for generating a relationship between measurement data of a sensor node and an object for the sensor node registered in the directory server DRS in FIGS. 16 and 17 will be described with reference to FIG. This process is performed by a sensor network service administrator.

  In FIG. 18, the service manager connects to the directory server DRS from the service manager terminal ADT-B, and calls the search interface of the device management unit NMG. Then, a desired sensor node is searched based on the ID or the like, and the sensor node that matches the search condition is returned to the service manager terminal ADT-B.

  The service manager terminal ADT-B outputs the sensor node search result received from the device management unit NMG to a display device (not shown).

  The service manager selects a desired sensor node from the sensor nodes displayed on the service manager terminal ADT-B, designates an object associated with the measurement data of the sensor node, and sends it to the model management unit MMG of the directory server DRS. sign up.

  For example, the object OBJ-1 “Mr. Suzuki's position” is registered as the object of the name tag type sensor node (MSN-1 in FIG. 5) with ID = 01 in the sensor information table STB shown in FIG. By this registration, a real world model list (MDL) indicating the relationship between the object and its information link is generated (FIG. 14).

  Then, for the object OBJ-1 "Mr. Suzuki's position", the model management unit MMG determines the position of the base station BST that received the sensor node MSN whose tag ID is TG-1 (Mr. Suzuki's identifier), for example, A predetermined command is instructed to the distributed data processing server DDS-1 so as to be stored in the distributed data processing server DDS-1.

  In the distributed data processing server DDS-1 that has received the command, when the event action control unit EAC receives the data of TG-1 in which the tag ID indicates Mr. Suzuki, a value for discriminating the position of the received base station BST Is registered in the database DB of the distributed data processing server DDS-1.

  Then, an information storage destination corresponding to the object OBJ-1 in the real world model list MDL is set for the data entity “Position of Mr. Suzuki” stored in the database DB of the distributed data processing server DDS-1. .

  Alternatively, for the object OBJ-2 “Mr. Suzuki”, the model management unit MMG sets the value “00” if the measured value of the wireless sensor node WSN-0 including the pressure sensitive switch as the sensor SSR is ON. The distributed data processing server DDS-1 writes data in the database DB, and if the measured value of the wireless sensor node WSN-0 is OFF, the information "01" is written in the distributed data processing server DDS-1 database DB. Command DDS-1.

  In the distributed data processing server DDS-1 that has received this command, the event action control unit EAC sets “00” or “01” (corresponding to ON / OFF, respectively) that is the measurement data value of the sensor node WSN-0. Then, a process of writing to the database DB of the data processing server DDS-1 is performed.

  In the same manner as described above, the information storage destination corresponding to the object OBJ-2 in the real world model list MDL is stored for the data entity “Mr. Suzuki” stored in the database DB of the distributed data processing server DDS-1. Is set.

  Thus, the object (information storage destination) set by the model management unit MMG and the position of the distributed data processing server DDS that actually stores information are set.

  As shown in FIG. 14, the model management unit MMG creates an object “Mr. Suzuki” OBJ-1 and stores the model name, data ID, and information storage location in the real world model list MDL. When the object registration is completed, the model management unit MMG transmits a completion notification to the service manager terminal ADT-B.

  The service administrator terminal ADT-B displays the received object generation completion notification, and in the case of generating an object, the above processing is repeated to generate a desired object.

<Definition of model binding list>
Next, setting of the model binding list MBL indicating the correlation between the plurality of objects OBJ-1 to n after generating a plurality of objects based on the definition of the model list MDL will be described with reference to FIG.

  In FIG. 19, the service manager connects from the service manager terminal ADT-B to the model manager MMG of the directory server DRS and calls the search interface of the model manager MMG. Then, a desired object is searched, and an object that matches the search condition is returned to the service manager terminal ADT-B.

  The service manager terminal ADT-B outputs the object search result received from the model management unit MMG to a display device (not shown).

  The service manager selects a desired object from the objects displayed on the service manager terminal ADT-B, and generates a common element as a model binding list in the model manager MMG of the directory server DRS. Request.

  For example, as shown in FIG. 15, a person name “Mr. Suzuki” is generated as a model binding list MBL-P, and Suzuki's position OBJ-1 and Mr. Suzuki's seated state OBJ-2 are added to this model binding list MBL-P. Associating objects such as Suzuki's temperature OBJ-3.

  The model management unit MMG associates the model binding list MBL-P with the information storage destinations of the objects OBJ-1 to OBJ-3 and stores them in the model binding list MBL.

  When the registration of the model binding list MBL is completed, the model management unit MMG transmits a completion notification to the service manager terminal ADT-B.

  The service manager terminal ADT-B displays the received model binding list generation completion notification, and further generates the desired model binding list by repeatedly executing the above process when generating the model binding list. .

