Page 96 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 5 – Internet of Everything

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 5

3.2 Reference IoT architecture be the most suitable and interesting for each user, based
on their pro ile and context: and enabling objects to be as
Recently, several studies have looked at the problems of much plug & play as possible.
managing and effectively using large numbers of hetero‑
The speci ic focus of this work will be on the interfaces be‑
geneous devices, and have found a solution in the use of
tween the users and IoT enriched environments. Indeed,
social networking principles and technologies. In [15],
a user‑friendly interface will be provided by means of the
the de inition of the Social IoT (SIoT) has been formalized, chatbot to users to assist them to create their pro ile and
and it is intended to be a social network in which each manage their services and objects. Based on their pro‑
node is an entity capable of forming social relationships ile and thanks to context‑awareness mechanisms, the VU
with other things on its own, according to the rules set will be able to suggest relevant services that are expected
by the owner. The proposed model is based on the Lysis to be the most suitable and interesting for each user, and
cloud SIoT architecture [16], which incorporates virtual settings will be automatically con igured.
objects as digital counterparts to physical objects to en‑
hance their capabilities in a transparent manner to users. 4.
Lysis architecture foresees a four level structure of inde‑ PROPOSED ARCHITECTURE
pendent modules. Its lowest layer is populated by Real The proposed solution is aimed at designing and exper‑
World Objects (RWO), i.e. physical IoT devices able to per‑ imenting a chatbot system that simplies the interaction
form basic tasks. On top of this, the virtualization layer, of the users with the VU in an IoT platform by means of
directly interfaces with the real world and is populated by text messaging. As previously explained in Section 3, the
Social Virtual Objects (SVO), which are VOs with socializa‑ VU is the virtualization of the user and takes decisions on
tion capabilities. The aggregation layer is responsible for their behalf for known activities; as such, it interacts with
composing several SVOs into entities with extended capa‑ all the modules of the Lysis IoT architecture [17]. The VU
bilities, called Micro‑Engines (MEs). Finally, at the appli‑ was not introduced speci ically for the Lysis platform. In
cation layer, user‑oriented macro‑services are provided fact, it follows the more general concept of virtualization
(APP).
in the IoT and of virtual objects. The concept of VU arises
Socialization algorithms implemented in the irst two lev‑ from the need to provide a virtualization element that
els allow for the creation of social relations as foreseen in constantly deals with the context of the user it represents,
the SIoT paradigm. The resulting social graph is exploited their interaction interfaces and their IoT services. In this
to ind the required resources. scenario, the VU is a standalone element in a distributed
virtualization system. The Lysis platform, which we use
3.3 User virtualization in IoT as a development environment, is precisely a distributed
system of elements that allows for the creation of a social
The widespread presence of connected objects through‑ network among virtual objects in order to facilitate their
out daily life has allowed the Internet of Things (IoT) to interaction. The VU could be used in a centralized system,
spread. The IoT vision forms a collaborative ecosystem possibly vertical; however, in this case it would not bring
for a multitude of heterogeneous objects with different all the advantages that characterize the implementation
connectivity and computing capabilities to achieve the in a distributed system. Furthermore, any IoT platform
common purpose of providing user services. that is a candidate for the integration of the VU and its in‑
At the current time, the IoT platforms seem to present terfaces, such as the chatbot, should provide open APIs
several pending issues that prevent a full spread of IoT that allow for full integration.
applications. Indeed, services are mostly con igured man‑ Fig.1 shows the components of the overall architecture
ually by users, according to preferences that could be according to the Lysis model. The VU communicates with
shared among similar or cross‑domain services (e.g., pref‑ all levels to provide user preference information to build
erences about ambient temperature at home and at work tailored IoT services. The chatbot system is a back‑end
to manage HVAC systems). Secondly, the users that ac‑ service for proxies the communication between the users
cess an IoT platform need to autonomously look for the the the VU.
required services among a plethora of them. The risk is Fig. 2 shows the architecture of the proposed solution
in a decrease in the quality and reliability perceived by that has been designed to address the requirements that
users, who therefore risk being discouraged from using have been previously discussed.
IoT applications. Our system has the objective of exploit‑ The upper layer implements the functionalities to re‑
ing the concept of Virtual User (VU)[17] to improve the ceive and transmit requests and data. The requests are
user experience and, at the same time, enhance the ef i‑ generated by either web apps or (proprietary / non‑
ciency and usability of the IoT platforms and services. proprietary) messaging services which are used by the
The VU is the virtualization of a user, and it is represented user for sending and receiving messages. Each request
by an agent that enables the following major bene its: contains the intent, i.e., the action that the user would like
providing users with a user‑friendly interface that en‑ to take, which is written in natural language. The intents
ables automatic or assisted setup of their pro ile, objects are then received by the chatbot API gateway to be sent
and services; proposing the services that are expected to to an AI‑based service that interprets the intents to

84 © International Telecommunication Union, 2021