Semantic Resource Discovery with CoAP in the Internet of Things

Ali Yachir

, Badis Djamaa

1,2

, Kheireddine Zeghouani

, Marwen Bellal

and Mohammed Boudali

Artificial Intelligence Laboratory, Military Polytechnic School, PO BOX 17, Bordj-El-Bahri, 16111, Algiers, Algeria

Centre for Electronic Warfare, Cranfield University, SN6 8LA, Shrivenham, U.K.

Keywords: Internet of Things, Smart Environments, CoAP, Device, Resource, Service, Smart Object, Semantic Web

Technologies, CoRE Link Format, Request Resolution, Semantic Resource Discovery, Resource Directory.

Abstract: The Constrained Application Protocol (CoAP) is a lightweight and power-efficient Internet standard

specifically designed for M2M communication in the Internet of Things (IoT). CoAP provides a set of

mechanisms for IoT interactions including request/response, publish/subscribe and resource discovery. For

the latter, a Resource Directory (RD) solution is proposed to register and store information about IoT resources

to be queried by users. Such a solution, however, only allows syntactic discovery. In this paper, we extend

CoAP with lightweight semantic-rich information by defining appropriate CoRE link format attributes

describing both IoT resources and user requests. Such an extension is integrated with the RD to facilitate

semantic resources discovery. Implementation and thorough evaluations of the proposed approach show

important performance enhancements when compared with the default RD solution.

1 INTRODUCTION

As a natural continuity of ubiquitous computing

(Weiser, 1991) and ambient intelligence (ISTAG,

2003), the Internet of Things (IoT) (Atzori et al.,

2010; Suresh et al., 2014) envisages a future Internet

architecture integrating both physical and cyber

worlds by combining sensing and actuation with

digital services. The challenge resides in ensuring

seamless interoperability among a huge number of

constrained heterogeneous IoT devices. Indeed, IoT

devices are typically resource-constrained objects

mainly characterized by limited memory, energy and

processing power. Additionally, such devices

communicates over Low-power and Lossy Networks

(LLNs), such as IEEE 802.15.4 imposing, thus, other

constraints on the reliability and amount of

exchanged data. Hence, seamless integration of such

devices into the Internet requires new lightweight and

power-efficient protocols. Among several

alternatives, the Constrained Application Protocol

(CoAP) (Shelby et al., 2014) is emerging as a

widespread standard that fulfil most of the above-

mentioned requirements.

In CoAP-enabled IoT applications, each device is

seen as an endpoint, exposing sensor readings,

actuating capabilities and internal information as

REST resources (Fielding and Taylor, 2002) that can

be queried by clients. Moreover, CoAP-based

systems usually use a CoRE Resource Directory (RD)

(Shelby et al., 2017) where resource providers

However, the RD solution only allows syntactic and

simplistic data-oriented registration and querying of

resources. For this reason, automatic and intelligent

discovery of required resources among huge

heterogeneous ones remains inadequate with the

native RD. Thus, semantic enhancement of CoAP

and, obviously of RD, is a key aspect for better

representing, storing, organizing, discovering and

providing information generated/consumed by IoT

entities. This challenge can benefit from the semantic

Web technologies (Barnaghi et al., 2012; Bonino and

Procaccianti, 2014) such as Resource Description

Framework (RDF), Ontology Web Language (OWL)

and Protocol And Rdf Query Language (SPARQL).

Extending RD with semantic-rich information

may follow three major steps. First, defining a

comprehensive semantic model describing all the

physical and virtual entities surrounding a device

such as locations, persons, appliances and resources.

Second, extending CoAP to support semantic

resources registration and querying of the RD. Third,

integrating semantics in the RD itself to allow

discovery and ranking of resources, which best match

user requirements and closely meet the specified

quality of service level.

Yachir, A., Djamaa, B., Zeghouani, K., Bellal, M. and Boudali, M.

Semantic Resource Discovery with CoAP in the Internet of Things.

