Vous êtes sur la page 1sur 14

International Journal of Computer Science Research & Technology (IJCSRT)

ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

Achieving Qos In Distributed Systems Using Software


Engineering Principles

J. Jeyachitra
Research Scholar
Dr. K. Iyakutti Dr. K. Alagarsamy
Dept. of Computer Professor Associate Professor
Applications Dept. of Physics and Dept. of Computer
School of Information Nanotechnology Applications
Technology SRM University School of Information
Madurai Kamaraj Chennai Technology
University Madurai Kamaraj
University

Abstract balancing scheme to clear the occurrence of


traffic in distributed system. Another one is
T adapting the task based validation scheme for
Usually, distributed systems
SR
security purpose. By utilizing these two
permit the set of applications to access the data
engineering principles, the security and the QoS
from both the local and remote database. In
of the distributed systems are achieved. The
IJC

distributed systems, a set of databases included


experimental performance of QoS
utilized for processing are viewed as a single set
accomplishment in distributed systems using
of source. In order to process the data properly,
software engineering principles [ASEP] is
distributed systems are in need of QoS support.
evaluated with Solar Flare Data Set from UCI
To maintain the quality properties of services in
repository against existing non-parametric
distributed systems, it is necessary to analyze the
approach for network traffic classification to
set of resources and the users in the
attain quality assurance, security, and
environment. For enhancing the quality in
communication cost. Results shows that the
distributed system, many research works have
proposed ASEP scheme provides 5-7% in
been presented and analyzed. Among the
security rate to achieve the QoS.
techniques, software engineering principles
plays a vital role in presenting the fine quality of Keywords: Distributed
service in distributed systems. This work systems, QoS, software engineering
describes the process of implementing the principles, load balancing scheme,
software engineering principles for enhancing
task based validation scheme, security.
the security and QoS in distributed systems. Two
principles are utilized here, one is effective load

IJCSRTV1IS050058 www.ijcsrt.org 1
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

1. INTRODUCTION inclination of a scheme from the paid for


functionality, and can be customized to hold an
enormous number of structure fundamentals.
Distributed system with task
processing computers, receive set of loaded tasks
The traffic engineering issues is to
in the server for transmitting the tasks with the
identify the effective paths over the distributed
specified path. The load imbalance in the
systems to deal intelligently with the difficulty of
distributed system provides the data
distributed computing systems. The main
communication services, for improving the
genuine objective is processed by the load
quality of distributed system. The software
balancing with all set of nodes in a distributed
engineering traffic techniques are utilized among
system. It is processed based on the interrelated
the set of load balancing techniques which is
nodes of the system without following any set of
applied to enhance the quality of the
procedures for transmission. To address this
communication across the distributed system.
issue, Hassle-free Curved Optimization (HCO)
traffic engineering principle is introduced.
Load balancing in
distributed computer systems defines the
HCO algorithmic solution optimize the
T
reallocation of the set of tasks to the specified
work load in distributed systems. It identified the
SR
centralized processors in the system to progress
assignment of load using heuristic algorithm via
the system presentation. Various study have
HCO principles. To control the user access, task
provided a information that sharing the work
IJC

based validation scheme is existed as a concept


load arbitrarily to the set of nodes in a
for security measure. The user access control
distributed system usually transformed the
over the distributed system has been processed
system presentation and utilizing the resources.
with deliberately and unintentionally.
Dynamic load balancing algorithms scruntizes
the changes occurrence over the system work
With regardless of, various security
load and redistribute the work load in a given
measures utilized for constructing a distributed
interval of time.
system for encryption and decryption techniques
to prevent any possibility of any third party
Security in computer systems is
attacks. Task based validation framework
concerned with resources from prohibited access
systematically spans the implementations
yet as building confident requests can be
standards for processing the tasks specified by
satisfied all the time. The present expansion of
the user.
computer systems uniformly in scale and
complexity fronts notable organization faces.
Policy-based systems association is a very clever 2. LITERATURE REVIEW
declaration in this state. It authorizes the Numerous online data study requests

detachment of the rules that supervise the actions need uninterrupted query operations for instance