<Search model binding list>
Next, an example of processing in which the sensor network user refers to the sensor node data using the model bind list using the model bind list MBL set as described above will be described with reference to FIGS. .

  The user terminal UST connects to the search engine SER of the directory server DRS and requests the model binding management unit MBM to search the model binding list MBL. This search request is made by, for example, keyword search or GUI as shown in FIG.

  The model bind management unit MBM responds to the requested search result to the user terminal UST, and displays the result of the model bind list that matches the search request on a display device (not shown) of the user terminal UST.

  In the user terminal UST, the user selects an arbitrary model binding list from the search results and requests information (STEP 110).

  Here, as shown in FIG. 15, the model binding list is configured by link destinations of a tree structure in which elements common to the objects OBJ are collected, and any link displayed in the model binding list at the user terminal UST. By selecting the destination, a request for information is made to the distributed data processing server DDS at the link destination.

  The distributed data processing server DDS accesses the measurement data or processed data requested from the user terminal UST, and returns the access result to the attribute-based semantic interpretation list management unit ATM of the directory server DRS.

  In the directory server DRS, the attribute-specific semantic interpretation list management unit ATM acquires the meaning for the return value of the attribute-specific semantic interpretation list ATL shown in FIG. 13 from the ID of the measurement data transmitted from the distributed data processing server DDS ( (STEP 112).

  Next, the search engine SER of the directory server DRS returns the meaning corresponding to the measurement data analyzed by the attribute-specific semantic interpretation list management unit ATM to the user terminal UST, and the user terminal UST distributes the response from the directory server DRS. Displayed instead of the reply from the data processing server DDS.

  For example, when the link destination LINK-1 of the model bind list MBL-P in FIG. 15 is selected, the measurement data of the distributed data processing server DDS-1 preset for the Mr. Suzuki position OBJ-1 is accessed from the user terminal UST. Is called. For example, if the link destination LINK-1 is associated with the data storage destination of the sensor information table STB shown in FIG. 12, the distributed data processing server DDS is the measurement data from the database DB corresponding to this data storage destination. The measurement data of the wireless sensor node MSN-1 is read and returned to the directory server DRS.

  The directory server DRS selects the location table ATL-p of the attribute-specific semantic interpretation list ATL from the data attributes stored together with the data, and acquires the meaning corresponding to the return value (measurement data). In this case, for example, if the return value = 02, the information of the link destination LINK-1 of the model binding list MBL-P is “A meeting room”. Therefore, the response to the object OBJ-1 “Mr. Suzuki” in the model binding list MBL-P has a significant meaning for the user “A conference room” from the value of the measurement data “02” of the sensor node MSN-1. It is converted into information and displayed (or notified) on the user terminal UST. In this example, the method in which the data attribute is acquired together with the data is shown. In this case, at the time of registering the sensor node, it is only necessary to reserve the data storage destination and specify the attribute for the distributed data processing server DDS that receives data from the sensor node. As another method for acquiring the data attribute, a method of specifying the attribute for the model when registering the real world model list MDL may be used.

  FIG. 22 shows the processing of FIG. 20 performed for “Mr. Suzuki's sitting state” LINK-2 in the model binding list MBL-P of FIG. Also in this case, the return value “00” from each of the wireless sensor nodes WSN-3 to 10 is read from the distributed data processing server DDS, and the return value = “00” in the attribute-based semantic interpretation list management unit ATM of the directory server DRS. Can be returned to the user terminal UST from the search engine SER.

  FIG. 23 shows the processing of FIG. 20 performed for “Mr. Suzuki's temperature” LINK-3 in the model binding list MBL-P of FIG. Also in this case, the return value “x” from the sensor SSR-1 of each wireless sensor node MSN-1 is read from the distributed data processing server DDS and returned by the attribute-specific semantic interpretation list management unit ATM of the directory server DRS. = X is calculated as the temperature y = f (x), and can be returned to the user terminal UST from the significant information search engine SER that says “Mr. Suzuki's ambient temperature is y ° C.”.

  FIG. 24 shows the process of FIG. 20 performed for “Meeting room member A” in the model binding list MBL-R of FIG. In this case, when the object OBJ-4 of the A meeting room is generated by the model management unit MMG, the predetermined distributed data processing server DDS-1 detects the object at the base station BST-1 corresponding to the A meeting room. The tag ID of the name tag node is read into the base station BST-1 as measurement data. Then, this value is stored in the information link destination (here, distributed data processing server DDS-1) shown in FIG. 14, which is preset as a data storage destination.

  The distributed data processing server DDS-1 collects the tag IDs of the wireless sensor nodes MSN-1 to MSN-n from the base station BST-1 at a predetermined cycle, and indicates a value indicating the member of the A conference room (the tag ID of the name tag node). Update). In FIG. 24, the wireless sensor nodes MSN-1 to n collected by the distributed data processing server DDS-1 indicate that employees with tag IDs “01” and “02” are detected in the A meeting room. .

