Vol 11 No 1 - March 2014

ISSN: 1694-2507 (Print)
ISSN: 1694-2108 (Online)
International Journal of Computer Science
and Business Informatics
(IJCSBI.ORG)
VOL 11, NO 1
MARCH 2014

Table of Contents VOL 11, NO 1 MARCH 2014
Energy- Aware QoS Based Routing Protocols for Heterogeneous WSNs: A Survey ...............................1
Sridevi S., Rumeniya G. and Usha M.
Optimization of Outsourcing ICT Projects in Public Organizations; Case Study: Public Center of Iranian
ICT Studies ................................................................................................................................................. 20
Majid Nili Ahmadabadi, Abbas Bagheri and Fariba Abolghasemi
An Optimized CBIR Using Particle Swarm Optimization Algorithm ......................................................... 40
Subhakala S., Bhuvana S. and Radhakrishnan R.
Study of Satisfaction Assessment Techniques for Textual Requirements .............................................. 56
K. S. Divya, R. Subha and Dr. S. Palaniswami
Survey of MAC Protocols for Heterogeneous Traffic in Wireless Sensor Networks ............................... 67
Sridevi S., Priyadharshini R. and Usha M.
Harnessing Social Media for Business Success. Case Study of Zimbabwe ............................................... 80
Musungwini Samuel, Zhou Tinashe Gwendolyn, Zhou Munyaradzi and Ruvinga Caroline
Quality Platforms for Innovation and Breakthrough................................................................................ 90
Dr. Hima Gupta
Development of Virtual Experiment on Waveform Conversion Using Virtual Intelligent SoftLab ...... 107
Bhaskar Y. Kathane
IJCSBI.ORG

International Journal of Computer Science and Business Informatics
IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 1
Energy- Aware QoS Based
Routing Protocols for
Heterogeneous WSNs: A Survey
Sridevi S.
Associate Professor, Department of Computer Science and Engineering,
Sona College of Technology,
Salem, India
Rumeniya G.
PG Scholar, Department of Computer Science and Engineering,
Salem, India
Usha M.
Professor& Dean, Department of Computer Science and Engineering,
Salem, India
ABSTRACT
WSNs (Wireless Sensor Networks) are a huge collection of sensor nodes which have
limited battery power and limited computational capacity. The power limitation causes the
nodes to premature dead so the node power should be used efficiently to prolong the
network lifetime. In time critical applications, the data should reach the destination within a
deadline and without any packet loss which means the QoS metrics such as reliability and
delay are very essential for delivering the data to destination. One of the vital challenges for
research in wireless sensor networks is the implementation of routing protocols which
achieve both Quality of Service (QoS) and energy efficiency. The main task of the routing
protocol is to discover and maintain the routes to transmit the data over the network. At
present, to increase the performance of the networks, to achieve load balancing and to
provide fault tolerance multipath routing techniques are widely used rather than single path
routing technique. We present a review on the existing routing protocols for WSN by
considering energy efficiency and QoS. We focus on the main motivation behind the
development of each protocol and explain the function of various protocols in detail. We
compare the protocols based on energy efficiency and QoS metrics. Finally we conclude
the study by giving future research directions.
Keywords
WSNs, Routing Protocol, Multipath Routing, Fault Tolerance, Cross Layer Module.
1. INTRODUCTION
Wireless sensor network consists of number of sensor nodes deployed in the
target area to gather information, collaborate with each other and send the
gathered data to the sink node in a multi hop fashion [1]. In traditional
methods sensor nodes send their data directly to the sink node in a single-

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 2
hop approach. This has many drawbacks such as expensive and faster
energy depletion since the target sensing nodes are far away from the sink
node [2]. To overcome this drawback, multi-hop based approach is carried
out over short communication radius which saves energy and reduces
communication interference. Due to the dense deployment of the nodes we
can have multiple paths for data transmission from the source nodes to the
sink [3].
Many of the applications require QoS like military applications, fire
detection and biomedical applications. On the battlefield, sensors can be
used to detect unfriendly objects, vehicles, aircraft, and personnel. On the
health care applications [4], [5] and [6], smart wearable and companionable
wireless devices can be attached to or the sensors can be implanted inside
the human body to observe the essential signs of the patient body. The
routing protocols are required to choose the best path that satisfies the QoS
requirements as well as improves the lifetime of the network. The
characteristics of WSNs are rapid deployment, self-organization, and fault-
tolerance which make them adaptable for real time and non-real time
applications [7].
2. MOTIVATION
The sensor nodes are having limited energy, storage capacity and
bandwidth. The energy of the sensor nodes are consumed while sensing,
processing and transmission. So energy of the node should be used
efficiently to avoid early dead. In recent years, WSNs are used in mission
critical applications. For example, in fire detection application when the
event has detected, immediately the sensor node must gather and transmit
the information about the event to the sink within the deadline and without
any packet loss. But in many cases, the packets failed to reach the sink
within deadline and without any packet loss. The main reason for this is the
limited functionalities and inaccurate observation or low reporting rate of
the sensor nodes.
Many of the applications require QoS delivery for the data transmission.
The known fact is that the QoS always conflicts with energy efficiency
since the designs require more energy to minimize packet errors or failures
and to reduce latency. There are many existing routing protocols which try
to minimize the packet errors by considering retransmission which requires
more energy and to find best routing path for real time data, it needs to
perform some operations that also consumes more energy. Hence, a
thorough study has to be made to learn about the trade-off between energy
efficiency and QoS. The purpose of this survey is to focus on how the
WSNs provide the QoS and energy efficiency for real time applications.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 3
3. DIFFERENT KINDS OF ROUTING SCHEMES
The routing protocols are classified into three types according to their
characteristics: Proactive, reactive and hybrid routing [8]. The routing
protocols can be classified according to their operations as follows: Route
construction, Network Structure, Communication Model, Number of paths
and QoS [9]. The routing protocols dependent to the network structure are
further classified into flat routing or hierarchical routing. The
communication model based routing protocolscan befurther classified into
three ways:Query-based, Coherent and non-coherent based andNegotiation-
based [9].
3.1 Classification of routing protocols according to route construction
The three different routing strategies are identified in wireless networks:
proactive, reactive, and hybrid. In case of proactive routing, all the paths are
constructed by periodically broadcasting control messages before they are
actually needed then these constructed paths information are stored on the
routing table of each node. In case of reactive routing, the paths are
constructed between source and destination only when needed and it is
dependent on dynamic route search. The hybrid routing strategy relies on
both proactive and reactive routing protocols to achieve stability and
scalability in large networks.
3.2 Classification of Routing Protocols based on Network Structure
The nodes in a sensor network can be organized in one of the following
three ways: flat, hierarchical based and location based. In flat routing
protocols all the nodes are treated in the same way and they have minimal
overhead to maintain the infrastructure between the interacting nodes. In
hierarchical routing strategy, the nodes are grouped into clusters. Each
member in the cluster sends data to the corresponding cluster head which
aggregates the data and forwards to the sink through multiple hops. The
election algorithm selects the cluster heads based on parameters like
residual energy and distance. The cluster head has the additional
responsibility of coordinating the activities of its members and forwarding
data from one cluster to another.
3.3 Classification of Routing Protocols based on communication model
The routing protocol based on communication model can be classified into
two types according to their operations: negotiation based routing and query
based routing. The negotiation based protocols tries to eliminate the
redundant data by including high level data descriptors in the data
transmission. In query based protocols, the sink node starts the
communication by distributing a query for data over the network [10].