IJCSRTV1IS050058 www.ijcsrt.org 2
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

association, aggregation, and straining to data question, a novel status system is constructed
streams in actual time. Disseminated stream upon the multivariate Bayesian assumption [4].
processing schemes permit network stream In recent times, the IEEE 1900.4 set precised a
processing to attain better scalability and quality- policy-based radio reserve organization (RRM)
of-service (QoS) condition. In [1], the author structure in which the choice making procedure
presented Synergy, a narrative dispersed stream is dispersed among network-terminal units.
processing middleware that presents habitual In [5], the author presented three main RRM
sharing-aware constituent symphony methods from diverse features, spanning
ability. Conventional admittance direct diagonally prognostic versus reactive to model-
representations acquire some presentation based approaches.
consequence.
Light-trail is planned as
In large-scale standard systems with a candidate to transmit IP traffic over wavelength
physically dispersed server clusters, accessing separation multiplexing networks agreed its
admittance control on every group can effect ability of allowing high-speed provisioning and
in long reply latency, if the customer request is obliging multi-granularity passage. With the
deprived of one location and has to be on warded current expansion on ultra-fast visual board up,
T
to another site. In [2], the author measured a the author proposed enhanced light-trail
SR
destructive admittance control strategy where construction, termed as opportunistic hyper
every agent might disclose more requests than its channel in [6]. The paper [7] presented a
IJC

owed bandwidth permits. combined mode of partitioning strategy for


obtaining latency underneath a threshold to the
Quality of service utmost number of users as
(QoS) provisioning is one of the mainly achievable in DVE systems.
significant measures in recently rising UWB
WPANs, as they are anticipated to hold up a In [8] a resource
broad range of applications from time- distribution algorithm unites packet arrangement
constrained, multimedia torrent to throughput and power obligation is proposed to accomplish
starving, content transport requests. In [3], the efficient resource consumption beneath
author behavioral a complete hypothetical QoS constraints. Network quality-of-service and
examination and demonstrated that with the security have been measured as split entities and
presently suggested parameter setting, EDCA study in these areas has mainly continued
cannot offer sufficient QoS. separately [9]. In [10], the author presented a
novel lining analytical structure to revise the
A basic crisis is, thus, tradeoff among the energy reduction and
how to choose consistent servers from a huge the QoS at a transmit node.
applicant pool. To respond this significant

IJCSRTV1IS050058 www.ijcsrt.org 3
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

The current study engineering principles, the quality and security


in wireless networks prompt the opportunistic of the distributed systems is achieved. The
communication that develops guide fluctuations architecture diagram of the proposed achieving
to develop the entire performance. In wireless ad QoS in distributed systems using software
hoc networks, nodes might contain packets engineering principles [ASEP] is shown in fig
intended to numerous adjacent nodes. The author 3.1.
in [11] measured an opportunistic setting up that
takes benefit of a time-varying communication
channel amongst diverse receivers to
progress system performance.

After analyzing the


existing work issues, this work presented
software engineering principles for resolving it.
The load balancing and security principles are
modeled here for enhancing the QoS in
distributed systems.
T
SR
3. PROPOSED METHODOLOGY
IJC

The ability of the system


to process the users‟ requests in more than one
interconnected processor to be operating in a From fig 3.1, it is being observed that the quality
minimal interval of time is referred as distributed in distributed systems is achieved by adapting
systems. The main objective of distributed the set of software engineering principles. The
systems is to allocate the set of required load balancing is achieved by implementing the
resources to the set of requested users to achieve optimization technique to remove the traffic
the task. The communication over the distributed occurrence over the network. The security is
systems is achieved via a set of messages over achieved by adapting the users‟ task based
the network link. So, it is necessary to provide validation scheme. By performing these two
the quality over routing by clearing the approaches, the quality in the distributed systems
occurrence of traffic and maintaining the is achieved.
effective load balancing across the network.
Since it is a distributed network environment, The software engineering principles
there is a highest chance of unauthorized users to for enhancing the quality over distributed
access the route path messages. To provide a systems is done using two phases. The first
security over the network, task based validation phase describes the load balancing scheme is
scheme is used. By these two set of software

