1. On Management, Abstraction & Semantics
Yongjing Zhang
Standard Research Lead, Carrier Software BU, Huawei Technologies Co., Ltd.
oneM2M

2. Agenda Concepts about M.A.S. The Management Capabilities in oneM2M
Architecture
Resource modeling
Protocol mapping
The Generic Abstraction & Semantic Capabilities in oneM2M
Interworking framework
Semantic enhancement
Evolution roadmap
Conclusion

3. Heterogeneous Networks, Devices, Interfaces
Why M.A.S.
IoT Applications
Home
Transport
Health
Energy
Retail …
ABSTRACTed APIs
Common Information Model & SEMANTICS
Common Service Functions (e.g. MANAGEMENT)
oneM2M
Heterogeneous Networks, Devices, Interfaces
ZigBee
KNX
BLE
ZWave
mBus
3GPP/2
OMA
BBF …

4. Concepts - Abstraction
Abstraction: generalizing the information model
 to hide the complexity of the specific technologies by providing a single format to represent devices and unified methods directly usable by the applications.
Interworking: mapping between two specific technologies
 to enable the information exchange between heterogeneous systems
Applications may still need to understand the native information model (e.g. Zigbee profile)
Interworking is the basis for Abstraction

5. Concepts - Semantics
Semantics: adding the meaning and relationships between concepts (e.g. data, devices, things) and their instances
 to enable machine understandable interoperability without a-priori agreement or configuration between communication parties
the formal specification of a conceptualization is done by 'ontology', which provides unambiguous vocabulary and model about objects, measurands, their properties and relationships.
Levels of meaningfullness
Abstraction vs. Semantic
Semantics is the evolution of Abstraction

6. Concepts - Management
Management: the management (configuration, monitoring, trouble shooting, upgrade, etc.) of devices (ADN/ASN/NoDN), applications (AEs) and common service entities (CSEs)
to provid 'Abstracted' unified & simplified management APIs for M2M applications.
Management is essentially a specific aspect of oneM2M Abstraction framework:
Data models: the resources describing the mgmt capabilities, properties and status
Operations: the actions performing mgmt tasks, e.g. download (firmware), get (status) or set (properties), execute (software installation)
Mgmt Applications
oneM2M Management Abstraction
(data models & operations)
OMA
BBF …
Management is a specific aspect of Abstraction

7. Management Capabilities
Device Management (DMG CSF)
Device Configuration (e.g. enable/ disable capabilities, provisioning)
Device Diagnostics and Monitoring (e.g. memory, battery, event logs, reboot)
Device Firmware Management
Device Topology Management (e.g. Area Network topology & characteristics )
Application & Service Layer Management (ASM CSF)
Configuration (e.g. CMDH Policy configuration)
Software Management (e.g. download/ install/ activation):

8. Management Architecture
IN-DMG-MA
Protocol Translation
Interaction with the Mgmt Server
Mgmt Server selection
Discovery of external mgmt objects
IN-AE
oneM2M abstracted mgmt APIs (HTTP/CoAP/MQTT)
Mca
Invoking underlying mgmt server functions,
Receiving mgmt results
Managing Area Nwk & Devices (technology specific)
Performing actual mgmt tasks by reusing existing techs (e.g. OMA, BBF)

9. Management Architecture
MN-DMG-MA or ASN-DMG-MA
Mapping between the DMG and Management Client
Interaction with the Mgmt Client
IN-AE
oneM2M abstracted mgmt APIs (HTTP/CoAP/MQTT)
Mca
Mgmt resource creation,
Service layer mgmt
Local mgmt objects discovery

10. Management Abstraction
oneM2M TS-0001 (ARC)/ TS-0004 (PRO)
Application &Service Layer Mgmt
<mgmtObj>
Generic Mgmt Resource Model
[software]
[cmdhPolicy]
 A specialized mgmt func. type (e.g. "software")
 Tech-specific mapping info
Device Mgmt
Specialization
 Link to sub-resources
[firmware]
[deviceInfo]
[deviceCapability]
 Attributes to be specialized for a specific mgmt func.
[eventLog]
[battery]
[memory]
[reboot]
[areaNwkInfo]
[areaNwkDeviceInfo]