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 4
3.4 Classification of Routing Protocols based on number of paths
Based on the number of paths used to route data from sensor nodes to the
sink node, routing protocols are divided into single path routing protocols
and multi path routing protocols. In single path routing one path is
constructed from source to sink to route the data. Due to this the nodes in
the selected path may die soon and the network lifetime is reduced. To
improve the network lifetime and reliability multi-path routing protocols are
proposed which construct multiple paths to achieve load balancing, fault
tolerance. The wireless sensor network routing can be made very efficient
and robust by incorporating different type of local state information such as
Link quality, distance between the nodes, Residual energy, Position
information etc. Disjoint Path routing protocols [11] construct multiple
disjoint paths between source and destination in one of two ways: Link-
disjoint path: The paths between source and destination have no common
link. Node-disjoint path: The paths between source and destination have no
common node. The both link disjoint path and node disjoint path have one
active path, and number of backup paths. A service flow will be redirected
to the backup path if the active path fails. Load balancing is another
important aspect to avoid network congestion and optimize network
throughput and to prolong the network lifetime.
3.5 Classification of Routing Protocols based on QoS
The Quality-of-Service (QoS) provisioning in WSNs is a challenging task,
because of two reasons. First, resource constraints, the dynamic network
topology, unbalanced traffic, data redundancy, scarcity of node energy,
energy consumption for computation and bandwidth pose challenges on the
design of QoS support routing protocol in WSNs [12]. Second, there exist
wide differences in traffic generation rate, latency and reliability amongst
the data packets. The QoS based protocols aims to achieve QoS metrics
such as reliability, delay, energy efficiency and throughput [13].
The rest of the paper is structured as follows. Section 2 describes the
taxonomy of recently proposed routing protocols for wireless sensor
networks. Section 3 compares the studied protocols based on QoS metrics,
energy efficiency and path selection criteria. Section 4 concludes and gives
future research directions.
4. TAXONOMY OF EXISTING ROUTING PROTOCOLS FORWSNS
4.1 Energy efficient and QoS based routing protocol (EQSR)
The Energy efficient and QoS based routing protocol (EQSR) [7] is
designed to satisfy the QoS requirements of real-time applications. . To
increase reliability EQSR uses multipath routing and XOR-based Forward

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 5
Error Correction (FEC) technique which provides data redundancy during
the data transmission.To meet delay requirements EQSR employs queuing
model which classify the traffic into real-time traffic and non-real-time
traffic through service differentiation technique. To find out the path EQSR
executes three phases: Initialization phase, Primary Path discovery phase,
Alternative Paths discovery phase. During the initialization phase each
sensor node broadcasts a HELLO message to its neighbor nodes. The
HELLO message includes fields for source ID, hop count, residual energy,
free buffer and link quality which are used to calculate the link cost as given
by equation (1).
α Eresd ,y + β Bbuffer ,y + γ Iinterferrence ,xy (1)
In Primary Path discovery phase, the sink node starts to find the routes
through sending RREQ message to its preferred neighbor chosen by the
equation (2). This process is continues until the source node receives the
RREQ message.
Next_hop = Maxy∈Nx {α Eresd ,y + β Bbuffer ,y + γ Iinterferrence ,xy }(2)
Where,Nxis the neighbor set of node x. Eresd,yand Bbuffer,y depicts the residual
energy and free buffer size at neighbor y, respectively. Iinterference,xy is
thesignal to noise ratio between node x and node y.
EQSR constructs node disjoint multiple paths during Alternative Paths
discovery phase. In this phase, the sink sends RREQ message to its next
most preferred one hop neighbor to construct alternative paths after the
construction of primary path. To construct node disjoint paths EQSR
restricts that each node should accept only one RREQ message. For that
reason each node accepts the first RREQ message and discards remaining
messages. The number of required paths k can be estimated according to the
need of successfully delivering a message to sink by using the equation (3).
k = xα. pi(1 − pi)N
i=1 + pi
N
i=1 (3)
Where, xα is the corresponding bound from the standard normal distribution
for various levels of α and piis the probability of successfully delivering a
message to sink.
EQSR calculates the transmission delay of paths by measuring the
propagation delay of RREQ message and gives the best paths for real-time
traffic and remaining paths for non-real-time traffic. The algorithm find out

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 6
k node disjoint paths, out of which l paths are used for sending real time
data and m paths are used for sending non real time data. Finally, Error
Correction Codes (ECC) for data packets is calculated by lightweight XOR-
based FEC algorithm. The EQSR improves the QoS metrics such as
reliability and delay, but it introduces high control overheadbecause of FEC
mechanismwhich performs the encoding and decoding operations.
Simulations are done in Ns2 and the results depict that the EQSR protocol
performs very well than MCMP protocol for real time traffic. But the
MCMP outperforms the EQSR protocol for non-real time traffic since
additional delay is introduced in EQSR due to the queuing model.EQSR
offers lower energy efficiency than MCMP since some energy is wasted for
calculating the FEC. The packet delivery ratio is increased in EQSR than
MCMP because the EQSR uses forward error correction (FEC) technique.
4.2 Localized Multi Objectives Routing protocol (LOCALMOR)
The new localized multi objectives routing protocol [14] differentiates the
data traffic according to their requirements of QoS metrics. It classifies the
traffic into critical packet, delay sensitive packet, reliable sensitive packet
and regular packet. For each data packet, this protocol tries to satisfy the
required QoS in an energy efficient way. To improve the reliability it
considers multi-sink single-path approach. The neighbor manager is
accountable for executing HELLO packet, implementing estimation
methods and running other modules. The neighbor table is updated by
HELLO packet which has the information related to node‟s current position,
residual energy, and estimated packet reception ratio and transmission delay
for each packet transmission. The sending node vi considers the time
window which is specified in terms of the number of packet transmitted and
the receiving node vjupdates its current window in terms of the number of
packet successfully received denoted as r and number of known packet
missed denoted as f. The number of transmitted and received packets can
be calculated with the help of sequence number of each packet. When the
current window size is equal to main window size then the link reliability
(or packet reception ratio) between node Vi and node Vj (prrvi,vj) is
calculated by using the estimator called Window Mean Exponential
Weighted Moving Average (WMEWMA) in regular time interval shown in
equation (4). The initial value of prrvi,vj is zero.
prrvi,vj = α. prrvi,vj + 1 − α
r
r+f
(4)
Here, α is a tunable parameter of the moving average. The delay can be
calculated by using equation (5) and (6) with the help of EWMA estimator.
To estimate the delay it considers both queuing delay and transmission

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 7
delay. This protocol uses several queues in which each type of packet is
inserted into a separate queue. The queuing delay is different for each
packet type and it is calculated through local time stamp in terms of exact
waiting time of each packet.
wVi packet. type = α. wVi packet. type + 1 − α . ω (5)
dtrVi = α. dtrVi + (1 − α)(tACK − size ACK bw − t0)(6)
Wherewvi is queuing delay, wvi[packet.type] is queuing delay for each type
of packet, dtrvi is transmission delay,t0 the time the packet is ready for
transmission, tACKthe time of the reception of acknowledgment (ACK)
packet, bw the bandwidth, and size(ACK) the size of the ACK packet. The
initial value of wvi[packet.type] and dtrvi is zero.
This protocol has different modules namely energy module, reliability
module and latency module. The energy module considers both
transmission cost and residual energy of routers to attain power efficiency.
For that, the min-max approach is used to find the energy efficient node.
Reliability module achieves the required reliability by sending a copy the
data packet to both primary and secondary sinks. When more than one node
has the same value for maximum reliability, the most power efficient node
is selected by energy module. Latency module calculates the required speed
by dividing distance by the time remaining to the deadline, rt. The
remaining time to deadline rt is calculated by equation (7).
rt = rtrec − (ttr − trec + size bw) (7)
Where trec represents the reception time, ttr the transmission time, rtrec is the
previous value of rt. If the incoming packet is delay sensitive packet then it
selects the node which meets the required deadline. If more than one node
satisfies the required deadline then the most energy efficient node is
selected. If the incoming packet is critical packet then it first calls the
reliability module then latency module and energy module. Finally the
queuing manager uses the multi-queue priority policy in which four
separated queues are used for each type of packet. Critical packet has the
highest priority than Delay sensitive packet and reliability sensitive packet
has lowest priority. To avoid starvation a time out policy is proposed for
each lower priority queue. When a packet arrives at a queue, a timeout value
is assigned and when the timer expires the packet is moved to the highest
priority queue.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 8
Simulation results depict that Packet reception ratio increases linearly from
86 to 87 percent for critical packets and 86 to 98 percent for reliable
sensitive packets whereas it is constant for delay-sensitive packets at the
interval of 80 to 83 percent. Moreover, above 96 percent of packets were
successfully transmitted with reasonable delay.The energy deviation is
small for low and moderate number of critical packets but the energy
deviation is gradually increases as the number of critical packet
increase.However, the LOCALMOR protocol achieves better lifetime than
any other existing protocols.
4.3 QoS-aware Peering Routing Protocol for Reliability Sensitive Data
(QPRR)
Zahoor et al. proposed a novel routing protocol in consideration of the QoS
requirements of body area networks (BAN) data. This QoS-aware Peering
Routing Protocol for Reliability Sensitive Data (QPRR) [4] protocol
improves the reliability of critical BAN data while transferring the data
from source to destination. For sending reliable sensitive packets (RSP), the
protocol calculates the reliability of all possible paths. These path
reliabilities can be obtained by using neighbor table information. The
routing table can hold up to three most reliable paths among all possible
paths.
To transmit any RSP data between source and destination it should consider
the following criteria. If the first path itself can accomplish the reliability
requirement then the source node transmits RSP through it. If the first path
reliability is lower than required reliability then QPRR aggregates the
reliability of two paths and then QPRR compares the required reliability
with two paths aggregated reliability. If the two paths aggregated reliability
is greater than required reliability then the copy of RSP packets transmitted
through two paths. If not QPRR aggregates three paths reliability then
compares it with required reliability. If the three paths reliability is greater
than required reliability then the copy of RSP packet transmitted through
three paths. Otherwise the packet is dropped. The path reliability between
source „i‟ to destination „Dst‟ is calculated by using the following equation
(8).
Rpath (i,Dst ) = Rlink (i,j) × Rpat h(j,Dst ) (8)
The link reliability between nodes „i‟ to node „j‟ can be calculated by using
EWMA (Exponentially weighted moving average) formula as follows:
Rlink (i,j) = 1 − α Rlink i,j + α. Xi(9)