IJCSRTV1IS050058 www.ijcsrt.org 4
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

distributed system by adapting the hassle free  Associate the nodes using the
curved optimization technique. The presence of network for efficient routing
traffic in the load imbalanced distributed system and the communication range
is identfied and rectified while minimizing the between the set of nodes over
communication overhead among different set of the network is similar.
distributed users.
 Each node in the network has
task t referrred as Tn.
The second phase describes the
process of providing secuirty to the distributed  Identify the number of users in
network environment. The task based validation the network.
scheme assigns an individual role to users in the
network based on their appropriate task over the  Assign the set of resource

network system. The role is assigned based on available over the network

the capabilities of the users and their limits to the


 Set some tasks to each of the
network boundary to avoid the network traffic.
user in the network for an
effective communication
System model
T
 Analyze the users‟
SR
transmission and the packet
The presentation of the
delivery to the destination
distributed systems is as follows. Assume a set
IJC

node in the network.


of number of nodes N which are associated by a
network where every node in the network Achieving QoS using load balancing
comprises a set of resources. Each task arrived to
scheme
a set of nodes could either arrived from outside
The main problem in
or from other nodes in the network. All nodes in
software engineering principles is to identify the
the network should assign the same average
new set of techniques to deprive the difficulties
arrival rate of tasks from the network. Identify
arise in the distributed systems and dealt with
the set of users for requistion of nodes and
intelligently. For equalizing the workload among
packets in the distributed network.
the set of nodes in the network, in this work, a
 Nodes of the distributed crucial role is played by the balancing of the
system are arranged in the work load composed of interrelated nodes that
form of 1 to N, where N is the processes the network without proceeding with
total number of nodes in the the set of rules.
distributed system. Consider a network with a
set of interrelated nodes in which a network is
 Assign unique id to each set of
processed with interrelated nodes. Each node is
nodes in the network.
processed with a set of resource to achieve the

IJCSRTV1IS050058 www.ijcsrt.org 5
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

specified task. Each node in the network will added into the resource pool. The capacity of
define its limit to process the assigned task with new arrival resource is determined as,
the required amount of resources. So, the main
objective of load balancing is to distribute the Cap( Ri )  P  MIPS ( P) ……………. Eqn 1
jobs to all the nodes in the network, it will: where
(1) Boost the processing rate of the network to
ti
achieve the assigned task; MIPS ( P) 
T  1,000,000
(2) Boost the number of nodes to increase the
quality of service over the network for
Set a counter C for each of
utilization.
the resource (both the existing and new arrival
For an effective set of
set of resources) to identify the number of
load balancing scheme, heuristic technique is
successful tasks in the system. Based on the
utilized here, an optimization method for
assignment of counter, the loads (set of tasks to
identifying the presence of hassles. The presence
be processed) in the distributed system are
of hassles is identified based on the resource
balanced. For instance, if the new task enters for
demand over the resource pool. When the
processing, then it will search for the suitable
demand occurs, the task assigned to the resource
T resource (based on the resource capacity and the
might get interrupted. So, a hassle free
task processing time) needed to process the task.
SR
distributed system is built for balancing the load
Before assigning the task to the particular
effectively.
resource, it is necessary to check with the
IJC