11. Management Abstraction
oneM2M TS-0001 (ARC)/ TS-0004 (PRO)
Application &Service Layer Mgmt
<mgmtObj>
Generic Mgmt Resource Model
[software]
[cmdhPolicy]
 A specialized mgmt func. type (e.g. "software")
 Tech-specific mapping info
Device Mgmt
Specialization
 Link to sub-resources
[firmware]
[deviceInfo]
[deviceCapability]
 Attributes to be specialized for a specific mgmt func.
[eventLog]
[battery]
[memory]
[reboot]
[areaNwkInfo]
[areaNwkDeviceInfo]
Mapping
oneM2M TS-0005 (OMA)
oneM2M TS-0006 (BBF)
OMA
DM 1.X
DM 2.0
LWM2M
BBF
TR069/TR181
FUMO
SCOMO
DevInfo
DiagMO …
Device Object
Firmware Update Object
Software Mgmt Object
DeviceInfo
SoftwareModules
X_oneM2M_...

12. Management Abstraction
oneM2M TS-0001 (ARC)/ TS-0004 (PRO)
Generic Resource Model for RPC-like mgmt commands (BBF TR069)
* Each execution creates a <execInstance> to maintain the execution status and result
<mgmtCmd>
<execInstance>
 RPC cmd type (e.g. "download")
 Cmd arguments (opaque)
 Cancel a pending cmd
 Trigger of execution (by 'UPDATE')
 Target device(s) (<node> or <group> URI)
 Execution mode(e.g. 'repeated')
 Inherit from the parent <mgmtCmd>
 In association with execution mode

13. Management Abstraction
oneM2M TS-0001 (ARC)/ TS-0004 (PRO)
Generic Resource Model for RPC-like mgmt commands (BBF TR069)
* Each execution creates a <execInstance> to maintain the execution status and result
<mgmtCmd>
<execInstance>
 RPC cmd type (e.g. "download")
 Cmd arguments (opaque)
 Cancel a pending cmd
 Trigger of execution (by 'UPDATE')
 Target device(s) (<node> or <group> URI)
 Execution mode(e.g. 'repeated')
 Inherit from the parent <mgmtCmd>
 In association with execution mode
Mapping
oneM2M TS-0006 (BBF)
BBF TR069
RESET
REBOOT
UPLOAD
DOWNLOAD
SOFTWAREINSTALL
SOFTWAREUNINSTALL

14. Management Example Flow
IN-CSE
MN/ASN-CSE
IN-AE
Mgmt Adaptor
Mgmt Server
Mgmt Client
Mgmt Adaptor
Registration, CREATE <ASN-node>
<IN-CSEbase>
<ASN-remoteCSE>
Resource creation
Local mgt objects discovery
<ASN-node>
[deviceInfo]
CREATE <deviceInfo>, <battery>, <firmware>…
[battery]
[firmware]
Resource creation
[…]
Keep in sync by UPDATE
RETRIEVE [battery]
Invoke battery monitoring
Mgmt Session (e.g. DM/TR069)
Battery info
[battery] representation
[battery] update
UPDATE [firmware]
Invoke firmware update
Mgmt Session (e.g. DM/TR069) OK
[firmware] update OK

15. Generic Abstraction/Semantics
Data collection
Application
Resource types <AE>, <container> & <contentInstance> are used for the abstraction of M2M applications, data collection and instances..
Data instance
Hierarchical data collection

16. Generic Abstraction/Semantics
Data collection
Application
Resource types <AE>, <container> & <contentInstance> are used for the abstraction of M2M applications, data collection and instances..
Attribute 'ontologyRef' is to provide the semantic annotation (meaning) for application and data. It's the rudimentary step towards semantic capability enablement.
Data instance
Hierarchical data collection