  The distributed data processing server DDS-1 transmits the processed data “01, 02” to the attribute-based semantic interpretation list management unit ATM of the directory server DRS.

  The attribute-based semantic interpretation list management unit ATM of the directory server DRS converts the received processed data from the predefined person name table ATL-m into significant information such as 01 = Suzuki, 02 = Tanaka, and the user terminal UST. Send to.

  As a result, the user terminal UST can obtain significant information such as “Suzuki and Tanaka are present in the A conference room” in response to the information request of being a member of the A conference room in the model bind list MBL-P.

  FIG. 25 shows the case where the process of FIG. 20 is performed for “number of people in the A meeting room” in the model binding list MBL-R of FIG. In this case, when the model management unit MMG generates an object OBJ-5, which is the number of people in the A meeting room, the predetermined distributed data processing server DDS-1 uses the number of people in the A meeting room, specifically for each time period. The number of IDs of the name tag nodes detected by the base station BST-1 corresponding to the conference room A or the number of ON of the seating nodes is calculated. This value is stored in the information link destination of FIG. 14 set in advance as the data storage destination of the object OBJ-5.

  The distributed data processing server DDS-1 collects the number x of IDs of the wireless sensor nodes MSN-1 to MSn-n from the base station BST-1 at a predetermined cycle, and obtains a value (processing data) indicating the number of persons in the A meeting room. Calculate and update. The distributed data processing server DDS-1 transmits the processed data x to the attribute-based semantic interpretation list management unit ATM of the directory server DRS.

  In the attribute-based semantic interpretation list management unit ATM of the directory server DRS, the received processing data is converted from the predefined number-of-people table ATL-n into significant information such as the number of people y = x, and the user terminal from the search engine SER. Send to UST.

  As a result, the user terminal UST can obtain significant information that “there are y people in the A meeting room” in response to the information request of the number of A meeting rooms in the model binding list MBL-P.

<Action control unit>
FIG. 26 is a block diagram showing details of the action control unit ACC of the directory server DRS.

  The action control unit ACC automatically performs a preset operation (action) based on the event occurrence notification received from the event action control unit EAC of the plurality of distributed data processing servers DDS.

  Therefore, the action control unit ACC analyzes the received action by the action receiving unit ARC that receives the action setting from the user terminal UST via the session control unit SES, and the directory server DRS and the distributed data processing server according to the analysis result An action analysis unit AAN that sets function (or load) sharing between DDSs, an action management unit AMG that manages definition and execution of actions, and a relationship between events and actions corresponding to setting requests from the user terminal UST are stored. An action table ATB, an event monitoring instruction unit EMN for sending a command to the distributed data processing servers DDS-1 to n so as to monitor events defined in the action table ATB, and each distributed data processing server DDS-1 to n Receive notification of events that have occurred An event receiver ERC that, based on the definition of the received event and action table ATB, composed of an action execution section ACE for executing a predetermined operation.

  The action registration will be described with reference to the timing chart of FIG. In FIG. 27, first, the user (or service manager) connects to the action control unit ACC of the directory server DRS from the user terminal UST or the like, and requests action setting. For example, as an example of the action, as shown in FIG. 28, a case will be considered in which the action of monitoring the seat of Mr. X and transmitting a pop-up notification to the user terminal UST with the IP address: A is considered.

  When the action reception unit ARC of the action control unit ACC receives the action setting request, the action reception unit ARC requests the action analysis unit AAN to set the action. The action analysis unit AAN selects, for example, the data ID to be monitored from Mr. X's seat, and determines how the measurement data of the sensor node will be generated. Here, in order to convert the real world event “Mr. X's seat” into a data ID, the “Mr. X's seat” is referred to by referring to the real world model list MDL and the attribute-specific semantic interpretation list ATL of the real world model table MTB. Search for the model.

  Here, as shown in FIG. 29, in the case of Mr. X = Mr. Suzuki, since the model is already defined in the real world model table MTB, information for storing data ID = X2 and data from the lists MDL and ATL. The storage destination (distributed data processing server DDS1) is acquired.

  Next, since the distributed data processing server DDS monitors the event occurrence of “Mr. X's seat” in the action management unit AMG, the “Ms. X's seating” is sent to the distributed data processing server DDS that manages the selected sensor node. Is sent to monitor the event occurrence. Then, the action management unit AMG sets an action “send pop-up notification to the user terminal UST of IP address: A” in the action table ATB, and sets the sensor node ID as an ID of an event for executing the action. To do.