Assume a system which is


counter value. If the task is completed with the
distributed in nature contains a numebr of tasks
assigned resource, then the counter value of the
referred as tn with the set of resources R, where n
resource will increase. The counter value
is the total number of tasks waiting in the queue
increases based on the number of successful
for requesting the resources to accomplish the
tasks processing in the particular resource are
task. The tasks in the waiting queue Q are
done. The possibility of assigning the new task
monitored by the user agent UA. If the user
with the resource Ri determined as,
agent UA wants to enter into the network, then
Ci (t ) a  init ( Ri )
the resource capacity is identified based on the P( Ri )  …………
two metrics,  (Ci (t ) a  init ( Ri ))
n
 Processing eqn 2
elements P and Where i and n is the
 MIPS rating of available set of resources
the processing Ci(t) – current processing
element task in the resource
With these metrics, the capacity of Init (Ri) – initial value of the
the new arrival set of resources are identified and resource i

IJCSRTV1IS050058 www.ijcsrt.org 6
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

a- intensity of the initial based validation is a technique which controls


performance values the system access to authoritative users. The task
By setting up the counter based validation scheme assigns an individual
value, the tasks are processed with a given set of role to users in the network based on their job
resources by specifying the tasks and resources over the system. The capabilities of the users and
required. When adapting this load balancing their limits to the network boundary are
scheme in traffic engineering principle, the identified to assign a suitable role for them to
occurrence of traffic through the tasks processing avoid the network traffic.
is resolved. For this case, heuristic technique is To achieve the defined operations for
applied as an optimization technique for the users in the network, specific roles are
resolving the data traffic in the distributed assigned. The essential view of task based
systems. validation scheme is that the access permissions
In the HCO based traffic engineering, over the system for processing the task are
the selection of node is done for processing the assigned based on their roles. Task based
task based on their service so as to plan the validation scheme provides limit by approving
traffic flows onto a distributed system domain. the set of users access to definite data or
HCO based traffic engineering algorithm resources. There are some crucial rules for task
T
processes a set of user requests for resources are based validation scheme which is defined as:
SR
received at the system domain. In each set of
users‟ requests, the amount of resources they  Assigning a role
IJC

needs and the service it provided are clearly to user


specified. Based on the information in the users‟  Authorizing the
query, the HCO technique processed the request role
and optimizes the specified path in the request  User access
query. The status of the users‟ query processed is permission
determined based on maintaining the table at authorization
which the nodes in the network are processed.
By the way, the load balancing is achieved.
The first rule describes the process
of assigning a role to the set of users based on
Task based validation scheme for processing the task in the network system. The
security
second rule depicts the process of user role
Even when the load balancing is authorization based on their successful
effectively achieved, the data in the network are accomplishment of task over the network. The
less in security level, because of distributed in third rule describes the process of depicting the
nature. So, for this case, a task based validation user access permissions for each set of individual
scheme is introduced in this section. A task- users which are independent to each other. The

IJCSRTV1IS050058 www.ijcsrt.org 7
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

task of the individual set of users is assigned the diverse set of user roles, the user requests in
with the administrative roles which are the traffic is analyzed and processed.
maintained as a list in the distributed systems.
The algorithm below describes the
The lists are,
process of enhancing QoS and security in
 Resource list, set
distributed systems is described as follows,
of resources the
authorized users
can access Input: Set of resources R, set of
permissions tasks T
 Service list, list
Assign U as a set of users
of services user
NewU is the new user and ExisU is the
roles are defined
existing user
and accessed.
R is the role assigned to the user
L is the limit given to the users U for security
 User role list, a
purpose
list of resources and
Perm is the permission given to the set of
services the users
T
//users U
have granted access.
Load balancing scheme
SR
The users in the distributed systems are assigned Begin
with the set of individual roles to protect the data Initialize the available set
IJC

in the traffic roles are assigned as supervisor, of resources in the distributed system
analyzer, Evaluator. All these set of roles Determine the processing
maintain the above set of lists to protect the data tasks P
in the network. Determine the MIPS
rating of the particular task tn
At first, the supervisor analyzes the
Identify the number of
users requests and resources for processing the
tasks to be processed as n
tasks assigned to them. The supervisor sets a
Do
state based on the assignement of resources. The
state is being positive if it successfully assigned Begin
the resources to the tasks in the distributed Choose the task ti from tn

system. The analyzer monitors the activities of Identify the resource Ri to

the users and its processed job over the assign the task ti based on the cap(Ri)

distributed system. By examining the assigned Assign the task

tasks for the respective user role, the If task successfully


implementation is assigned with it. After completed
analyzation, the evaluator evaluates the data Flag=flag+1
packets of the assigned set of users. So, based on Else