17. Interworking with non-oneM2M
(Informative)
The Inter-working Proxy Application Entity (IPE) abstracts and maps the non-oneM2M data model to the oneM2M resources exposed via the Mca reference point
IPE
 <AE>
ZigBee.
DLMS/COSEM.
Zwave.
BACnet.
ANSI C12.
mBus. 1 n
M2M Area Network
Sensor/Meter
CSE
Mbus/COSEM
IPE
Mca
Utily Application
Zigbee telco Profile
Application Service Node
Mcc
Infrastructure Node
specific
Data model
Awareness
Common 1 n
Device 1 n
<AE>,
<container>,
<contentInstance>
Application 1 n
Interface 1 1 n n
Data Field
Method
Generic data modeling of interworking
Translation of non-oneM2M Data Model to oneM2M Specific Data Model

18. Interworking Enhancement with Semantics
(Informative)
 <AE>
 *Semantic URI
 <container>
 *Semantic URI
 <container>
 *Semantic URI
 <contentInstance>
 <container>
* Attribute 'ontologyRef' links to the ontology definition of each concept (e.g. 'FunctionBlock').
A generic semantic concept model (ontology) representing an Area Network
An example of mapping to oneM2M resources

19. Roadmap to Semantic Enablement
(Informative)
3. Use of Semantic Technologies for specific Platform Functionalities
1. Semantic Modeling (Ontology)
4. Full Semantic Platform
2. Semantic Annotation

20. Roadmap to Semantic Enablement
(Informative)
1. Semantic Modeling (Ontology)

21. Roadmap to Semantic Enablement
(Informative)
2. Semantic Annotation

22. Roadmap to Semantic Enablement
(Informative)
3. Use of Semantic Technologies for specific Platform Functionalities

23. Roadmap to Semantic Enablement
(Informative)
4. Full Semantic Platform

24. Conclusion Common Service Layer oneM2M Applications Semantics
Annotation
Ontology
Query
Reasoning
Mash-up …
Common Service Layer
oneM2M
Abstraction
[battery], [memory], [deviceInfo], [software], [firmware], …
<AE> (e.g. IPE), <container>, <contentInstance>, …
Interworking
<mgmtObj>, <mgmtCmd>
Technology-specific
ZigBee
KNX …
OMA
BBF …
Management
Abstraction
Semantics

25. Join us at the oneM2M showcase event

27. backup

28. Concepts - Abstraction
Examples of existing work study:
ETSI M2M ZigBee Interworking
HGI Smart Home Abstraction Layer (SHAL)
High-level architecture for supporting device abstraction
(Ref: ETSI TS : "Machine-to-Machine communications (M2M); Functional architecture".)
Mapping of an abstract device to the ETSI M2M resource architecture using the <subcontainer> resource
Abstract Interface Descriptions
(e. g. XML)
SOAP
REST
A high-level conceptual HGI architecture
The abstract appliance interface descriptions should be mappable to various environments
(Ref: HGI02029: "Smart Home Architecture and System Requirements")

29. Concepts - Semantics Examples of existing work study:
OGC Best Practice for semantic annotation
W3C Semantic Sensor Network (SSN) Ontology based on OGC SWE information model
(Ref:　Open Geospatial Consortium Best Practice, Semantic annotations in OGC standards.)
Semantic annotations at levels of: service metadata, data model & data entities.
(Ref:　Semantic Sensor Network XG Final Report, W3C Incubator Group Report 28 June 2011.)
Overview of the Semantic Sensor Network ontology classes and properties

30. An Example Case using Semantics
User request for high level info (e.g. Get Discomfort Index of Room X)
Virtual resource created by semantic mash-up
Semantic annotated data
Raw data from physical devices
An example of Home Environment Monitoring Service using semantic mash-up