IEEE Consumer Electronics Magazine - May 2018 - 67

The middleware receives the data flow sent from the sensory layer using push (sensors transmit their values based
on timer or change) or pull (sensors are queried from the
middleware to transmit their values) techniques. The transmitted values for each sensor may also be tagged with
descriptive metadata such as identification (ID), location,
interaction protocol, and so on. The middleware should be
able to catalog the SOs, keeping an updated list of all those
placed within its range of responsibility, a task that raises
many challenges. In particular, the middleware system
should be able to identify which SOs are in its area, their
IDs and functions, their last known location, and other
metadata. When a new SO enters or an old SO leaves the
area, the system should be able to track this change and
update the database.
Let us consider a typical scenario in which Lane enters
into the university library with her user device (e.g., smartphone). The middleware residing in the library will track
the entrance of a new SO and initiate communication protocols. The user's device responds by transmitting the
user's profile (Lane's privileges) to the middleware, which
directs the information to the SO management module.
This module has stored an up-to-date list of SOs in the
area, e.g., all cameras, detectors (smoke, sound, motion),
and other sensors. Based on the user profile, the module
sends a filtered list of all of the SOs so that Lane can view
or manipulate them.

The CA module of Lane's smartphone receives the list
and performs CA computations based on the user's profile
(e.g., the type of books Lane prefers to borrow), further personalizing the list of SOs that will be of interest to her. The
output of this process is sent to the AR component, which is
responsible for visualizing it to the device's screen and
allows Lane to interact with the objects.
Although many applications and services of the IoT will
have a constrained space (e.g., smart home, office, hospital or
museum), many great possibilities will arise when the user
leaves this space. It is important to define the area of control
attributed to each middleware; for instance, a geospatial
approach could assign middleware to all SOs within a perimeter, thus allowing overlapping areas that may add complexity,
but it also enhances the SO management. The middleware
system should be able to identify which SOs are in its area,
their IDs and functions, their last known location, and other
metadata, personalizing the IoT devices and services for people wandering around the area. The user device should be able
to hop from middleware to middleware, sometimes merging
lists from different sources to eventually present to the user
the actual SOs of interest in the area (Figure 2).
With the introduction of more complex procedures such
as those previously described, new issues may arise. The
initiation of communication between the user device and the
IoT system is challenging in public areas with multiple
zones of influence. It may be costly to the device's battery

Interaction

Adapt

Sense

Adapt
Sense

Profile
Number ID 3

Sense

Profile
Number ID 1

AR
Application-3-D
Representation of SOs

Understand
Profile
Number ID 2

Understand

Adapt

Personalized Data
Middleware

Sensor Data

Location Data

Sensors SOs

FIGURE 2. The three layers in our architecture.

MAy 2018

IEEE Consumer Electronics Magazine

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - May 2018