IJCSRTV1IS050058 www.ijcsrt.org 8
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

accomplished. By following the subsequent


steps, the security and quality is achieved.
The performance of the proposed
Keep the task in the task set T
ASEP is measured in terms of quality assurance,
Wait for the appropriate resource secuirty and communication cost.
End If
While (tn = NULL) EXPERIMENTAL EVALUATION
End do
Experimental evaluation is done to
End
estimate the performance of the proposed QoS
accomplishment in distributed systems using
// Task based validation scheme
software engineering principles [ASEP].
If NewU enters into the network Experiments are conducted with the Solar Flare
Identify the L assigned to the user Data Set from UCI repository against existing
Fig 3.2 Process of achieving QoS in
If (L=1) non-parametric approach for the process of
distributed systems
providing security and quality to the distributed
Continue to process
systems.
T
Exit
The description of solar flare data set is
SR
Else defined as follows (Table 1):
Abort the user requests The solar flare database contains 3
IJC

End If potential classes, one for the number of times a


certain type of solar flare occurred in a 24 hour
For each users U
period. Each instance represents captured
Identify the R (U)
features for 1 active region on the sun. The total
End For
number of instances used here are 323 and
If R (U) has authorized set of packets second set of data instances are measured to be
Permit the user to its 1066. The total number of attributes used here
assigned limits are 13 with 3 set of class attributes.
Else
Check with the limits of the actual assigned
Dataset name Solar flare dataset
tasks
No. of instances 323, 1066
End If
Total number of 10
End
attributes
Number of class 3
Fig 3.2 describes the process of achieving the attributes
QoS in distributed systems is effectively Table 1 Details of solar flare dataset

IJCSRTV1IS050058 www.ijcsrt.org 9
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

The values of the attribute information are With these set, the experiments are
described in the subsequent table (table 2). conducted to estimate the performance of the
proposed ASEP scheme.

Attribute Values
Quality assurance
Code for class (A,B,C,D,E,F,H)
Quality assurance (QA) is defined as
the engineering behaviors for providing a quality
Code for largest (X,R,S,A,H,K)
spot size distributed system so that task for the specified
set of service is monitored. It is the methodical
Code for spot (X,O,I,C)
dimension, and a connected loop that presents
distribution
error preclusion.
Activity 1-reduced, 2-
unchanged Security
Security is the extent of confrontation
Evolution 1-decay, 2 – no
growth, 3 - growth to, or fortification from the unauthorized access
to users in the distributed systems. Security
24 hour flare 1-nothing as big as
achieves the secure process of tasks assigned
activity code an m1, 2- one M1
T with the users in the network without any
Historically 1-yes, 2- No
SR
complex interruption.

Area 1-small, 2- large


Execution time
IJC

Area of the largest 1 = < =5, 2 = >5 The time taken to achieve the
spot
reliable security and QoS (considered both the
load balancing and task based validation scheme)
Table 2 Attribute description
to process the data in the dataset is referred as
From all these predictors three set of flare classes execution time. It is measured in terms of
are predicted, which are represented in the last seconds.
three columns. 5. RESULTS AND DISCUSSION
ASEP scheme in distributed
C - Common flares - Class flares production in system is compared with the existing distributed
24 hours; admission control model [2] for measuring the
M - Moderate flares - Class flares production in rate of security, quality assurance and execution
24 hours; time. The below table and graph describes the
X - Severe flares - Class performance of the proposed ASEP Scheme in
flares production in 24 hours; distributed systems and existing self protection
scheme in clustered distributed systems.

