Information Technology Reference
In-Depth Information
could also help in discovery of the services. As
newer technologies such as semantic technologies
evolve, there is hope for using them to aid in the
discovery process.
As the demand for computationally intensive
applications such as for providing augmented
reality, on-the-fly decision making, and learning
grows, there will be increasing demand on the
cloud to provide real-time, scalable compute re-
sources. Energy critical mobile devices will have
to depend on Cloud Computing (Kumar and Lu
2010) for machine cycles. Computation offloading
seems to be the way to go, to give mobile devices
access to applications that can revolutionize qual-
ity of life. Response time is extremely critical for
many such applications. There is already research
to address this issue. Virtual cloudlets (Satyana-
rayanan et al. 2009) and Ad hoc cloud computing
can be possible solutions to this issue.
FUTURE RESEARCH DIRECTIONS
There are a number of interesting developments
lately in the area of cloud computing, that provide
a peek into the future of this exciting technology.
There is talk about “Sky Computing” where mul-
tiple clouds work as one (Fortes 2010) to harness
applications and services spread across different
clouds. Users can then mix and match what the
various clouds have to offer them and use the
Virtual Cloud.
In the authors' view, research drivers for the
cloud can be represented by the following research
vector tuple:
Intelligence
There are multiple reasons for building intelligence
in the cloud. The very idea of elasticity in the
cloud requires Artificial Intelligence techniques
to control computation. Machine Learning is
needed to train the load balancing modules in
the cloud to predict and handle demand elasticity.
This is a crucial aspect to the elastic and dynamic
nature of cloud computing. Probabilistic graphic
models could possibly be applied to abstract load
balancing and used for effective prediction. This
will also improve resource utilization. Having a
quantitative model of the cloud is quite important
for the end customer (Durkee 2010), when signing
the contracts. Artificial Intelligence techniques
are expected to play a role in modeling the cloud
behavior.
Applications themselves will need to demon-
strate intelligence as the dependence on the web
for many high end needs increases. Web 3.0 has
already been a step in this direction. Combined
with mobility and intelligence, cloud computing
provides a promising platform for useful appli-
cations (Pendyala and Holliday 2010). Semantic
technologies already provide for reasoning and
deduction. This area needs to further consolidate
and cover increasing needs for intelligence, such
R = (M, I, A, R, T)
Where M = mobility, I = intelligence, A =
Architecture, R = Robustness and Security, and
T = Trust and Privacy
These aspects are further discussed below.
Mobility
Computation is increasingly becoming mobile
with the proliferation of the relatively inexpensive
and networked mobile devices. Mobile devices
will continue to be the only computing devices
accessible to populations in developing countries.
The global market for mobile devices is bound
to grow in leaps and bounds, as more and more
echelons of the global society get added to the
economic mainstream. The Ubiquitous Computer
that the cloud represents, needs to scale to this
need very quickly. Any paradigm shift in comput-
ing cannot afford to overlook mobility aspect to
be successful.
Search WWH ::
Custom Search