We use proprietary and third party's cookies to improve your experience and our services, identifying your Internet Browsing preferences on our website; develop analytic activities and display advertising based on your preferences. If you keep browsing, you accept its use. You can get more information on our Cookie Policy
Cookies Policy
FIWARE.ArchitectureDescription.IoT.Backend.ThingsManagement - FIWARE Forge Wiki

FIWARE.ArchitectureDescription.IoT.Backend.ThingsManagement

From FIWARE Forge Wiki

Jump to: navigation, search

Contents

Copyright

Legal Notice

Please check the following Legal Notice to understand the rights to use these specifications.


Overview

Data model outline

The IoT Things Management GE (TM GE for short) is the part of the IoT Backend which is responsible for Thing-level interaction. The underlying data model of this GE is based on the OMA NGSI Context Management Information Model which relies on the concept of context entities, which are generic entities whose state is described based on values of attributes and associated metadata. In the context of IoT, context entities and context entity attributes can be used to model IoT resources and the variables they measure, respectively, but also - and more importantly - arbitrary physical objects (Things) like rooms, persons, etc. and their attributes like temperature, geo-location, etc.

Functionality outline

The TM GE is the point of contact for accessing information about entities and their attributes. The northbound interface supports one-time queries and continuous subscriptions for that kind of context information. Applications or other GEs can further receive unsolicited information updates from the TM GE about both the availability of context information and actual context information. In the FI-WARE architecture, the Publish/Subscribe Broker GE takes the role of the Application consuming context information from the TM GE.

Using the FI-WARE NGSI-9 interface that the TM GE provides, applications will be able to register, query for and subscribe to updates on context availability information, that is:

● Information about IoT resources and the variables they measure

● Information about Things and their attributes

● Information about Associations establishing how attributes of Things can be derived from attributes of other Things or from variables measured by IoT resources

The TM GE implements the subset of the FI-WARE NGSI interfaces required to support integration with the FI-WARE Publish/Subscribe Broker GE so that context information about Things becomes accessible to applications. Actually, using the FI-WARE NGSI-10 interface that the Publish/Subscribe Broker GE provides, applications will be able to:

● Query about attributes of Things

● Subscribe about updates on attributes of Things

Note that the Publish/Subscribe Broker GE may manage context information not necessarily provided by the TM GE, therefore linked to the Internet of Things, but gathered from other parts of the application.

The context information and context availability information provided by the TM GE is either (a) forwarded from IoT Agents exporting FI-WARE NGSI interfaces, or (b) derived from lower-level context information inside the TM GE. In case (a), IoT Agents can be both the Data Handling GE in IoT Gateways, or the Backend Device Management GE. It is also possible that the context information is provided by other IoT Backend systems. In case (b), the TM GE applies basic mapping rules derived from associations registered through the FI-WARE NGSI-9 interface that the TM GE exports. It may also apply more complex rules based on inference engine capabilities it will support.


Things Management GE and the interacting components

Basic Concepts

The Things Management GE is based on the OMA NGSI context data model.

FI-WARE NGSI

The TM GE the FI-WARE NGSI Context Management specifications (FI-WARE NGSI specifications for short).

FI-WARE NGSI Context Management specifications are based on the NGSI Context Management specifications defined by OMA (Open Mobile Alliance). They take the form of a RESTful binding specification of the two context management interfaces defined in the OMA NGSI Context Management specifications, namely NGSI-9 and NGSI-10 (see FI-WARE NGSI Open RESTful API Specification (PRELIMINARY)).They also solve some ambiguities in the OMA specs and extend them when necessary to implement the FI-WARE Vision.

You can visit the FI-WARE NGSI Context Management tutorial to learn the main concepts underlying FI-WARE NGSI Context Management specifications.

Associations in FI-WARE NGSI-9

Background

OMA NGSI-9 is an interface for exchanging information on the availability of context information. The central information container of NGSI-9 is a data structure called ContextRegistration. Each ContextRegistration instance contains a list of entities, a list of attributes, the URI of an application where context information about these entities/attributes is available, and a list of metadata.

In NGSI-9 is implicitly assumed that the providing application exposes an NGSI-10 interface, where the context information can be queried. It is also implicitly assumed that the context information is provided using the same entity/attribute combinations. For example, a ContextRegistration instance containing the entity "Car_2", attribute "speed", and URI "http://cars.info" represents the information that the speed of "Car_2" can be queried at "http://cars.info" by an NGSI-10 queryContext operation asking for attribute "speed" of entity "Car_2".

Enhanced Associations

In the following paragraphs we describe a generalization of this principle. We introduce a special type of metadata which modifies the query that has to be invoked at the providing application in order to obtain the desired information. In NGSI, each piece of metadata has a name, a type and a value. The metadata introduced at this point has name "knownAs" and type "QueryContextRequest". The metadata value represents the necessary query message for the providing application: When invoking the query context operation on the providing application using this message, the returned result will be the context information described in the ContextRegistration structure.