  In the distributed data processing server DDS that has received a command from the action management unit AMG of the directory server DRS, as shown in FIG. 30, for the data ID = X2 acquired from the real world model list MDL, the seating acquired from the attribute semantic interpretation list ATL The action control unit ACC of the directory server DRS is registered in the condition “00” and the notification destination of the event to be performed as an action. The notification to the directory server DRS performed by the distributed data processing server DDS-1 is an action in the distributed data processing server DDS-1.

  That is, in the event table ETB of FIG. 30, the sensor node ID = X2 of the pressure sensor indicating “Mr. Suzuki's seat” is set in the data ID column indicating the ID of the measurement data, and the event condition column is set in the event condition column. “00” indicating seating is set, and an action of notifying the action control unit ACC of the directory server DRS is set in the action column of the distributed data processing server DDS-1.

  Further, in the action table ATB of FIG. 31, a sensor node ID = X2 indicating “Mr. Suzuki's seating” is set in the data ID column indicating the event ID to be monitored, and the distributed data is displayed in the event condition column. The event occurrence reception from the processing server DDS-1 is set, the pop-up notification to the user terminal UST is set in the action column executed by the directory server DRS, and the A of the user terminals UST is set in the action parameter column. An IP address indicating is set.

  The action registered by the action management unit AMG in the action table ATB is, as shown in FIG. 31, an event condition that the event of data ID = X2 has been received, and an action called pop-up notification, an address ( Here, the setting is made to be executed for the terminal having the IP address A).

  The action setting request screens of FIGS. 28 and 29 are provided to the user terminal UST by the action receiving unit ARC of the directory server DRS, and the real world model list MDL is associated with the name pull-down menu. The pull-down menus of “sitting”, “meeting”, and “return home” are associated with the attribute semantic interpretation list ATL, and the actions to be executed by the directory server DRS are set in the “pop-up” and “mail” pull-down menus. ing.

  As described above, a single action is performed from one event occurrence and one action is set as described above with reference to FIG. That is, an action setting request is made from the user terminal UST to the action control unit ACC of the directory server DRS, an action analysis and event monitoring instruction is generated by the action control unit ACC, and the event action of the distributed data processing server DDS An event table ETB is defined in the control unit EAC. Thereafter, the action management unit AMG of the action control unit ACC instructs the event reception unit ERC to monitor the set event (data ID = X2). As a result, the action control unit ACC notifies the user terminal UST that a series of action settings has been completed.

<Execution of action>
FIG. 33 is a time chart showing the execution of the actions set in FIG. 28 and FIG.

  When the measurement data of the monitoring target sensor node changes to “00” as the event generation condition and the seating of Mr. X is determined, the distributed data processing server DDS-1 generates an event notification regarding the data ID = X2. .

  This event occurrence is notified from the distributed data processing server DDS to the directory server DRS, and is received by the event receiving unit ERC in FIG.

  The action management unit AMG of the directory server DRS searches the action table ATB of FIG. 31 from the received event ID, and determines the presence or absence of the corresponding action. Since the received event with ID = X2 is defined in the action table ATB, the action management unit AMG notifies the action execution unit ACE of the action and parameters of the action table ATB.

  The action execution unit ACE transmits a pop-up notification to the user terminal UST with the IP address: A instructed from the action management unit AMG.

  A pop-up notification is transmitted to the user terminal UST of the IP address: A, and it can be confirmed that Mr. X's seating has been detected.

<Setting and executing multiple actions>
In FIG. 28, FIG. 29 and FIG. 33, the example of performing one action when one event occurs has been described. However, as shown in FIG. 34 to FIG. Can be set.

  34 and 35 are multiple action setting request screens. This setting request screen defines a pull-down menu that allows selection of a status such as “sitting” for each of the two name fields. The event conditions corresponding to these two names are associated with the real world model list MDL and the attribute-specific semantic interpretation list ATL in the real world model table MTB, as in FIGS.

  Furthermore, a pull-down menu for setting a logical expression (and / or) of these two event conditions is added.

  Then, like the single action, an action (pop-up notification, mail transmission) executed by the directory server DRS and a parameter column (address, etc.) necessary for executing the action are set.

  Here, when both the event occurrence of the distributed data processing server DDS-1 “Mr. Suzuki's seating” and the event occurrence of “Mr. Tanaka's seating” from the distributed data processing server DDS-2 are established, the mail is transmitted. An example of an action to do will be described.

  First, the event “Mr. Suzuki's seating” is set in the same manner as in FIG. 28 and FIG. 29, and the event table ETB of the distributed data processing server DDS-2 that monitors Mr. Suzuki's seating is shown in FIG. An event is set. FIG. 39 shows a time chart for setting the action table at this time.

  Next, as for the event “Mr. Tanaka's seating”, the sensor node ID = Y2 for detecting Mr. Tanaka's seating is set as the data ID, and the seating is indicated from the semantic interpretation list ATL, as in FIGS. “00” is set as an event condition, and when the event condition is satisfied, an action of notifying the action control unit ACC of the directory server DRS is set in the event table ETB of the distributed data processing server DDS-2 as shown in FIG.