DOI: 10.5220/0006419400750082

In Proceedings of the 14th International Joint Conference on e-Business and Telecommunications (ICETE 2017) - Volume 6: WINSYS, pages 75-82

ISBN: 978-989-758-261-5

Although some work has been proposed for

semantic modeling of things in the literature, the three

aforementioned steps are either addressed partially or

are at a preliminary stage and need to be studied more

deeply. Accordingly, in this paper, we propose a new

semantic support for CoAP by defining appropriate

CoRE Link Format (CLF) (Shelby, 2012) attributes

describing both IoT devices and user requests. This

work is in the continuity of our previous work

presented in (Yachir et al., 2016a; Yachir et al.,

2016b).

The rest of this paper is organized as follows.

Section 2 discusses related work applying semantic

web technologies in IoT. Section 3 provides an

overview of the proposed semantic model in our

previous work. Section 4 defines appropriate

attributes for mapping between the designed semantic

model and CoAP. This is followed by the description

of the proposed framework for device registration and

interrogation via the resource directory (RD). Section

6 is devoted to assess the performance of the proposed

mechanisms. The paper concludes in section 7.

2 RELATED WORK

Various research work have been proposed recently

to improve interoperability between heterogeneous

IoT devices using semantic web technologies. In

(Yuan et al., 2013), a general tree-based metadata

model is proposed to describe IoT entities along three

clusters of information: resource, service and context

information. In (Taccari et al., 2015), IoT entities and

features characterizing an earthquake scenario are

described using earthquake emergency management

ontologies (Spalazzi et al., 2014). In (Wang et al.,

2015), a framework for multisource heterogeneous

information fusion in the IoT is designed. The

collected sensor data are modelled using the Semantic

Sensor Network (SSN) ontology (Compton et al.,

2012). In (Sun and Jara, 2014), the authors propose a

semantic model for IoT information organizing where

object and event layers are represented using a

semantic link network model. In (Jara et al., 2014),

the aspects of the Semantic Web of Things (SWoT)

are presented and discussed along with analysing

their impact on the performance of the IoT resources.

In (De et al., 2011), based on the SENSEI project and

the SSN ontology, a semantic annotation framework

for IoT components is proposed.

In (Ruta et al., 2013), a novel framework for

SWoT based on a backward-compatible extension of

CoAP is proposed. In (Kovatsch et al., 2015), a

semantic description of IoT devices, based on the

RESTdesc format (Verborgh et al., 2012), is designed

to deal with self-configurable service composition in

resource-constrained environments using CoAP. In

(Yachir et al., 2016b), a service-oriented, user-

centered and event-aware Framework for service

discovery and selection in IoT is proposed. In (Ayari

et al., 2015), a semantic approach is proposed for

robots interaction with humans. In (Han and Crespi,

2017), a service provisioning architecture for smart

objects with semantic annotation is proposed to

enable the integration of IoT applications into the

Web. In (Urbieta et al., 2016), an adaptive service

composition framework for IoT-based Smart Cities is

proposed. In (Roffia et al., 2016), a publish-subscribe

architecture, based on a generic SPARQL endpoint,

is designed for interoperability in IoT.

The works discussed above show that various

semantic models are proposed to describe things, but

they lack three important factors. First, some of the

proposed semantic models deal only with sensors

capabilities using SSN ontology without addressing

actuators and the more general notion of thing

ignoring its relationships with either the physical and

digital worlds. These relationships are important to

identify the entities to which a device is attached or it

might control, the space where is located and services

(context, event and/or action) it might provide.

Second, the notion of quality of service (QoS) is taken

into account partially without including either user

preferences, the quality of the physical device or the

quality of its hosted software resources. Finally,

integration of the proposed models in CoAP and RD

is not considered or is at a preliminary stage. Indeed,

the mapping from concepts in the semantic model to

CoAP protocol is not clearly defined. As a result, RD

still performs simplistic resource discovery without

semantic matching of user requirements and the

specified quality of service parameters.

3 PROPOSED SEMANTIC

MODEL OVERVIEW

This section gives a brief overview of our proposed

semantic model for IoT (Yachir et al., 2016a), which

describes both IoT resources and user requests along

with their resolution.

3.1 Resource Description

An IoT resource includes the physical device and its

hosted software services. Accordingly, its description

WINSYS 2017 - 14th International Conference on Wireless Networks and Mobile Systems

comprises two main parts: device description and

service description, as shown in Figure 1 bellow.