Before formally defining this special type of ContextRegistration instances, let us walk through an example. Assume that a ContextRegistration contains one entity ID "Car_2", one attribute "speed", providingApplication is "http://cars.info", an there is a piece of metadata as described above. The metadata value is an instance of QueryContextRequest which contains one entityId "speedometer_38" and one attribute "measurement". The information described by the ContextRegistration instance just described is as follows: in order to find out the speed of Car_2, one has to go to "http://cars.info" and query for the "measurement" of "speedometer_38".

An instance of ContextRegistration as the one just described is called an Association. Formally, a ContextRegistration instance is an Association, if

  • It contains exactly one entity ID and exactly one attribute. This attribute is not a domain name, and the entity ID is no pattern.
  • It contains a piece of metadata with name "knownAs" and type "QueryContextRequest".
  • The value of the "knownAs" metadata is an instance of QueryContextRequest which contains exactly one entity ID and one attribute. The entity ID is no pattern, and the attribute is no domain.

TM GE Architecture

This section describes the internal architecture of the FI-WARE TM GE implementation.

IoT TM GE internal architecture

The main components of the TM GE are described in the following sections.

IoT Broker

The IoT Broker is the point of contact for accessing Thing-level information via the northbound NGSI-10 interface.

When requests arrive through this interface, the IoT Broker first discovers the IoT resources that can provide the required information. The discovery is typically realized by interacting with the Configuration Management via the NGSI-9 interface. Following the discovery, the IoT Broker queries the discovered information sources through its southbound NGSI-10 interface.. In the FI-WARE Architecture, these information sources are Data Handling GE instances on IoT Gateways or the information of interest is provided by the Backend Device Management GE. Subscriptions through the northbound interface are handled similarly.

The IoT Broker can also receive NGSI-10 updates via the southbound interface. In this case, the received information is forwarded through the northbound interface to a fixed information sink. In the FI-WARE architecture, this information sink is the Publish/Subscribe Broker GE.

Configuration Management

The Configuration Management component implements a context information registry in which context provider applications can be registered. In addition, components interacting with the Configuration Management can perform discovery operations on that context registration information or subscribe to changes on it.

It detail, the component provides the following functionality;

● Allow IoT Broker discovery and subscription of context information, through the internal NGSI-9 interface exposed by the Configuration Management..

● Announcing the availability of context information in the towards the Publish/Subscribe Broker GE (or any other component supporting the NGSI-9 interface) through the northbound NGSI-9 interface.

The Configuration Management component stores the context information in a Configuration Repository described in next section. This repository is used to answer NGSI-9 request in the exposed interfaces described above.

Receiving registrations from IoT Gateways and the Thing-level Adapter through the NGSI-9 southbound interface and storing this information in the Configuration Repository , thus updating the context information registry.

Configuration Repository

The Configuration Repository stores information on the availability of context information and can be accessed through the Configuration Management. When the IoT Broker receives a context query, its first step is to use the queryContextAvailability operation on the Configuration Management in order to find out where the desired context information can be found. The Configuration Management returns a number of ContextRegistration instances, which can possibly be associations.

By this approach, the TM GE can maintain higher-level context information that is not available in the IoT Gateways or Devices. For example, a Gateway might not know the concept of cars, but only maintains a list of sensors and their measurements. The information about which sensors provides information about which car could be maintained only by the TM GE (moreover, the relationship between sensors and cars could have been automatically derived by the Discovery Engine component described in next section). Still taking as example the cars, when an application queries the status of a certain car, the TM GE determines what sensors provide that information, queries those sensors' measurements, and returns the measured values to the application.

Discovery Engine (PRELIMINARY)

The discovery engine is a component that makes use of machine learning (ML) techniques for resolving object descriptions, whereby objects can refer to Things or Resources. The component uses probabilistic ML techniques to group objects with similar descriptions. Its purpose is to allow automated discovery, ranking and recommendation of objects. The probabilistic machine-learning approach is completely unsupervised. It provides an efficient mechanism for publication and discovery by: · Automatically clustering all the object descriptions within a repository, hence allowing the Discovery Engine to be scalable to large object repositories. · And automating the ranking of results in order of relevance. It is also Interoperable between different types of description technologies as it bases its technique on text analysis, as long as wrappers are provided for a particular description format. With regards to the Discovery Engine’s role in the Things Management GE, it will first retrieve available NGSI-9 descriptions from the repository, which in this case is the Configuration Repository. The Discovery Engine will use the NGSI-9 descriptions to extract the link for a semantic description of the object. This is description is termed as the Advanced Description. The Advanced Description is essentially an RDF description based on W3C SSN and IoT-A IoT models. It will be used for extracting semantic concepts which will be used for training the engine. The discovery engine will also host a repository based on the “Linked IoT Data Platform” for storing the Advanced Descriptions. The Advanced Description will also enable the dynamic binding of spatial, temporal, and thematic associations between Things and Resources.

When upon startup, there are no descriptions; the Discovery Engine will wait until an acceptable number of descriptions have been stored in order to start the training process and defining the clusters. In the case when discovery requests are received by the Configuration Management from the IoT Broker via NGSI-9, the request can be forwarded to the discovery engine to provide a recommended, ranked set of results, especially in the case where a request providing an ID/attribute is not exactly matched with the descriptions in the repository.