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 9
The average probability of successful transmission is calculated by using
equation (10).
Xi =
NAcks
NTrans
(10)
Where,Rpath(i,Dst) is the path reliability between node „i‟ to destination.
Rlink(i,j) is the link reliability between node „i‟ to node „j‟. Rpath(j,Dst) is the
path reliability between node „i‟ to destination. α is the average weighting
factor that satisfies 0< α ≤ 1.This protocol takes α as 0.4. NAcksis the number
of acknowledgement received and NTrans is the number of packets
transmitted.
Simulation results show that the QPRR reliability is above 75% for low
dense nodes and above 74% for high dense nodes and it uses low
transmission power which provides better transmission rate.The QPRR
provides better reliability but the drawback is the network traffic load is
increased.
4.4 Energy Efficient Node Disjoint Multipath Routing Protocol
(EENDMRP)
The Energy Efficient Node Disjoint Multipath Routing Protocol
(EENDMRP) [15] provided for the reliability analysis of route redundancy
in WSN. EENDMRP concentrates on route redundancy in a single node
level redundancy over a single path, single node level redundancy through
multi node over single path, and single node level redundancy through
multiple level multiple nodes in a single path. EENDMRP is a proactive
protocol and it considers number of stages between source and destination.
The sink node is at stage zero. The one hop neighbors of sink node are stage
1likewise for each node a stage is assigned towards source node. This is
done for avoid the construction of path with loops. It considers the node
which has residual energy greater than threshold energy during path
construction in the WSNs.
To construct the route each node exchanges the route construction (RCON)
packet. If the RCON packet is received by node which is not in the route
that reaches the sink then the node processes the RCON packet. If the
RCON packet is received by node which is already in the route that reaches
the sink then it compares the node‟s hop count value with packet‟s hop
count value. If the node‟s hop-count value is greater than packet‟s hop-
count value and the node‟s residual energy greater than threshold energy
value then RCON is processed. If not, it drops the packet. Each node‟s
routing table is updated while receiving RCON packet which has the fields
such as node id and hop-count value. Finally, all possible node disjoint
paths are constructed between source and destination. If any node in the

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 10
path failed to transmit the packets due to node death or node dislocation,
then EENDMRP reports the source node by sending route error packet
(RERR). The source node removes the failed path from the routing table
and calls the route maintenance phase then the alternate path is provided
between the node which created the RERR packet and sink node.
4.5 Lifetime Maximizing Dynamic Energy Efficient Routing Protocol
In [16], the authors proposed Energy efficient routing protocol to balance
the energy consumption among nodes and to avoid the premature death of
nodes. The proposed energy efficient routing protocol has three phases
namely initialization phase, selection of next hop and generation of DEERT
phase and tree maintenance phase. During the initialization phase a level is
assigned to each node based on the hop distance from the sink node which
at level 0. A node can selects its next hop from lower level or in the same
level. The data packets are transmitted from higher level node to lower level
node. Every node selects the next hop neighbor based on the cost of the link
between itself and its neighbor and the load of the neighbor. The link cost
between the nodes u and v is calculated by equation (11).
Cuv = min⁡{REu − Etx , REv − Erx }(11)
Where, Etx is transmission cost of node. Erx is reception cost of node and
REu and REv is the residual energy of nodes u and v respectively.
The load of node is calculated based on the sum of energy consumed for
transmission of a packet to a neighbor node and energy consumed for
receiving a packet from the children nodes and energy used for overhearing.
In the tree construction phase, a distinct energy efficient routing tree rooted
at the sink node is constructed based on the link cost for efficiently routing
the data. After a fixed amount of time, the tree is reconstructed again.
The tree maintenance algorithm reconstructs the tree in the
following cases:
 If there is no response from neighboring nodes then that node is
considered as dead node.
 If the residual energy of the neighbor node is lower than threshold
value.
 If there is no appropriate next hop node then the source node
transmits its data directly to the sink node and updates its level and
other parameters consequently.
Simulation results depict that the DEERT has a better performance than
SBT, DEBR and aggregation tree based routing in terms of number of nodes
alive after certain number of rounds thus improving the lifetime of the

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 11
network. In the beginning, the DEBR end to end delay in terms of hop count
is lesser than proposed protocol whereas the end to end delay of DEBR is
increased when the number of rounds increases. This protocol concentrates
only on energy efficiency but does not support for QoS.
4.6 Braided multipath routing protocol
The braided multipath routing protocol [17] is provided for transmitting the
data packets from source to destination and giving the network ability to
adapt to fluctuations or failures. The source node constructs the path after
detecting the target. Once the target is detected then the node sends its ID in
a packet declaring that it has attained a target. The node which receives this
packet will response with its own ID to the source node declaring that the
preceding node as part of its path. Then the new node broadcasts its ID to
the next hop which response and forward the message likewise the process
is continued until it reach the sink. Then several paths will be created from
source to destination. For that, the destination node will give priority
numbers to the paths and they select the path which has the minimum
number of hops to the source and the nodes in that path are informed to
selected backup nodes.
To save the energy of all other nodes in the network, the nodes are entering
into an energy saving mode and activating from time to time to check the
ups and downs in the network. If any packet is transmitted from target to
sink then the sink will check its own route to this target and update its path
when the received one has minimum number of hops than the stored one.
Simulation results show that the braided algorithm uses backup nodes which
improve fault tolerance in the network. It is possible that only one or two
backup nodes can be established by path, leaving the other nodes without
backup thus making the path susceptible. In networks of higher density the
backup nodes improve fault tolerance at low costs.
4.7 Link Quality estimation based Routing protocol (LQER)
The LQER (Link Quality estimation based Routing) protocol [18] is
designed to improve reliability and energy efficiency in WSNs. It
incorporated minimum hop count field and dynamic window concept (m;
k). A path is constructed between the source and the sink nodes based on the
hop-count value. The sink node broadcast an advertisement (ADV) message
to its neighbors by setting the hop-count value as zero. For other nodes in
the networks, the hop count value is calculated based on the number of hops
of that node to the sink.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 12
If the current hop-count value is same or greater than next hop-count value
then that node is added as a forwarding node in the path. Or else the
message is rejected. Here, m is the number of data packets successfully
transmitted and k is the total number of packets transmitted. The dynamic
window concept is used to record the historical link status of data packets
based on m and k. The sufficient reliability can be achieved by using
historical link status information which keeps the word of k bit. If the data
transmission is not successful then that bit is represented as 0. Otherwise it
is represented as 1. The leftmost bit is oldest bit while the right most bit is
newest bit. When the new packet is transmitted, all the packets in the word
of k bits are shifted one position to the left and one bit is added in right most
position to indicate the current status. The quality of link p is calculated by
equation (12).
p = m
k (12)
The historical link table can be updated dynamically with a low computing
cost and complexity. When the routing data is ready to transmit, LQER lists
all the neighbor nodes of current node and chooses the path with largest
value of p to transmit routing data.
Simulation results show that Successful transmission rate in LQER is
greater than that in MHFR and MCR. When the number of nodes increases,
the deviation is small in LQER, which specifies a good scalability of data
delivery effectiveness whereas the successful transmission rate decreases
rapidly in MHFR and MCR.
4.8 QoS-aware Peering Routing Protocol for Delay Sensitive
Data(QPRD)
The QoS-aware Peering Routing Protocol for Delay Sensitive Data (QPRD)
[5] is provided for handling delay-sensitive packets. It calculates the node
delay and path delay of all constructed path between source and destination
and finds the best path among all possible paths according to the delay
requirement. Each node has a routing table which contains information of
next hop with the lowest end to end delay. A delay sensitive packet (DSP) is
transmitted in a path if the latency of the path is less than or equal to the
delay requirement of the packet.
QPRD has other modules to choose the best path for transmitting the packet.
They are MAC receiver module, Delay module, Packet classifier module,
Hello protocol module, Routing service module, QoS-aware queuing
module and MAC transmitter. The MAC receiver forwards the packets only