Figure 1: Proposed semantic model for device and service

description.

Firstly, the device description part is structured

around four main entities: space, person, appliance

and the device itself. Device is the key entity that

takes control of the three other entities. It refers to a

physical object that can be a sensor or an actuator and

it has four main direct relationships: a device may be

worn by a person or embedded on an appliance. It can

also control an appliance and obviously, it is located

at a given space. Moreover, each entity is described

in its own reference ontology. Accordingly, four

reference ontologies are distinguished: Device

Reference Ontology (DRO), Space Reference

Ontology (SRO), Person Reference Ontology (PRO)

and Appliance Reference Ontology (ARO). Property

attributes of such ontologies are defined in a well-

known ontology denoted as a Reference Ontology

(RO). This ontology plays the role of a global

dictionary where entities/services share vocabulary.

Secondly, the service description part is

structured around a core concept called ambient

service (as), which represents a software resource

provided by a device. An ambient service can provide

output parameters, observed parameters as well as

effects with some quality of service level. Output and

observed parameters are annotated in the reference

ontology (RO) whereas the produced effect is

represented as an RDF statement, in a Subject-

Predicate-Object structure. Statements are triplet that

consist of an Entity (“subject”), a Property

(“predicate”), and the value of the Property

(“object”), where this value can be another entity, an

attribute or a literal. Regarding the quality of service

(QoS), we distinguish the quality characterizing an

ambient service (SQoS) and the quality of the device

(DQoS) hosting such a service.

3.2 Request Description and Resolution

A user request is described by four main components

as follows: requested subject, requested subject

property, required QoS level and required QoS

parameters. Requested subject is formalized as a

concept, a sub concept or an individual from one of

the aforementioned reference ontologies. Requested

subject property is a specific characteristic

formulized as an ontological property of the requested

entity. Required QoS level specifies a minimum score

threshold required for the quality of service. To

compute a score of an ambient service, user should

specify in its request the relative importance accorded

for each required QoS parameter.

Candidate ambient services that satisfy functional

requirements (i.e. requested subject and requested

subject property) are inferred using the Request

Resolution Rule (R3) depicted in Figure 2 using

SPARQL. The inferred services are then evaluated

and ranked according to the quality of service

including SQoS, DQoS and user’s preferences. In this

paper, the R3 rule is implemented in the RD side and

the proposed semantic model is mapped to the CoAP

protocol and implemented under Contiki OS.

Figure 2: Request Resolution Rule (R3) described in

SPARQL.

4 SEMANTIC-ENHANCED COAP

AND CORE LINK FORMAT

CoAP follows the REST architectural style for

making data and resources accessible. In fact, every

resource in CoAP is identified by a URI (Uniform

Resource Identifier). Clients may access resources via

synchronous request/response interactions, using

HTTP-derived methods GET, PUT, POST, and

DELETE. Furthermore, CoAP provides a mechanism

for registering, discovering and advertising resources

that a given CoAP server is making available. Such

SELECT ?device ?as WHERE

{ ?device :hasRelationshipProperty :Subject . ?as

:isProvidedBy ?device.

?as :hasOutputParameters : Parameter }.........(a)

SELECT ?device ?as WHERE

{ ?device :hasRelationshipProperty :Subject . ?as

:isProvidedBy ?device.

?as :hasObservedParameters : Parameter }......(b)

SELECT ?device ?as WHERE

{ ?device :hasRelationshipProperty :Subject . ?as

:isProvidedBy ?device.

?as :hasEffects ?eff. ?eff :hasSubject :Subject.

?eff :hasObject :Parameter}..............................(c)

Semantic Resource Discovery with CoAP in the Internet of Things

discovery protocol uses the CoRE Link Format (CLF)

specification as default, where a client can access the

reserved /.well-known/core URI path on the server

with the POST method to register a resource, or with

GET to discover available ones. GET requests can

include specific attributes in the URI-query field to

filter a particular resource to retrieve. Some standard

attributes are defined in (Shelby et al., 2017). This

work reuses existing standard attributes as much as

possible along with defining others, when necessary,

to fulfil the requirements of mapping our semantic

model to CLF.