  In the action control unit ACC of the directory server DRS, as shown in FIG. 38, two conditions are combined and set in the action table ATB with a logical expression “AND”.

  For the two conditions of the action table ATB joined by “AND”, “email transmission” is set in the action column, and the transmission destination address is set in the parameter column.

  Similarly to FIG. 32, the user terminal UST requests the action control unit ACC to set actions related to Mr. Suzuki and Mr. Tanaka, and the event monitoring instruction unit EMN sends the distributed data processing server DDS- 1 is set to notify an event when the measurement data of the sensor node with data ID = X2 reaches a predetermined condition (Mr. Suzuki's seat), and the event monitoring instruction unit EMN sends it to the distributed data processing server DDS-2. On the other hand, an event is set to be notified when the measurement data of the sensor node with the data ID = Y2 reaches a predetermined condition (Mr. Tanaka's seat).

  A new event is added to the event table ETB in each of the distributed data processing servers DDS-1 and 2, and the event occurrence determination unit EVM of each of the distributed data processing servers DDS-1 and 2 starts event monitoring for the measurement data.

  In the action management unit AMG of the action control unit ACC, the event reception unit ERC is instructed to monitor the event with the data ID = X2 and Y2, and the setting ends.

  Next, FIG. 40 is a time chart showing how actions are executed.

  First, the distributed data processing server DDS-1 generates an event with data ID = X2 with Mr. Suzuki's seating. Although the action control unit ACC receives the event with the data ID = X2, the action table ATB holds the action because it cannot be executed unless Tanaka is seated.

  Next, the distributed data processing server DDS-2 generates an event with data ID = Y2 along with Mr. Y's seat. The action control unit ACC receives the event with the data ID = Y2, and since the AND condition of the data ID = X2 and Y2 is satisfied in the action table ATB, the action is executed and the mail is transmitted to a predetermined mail address.

  In this way, an action can be executed on the condition of a plurality of events, and only a response necessary for the user can be obtained from a large number of sensors. As a result, even when there are an enormous number of sensor nodes, the user can detect desired information (or a change in information) almost in real time, and the information of the sensor nodes can be used effectively.

Second Embodiment
FIGS. 41 to 45 show the execution of a single action on the distributed data processing server DDS side. The action execution unit ACE is added to the event action control unit EAC of the distributed data processing server DDS shown in FIG. The event table ETB in FIG. 9 is replaced with the event action table EATB, and the other configurations are the same as those in the first embodiment.

  In FIG. 41, the event action control unit EAC of the distributed data processing server DDS includes an event action table EATB that associates measurement data collected from the base station BST with the event action via the directory server interface DSI.

  As shown in FIG. 43, the event action table EATB includes a data ID assigned to each sensor node and given to the measurement data, an event content column indicating a condition of measurement data for generating the event, and distributed data when the event occurs. An action field indicating the content of the action executed by the processing server DDS, a parameter field for storing a value necessary for executing the action, and a data storage for determining whether or not measurement data is stored in the database DB when an event occurs One record is composed of DHL.

  For example, in the figure, the measurement data with the data ID X1 generates an event when the value is “02”, sends a mail to the address specified in the parameter column, and writes the measurement data to the disk device DSK. Is set to not.

  In the distributed data processing server DDS, the measurement data received from the base station BST is first received by the sensing data ID extraction unit IDE, and the ID assigned to each sensor node is extracted from the measurement data, and this ID is used as the data ID. . Further, the sensing data ID extraction unit IDE sends the measurement data to the latest data memory LDM.

  The extracted data ID is sent to the event search unit EVS, and the event action table EATB is searched. If there is a record with a matching data ID, the event content and measurement data of the record are sent to the event occurrence determination unit EVM.

  In the event occurrence determination unit EVM, the value of the measurement data is compared with the event content EVT, and if the condition is satisfied, it is sent to the action execution unit ACE to execute the set action. Then, the action execution unit ACE sends the event occurrence to the latest data memory LDM and the DB control unit DBC.

  The DB control unit DBC writes the data for which the data storage DHL of the event action table EATB is YES to the disk device DSK.

  The data access receiving unit DAR is the same as in the first embodiment. If the access request is the latest data, the measurement data corresponding to the data ID included in the access request is read from the latest data memory LDM, and the directory server interface Return to DSI.

  FIG. 44 shows a time chart when setting actions in the distributed data processing server DDS. FIG. 42 shows an example of an interface transmitted by the action control unit ACC of the directory server DRS to the user terminal UST when setting actions. Indicates. When a single action is set, the directory server DRS communicates with the distributed data processing server DDS, and an action setting request from the user terminal UST is sent to the distributed data processing server DDS corresponding to the specified sensor node ID. To set.