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 13
if the packets MAC address matches with its own MAC address. The delay
module calculates node delay by using the equation (13).
DLnode i = DLtrans (i) + DLqueues +channel + DLproc (13)
Where, DLqueue+channel are queuing and channel delay, DLtrans(i) is
transmission time of a packet, DLproc is processing delay of a node. The
transmission time is calculated by dividing the total number of bits in each
packet by data rate. Exponentially Weighted Moving Average (EWMA)
formula is used to estimate queuing and channel delay. The path delay of
node i to destination DLpath(i,Dst) is calculated by using equation (14).
DLpath (i,Dst ) = DLnode (i) + DLpath (j,Dst ) (14)
The packet classifier differentiates data packet and Hello packets and the
packets are processed according to the type. The hello packet is broadcasted
to each neighbor node. In hello protocol module, the neighbor table
constructor constructs the neighbor table based on the node delay and path
delay. The routing services module is accountable for creating the routing
table and classifying the data packets into Delay-Sensitive Packets (DSPs)
and Ordinary Packets (OPs). For DSP, it chooses the path with minimum
end to end delay. For ordinary packet (OP), it chooses the energy efficient
next hop. The QoS-aware Queuing Module (QQM) separates the data
packets into DSP and OP. It maintains separate queue for each type of data
packet. The DSP has the highest priority than OP. The OP queue can
transmit its data only if the DSP queue is empty. For fair treatment of lowest
priority data, a timeout policy is used. Finally the MAC transmitter receives
all packets and stores it in queue. It transmits the packet in first in first out
policy.
Simulation results show that in static environment 94% of the DSPs are
transmitted within the deadline limits and in mobile environment it provides
an improvement of 35% than DMQoS.
4.9 Energy aware peering routing protocol (EPR)
The energy aware peering routing protocol (EPR) [6] is designed to provide
a reduced network traffic load, improved energy efficiency and improved
reliability. It selects the next hop which has higher battery power and
shorter distance to the sink. It has three main parts namely hello message
module, neighbor table construction module and routing table creation
module. The hello message module is used to update the neighbor node
information such as destination location, destination ID, sender node‟s ID,
distance from next hop to destination and residual energy of neighbor node.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 14
The neighbor node information will be added in the sender node‟s neighbor
table by using hello protocol. If a node does not receive any hello message
from its neighbors for a particular time then it assumes that the neighbor has
moved away or the link to the neighbor has broken down. The distance
between the nodes i and DST can be calculated by the following equation
(15).
D(i,DST ) = (Xi − XDST )2 + (Yi − YDST )2 (15)
Where, Xi, Yi denote the X, Y coordinates of node i. XDST, YDST denote the
X, Y coordinates of the destination. The communication cost can be
calculated by using the parameters such as distance between two nodes and
node‟s residual energy. The routing table will selects the neighbor node with
lower communication cost from the neighbor table.
Simulation results show that about 34% of average traffic load is decreased
and about 23% of data transmission rate is increased than other similar
protocols.
4.10 Integrated link quality estimation-based routing Protocol(I-LQER)
I-LQER (integrated link quality estimation-based routing protocol) [19] is
designed to provide quality of service and to reduce power consumption. I-
LQER assigns different weights for the link quality records and link
stability is calculated based on this value. The link quality is estimated by
considering weighted factor along with m / k. Here, m is the number of data
packets successfully transmitted and k is the total number of packets
transmitted. It selects the node which has the greatest link quality.
It believes that the nearest period of transmission has relevance with current
transmission. If the node has a high probability to maintain the current link
quality then that node is taken as a good stability node. If the node has a low
probability to maintain the current link quality then that node is taken as a
low stability node. It compares the nodes record status in the nearest period.
Based on that, it selects the best node to forward the data. For example, if
we consider two nodes P and Q with link quality record status as 00 0011
1111 1111 and 11 1111 0100 0000 respectively where 1 denotes the good
link quality and 0 denotes the bad link quality, then the node P has a better
link quality stability than node Q.
Simulation results depict that the performance of I-LQER is superior to
LQER protocol in terms of end to end delay. For a network with 10 sensor
nodes, I-LQER gives an average delay of 9.00 ms and LQER gives an
average delay of 10.63 ms.when the number of nodes is increased to 100,
then I-LQER offers and average delay is 19.80 ms and LQER gives an

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 15
average delay of 28.03ms. This shows that I-LQER has a better scalability
than LQER.
5. COMPARISON OF DIFFERENT ROUTING PROTOCOLS
We compare the studied protocols based on reliability, delay, energy
efficiency and load balancing issues. Maximum number of protocols studied
in this paper construct single path to deliver data from the source to the sink.
Some protocols construct multiple paths for data delivery. All the protocols
use different criteria for the path selection. Almost all the protocols studied
focus on energy efficient routing. Table 1 and 2 gives the results of our
comparison. Only few protocols like LOCALMOR and EQSR provides
QoS support for heterogeneous traffic based on the type of traffic.
Table 1.Comparison of the routing protocols based on energy efficiency and QoS.
Table 2.Comparison of the routing protocols based on multipath support.
Scheme Reliability Delay (timely
delivery)
Energy
efficiency
Traffic
differentiation
Mobility
support
QPRR Yes No Yes OP, RSD Good
QPRD No Yes Yes OP, DSP Good
LOCALMOR Yes Yes Yes CSP, DSP, RSP Low
EQSR Yes Yes Yes Real time, Non
Real time
No
DEERT No No Yes -
-
EPR No No Yes OP Good
I-LQER Yes No Yes
-
Yes
Braided
multipath
algorithm
Yes No No
- -
EENDMRP Yes No Yes
- -
LQER Yes No Yes - -
Scheme Number
of Paths
Path
reconstruction
Path metric Load
balancing
Path
chooser
QPRR Up to
three
paths
No End to end
reliable path
- Source
node
QPRD Single
path
No Least end to end
delay path
- Source
node
LOCALM
OR
Single
path
No Minimum delay,
maximum
Yes Source
node

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 16
6. ERROR RECOVERY SCHEMES
6.1 Automatic Repeat Request (ARQ)
ARQ is the error recovery mechanism which uses the cyclic redundancy
check(CRC) technique to find error packet and it can retransmit the error
packet until the packet becomes error free at receiver side. If the packet is
successfully received by the receiver then it will send the positive
acknowledgement (ACK) to sender, otherwise it will send the negative
acknowledgement (NACK). If the ACK is not received by the sender within
the timeout frame then it will retransmit the packet. The drawback of ARQ
is retransmission which induces the additional cost.
6.2 Forward Error Correction (FEC)
FEC mechanism is mostly preferable in multi-hop WSNs to control the
packet transmission errors by adding the error correcting codes (ECCs) with
the sending data. The receiver can detect and correct the amount of bit
reliability and
maximum
residual energy
according to the
packet
requirement
EQSR Multi
path
No minimum end to
end delay path
for real time
traffic
Yes Source
node
DEERT Single
path
Yes Maximum
battery power
Yes Source
node
EPR Single
path
No End to End
Energy efficient
path
Yes Source
node
I-LQER Single
path
No Reliable path by
considering link
quality and link
stability
No Source
node
Braided
multipath
algorithm
Multi
path
No The path with
minimum
number hops
Yes Sink node
EENDMRP Multi
path
Yes The path with
minimum
number of hops,
maximum
residual energy
and maximum
path cost
Yes Source
node
LQER Single
path
No Reliable path by
considering link
quality
- Source
node