4.1 CLF based Resource Description

The proposed semantic device description in CLF

contains both standard and new added attributes. For

instance, the standard attribute endpoint (ep) is used

for device name while a new attribute called entity

(ent) is used to map the three main other entities of

the model, namely person, space and appliance.

Table 1 presents the proposed mapping of our

semantic model in CLF. The mapping reuses standard

attributes as much as possible. It may give new

semantics to some standard attributes such as rt in a

way that is backward compatible. The mapping also

introduces a few new attributes necessary to describe

the semantics of our model. All attributes, their

necessity and meanings are discussed below.

 ep := endpoint (mandatory). The name of the

registering device, unique within that domain. Its

maximum length is 63 bytes.

 et := endpoint type (mandatory). The URI of the

device domain conceptualization named in our

model as Device Reference Ontology (DRO);

 rt := resource type (mandatory). The functional

parameters of the ambient service provided by the

device in et. As mentioned in section 3, a

functional parameter can be an output/observed

parameter or an effect. Hence, rt is formatted to

support both. First, an output/observed parameter

is described in rt as follow:

where Service indicates the name of the ambient

service; Output indicates the name of the output

parameter; and Output_RO indicates the reference

ontology where the output parameter is described.

Second, an effect is described by six fields in rt as

follow:

where Service and Effect indicate the name of the

ambient service and the effect respectively;

Subject and Subject_RO indicate respectively the

name and the reference ontology of the subject of

the effect. Finally, Predicate and Object indicate

the predicate name and the object of the effect.

 obs := observable (optional). Indicates whether

the described output parameter in an rt field is

observable. So, obs is combined with rt

describe an observed parameter. A resource is

observable if its obs equals 1.

 ent := entity (mandatory). Indicates the name of

the entity (space, appliance and/or person) which

has a relationship with the device referenced in ep

and et. Accordingly, ent is structured on three

fields as follow:

where Space is the name of the location where the

device is situated; Appliance is the name of the

appliance controlled by the device. Embedded

indicates whether the device is embedded on the

appliance or not. Finally, Person is the name of

the person wearing the device. It should be noted

that a given device could be only on one of the

three flowing states at the same time: worn by a

person; embedded on an appliance or free.

 entro := entity reference ontology (mandatory).

Indicates the URIs of the domain

conceptualization of the entities specified in the

ent field. In other words, it contains the three

above mentioned reference ontologies namely:

SRO, PRO and ARO. Accordingly, entro is

structured as follow:

 sqos := service QoS (optional). Contains the

quality of service (QoS) parameters of the

ambient service referenced in rt. It is structured on

n fields such as n is the number of QoS

parameters. Each field is divided on three sub

fields indicating the name of the quality

parameter, its current value and a flag (min/max)

indicating whether the parameter is to maximize

(min/max=1) or to minimize (min/max=0). Thus,

sqos is structured as follow:

 dqos := device QoS (optional). Contains the

parameters of quality of service (QoS) of the

device referenced in ep and et. dqos has exactly

the same structure as sqos where the key word

dqos is used instead of sqos.

rt= "Service|Output|Output_RO"

rt="Ser

ice

Effect

Sub

ect

Sub

ect_RO

Predicate

ect"

ent="Space|Appliance:Embedded|Person"

entro= "SRO

ARO

PRO"

sqos="sqos

:val

: max/min|…|sqos

:val

:max/min"

WINSYS 2017 - 14th International Conference on Wireless Networks and Mobile Systems

Table 1: Resource Description Mapping to CLF.

Concept Concept

attribute

CoAP

attribute

Attribute

type

Device

Device name Endpoint (ep)

Standard

attributes

Device Reference

Ontology

Endpoint Type

(et)

Ambient

Service

Ambient Service

parameters

Resource Type

(rt)

Observable Observable

(obs)

Space,

Appliance,

Person

Entity Names Entity (ent)

Added attributes

Entity Reference

Ontologies

Entity

reference

ontology

(entro)

QoS

Service QoS

sqos

Device QoS

dqos

Having introduced and detailed the necessary

mapping attributes, let us see the corresponding

semantic description of a device named

“Imote2Sensor” located in the “Kitchen” and

characterized by energy level and reliability as dqos

parameters. This device provides a service

“getTemperature” having “temperature” as an output

parameter that can be observed. In addition, the

provided service is characterized by a response time

and energy cost as sqos parameters. The CoAP

message corresponding to such description is:

4.2 CLF based Request Description

Having presented the proposed mapping of our

semantic resource description into the CoRE Link

Format carried in CoAP messages, this section

follows the same approach to map user requests.