IJCSRTV1IS050058 www.ijcsrt.org 10
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

No. of Quality assurance (%) proposed ASEP model provides good quality of
assurance. Because, the proposed ASEP model
tasks Proposed Existing
balanced the task loads in the distributed systems
ASEP distributed efficiently by implementing the load balancing
scheme admission scheme in the distributed systems. So, the data
control model traffic rate occurrence is less in the proposed
ASEP model, contrast to the existing distributed
25 62 40
admission control model. The variance in the
50 69 46
quality assurance is 10-155 high in the proposed
75 73 52 ASEP approach.
100 76 58
125 81 62 No. of Security (%)
150 85 66 users Proposed Existing
200 89 76 ASEP distributed
scheme admission
Table 5.1 No. of tasks vs. quality control model
T
assurance 50 54 40
SR
The quality assurance is measured based on the 100 62 46
number of tasks present in the distributed
IJC

150 69 52
systems. The value of the proposed ASEP
scheme is compared with the existing distributed 200 75 58
admission control model is illustrated in table
250 82 64
5.1.
300 86 72
100
Quality assurance (%)

80

60
350 89 79
40

20

0
25 50 75 100 125 150 200 Table 5.2 No. of users vs. security
No. of tasks

Proposed ASEP scheme Existing distributed admission control model

The security of the users‟ data is measured based


Fig 5.1 No. of tasks vs. quality
on the number of users in the distributed
assurance
systems. The value of the proposed ASEP
Fig 5.1 describes the quality assurance is
scheme is compared with the existing distributed
measured based on the number of tasks present
admission control model is illustrated in table
in the distributed systems. Compared to the
5.2.
existing distributed admission control model, the

IJCSRTV1IS050058 www.ijcsrt.org 11
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

No. of Execution time (seconds)


100
tasks Proposed Existing
80
Security (%)

ASEP distributed
60
scheme admission
40

20
control

0 model
50 100 150 200 250 300 350 50 14 20
No. of users
100 20 26
Proposed ASEP scheme Existing distributed admission control model
150 23 31
200 28 35
Fig 5.2 No. of users vs. security 250 35 40
300 39 43
Fig 5.2 describes the security level of the users 350 46 50
and their appropriate data in the distributed
Table 5.3 No. of tasks vs. execution
systems. Compared to the existing distributed
T
time
SR
admission control model, the proposed ASEP
The execution time is measured based on the
scheme provides high level of security. Since the
number of tasks present in the distributed
IJC

ASEP scheme is processed based on the


systems. The value of the proposed ASEP
assignment of role for the set of users involved
scheme is compared with the existing distributed
in the distributed systems, with the specified
admission control model is illustrated in table
limits are known to all the participants. The
5.3.
limits assigned to the user could not been crossed
with the limits. So the chance of loss of packets
Execution time (seconds)

60
50
in the distributed systems is less contrast to the
40
existing distributed admission control model. 30
20
The variance in the security level is 11-13% high
10
in the proposed ASEP scheme. 0
50 100 150 200 250 300 350
No. of tasks

Proposed ASEP scheme Existing distributed admission control model

Fig 5.3 No. of tasks vs. execution time

Fig 5.3 describes the execution time is measured


based on the number of tasks present in the
distributed systems. Compared to the existing