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 17
errors with the help of error correcting codes. However, the cost of
retransmission is very high since the FEC performs the encoding and
decoding operations which consume more energy.
7. CROSS LAYER MODULE
The concept of cross layer module is incorporating different classical layer
functionalities into a single functional protocol whereas the classical layer
structure is preserved i.e., the functionalities of each layer still remains
unbroken. Many cross layer module have been implemented to improve the
communication reliability, to improve energy efficiency and to avoid load
congestion. Most of the existing research integrates the MAC and physical
layers to reduce energy consumption and improve reliability, the MAC and
routing layers are integrated to extend the network lifetime, the routing and
physical layers are integrated to optimize the network throughput, the
transport and physical layer are integrated to control congestion [20] and the
application and MAC layer are integrated to provide QoS [21]. The cross
layer module improves the network performance and reduces the
implementation complexity and also outperforms the classical layer model
[22]. The network performance can be further improved while combining
multipath routing, FEC mechanism and cross layer module.
8. CONCLUSION AND FUTURE DIRECTIONS
The invention of smart, light-weight sensors makes the wireless sensor
network popular. Regarding the routing protocols, the reduced energy
consumption, the QoS, the scalability and the fault tolerance are the main
limitations in wireless sensor networks. This paper presents a study in what
way the recently proposed routing protocols are adapted to these
characteristics in WSNs. Although in the past years the energy efficient and
QoS based routing has been examined through various studies, yet there are
numerous significant research issues that should be further explored. The
Promising areas can be shortened as follows: 1) much research work has to
be done on multipath routing protocol to support both energy efficiency and
QoS 2) The cross layer module and the multipath routing with forward error
correction (FEC) technique can be used to increase the network
performance.
REFERENCES
[1] FengShan,WeifaLiang,JunLuo, and Xiaojun Shen,“Network lifetime maximization
for time-sensitive data gathering in wireless sensor networks,”Trans. on ELSEVIER,
Computer Networks, pp.1063-1077, 2013.
[2] Kazem Sohraby, Daniel Minoli and Taieb Znati,“Wireless Sensor Networks
Technology, Protocols, and Applications,” WILEY, A john wiley& sons, ISBN:
978-0-471-74300-2, 2007.
[3] AHMED E.A.A. Abdulla, HirokiNishiyama, and NeiKato,“Extending the Lifetime
of Wireless sensor networks: A hybrid routing algorithm,”Trans. on ELSEVIER,
Computer Communications, vol 35, pp.1056-1063, 2012.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 18
[4] ZahoorA.Khan,ShyamalaSivakumar, William Phillips, andBill Robertson,“A QOS-
aware Routing Protocolsfor Reliability Sensitive Data in Hospital Body Area
Networks,” Trans. on ELSEVIER, in proc. ANT, pp. 171-179, 2013.
[5] ZahoorA.Khan,ShyamalaSivakumar,William Phillips, and Bill Robertson,“A QOS-
aware Routing Protocol for Delay Sensitive Data in Hospital Body Area
Networks,”Trans. on IEEE, in proc. BWCCA, pp. 178-185, 2012.
[6] Zahoor A.Khan, Nauman Aslam, Shyamala Sivakumar, and William Phillips,
“Energy-aware Peering Routing Protocol for Indoor Hospital Body Area Network
Communication,”Trans. On ELSEVIER, in proc. ANT, pp. 188-196, 2012.
[7] Jalel Ben-othman and Bashir Yahya,“Energy Efficient and QOS Based Routing
Protocol for Wireless Sensor Networks,” Trans. on ELSEVIER, J.Parallel
Distrib.Comput.70, pp.849-857, 2010.
[8] MarjanRadi, BehnamDezfouli, Kamalrulnizam Abu Bakar, and
MalreyLee,“Multipath Routing in Wireless Networks: Survey and Research
challenges,”Trans. OnSensors, pp. 650-685, 2012.
[9] Nikolaos A. Pantazis, Stefanos A. Nikolidakis, and Dimitrios D. Vergados,“Energy-
Efficient Routing Protocols in Wireless Sensor Networks: A Survey,”Trans. on
IEEE COMMUNICATIONS SURVEYS & TUTORIALS, vol. 15, no. 2, pp.551-591,
2013.
[10] Mustafa IlhanAkbas and DamlaTurgut,“Lightweight routing with dynamic interests
in wireless sensor and actor networks,”Trans. On ELSEVIER, Ad- Hoc Networks,
pp.1- 15, 2013.
[11] YuchunGuo, Fernando Kuipers, and Piet Van Mieghem,“Link-disjoint paths for
reliable QoS routing,”Trans. OnINTERNATIONAL JOURNAL OF
COMMUNICATION SYSTEMS, pp.779–798, 2003.
[12] Srihari Nelakuditi , Srivatsan Varadarajan,and Zhi-Li Zhang,“On Localized Control
in QoS Routing,” IEEE Trans. Automatic Control, vol.47, no.6,pp.1026-1032, 2002.
[13] Mohammad Hammoudeh and Robert Newman,“Adaptive routing in wireless sensor
networks: QoSoptimisation for enhanced application performance,”Trans. on
ELSEVIER, Information Fusion, pp.1-13, 2013.
[14] DjamelDjenouri and Ilangko Balasingham,“Traffic- Differentiation-Based Modular
QOS Localized Routing For Wireless Sensor Networks,”IEEE Transactions on
Mobile Computing,vol.10,no.6,pp.797-809, 2011.
[15] Shiva Murthy G, R.J.DSouza, and Varaprasad G, “Reliability Analysis of Route
Redundancy Model for Energy Efficient Node Disjoint Multipath Routing in
Wireless sensor networks,”Trans. on ELSEVIER, in proc. ICMOC,pp.1487-1494,
2012.
[16] SanghitaBhattacharjee and SubhansuBandyopadhyay, “Lifetime Maximizing
Dynamic Energy Efficient Routing Protocol For Multi Hop Wireless Network,”
Trans. on ELSEVIER, Simulation Modeling Practice And Theory,pp. 15-29, 2013.
[17] Carlos Velasquez-Villada and YezidDonoso,“Multipath Routing Network
Management Protocol For Resilient And Energy Efficient Wireless Sensor
Networks,”Trans. on ELSEVIER, in proc. ITQM,pp. 387-394, 2013.
[18] Jiming Chen, Ruizhong Lin, Yanjun Li, and Youxian Sun,“LQER: A Link Quality
Estimation based Routing for Wireless Sensor Networks,”Trans. on SENSORS,
pp.1025-1038, 2008.
[19] Wei Quan , Fu-Teo Zhao,Jian-Feng Guan, Chang-Qiao XU, and Zhang Hong-
ke,“An Integrated Ling Quality Estimation-Based Routing For Wireless Sensor
Networks,”Trans. on ELSEVIER, pp.28-33, 2011.
[20] Tommaso Melodia, Mehmet C. Vuran and Dario Pompili,”The State of the Art in
Cross-layer Design for Wireless Sensor Networks,”Trans. on SPRINGER, pp.78-92,
2006.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 19
[21] Lucas D.P. Mendesm and Joel J.P.C. Rodrigues,”A survey on cross-layer solutions
for wireless sensor networks,” Trans. on ELSEIVER, in proc. Journal of Network
and Computer Applications, pp.1-12, 2010.
[22] Ian F.Akyildiz, Mehmet C. Vuran and Ozgur B. Akan,”A Cross-Layer Protocol for
Wireless Sensor Networks,”Trans. on IEEE, pp.1102-1107, 2006.
This paper may be cited as:
Sridevi S., Rumeniya G. and Usha M., 2014. Energy-aware QoS Based
Routing Protocols for Heterogeneous WSNs: A Survey. International
Journal of Computer Science and Business Informatics, Vol. 11, No. 1, pp.1 -
19.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 20
Optimization of Outsourcing ICT
Projects in Public Organizations; Case
Study: Public Center of Iranian ICT
Studies
Majid Nili Ahmadabadi
Department of Management,
Qom University, Qom, Iran
Abbas Bagheri
Department of Industrial engineering,
Islamic Azad University, Firuzkuh Branch,
Tehran, Iran
Fariba Abolghasemi
Department of Management
Payam Noor University, Tehran Branch,
Tehran, Iran
ABSTRACT
Outsourcing is a strategic decision and consequently, it has a fundamental impact on the
performance and costs of an organization. If the spiritual and economic costs are not
allocated appropriately, it will be led to competitive advantages and will drag the
organization on the verge of destruction. In this paper by using over a decade outsourcing
experience of outsourcing in a national center as well as presented models in this regard, an
integrated model has been presented which could be of great help in remarkable cost
reduction and it would result in high productivity in national projects. This contribution is
based on a knowledge management module. In this paper, Momma, J. and Hvolby’s four-
stage model will be introduced for outsourcing as the base model. Then, with an interview
to outsourcing of the research projects in Public center of Iranian ICT studies and its
analysis, the results and knowledge obtained will be discussed in a model framework for
the research project outsourcing. In the end, the required points will be introduced for using
proposed model and the benefits of its usage.
Keywords
Outsourcing, Public center, Research management, Research projects.
1. INTRODUCTION
Moving from an industrial society to the information society, turning
national economy into the global economy, centralization to
decentralization, and finally, hierarchical structure to the network system are
obvious signs of fundamental changes in today's environment which
undoubtedly answering them requires new solutions and strategies. One of