Hence, as mentioned in section 3.1, a user request is

described by four main components, namely:

requested subject, requested subject property,

required QoS level, required QoS parameters.

In the proposed mapping of Table 2, Requested

subject maps to the newly introduced link format

attribute entity (ent), whereas requested subject

property is mapped to the resource type (rt) attribute.

Moreover, required QoS parameters are mapped to

the defined sqos and dqos attributes in CoRE link

format. Finally, required QoS level is mapped to a

semantic threshold (sr) attribute similar to that

defined in (Ruta et al., 2013).

Table 2: User Request Description Mapping to CLF.

Request

attributes

CoAP

attributes

Attribute

type

Required QoS

Level

Semantic threshold (sr)

Standard

attributes

Requested

subjet param.

Resource Type (rt)

Observable (obs)

Requested

subject

Entity

(ent)

Added

attributes

Required QoS

Param.

Service QoS (sqos)

Device QoS (dqos)

Using these attributes, a user requesting the

“temperature” of the “kitchen” with high energy

level, very high reliability, a medium response time

and a low energy consumption as QoS preferences

can retrieve good matches above a threshold of 0.6 by

issuing the following CoAP query:

5 PROPOSED ARCHITECTURE

Having presented our model for extending CoAP’s

resource discovery with semantics, we have designed

and implemented a framework integrating the

proposed semantic model along with its

representation in CoAP. The architecture of such a

framework, shown in Figure 3, is structured around

five main components namely: Resource Directory

(RD), Device, Border Router, User Interface and

Reference Ontologies Server (ROS). RD, Device and

Border Router components are implemented under

Contiki 3.0 OS using Cooja simulator whereas User

Interface and ROS are implemented respectively in

standards web browser and web server.

Figure 3: Framework for Resource Description, Discovery

and Retrieving.

</pathRes1>;ep="Imote2Sensor";et="http://emp.org/Ontologie

s/Device.owl";ent="Kitchen";

entro="http://emp.org/Ontologies/ Space.owl;

dqos="Energy_level:70:1|Reliability:0.6:1"; rt="

getTemperature|temperature|http://emp.org/Ontologies/

coap://adressRD? ent="kitchen”; rt=”temperature”; dqos=

"Energy_level: high | Reliability: very high"; sqos=

"Response_Time:medium | Energy_Cost: low" ; sr=”0.6”

Semantic Resource Discovery with CoAP in the Internet of Things

Firstly, an RD is a web entity used as a repository

that stores information about web resources (devices

and services) and implements two main REST inter-

faces POST_Handler and GET_Handler. The first in-

terface is dedicated for resources registration whereas

the second one is dedicated for lookup of those resour-

ces using the R3 rule with SPARQL (see Section 3.2).

Secondly, a device in the proposed framework can

be a sensor and/or an actuator with resources

encapsulated as ambient services. A device can send to

the RD either its description via POST or a request

through a GET message. Thirdly, the Border Router is

a node that relies a PC or a Smartphone to the network

through a Serial Line Internet Protocol (SLIP)

(Romkey, 1988) or any available network interface.

Inside the constrained network, packets are routed

using RPL (Winter, 2012) or any routing protocol

deployed within the network.

Fourthly, Reference Ontologies Server (ROS) is a

web server that stores and provides access to all the

reference ontologies, namely: ARO, DRO, PRO, SRO

and RO. Information from such ontologies can be

retrieved or updated using the SPARQL language. An

example scenario using concrete ontologies is given in

(Yachir et al., 2016a).