Geo-Discovery (PRELIMINARY)

Geo-Discovery is an optional component which is specialized to perform fast and robust discovery of all entities (possibly of a certain type) within given geographic scopes. To this end it communicates with the IoT Broker and/or the Configuration Management. Details are still under discussion.

Additional Concepts

Enhanced Associations by OMA NGSI-9

Background

OMA NGSI-9 Normal 0 21 false false false FR ZH-CN AR-SA

is an interface for exchanging information on the availability of context information. The central information container of NGSI 9 is a data structure called ContextRegistration. Each ContextRegistration instance contains a list of entities, a list of attributes, the URI of an application where context information about these entities/attributes is available, and a list of metadata.

In NGSI-9 is implicitly assumed that the providing application exposes an NGSI-10 interface, where the context information can be queried. It is also implicitly assumed that the context information is provided using the same entity/attribute combinations. For example, a ContextRegistration instance containing the entity "Car_2", attribute "speed", and URI "http://cars.info" Normal 0 21 false false false FR ZH-CN AR-SA represents the information that the speed of "Car_2" can be queried at Normal 0 21 false false false FR ZH-CN AR-SA "http://cars.info" by an NGSI-10 queryContext operation asking for attribute "speed" of entity "Car_2". .

Enhanced Associations

In the following paragraphs we describe a generalization of this principle. We introduce a special type of metadata which modifies the query that has to be invoked at the providing application in order to obtain the desired information. In NGSI, each piece of metadata has a name, a type and a value. The metadata introduced at this point has name "knownAs" and type "QueryContextRequest". The metadata value represents the necessary query message for the providing application: When invoking the query context operation on the providing application using this message, the returned result will be the context information described in the ContextRegistration structure.

Before formally defining this special type of ContextRegistration instances, let us walk through an example. Assume that a ContextRegistration contains one entity ID "Car_2", one attribute "speed", providingApplication is "http://cars.info", an there is a piece of metadata as described above. The metadata value is an instance of QueryContextRequest which contains one entityId "speedometer_38" and one attribute "measurement". The information described by the ContextRegistration instance just described is as follows: in order to find out the speed of Car_2, one has to go to "http://cars.info" and query for the "measurement" of "speedometer_38".

An instance of ContextRegistration as the one just described is called an Association. Formally, a ContextRegistration instance is an Association, if

● It contains exactly one entity ID and exactly one attribute. This attribute is not a domain name, and the entity ID is no pattern.

● It contains a piece of metadata with name "knownAs" and type "QueryContextRequest".

● The value of the "knownAs" metadata is an instance of QueryContextRequest which contains exactly one entity ID and one attribute. The entity ID is no pattern, and the attribute is no domain.

Usage in the IoT TM GE

The Configuration Repository stores information on the availability of context information and is made accessible via the Configuration Management. When the IoT Broker receives a context query, its first step is to use the queryContextAvailability operation on the Configuration Management in order to find out where the desired context information can be found. The Configuration Management returns a number of ContextRegistration instances, which can possibly be associations.

By this approach, the TM GE can maintain higher-level context information that is not available in the IoT Gateways or Devices. For example, a Gateway might not know the concept of cars, but only maintains a list of sensors and their measurements. The information about which of the sensors provides information about which car could be maintained only in the backend, automatically derived by the Discovery Engine. When applications query the status of a certain car, the TM GE will find determine the sensors providing that information and, query for these sensors' measurements, and return the measured values to the application.

More examples of how associations are used can be found below .

Main Interactions

The TM GE has a number of interfaces for interacting with applications, users, and other GEs of the FI-WARE architecture. In that context, the GE communicates with the Publish/Subscribe Broker GE via the Northbound Interface and with the IoT Agents on the Southbound Interface. The role of IoT Agent can be played by either the Backend Device Management GE, or by the Gateway Data Handling GE.

Reception of RegisterContext operations from Agents

In order for their information on Things and/or Resources to be available at the Backend, IoT Agents need to register their information to the TM GE. This is done via the RegisterContext operation. Note that the registration can be performed on various levels of granularity: registering specific Entity/Attribute combinations, registering that information about certain entity type is available, or even in general registering that some entity information is available (without getting more specific) are all possible registrations. Also note that in absence of a Publish/Subscribe GE instance the registration is still stored in the TM GE. File:registration.jpg

Query Handling

Queries are one-time requests for information. They are realized by the queryContext operation of OMA NGSI-10. File:Query.jpg

Subscription Handling

Subscriptions are requests for information updates the issuer wishes to receive under conditions that have to be specified in the request message. The picture below shows the interaction diagram for the OMA NGSI-10 subscribeContext operation, but the same interaction pattern applies to the updateContextSubscription and unsubscribeContext operations. File:Subscription.jpg

Notification

Notifications are the counterpart of subscriptions. A notification is sent whenever the condition for it (that has been specified in the subscription) is satisfied. File:Notification.jpg

Personal tools
Create a book