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 21
these solutions is outsourcing that takes a wider range every day so that not
only the government but also the private sector have taken the approach of a
warm welcome. Given the key role of outsourcing in developing and
enhancing the skills, predicting valid statistical and scientific centers
suggest that the volume of global outsourcing markets in the current year
will reach one trillion dollar. Moreover, given the predictions made, volume
of design work and outsourcing research and development from 2004 to
2009 has reached from 179 billion dollars to 345 billion dollars [1].
According to Gartner research study group, global market for IT
outsourcing in 2008 was equivalent to U.S. $ 253.1 billion and will grow
7.2% per year. Meanwhile, Forrest predicts that European organizations will
cost more than 238 billion Euros in the field of IT outsourcing in 2008.
Outsourcing enjoyed particular complexities with the growth and evolution
of other management issues. Therefore, the concept of outsourcing becomes
a complex issue in conjunction with the organizational growth so that not
only in terms of financial savings, but also in different prospects, including
remaining in a competitive market, the need to join the global markets,
rising customer expectations and market competitiveness are raised and
outsourcing is becoming a need, a requirement and a pressing issue for an
organization [4].
Outsourcing has its own characteristics and different sectors of activities can
be outsourced in different industries. Thus, this perception that outsourcing
is a strategy, a guideline and a method, refers to this issue that what
characteristics have the things being transformed and what are the outcomes
of the outsourcing for that organization? Therefore, according to the issue of
outsourcing, its characteristics can be studied and an appropriate name can
be given to. In most current industrialized and manufacturing companies, a
part of production process has been usually outsourced. But there is a
bilateral cooperation between research centers for outsourcing. The
prerequisite for large and innovative companies is having an open and a
competitive economy. On the other hand, outsourcing occurs in an
organization when a management need is felt within. If there is no
management belief and resource allocations, outsourcing will not occur as it
should. Iran has various experiences of outsourcing, but this kind of
outsourcing is basically different from what is being done in the companies
such as Cisco, IBM and Microsoft and has changed from product-oriented
to a service-oriented company [4]. On the other hand, without a previous
planning for the entire process and its consequences, some measures have
been taken for changing organizational strategy, but poor results have been
obtained.
In this proper, outsourcing will be first studied as a strategic activity in the
global organizations. Since the presented models for outsourcing are very

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 22
rare, one of them which is more consistent with Iran is Alpha after its
introduction. Then, the experiences of this center for a decade of
outsourcing with various strategies will be analyzed and finally, a developed
model of outsourcing activities for research centers will be presented. In the
end, the parameters and inputs and outputs of the presented model and the
benefits of its usage will be presented.
2. BACKGROUND STUDY AND AN OVERVIEW ON RESEARCH
OUTSOURCING AND ITS GLOBAL ADVANTAGES
2.1. Definitions
Outsourcing means organizing that part of one’s redundant activities that are
not involved in the organization’s chain value and should be transferred to
external parts of an organization. Outsourcing in its particular concept
means organizing one’s tasks and activities that are involved in the
organization’s value chain [4].In the field of research, research management
means managing in doing one or more research project. To manage a
research project in the organization, it requires that major research
approaches to be defined along with organization’s major goals as well as
multiple smaller research projects to be defined along with each approach
and required budgets are allocated to them. Then, one of the capable staff
should take the responsibility of research management of this approach and
takes actions towards the activities related to projects outsourcing with a
group of colleagues[5].
2.2. Research Outsourcing
The outsourcing of R&D activities became popular from the late 1990 by
pharmaceutical companies. One reason for this issue is filling the research
gap among organizations. Another reason in using R&D outsourcing is that
producing new products requires a long time in the process of innovation
and market introduction. Other incentives that make the outsourcing of
research and development by the company include:
- R&D productivity is increased (reduced costs and increased revenues)
- The success of similar companies in the field of outsourcing research and
development
- The acquisition of knowledge generated by the outsourcing partner
- Multilateral engagement of the organization with colleagues, competitors
and customers and thus more information in doing research
- Ability to access the above-mentioned data through a partner organization
outsourcing
In doing such outsourcings, the general process of outsourcing has been
used and a specific principle or stage is not added or removed [6].

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 23
The first advantage of state’s research outsourcing is reducing costs and
time for such projects. The second advantage of this work is that the
organization performing the project is obligated to enforce laws that have
been enacted for public agencies. Since the number of these laws is led to
reducing the speed of these issues in state organizations and rising some
problems in the field of budget allocation, the outsourcing of these projects
is a way to overcome these legal problems, The static nature of government
organizations, people's lack of commitment to this organization, the need for
mobility in research projects, the need for coordination and integration and
minds of people in such projects, little and slow governmental monitoring
(which often takes the form, not the content) and problems alike can be
solved by research projects outsourcing in governments [7].
3. RELATED WORKS
In 2003, a research has been done by the students of research in operations
that integrated strategic management and organizational theory and used it
for inter-organizational communications [16]. Holcomb, H. and Hitt, M
proposed a theoretical model for strategic outsourcing with the resource-
based and transaction-based approaches in 2007 using the above research. In
his article written in 2008, McIvor, R presented the strategic decisions
towards outsourcing or lack of outsourcing of these activities as a model
[18].
Several theories have been presented previously for outsourcing, but it has
not often been led to a structural model. Existence a scientific model can be
effective in a successful outsourcing. The stationary and non-intelligent
nature of information that are produced through current tools and techniques
neither support management and control of dynamic processes nor the key
activities for operational and tactical levels.Therefore, they should use
trainings in the time of working, tacit knowledge and the colleagues’
suggestions, expert advices and finally trial and error method. Many authors
have discovered the impacts of outsourcing over flexibility and value
creation in that flexibility. Value creation can be obtained through the
combination of strategy, economy, technology and human factors [19, 20,
and 21]. A significant issue in doing outsourcing is that doing a successful
outsourcing, production characteristics and market considerations should be
consistent with the legal personality, functional strategies and company’s
business. The evidences indicate that a small number of researchers have
been attempted to develop the scientific model in an experimental level in
that their aim was to publicize their results in the books, scientific and non-
functional communities [22, 23]. Momma, J. and Hvolbyhave indicated that
there is no outsourcing framework that is consistent with the harsh
conditions of a real atmosphere in business and they themselves presented a
model and framework for outsourcing [24].

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 24
4. METHODOLOGY
In this article, the model and framework introduced by Momma, J. and
Hvolbywill be introduced as the base model and will be criticized and
thenitwill be developed given the experience of alpha center subsequently.
This model includes a wide range of tools and techniques to search for
suitable suppliers, monitoring and improving them as well as performing
outsourcing projects and communications management in this project.
Overall, the tools and techniques help implement the models and collecting,
structuring and accessing to the required data will help its stages. The
decisions of outsourcing companies for outsourcing are based on three
criteria:
- The company outsources the product which itsproduction is the
vital resources and capabilities.
- Outsourcing will be assigned to the suppliers who create
competitive advantages (such as a larger scale, lower costs or
greater efficiency)
- Sometimes, outsourcing is a way to improve production efficiency,
creating the staff’s commitments and as a result increasing the
competitiveness and profitability of the company.
Figure 1 shows the stages and outsourcing methods in Momma, J. and
Hvolby’smodel.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 25
Figure 1: Momma-, J. and Hvolby’s four phase model [24]
According to this model, the characteristics of suppliers along with each
characteristic are collected. Important information in the field of history,
price, quality, roofing production capacity, production and delivery time,
warranty, and all such information from the organization can be an
important determinant. The next stage is like crossing the suppliers of the
primary filter. At the next stage, the agreements on the details will be done
and some candidates are eliminated and one or more final supplier is
selected. The last stage of this model indicates that using the performance of
the supplier and observing his/her working process, a decision can be made
about continuing the coordination or replacing him/her.
4.1Review of a decade of outsourcing in Public center of Iranian ICT
studies
In this part, outsourcing activities of the Center, hereinafter briefly called
center - is examined since 79 to 86. The Center's activities can be divided
into three periods. In the first period, some of the main activities of the
Center are removed from the main chain of activities and are outsourced by
changing the strategy aiming with privatization of the activities (such as
PCBs and workshops). In addition, some major activities such as the control
center for research projects, human resource management and etc. were also
outsourced. In the second period of strategy, the center was wholly changed
from project conducting to project steering.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 26
Year
Center’s
macro
strategy
Outsourcin
g issue
Outsourci
ng aim
The process and
working method
Results and
outcomes
2000-
2003
Conducting
Ministry of
ICT
research
projects and
subsidiaries,
Conducting
the projects
with an
emphasis to
fundamental
projects by
center
personnel
cooperated
with
universities
Outsourcing
contract side
activities
such as
transportatio
n, services,
restaurants
Insourcing
the Digital
Research
Project (a
project
conducted
by the
personnel of
contractor)
outsourcing
main
activities
such as
contractual
human
resource
managemen
t, circuit
board,
controlling
the projects,
consultancy
and doing
projects
abroad
Privatizatio
n,
Reducing
manpower
associated
with the
center,
enhancing
Internation
al relations
Financial assistance
for the
establishment of
satellite companies
by staff and other
internal and
external people,
Selling stock
repository and
material,
Selling various
workshops,
Transferring
transportation,
restaurant, service
contractor,
Assignment of
project
management,
informal human
resources
management to the
contractor,
Obtaining advice
and carrying out
joint projects
abroad.
The center
lost the
independent
conduction of
applied
projects; the
center lost its
workshops
and facilities
and thus
requires
assigning the
activities
outside the
center,
causing wide
disappointme
nts in the
areas of
taxation to the
projects
personnel
from the
project
authorities.
Secon
d
period
(2004
-
2005)
The
emphasis on
conducting
effective
projects for
ICT in the
country
along with
determining
the
ministry's
guidelines in
purchasing,
assigning,
decision-
making,
policy-
All research
projects are
almost
outsourced
and the
colleagues
only take
the
managemen
t over
researches.
The field of
study in this
time is
doing
research
activities
Attracting
the macro
research
budgets
and
projects
outsourcing
independen
t from the
fact that
there are
final
customers
and results
demander
in the
ministry
Projects and
designations
outsourcing in a
wide and macro-
level and domestic
companies/outsourc
ing research
projects to the
external companies,
administering
contracts with
formal and informal
personnel of the
center, voiding the
precedence of the
previous satellite
companies,
Existing some
interested
personnel in
fundamental
and
developmenta
l projects
from the
center,
creating
dissatisfaction
among the
personnel for
special
payments to
people, doing
cumulative