Finally, user interface is developed based on the

Copper plugin (Copper CoAP, 2016), which is a CoAP

agent for Firefox. In this work, we have developed two

new plugins/apps for Google Chrome and Android OS

allowing seamless interactions between users and

objects using smartphones. Such applications are

developed using HTML5/CSS3, Bootstrap, JQuery

and Chrome APIs. The developed application provides

dedicated interfaces for both device registration and

user request specification. The developed application

and a simple use case scenario implemented in Contiki

OS using Cooja are reported in the following link:

https://www.dropbox.com/s/cnbyzpzmarrw5f6/Chrom

eAndroidCopper.mp4

6 PERFORMANCE EVALUATION

6.1 Experimental Model and

Performance Metrics

The proposed semantic model over CoAP is abbrevia-

ted SemRD for Semantic Resource Directory. To eva-

luate the performance of such a model, we imple-

mented it in Contiki OS. To put results into context,

SemRD was compared with the standard RD solution.

A scenario comprising a simulated network composed

of thirty (30) emulated Sky motes and a resource

directory was used in our evaluations. The topology is

depicted in Figure 4. Each mote can play the role of

either a client requesting available resources from the

RD/SemRD or a provider registering its resources at the

RD/SemRD or both roles. In all cases, a node is

considered aware of the RD before accessing it.

For the sake of this evaluation, we developed an

application running the two RD solutions above UDP

in a constrained 6LoWPAN network in Contiki. At the

routing layer, the RPL protocol is deployed to ensure

routing between RDs and other network nodes. At the

link layer, simulations are conducted both with the

ContikiMAC (Dunkels, 2011) Radio Duty Cycling

(RDC) protocol and without using RDC (NullRDC). Si-

mulation configuration parameters are given in Table 3.

Figure 4: Simulated network topology.

To be able to draw conclusions on the time/cost

performance of evaluated approaches, Registration

time for devices, Response time for requests, Average

Response size and the Radio duty cycle, as a proxy of

energy consumption, were measured when varying

the number of clients and providers. Registration

(response) time is measured as the time spent from

sending a description (request) until receiving a

confirmation (response), averaged over all successful

registrations (requests). The packet size is measured

at the routing layer and averaged over all sent packets

per node. The network duty cycle, as an indicator of

energy consumption, is measured using Contiki's

power profiler (Dunkels et al., 2011).

Table 3: Simulation configuration parameters.

Parameter Value

Number of nodes 31

Type of nodes Wisemote

Network 300m × 300m

Simulation time 600 secondes

Routing protocol RPL

Transmission range 60 m

RDC / MAC / adaptation

layer

ContikiMAC Null RDC /

CSMA / 6LoWPAN

OS Contiki 3.0/Cooja

WINSYS 2017 - 14th International Conference on Wireless Networks and Mobile Systems

6.2 Results and Discussion

Figure 5: Performance when varying provided resources

(number of clients = 5).

Figure 6: Performance when varying number of clients

(number of resources = 16).

As we can see in Figure 5.a, the average registration

time of a resource increases with the number of

provided resources in the network. This increase is

more visible in networks deploying RDC. When

comparing the two approaches, SemRD takes more

time in resource registration. This is because SemRD

packet size is bigger than that of RD, due to the

embedded semantic attributes. However, when it

comes to the response size (Figure 5.b), SemRD

showed a clear amelioration, with responses of about

half the size of that of RD, thanks to the embedded

semantics that allowed fine-grained filtering of

responses. This in turn translated into a decrease in

response time (results not showed for space reasons).

Figure 7: Energy consumption (duty cycle).

Figure 6.a and 6.b present the average registration

and response times of the evaluated solutions when

varying the number of clients. These figures show a

similar behavior to that of Figure 5. Indeed, while the

registration time is fairly independent from the

number of clients (Figure 6.a), SemRD registered

higher registration times than RD because of the

additional semantic attributes. This, however, has the

advantage of lowering response times, achieved by

SemRD, because of minimizing their sizes as can be

seen in Figure 6.b. It should be noted, from this

figure, that response times of both SemRD and RD

increase with the number clients.

Finally, Figure 7 presents energy consumption

topography of SemRD and RD when varying both

providers and clients. As can be seen in this figure,

nodes’ duty cycles are equitably balanced (similar

color-heat per node) for both RD and SemRD. The

higher heats (yellow to red) observed in some nodes

can be due to their location (i.e. neighboring the RD

node). Overall, Figure 7 clearly show the load

balancing aspect of both approaches along with the

lower duty cycle (lower energy consumption).