IJCSRTV1IS050058 www.ijcsrt.org 12
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

distributed admission control model, the assurance with high security by consuming
proposed ASEP scheme consumes less time to minimal communication cost.
achieve the quality over distributed systems.
Since, the load balancing is achieved before
REFERENCES
providing security to the specified tasks of the
[1] Repantis, T. , Xiaohui Gu ; Kalogeraki, V,.,
users in the distributed systems, the time
„QoS-Aware Shared Component Composition
consumption is less.
for Distributed Stream Processing Systems‟,
Finally, it is being observed that the proposed
Parallel and Distributed Systems, IEEE
ASEP scheme provides reliable Qualtiy of
Transactions on (Volume:20 , Issue: 7 ), 2009.
service over the distributed systems by timely
adapting the software engineering principles.
[2] Zhonghang Xia, Wei Hao ; I-Ling Yen ; Peng
Li, “A distributed admission control model for
6. CONCLUSION QoS assurance in large-scale media delivery
In this paper, software systems”, Parallel and Distributed Systems,
engineering principles is presented for achieving IEEE Transactions on (Volume:16 , Issue: 12 ),
the QoS and security over the distributed T 2005
systems. Load balancing and security are the two
[3] Chunyu Hu ET. AL.,, “A Distributed
SR
principles utilized for achieving the QoS in
distributed systems that provides the active Approach of Proportional Bandwidth Allocation
environment for packet transmission and for Real-Time Services in UltraWideBand
IJC

assignment of limits and capabilities to a user of (UWB) WPANs”, Parallel and Distributed
the database. For effective load balancing, Systems, IEEE Transactions on (Volume:21
heuristic optimization technique is presented , Issue: 11 ), 2010
with given communication pattern to produce an
optimal solution. The optimization technique [4] Yanchao Zhang, Yuguang Fang, “A Fine-
provides hassle free HCO algorithm used here Grained Reputation System for Reliable Service
for identifying the presence of hassles based on Selection in Peer-to-Peer Networks”, Parallel
the trial and error method. For security, the task and Distributed Systems, IEEE Transactions
based validation scheme is presented based on on (Volume:18 , Issue: 8 ), 2007.
the assignment of role and user access
permission limits supported with the [5] Eng Hwee Ong ; Khan, J.Y. ; Mahata, K.,
environment. Experimental evaluation is done “Radio Resource Management of Composite
with the dataset extracted from UCI repository to Wireless Networks: Predictive and Reactive
analyze the performance. Results revealed that Approaches”, Mobile Computing, IEEE
compared to the non parametric approach, the Transactions on (Volume:11 , Issue: 5 ), 2012
proposed ASEP provides 5-6% reliable quality [6] Jing Chen et. Al., “Opportunistic Optical
Hyperchannel and Its Distributed QoS Assuring

IJCSRTV1IS050058 www.ijcsrt.org 13
International Journal of Computer Science Research & Technology (IJCSRT)
ISSN: 2321-8827
Vol. 1 Issue 5, October - 2013

Access Control”, Parallel and Distributed


Systems, IEEE Transactions on (Volume:20
, Issue: 11 ), 2009.

[7] Morillo, P. et. Al., “A Latency-Aware


Partitioning Method for Distributed Virtual
Environment Systems”, Parallel and Distributed
Systems, IEEE Transactions on (Volume:18
, Issue: 9 ), 2007

[8] Huang, V. et. Al., “QoS based fair resource


allocation in multi-cell TD/CDMA
communication systems”, Wireless
Communications, IEEE Transactions
on (Volume:5 , Issue: 2 ), 2006.

[9] ZhengMing Shen ; Thomas, J.P., “Security


T
and QoS Self-Optimization in Mobile Ad Hoc
SR
Networks”, Mobile Computing, IEEE
Transactions on (Volume:7 , Issue: 9 ), 2008.
IJC

[10] Fallahi, A. et. Al., “QoS and energy trade


off in distributed energy-limited mesh/relay
networks: a queuing analysis”, Parallel and
Distributed Systems, IEEE Transactions
on (Volume:17 , Issue: 6 ), 2006.

[11] Qing Chen et. Al., “QoS-Aware


Cooperative and Opportunistic Scheduling
Exploiting Multiuser Diversity for Rate-
AdaptiveAd Hoc Networks”, Vehicular
Technology, IEEE Transactions on (Volume:57
, Issue: 2 ), 2008.

IJCSRTV1IS050058 www.ijcsrt.org 14

Vous aimerez peut-être aussi