Database Reference
In-Depth Information
other enabling interoperability. Emerging technologies and systems development of the
IT sector work for this interoperability challenge, additionally in this complex scenario,
it is highly desirable and necessary endow the IoT system(s) with self-monitoring and
self-managing capacity.
The deployment and management of IoT platform(s) in the large scalable, using
cloud platforms, has brought even a bigger challenge. Starting from the initial
deployment, the IoT system or service has to go through the learning phase of different
scenarios, while serving its end users with maximum output with minimum cost
possible. Currently available cloud architecture can itself be self-managing, but the
complexity and dynamically changing requirements of the implementation of IoT
system in the same form is much more complex and challenging than that of any
general software application.
In the Internet of Things area, it is becoming a common practice using agent(s) for
con
guring and monitoring operations.
IoT systems rely on the principles for
deployment and con
guration based on autonomic service delivery models. In auto-
nomic systems, agents analyse the requirements, dependencies and resolves them
following self-organizing principles and also, by design, it deploys the IoT service
delivery model into the desired platforms. While the service control loop for distributed
platforms has already been proposed, by means of autonomics [
1
], the dynamic
deployment of the different modules of an IoT system is still due. At this moment the
deployment of IoT service delivery models is manual, thus, much complex to handle
for a non-technical user, as well as time consuming and error prone.
In the area of Internet of Things, autonomic agents can dynamically evaluate the
system requirements and deploy dependencies of an IoT service delivery model.
Autonomic agent(s) can also extensively be used on cloud optimization, for example
using event data distribution or event splitting mechanism to control service infra-
structures on demand. Autonomic in cloud services have been already proposed within
the framework of the OpenIoT project (
www.openiot.eu
)
, where the self-tuning or self-
con
guration facility of the IoT service delivery model is a novel approach, this
approach is explained.
In this paper a generic approach or framework to support the administrator/user
deploying IoT platform(s) with an autonomic guideline [
1
] taking care of all the
underlying details of the existing framework is intruded and explained. The framework
functionality analyses the requirements, dependencies and resolve them
“
autonomi-
cally
and as main objective deploy the IoT service delivery model into the desired
platform by means of cloud or local machine. As example an open source platform
(
www.superstreamcollider.org
) to demonstrate cloud or local machine instantiation is
used. Some experiments are performed under the SSC-Fed4FIRE experiments project
to demonstrate the use of autonomic agents in IoT environments.
As a second part in this paper, an autonomic framework for continuously monitoring
and optimizing the IoT test-bed, regardless the infrastructure on which the platform is
hosted is discussed. Although many cloud monitoring services (e.g. monitis, nagios) are
available in the market, the IoT platform(s) needs to extract the information and map
those information into customized requirements that can serve into the tuning of the
platform itself. Moreover, the proposed autonomic agent not only extracts and maps the
information, but also it is able to tune-up the underlying IoT framework.
”
Search WWH ::
Custom Search