7 CONCLUSION

This paper proposed a concrete lightweight mapping

and implementation of our previous semantic model

(Yachir et al., 2016a) in CoAP. This mapping is

achieved by defining appropriate CoRE Link Format

attributes describing both IoT devices/resources and

user requests. An RD-centered framework was also

designed to facilitate IoT resources publication and

retrieving using appropriate user interfaces

communicating through semantic-enhanced CoAP.

Simulation results have shown the performance of

such mechanism when compared with the default RD

solution.

Future work consists on more simulations and

testbed experiments to validate the model at a larger

scale. Enhancing the CoAP mapping along with

considering other description formats is also planned.

Currently, the authors are actively working to

integrate, in the proposed semantic model, the

publish/subscribe mechanism.

REFERENCES

Atzori, L., Iera, A., Morabito, G., 2010. The Internet of

Things: A survey. In Computer Networks 54 (2010)

2787–2805.

Ayari, N., Chibani, A., Amirat, Y., Matson, E. T., 2015.

Semantic Resource Discovery with CoAP in the Internet of Things

A novel approach based on commonsense knowledge

representation and reasoning in open world for intelligent

ambient assisted living services. In Proc. IEEE/RSJ Int.

Conf. Intell. Robot. Syst. (IROS’15), pp. 6007-6013.

Barnaghi, P., Wang, W., Henson, C., Taylor, K., 2012.

Semantics for the Internet of Things: Early Progress and

Back to the Future. In International journal on Semantic

web and information systems. Vol. 8, Issue 1.

Bonino, D., Procaccianti, G., 2014. Exploiting semantic

technologies in smart environments and grids: Emerging

roles and case studies. In Science of Computer

Programming 95, 112–134, Elsevier.

Compton, M., Barnaghi, P., Bermudez, L., Garcı´a-Castro,

R., Corcho, O., Cox, S., Taylor, K., 2012. The SSN

ontology of the W3C semantic sensor network incubator

group. In Web Semant Sci Serv Agents World Wide Web

17:25–32.

Copper CoAP, 2016. “Copper (Cu) CoAP user-agent -

JavaScript CoAP Implementation.”[Online]. Available:

http://people.inf.ethz.ch/mkovatsc/copper.php.

[Accessed: 21-Apr-2016].

De, S., Barnaghi, P., Bauer, M., Meissner, S., 2011. Service

modelling for the Internet of Things. In Federated

Conference on Computer Science and Information

Systems (FedCSIS), Szczecin, Poland.

Dunkels, A., 2011. The ContikiMAC Radio Duty Cycling

Protocol. Swedish Institute of Computer Science (SICS),

Technical Report T2011:13, ISSN 1100-3154.

Dunkels, A., Eriksson, J., Finne, N., Tsiftes, N., 2011.

Powertrace: Network-level power profiling for low-

power wireless networks. In Swedish Institute of

Computer Science, Technical Report T2011:05.

Fielding, R. T., Taylor, R. N., 2002. Principled design of the

modern web architecture. In ACM Trans. Internet

Technol. 2 (2) 115-150.

Han, S. N., Crespi, N., 2017. Semantic service provisioning

for smart objects: Integrating IoT applications into the

web. In Future Generation Computer Systems,

http://dx.doi.org/10.1016/j.future.2016.12.037

ISTAG, 2003. Ambient intelligence: from vision to reality.

In Eur Comm Rep 1–31.

Jara, A. J., Olivieri, A.C., Bocchi, Y., Jung, M. Kastner, W.,

Skarmeta, A.F., 2014. Semantic Web of Things: an

analysis of the application semantics for the IoT moving

towards the IoT convergence. In Int. J. Web and Grid

Services, Vol. 10, Nos. 2/3, pp.244–272.

Kovatsch, M., Hassan, Y.N., Mayer, S., 2015. Practical

Semantics for the Internet of Things: Physical States,

Device Mashups, and Open Questions. In 5th IEEE

International Conference on the Internet of Things (IoT),

Seoul, South Korean, P 54-61.