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 27
Year
Center’s
macro
strategy
Outsourcin
g issue
Outsourci
ng aim
The process and
working method
Results and
outcomes
making, etc. related to
ICT at the
national
level (other
ministries
and three
branches
obtaining the
consultation and
doing shared
projects with
abroad, doing
projects for The
Judiciary, banks,
etc., doing
outsourcing
activities as a
special activity
through the center
personnel (with
special and
unstructured paying
, wide outsourcing
of the projects and
attracting the
budgets and wide
participation of the
universities, etc. in
doing the projects
conducted or
semi-
conducted
projects
without
customer,
existing
insufficient
transparency
in payments,
existing
multiple
managed or
unmanaged
costs,
dissatisfaction
of the
receptor
companies o
outsourcing
for the
insufficiency
of
outsourcing
management
Third
period
(2006
)
Doing
various
strategic,
fundamental
and
developmen
tal projects
for ICT
ministry and
subsidiary
organization
s
Doing
research
managemen
t with new
organizing,
outsourcing
research
activities
only for the
time of
existing
requests and
final
customer in
the
subsidiary
organization
Attracting
research
budgets of
the center,
doing
requested
projects by
the
ministry
and
subsidiary
organizatio
ns in
meeting
their needs
Projects
outsourcing and
designs in a more
limited level to
universities and
domestic
companies, doing
outsourcing
activities as a
special activity by
the center personnel
(with special, and to
some extent
structured,
payments, building
elite center for the
macro management
of the company's
strategies
(comprised of the
ordinary personnel
of the center,
though with
different
Continuing
existing of
some of the
personnel
interested in
fundamental
and
developmenta
l designs from
the center,
creating
dissatisfaction
among the
personnel for
the sake of
special
payments to
people,
inability in
attracting
budgets and
conducting
required
research

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 28
Year
Center’s
macro
strategy
Outsourcin
g issue
Outsourci
ng aim
The process and
working method
Results and
outcomes
regulations),
outsourcing
research projects to
the external projects
projects for
applying
harsh
regulation in
the customer
and obtaining
the costs of
working from
them, existing
a couple of
strategic
activities
without any
result by the
center's
personnel and
multiple
office works,
exhaustion of
research
personnel
from doing
their works
without any
result,
existing
unmanaged
costs, limited
projects
outsourcing
and wide
attracting of
the budgets
and limited
participation
of universities
and
companies in
doing the
projects due
to reducing
the reliance
caused by
outsourcing
management
insufficiency

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 29
In table (2), the condition of Public center of Iranian ICT studies has been
considered and analyzed in terms of 8 important factors in projects
outsourcing. Considering 6 indicated aspects in Momma, J. and
Hvolby'smodel in Table (2) shows that the research center did not follow a
specific pattern and just some concise regulations were enough and trial and
error methods relying on traditional ways are applied, therefore, multiple
problems were arisen.
Table 2: studying Public center of Iranian ICT studies in terms of the important
factors affecting Momma, J. and Hvolby's outsourcing process
Important factors
in outsourcing
process
Period 1 Period 2 Period 3
The prospect and
written
organizational
strategy
Not exis
It has been provided
for the first time
Existed but not
updated
Organizational
memory/continuous
improvement
Not exist Not exist Not exist
Competitive
analysis
It is done in the minds
of managers
In is done in the
minds of managers
and the experience
is not delivered to
other levels of
organization
In is done in the
minds of managers
and the experience
is delivered to other
levels of
organization
Evaluation and the
approval of
expositors
It is done only at the
level of managers and
does not relate to the
staff
It is done with an
expertise working
(by the scholars
who have no
expertise in this
regard and have
been not trained)
and some decisions
made at the
management level.
A primary guideline
has been prepared
for this issue
It Is done with a
scholarly working
(by the scholars who
have the previous
experience in this
regard and have
been not trained as
well) and some
decisions at the
management level.
A vague guideline
has been prepared
for this issue
Negotiation for
contract
It is done only at the
management level and
does not relate to the
staff
The significant
contents of the
contract are
provided to the
service presenters at
the scholar level,
but the decisions
are made at the
level of top
managers
The significant
contents of the
contract are
provided to the
service presenters at
the scholar level,
but the decisions are
made at the level of
top managers
Project The project is done, The project is done, The project is done,

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 30
Important factors
in outsourcing
process
Period 1 Period 2 Period 3
implementation and
knowledge transfer
but knowledge
management is not
transparent
but knowledge
management is not
transparent or it
some
inattentiveness
arises
but knowledge
management is not
transparent or it
some inattentiveness
arises
Communications
management
Communications
management is also
outsourced and is
done by project
control unit. Project
control is done only at
the temporal control
of the project
(reporting the retards
to the management)
without regarding the
project contents and
the backwardness
reasons and its
analysis
Two organizational
units perform the
communications
management
imperfectly under
the research branch
known planning
center and studying
research designs
and the office for
supervision and
evaluation.
Two organizational
units perform the
communications
management
imperfectly under
the research branch
known planning
center and studying
research designs and
the office for
supervision and
evaluation.
Contract
termination
It is done with the
management opinion.
It is done with the
opinion of research
management and
middle centersin
research ministry
and finally by the
approval of top
management
4.2 The suggested model for outsourcing for the research projects
With respect to what are indicated in part 4, it was observed that Public center
of Iranian ICT studiesmanagement has made some wrong decisions in the field of
outsourcing which the result of their implementation in the past was not
satisfactory. One reason is the instability of top managers in this center, for
they are prone to change the managers in a short period of time and the new
managers were in lack of management skills which the previous managers
had in their management time. Moreover, the lack of organizational memory
in this center causes that despite the tendency of new managers for using the
experiences of previous managers, they could not get these experiences as
they must and as a result people's memory are used in this regard.
Obviously, this method was replete with many mistakes and thus was not
fruitful. Existing a unit named organizational memory in an organization
can preserve the old useful information, it can makes it possible for the top

