Multi-agent cognitive architecture-enabled IoT applications of mobile
Vikas S. Shah
Received: 9 May 2017 /Accepted: 4 May 2018
Institut Mines-Télécom and Springer International Publishing AG, part of Springer Nature 2018
Mobile edge computing (MEC) offers cloud capabilities and service environment for Internet-of-Things (IoT) applications at the
edge of the mobile network. New services for a specific set of IoT application programming interfaces (APIs) of diversified
industry verticals can efficiently be deployed. As the set of APIs based on the IoT continues to thrive, reliance on the underlying
cloud computing to deliver real-time and context-based services is also prominent in MEC. Due to the increased level of mutual
interference of IoT APIs, the complexities are also multiplied. The set of IoT APIs competes for available resources within
heterogeneous clouds (HCs). It is further aggravated by the unprecedented number of connected devices. We are presenting a
cognitive agent (CA)-oriented approach to proactively subordinate IoT APIs with HCs. This paper contributes to establish an
empirical multi-agent cognitive architecture framework (MCAF) for continuous transition of IoT APIs. We identified an adapt-
able composition method and a classification technique for CAs during this research effort. It introduces rationalization of CA
creation, migration, and control. The approach achieves transparency in the heterogeneity and distribution of IoT APIs. It
provides dynamism of cloud computing environments. The paper also illustrates the architecture impact analysis with the
indicative capabilities of auto-scaling, predictability in HCs’ resource use, and responsiveness and attainment of IoT APIs.
Keywords Application programming interfaces (APIs)
Cognitive agents (CAs)
Heterogeneous cloud services (HCSs)
of Things (IoT)
Mobile edge computing (MEC) is engaging smart electronics,
wearable devices, and many other gadgets to edge computing.
As these devices are fabricated at the edge, diversified appli-
cations with different goals are also meshing at the edge .
The pervasive paradigms of such applications of Internet of
Things (IoT) increase the value of information generated by
the number of interconnections between edge users and the
devices . The information is denoted by things. The proc-
essed information is transformed into insight knowledge .
IoT will usher in an extensive range of smart services and
applications to cope with many of the challenges that people
and organizations face in their everyday lives. It is primarily
due to advances in digital technologies that connects humans
or things with either anyone or anything, in any place, and at
any time [4, 5].
MEC leverages conceptualization of centralized radio ac-
cess network (RAN) that concentrates different processing
resources together to form a pool in a central data center.
This aggregation of resources in shared locations reduces the
deployment costs and endorses lowest latency connections
between different RAN processing units enabling a series of
enhanced capabilities such as advanced radio resource man-
agement (RRM) coordination and interference management
When these resources run over virtual infrastructures,
adding flexible and scalable resource management capabili-
ties, it transforms to the heterogeneous cloud (HC) computing
environment. The environment enables users to publish IoT
applications which allow data processing and storage to move
from the devices to a centralized computing platform in HC.
These applications are then accessed over the wireless con-
nection based on a thin native client or web browser by other
users. Essentially, it is an infrastructure where the IoT APIs,
data storage, and data processing occur outside of the devices.
* Vikas S. Shah
Connected Enterprise Services, Wipro Technologies, 2 Tower Center
Boulevard, Suite 2200, East Brunswick, NJ, USA
Annals of Telecommunications