Roffia, L., Morandi, F., Kiljander, J., D’Elia, A., Vergari, F.,

Viola, F., Bononi, L., Cinotti, T. S., 2016. A Semantic

Publish-Subscribe Architecture for the Internet of

Things. In IEEE Internet of Things Journal, DOI

10.1109/JIOT.2016.2587380.

Romkey, J. L., 1988. Nonstandard for transmission of IP

datagrams over serial lines: SLIP.

Ruta, M., Scioscia, F., Pinto, A., Di Sciascio, E., Gramegna,

F., Ieva, S., Loseto, G., 2013. Resource annotation,

dissemination and discovery in the Semantic Web of

Things: a CoAP-based framework. In IEEE Internet of

Things and Cyber, Physical and Social Computing. DOI

10.1109/GreenCom-iThings-CPSCom.

Shelby, Z., 2012. CoRE Link Format. IETF RFC 6690.

Shelby, Z., Hartke, K., Bormann, C., 2014. The Constrained

Application Protocol (CoAP). IETF RFC 7252.

Shelby, Z., Koster, M., Bormann, C., Van Derstok, P., 2017.

CoRE Resource Directory. Internet Draft IETF.

Spalazzi, L., Taccari, G., Bernardini, A., 2014. An iot

ontology for earthquake emergency evaluation and

response. In International Symposium on Collaborative

Technologies and Systems (CTS), IEEE.

Sun, Y., Jara, A. J., 2014. An extensible and active semantic

model of information organizing for the Internet of

Things. In Pers Ubiquit Comput, 18:1821–1833. DOI

10.1007/s00779-014-0786-z. Springer-Verlag London.

Suresh, P., Daniel, J. V., Parthasarathy, V., Aswathy, R. H.,

2014. A state of the art review on the Internet of Things

(IoT) history, technology and fields of deployment. In

Proc. IEEE Int. Conf. Sci., Eng. Manage. Res.

(ICSEMR’14), Chennai, India, pp. 1-8.

Taccari, G., Bernardini, G., Spalazzi, L., D’Orazio, M., 2015.

Earthquake Emergencies Management by Means of

Semantic-Based Internet of Things. In IoT360, Part II,

LNICST 151, pp. 318–327. DOI: 10.1007/978-3-319-

19743-2 43.

Urbieta, A., Gonz´alez-Beltr´an, A., Ben Mokhtar, S., Anwar

Hossain, M., Capra, L., 2016. Adaptive and context-

aware service composition for IoT-based smart cities. In

Future Generation Computer Systems,

http://dx.doi.org/10.1016/j.future.2016.12.038

Verborgh, R., Haerinck, V., Steiner, T., Van Deursen, D.,

Van Hoecke, S., De Roo, J., Van de Walle, R., Vallès, J.

G., 2012. In Functional Composition of Sensor Web

APIs. In Proc. SSN, Boston, USA.

Wang, F., Hu, L., Zhou, J., Hu, J., Zhao, K., 2015. A

semantics-based approach to multi-source heterogeneous

information fusion in the internet of things. In Soft

Comput. DOI 10.1007/s00500-015-1899-7, Springer-

Verlag.

Weiser, M., 1991. The computer for the 21st century. In Sci

Am 265:94–104.

Winter, T., 2012. RPL: IPv6 routing protocol for low-power

and lossy networks.

Yachir, A., Djamaa, B., Mecheti, A., Amirat, Y., Aissani, M.,

2016a. A comprehensive semantic model for smart

object description and request resolution in the internet

of things. In The 7th International Conference on

Ambient Systems, Networks and Technologies (ANT

2016). Procedia Computer Science 83, 147 – 154.

Yachir, A., Amirat, Y., Chibani, A., Badache, N., 2016b.

Event-Aware Framework for Dynamic Services

Discovery and Selection in the Context of Ambient

Intelligence and Internet of Things. In IEEE Trans.

Automation Science and Engineering 13 (1): 85-102.

Yuan, Y., Wu, X., Lu, Y., 2013. A Flexible Framework for

Representations of Entities in the Internet of Things. In

ISCTCS, CCIS 320, pp. 295–302.

WINSYS 2017 - 14th International Conference on Wireless Networks and Mobile Systems