  First, the user (or service manager) connects to the action control unit ACC of the directory server DRS from the user terminal UST or the like, and requests action setting. For example, as an example of the action, as shown in FIG. 28, an action is set in which a popup notification is transmitted to the user terminal UST of the IP address: A after monitoring the position of Mr. X and entering the conference room A. Consider the case.

  When the action reception unit ARC of the action control unit ACC receives the action setting request, the action reception unit ARC requests the action analysis unit AAN to set the action. The action analysis unit AAN selects, for example, the data ID to be monitored from Mr. X's seat, and determines how the measurement data of the sensor node will be generated. Here, in order to convert the real-world event of “Mr. X's position” into a data ID, “Mr. X's seat” is referred to by referring to the real-world model list MDL and attribute-based semantic interpretation list ATL of the real-world model table MTB Search for the model.

  Here, as shown in FIG. 29, in the case of Mr. X = Mr. Suzuki, since the model is already defined in the real world model table MTB, the data ID = X2 and the data are stored from the lists MDL and ATL. An information storage destination (distributed data processing server DDS1) is acquired.

  Next, the action management unit AMG determines whether or not the request from the user terminal UST is a single action. In the case of a single action, the requested action is assigned to the distributed data processing server of the information storage destination. Set to let DDS perform.

  The distributed data processing server DDS monitors the occurrence of an event relating to “Mr. X's position” and performs an action associated with the occurrence of the event on the distributed data processing server DDS. A command is sent to monitor the occurrence of an event that “the location of the event” matches “A meeting room”. Further, the action control unit ACC of the directory server DRS sets an action “send mail to the user of mail address: mailto_b@xyz.com” in the event action table EATB for the distributed data processing server DDS, and the action The sensor node ID is set as the ID of the event for executing the above.

  In the distributed data processing server DDS that has received a command from the action management unit AMG of the directory server DRS, as shown in FIG. 43, for the data ID = X1 acquired from the real world model list MDL, the A acquired from the attribute semantic interpretation list ATL. The e-mail address is registered in the condition “02” of the conference room and the event notification destination to be performed as an action.

  As shown in FIG. 43, the action registered by the action management unit AMG in the distributed data processing server DDS is set to execute the action of mail transmission to the address described in the parameter when the event of data ID = X1 occurs. To do.

  In this way, when there is a single action setting request from the user terminal UST, the action control unit ACC of the directory server DRS does not set the own action table ATB, but the corresponding distributed data processing server DDS. Settings are made, and both events and actions are set in the event action table EATB of the distributed data processing server DDS.

  The event action in the distributed data processing server DDS is executed as shown in FIG. 45. When Mr. X enters the conference room, the value of data ID = X1 becomes “02” and is defined in the event action table EATB of FIG. Event occurrence monitoring and associated actions are implemented. By executing the action, a predetermined mail address is notified that Mr. X has entered the A meeting room.

  In this case, the directory server DRS only sets the action for the distributed data processing server DDS, and does not need to monitor the actual event occurrence. For this reason, the collection of data and the execution of a single action are left to the distributed data processing server DDS, and the directory server DRS only needs to perform a search request from the user terminal UST and monitor a plurality of actions. When it is large, the load on the directory server DRS is prevented from becoming excessive, and the sensor network can be smoothly operated.

<Modification 1>
46 and 47 are views showing a first modification of the first or second embodiment. In this first modification, measured values from a certain sensor node are stored as raw data A in a predetermined distributed data processing server DDS, and measured values from different sensor nodes are stored as raw data B with predetermined distributed data. Store in the processing server DDS.

  In each distributed data processing server DDS, raw data A and B are processed (for example, an average value of unit time), and the processed results are further stored as processed directory server DR data A and B, respectively. The processing timing of the raw data A and B may be implemented as an action based on a predetermined condition (elapse of time) in the directory server DRS or each distributed data processing server DDS.

  Further, each distributed data processing server DDS calculates secondary data C from the processed data A and B subjected to processing as a predetermined action, and stores it as new processed data in the predetermined distributed data processing server DDS. The processed secondary data C is further stored as tertiary data C '.

  For example, when the raw data A is temperature and the raw data B is humidity, the processed data A and B are respectively the average temperature and the average humidity of unit time. Further, the discomfort index obtained from the average temperature and the average humidity can be obtained as the secondary data C, and the average value of the unit time of the secondary data C as the tertiary data C ′.

  In the first or second embodiment, the event occurrence is the measurement data. However, the generation of the event or the execution of the action can be performed from the processing data A, B, the secondary data C, or the tertiary data C ′ as described above. it can.

