Abstract:
Cloud computing has created a buzz in the field of computing due to its service oriented
architecture. It allows individuals and businesses to use software and hardware
managed by third parties at distant locations.
However the growth of cloud computing has brought about the increase in cloud
vendors and service providers offering variety of services to end users. The diversity in
cloud platforms and services has made it difficult to decide which service to choose
depending upon varying needs and constraints.
Quality of Service (QoS), attributes help us in examining the quality of a service that a
vendor provides. However selecting the best service under certain conditions is still a
cumbersome task. The Process is considered to be NP-hard (Non-deterministic
Polynomial-time hard) problem as there is no ultimate solution.
This thesis addresses the challenge of finding near-optimal solutions of the problem,
with reduced computational complexity, and is suitable for selecting services in realtime.
In order to achieve this goal, user requirements are given top priority and
happiness measure of user is calculated upon which services are filtered. Simulated
Annealing algorithm along with greedy approach is used to determine to find the near
optimal solution with time efficiency.
A formal model is used to calculate weights for each service according to user
requirements and priority for each service class which plays a major role in increasing
the satisfaction level of a user and helps in filtering the services that makes the
algorithm work fast.
Enhanced service selection algorithm can reduce latency and help achieve higher level
of user satisfaction in cloud computing environment.