  As shown in FIG. 47, the measurement data (raw data) and the processed data are stored in one distributed data processing server DDS. In this case, the processed data A and B are transferred to a distributed data management server C different from the distributed data management server that manages the processed data, the processed data A and B are treated as raw data, and the tertiary data C ′ is processed data. Can be treated as

<Modification 2>
FIG. 48 shows a second modification in which the network connecting the distributed data processing server DDS is divided into a plurality of parts in the first or second embodiment, and the other configurations are the first or second. This is the same as the embodiment.

  In this example, the network to which the distributed data processing server DDS is connected is set differently depending on the frequency of referring to the measurement data.

  A directory server DRS (not shown) that has a high reference frequency of measurement data or processed data is connected to the same network 1 as the directory server DRS. Then, a distributed data processing server DDS that stores matching processing data D with relatively low reference frequency is connected to the network 2, and a distributed data processing server DDS that stores processing data E that is hardly referenced is connected to the network 3. And each network 1-3 is connected by the gateway which is not illustrated.

  By arranging the distributed data processing server DDS in this way, the access speed to the distributed data processing server DDS having frequently referenced data can be improved.

<Modification 3>
In the first or second embodiment, the storage destination of the data corresponding to the model name is set as the information link destination in the real world model list MDL of the directory server DRS. However, the response speed may be fast. If necessary, the latest value of data may be stored in addition to the information link destination.

  In this case, the data traffic between the directory server DRS and the distributed data processing server DDS increases according to the number of objects, but data from each sensor is collected in the distributed data processing server DDS at a predetermined cycle. Although the load on the network NWK-1 is increased as compared with the embodiment described above, it is possible to respond quickly to a data request from the user terminal UST, and an improvement in response can be expected.

  As described above, according to the present invention, the directory server centrally manages the location of data and includes a plurality of distributed data processing servers that collect data from sensor nodes in real time, and is distributed over the network. Data from an enormous number of sensor nodes can be accessed quickly and easily, and can be applied to a sensor network including a large number of sensor nodes.

1 is a system configuration diagram of a sensor network showing a first embodiment of the present invention. The functional block diagram of a sensor network. The block diagram which shows an example of the wireless sensor node WSN. It is a graph which shows the operation state of a wireless sensor node, and shows the relationship between time and current consumption. Explanatory drawing which shows an example of arrangement | positioning of a wireless sensor node. It is a block diagram which shows the relationship between the measurement data of an object and a sensor node, and shows the start time of measurement. It is a block diagram which shows the relationship of the measurement data of an object and a sensor node, and shows the state which predetermined time passed from the measurement start. The graph which shows the relationship between the data amount of an object, the data amount of measurement data of a sensor node, and time. The block diagram which shows the event action control part of distributed data processing server DDS. Explanatory drawing of an event table. The block diagram which shows the principal part of the directory server DRS. Explanatory drawing of a sensor information table. Explanatory drawing of the semantic interpretation list classified by attribute. The block diagram which shows the relationship between a real world model list and the distributed data processing server DDS. Explanatory drawing of a model binding list. The time chart which shows the mode of sensor information registration. Data format for sensor node registration. Time chart showing how the real world model list is registered. A time chart showing how a model bind list is registered. The time chart which shows an example of the response with respect to access to a model binding list. Explanatory drawing of a process when Mr. Suzuki's position is designated from a model binding list. Explanatory drawing of a process at the time of designating Mr. Suzuki's seating state from a model binding list. Explanatory drawing of a process when Mr. Suzuki's temperature is designated from a model binding list. Explanatory drawing of a process at the time of designating the member of A meeting room from a model binding list. Explanatory drawing of a process at the time of designating the number of A meeting rooms from a model binding list. The block diagram of action control part ACC of directory server DRS. The timing chart which shows the procedure of registration of an action table. Explanatory drawing of the action setting screen displayed on the user terminal UST at the time of registration of an action table. Similarly, an explanatory diagram of an action setting screen. Explanatory drawing which shows the entry of the event table of the distributed data processing server DDS. Explanatory drawing which shows the entry of the action table of the directory server DRS. A time chart showing the flow of setting a single action. A time chart showing the response flow of a single action. Explanatory drawing of the action setting screen displayed on the user terminal UST at the time of registration of multiple actions. Similarly, explanatory drawing of the action setting screen displayed on the user terminal UST at the time of registration of a plurality of actions. Explanatory drawing which shows the entry of the event table of distributed data processing server DDS-1. Explanatory drawing which shows the entry of the event table of distributed data processing server DDS-2. Explanatory drawing which shows the entry of the action table of the directory server DRS. A time chart showing the flow of setting multiple actions. The time chart which shows the flow of response of a plurality of events. The block diagram which shows 2nd Embodiment and shows the event action control part of distributed data processing server DDS Explanatory drawing of the action setting screen displayed on the user terminal UST at the time of action registration. Explanatory drawing which shows the entry of the event action table of distributed data processing server DDS. A time chart showing the flow of action settings. The time chart which shows the flow of the response of a some event and action. The system block diagram of the sensor network which shows a 1st modification. Similarly, the system block diagram of the sensor network which shows a 1st modification. The system block diagram of the sensor network which shows a 2nd modification.

DRS directory server DDS distributed data processing server WSN, MSN, FSN sensor node BST base station MMG model management unit MTB real world model table NWK-1, 2-n network

Claims (4)

A distributed server for storing data transmitted from a plurality of sensor nodes;
A plurality of distributed servers and a management server connected to user terminals via a network,
The distributed server is
A data processing unit that receives a plurality of types of sensor data transmitted from the plurality of sensor nodes, and processes the sensor data in a predetermined processing method;
An accumulation unit that accumulates machining data processed by the data processing unit;
The management server
A model list storing a plurality of model names set in advance, and an information storage destination for specifying which distributed server is a storage destination of data corresponding to the model name, and data of the sensor node Common to a plurality of model names among elements constituting the model name and a semantic interpretation list in which conversion definitions for converting the information into information understandable by a user using the user terminal are set corresponding to the model name A model management unit for managing a model binding list for storing definitions associated with information storage destinations of the model list for different types of data,
A search unit that receives a search request from the user terminal and responds to the user terminal with an information storage destination of data including the model name included in the search request;
A semantic interpretation unit that converts the data received from the distributed server into meaningful information that can be understood by a user using a user terminal, and responds to the user terminal on behalf of the distributed server;
The search unit
The search request is received from the user terminal, the model name element included in the search request is searched from the model binding list, the corresponding information storage destination is extracted, and the model name element included in the search request is Extracting the matching conversion definition from the semantic interpretation list, requesting data corresponding to the model name to one of the plurality of distributed servers based on the information storage destination;
One of the plurality of distributed servers is:
Based on the request from the search unit, the processing data is acquired from the storage unit and transmitted to the management server,
Is the semantic interpretation of the previous SL management server,
A sensor network system, wherein data received from one of the plurality of distributed servers is converted into significant information based on the extracted conversion definition, and the significant information is transmitted to the user terminal. The sensor network system according to claim 1,
The management server
Among the sensor nodes, a sensor information table for setting the distributed server for storing data in the storage unit ,
Instructs the distributed server set as the storage location of the sensor node data in the sensor information table to accumulate the sensor node data,
A distributed server for storing the data is allocated to the sensor node;
The sensor node transmits the data to the distributed server of the storage destination,
The sensor network system according to claim 1, wherein the distributed server receives the command and accumulates data only from the sensor nodes set in the sensor information table. A distributed server for storing data transmitted from a plurality of sensor nodes, and a management server connected to the plurality of distributed servers and user terminals via a first network, and transmitted from the plurality of sensor nodes. To retrieve data from a user terminal,
A first procedure of storing data transmitted from the plurality of sensor nodes in a storage unit of a distributed server;
The management server sets a model list that stores a plurality of preset model names and an information storage destination that identifies which distributed server is a storage destination of data corresponding to the model name A second procedure to
A third procedure in which the management server sets a semantic interpretation list in which a conversion definition for converting data of the sensor node into information understandable by a user using the user terminal is set corresponding to the model name; ,
Model binding in which the management server stores definitions that include elements common to a plurality of model names among the elements constituting the model name and that are associated with the information storage destination of the model list for different types of data A fourth step to set up the list;
The distributed server receives a plurality of types of sensor data transmitted from the plurality of sensor nodes, processes the sensor data using a predetermined processing method, and stores the processed data in the storage unit. A fifth procedure to accumulate;
The management server accepts a search request from the user terminal, searches the model binding list for an element of the model name included in the search request, extracts a corresponding information storage destination, and is included in the search request A sixth procedure for extracting the conversion definition having the same model name element from the semantic interpretation list and requesting data corresponding to the model name to one of the plurality of distributed servers based on the information storage destination; ,
One of the plurality of distributed servers, based on a request from the management server, obtains the processed data from the storage unit and transmits the processed data to the management server;
Eighth steps before SL management server, the data received from one of said plurality of distributed servers based on the conversion definition the extracted and converted into meaningful information, and transmits the significant information to the user terminal When,
A method for retrieving sensor data, comprising: A method for retrieving sensor data according to claim 3,
The management server has a sensor information table for setting the distributed server for storing data in the storage unit among the sensor nodes, and the distributed server set in the sensor node data storage destination in the sensor information table To instruct the sensor node to accumulate data of the sensor node,
The management server assigns a distributed server for storing the data to the sensor node;
Further including
The fifth procedure includes:
The sensor node transmits the sensor data to the distribution destination distributed server, and the distributed server receives the command and receives data only from the sensor node set in the sensor information table. Search method for sensor data.

Images