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 31
managers to devise more useful strategies in the field of outsourcing for the
organization which is led to a continuous improvement in this regard.
In the Momma, J. and Hvolby’s four-stage model, it is focused on the
customer and useful parameters are considered for primary identification
and evaluation, chose and approval of outsourcing service supplier,
approving contracts and performing the projects and finally evaluation and
improvement, but there is no relationship between outsourcing
implementation phases with organizational strategy and organizational
memory in this model. In addition, the continuous improvement should be
occurred in all sectors and processes of working in an organization;
however, it has not paying attention to the continuous improvement in
phases 1 and 2. These cases are considered as the main weaknesses of the
above model. Also in this article, Momma, J. and Hvolbypresented a
process as the framework of projects outsourcing that was proposed by
Laudon in 1998 for activities outsourcing and has been completed by
Momma in 2001. This framework comprises 6 continuous phases as 1)
competitive analysis, 2) evaluation and approval, 3) negotiation for contract,
4) project implementation and knowledge transfer, 5) communications
management, 6) contract termination. It has also been specified that this
process has also been presented in the four-staged model, but it is not in a
repetitive cycle in this process and on the other hand, it has not been paid
attention to organizational strategy and organizational memory in this
model. It is obvious that if the working results are not effective in changing
and modification of organizational strategies, continuous improvement will
not be obtained.
The developed model for outsourcing of the research projects according to
Figure (3) is of 6 stages in that some of key activities along with the related
assessors and desirable outputs (that are known as decision variables). These
6 stages are obtained through outsourcing competitive thinking and
implementing ideas of Bragg et.al [22] in that as previously been mentioned,
it has been presented in the framework suggested by Momma, J. and Hvolby
[24]. These 6 factors are proposed as the main components of outsourcing
process and have been shown in Figure (2). On the other hand, two factors
namely organizational prospect and strategy as well as organizational
memory can also be observed in this figure. Organizational memory plays
the main role in being a learner and maintaining the organizational records
and is related to the whole units and stages in order that both new obtained
information and knowledge are preserved in their memory and give the
previous records (information and knowledge created in the organization) to
the associated units for improving the application of current issues. Also,
this unit can give new created knowledge to the top managers in order to be
used in improving macro organizational prospects and strategy. Since the
outsourcing projects can be the source of knowledge creation and added

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 32
value for the organization, outsourcing projects are also considered in this
model, but with respect to the fact that this issue does not relate to
outsourcing, a general block is sufficient in this regard. In continuous,
outsourcing processes, key activities, performance assessors and expected
results will be explicated:
Figure (3): the developed model of outsourcing for the research projects
The process of research projects outsourcing: this process comprises 6
stages as follows:
Stage 1: competitive analysis
In this stage, other research organizations in the regarded scientific area will
be considered and some information is gathered. Key activities,
performance indices and expected outputs in this stage are as follows:
- Key activities: strategic analysis, SWOT analysis, vital and non-vital
competencies, mapping, etc.
Compet
itive
analysis
Evaluation
and
supplier
approval
Negotiation
for making
contract
Project
implementa
tion and
knowledge
transfer
Communicat
ion
management
with the
supplier
Organizational memory/controlling standards
Designi
ng the
perspect
ive and
organiza
tional
strategy,
planning
and
tactics
Results
evaluation
and contract
termination
Insourcing processes
Outsourcing Process

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 33
- Performance indices: increasing competitive advantages, amount of
added value and the information related to final consumers, level of
agreement in the field of outsourcing in organization, etc.
- Expected outputs: suitable strategic steering, amount of knowledge
increase in competitiveness and profitability, level of creating a
shared comprehension regarding probable advancement
opportunities
Stage 2: evaluation and approval of suppliers
In this stage, using the indices such as the number of organization's
researchers, the number of articles, inventions and other achievements and
indices such as the organization's record in the research issues,
organization's scientific grade and rank and other indices, the specified
candidates in the last stage are evaluated. Key activities, performance
indices and expected outputs in this stage are as follows:
- Key activities: defining the vital indices for evaluation (quantitative
and qualitative), surveying the details of supplier features, the
evaluation of real performance etc.
- Performance indices: the possibility of assessment criteria, the
number of approved suppliers, the closeness of suppliers'
geographical location in comparison with the customers' place,
determining a well-informed and experienced observer, the
agreement of the observer regarding project contents, etc.
- Expected outputs: reduced risks in choosing the supplier, better
review and the access to vital production competencies, level of
cognition and understanding the customers' opinions through the
suppliers and their higher performance, being sure of obtaining
suitable results with the opinion of observer, enhancing internal
processes of suppliers' evaluation and related guidelines, etc.
Stage 3: negotiation for making a contract
In this stage, a number of meetings are hold with the research organization's
representatives in that they are trying to reach a primary agreement. Key
activities, performance indices and expected outputs in this stage are as
follows:
- Key activities: defining the projects and legal and business
regulations, negotiation about the duration of contract and the time
of its delivery and bilateral agreements etc.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 34
- Performance indices: level of agreement about the terms, legal and
business provisions, the ability to determine delivery condition, the
inclination and effort of the supplier for cooperation during the
negotiation.
- Expected outputs: close, intimate and long-term relationship,
bilateral agreement regarding contract provisions, bilateral will for a
fair cooperation for the two sides, paying attention to the secrecy of
working results, etc.
Stage 4: project implementation and knowledge transfer
In this process, some research has begun. Being assured of shared
comprehension about the expected outputs are among the important points
that should be taken into consideration in this stage. Otherwise, project
initiation will be faced with some deviations and its revision requires paying
financial and temporal costs. Key activities, performance indices and
expected outputs in this stage are as follows:
- Key activities:establishment of the suppliers' incorporation
principles, defining the method of interactions, the compatibility of
organization with the suppliers' performance, etc.
- Performance indices: the ability to perform changing processes, level
of profitability and flexibility capacity, the ability in defining
interaction costs (cost curve), etc.
- Expected outputs: more capital and more accessible resources,
increasing further engineering benefits, logical balance between
domestic production and outsourcing
Stage 5: relationship management with the supplier
Choosing an observer that is competent both scientifically and executively
and have enough supervision on the research outputs in its due time and
based on existing services specified in the contract is so vital in this stage. In
some research projects which more than one expertbeing used within them,
multiple observers should be used as well. Key activities, performance
indices and expected outputs in this stage are as follows:
- Key activities: creating communications, supervising systems and
information, the relationship among developmental projects,
continuous evaluation of performance, etc.
- Performance indices: the ability to measure the minimum impacts in
the relations, the curve of the product's life time and a later time to
be entered the market, innovation and changing the customers'

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 35
habits, costing structures, the ability to deliver and final quality of
the product.
- Expected outputs: the percentage of the final quality of the product,
reduced cost of construction relative to other samples or a better
control on the costs, less time in presenting new products, etc.
Stage 6: results evaluation and contract termination
One of the important points in this stage is to transfer new knowledge in the
project from the supplier to the organization. This knowledge is not merely
the project documentations and it sometimes requires that some skills are
transferred with them.For this condition, training periods and knowledge
management guidelines can be used. Key activities, performance indices
and expected outputs in this stage are as follows:
- Key activities: the evaluation of the options of contract continuance,
changing the supplier of domestic production, making reviewing
bases in the competitive strategy, etc.
- Performance indices: the need to achieve determined objectives in
the outsourced competitive area, the ability to obtain assurance in
terms of passing the critical stage, the ability to replace supplier or
outsourcing, etc.
- Expected outputs: being aware of the time for making long-term
contracts, replacing the supplier or regarding outsourcing, a better
way for strategic planning, etc.
4.3 Organizational Memory Unit
The databases save information related to capabilities and the process of
improving the suppliers during their cooperation with the company in the
current and past projects. These databases help distinct between the
suppliers among a wide number of raw materials suppliers, product and
technology suppliers. Furthermore, it records a background of the potential
outsourcing shareholders that are useful to work with and will be useful in
the future as well. Existing organizational memory unit in the process of
research projects outsourcing paves the way for preserving decision making
skills, negotiation, and dominance over concepts and methods in that
organization during this process and will be developed in the period of time.
Therefore, experienced and sapient people who are well-aware of the
methodologies and concepts are involved in the process of outsourcing;
organizing and managing can learn at the end.A capable organizational
memory can help the organization in the following areas:
 Risk management

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 36
 Determining the key capability of the receptor company
 Determining the key capability of the company giving services
and making balance between the two key capabilities
 Outsourcing project management (i.e. according to PMBOK
standard)
 Changing and revising the prospect of the organization's
managers
 Surveying the suggestions an determining the best one in terms of
an independent supervisor
 Determining the criteria for professional ethics in preserving
secrete information from the part of service giver
 Determining the new paradigms of the organization giving
services in the field of service outsourcing
 Help revise organizational culture suitable to outsourcing
 Reviewing the political aftermaths in outsourcing (especially
international outsourcing)
 Determining the indices of technology alignment between the one
giving services and receiving services
5. The Impacts of using a model suggested for outsourcing management
of research organizations
The results of performance indices of above processes along with the access
to expected outputs are sent for the managers and authorities. Given the
obtained results, these people can change organization's strategy, revise the
prospect or issue some orders relative to operational revising measures. The
following issues are presented as some evidences in this regard:
Improving the determination of the competence for the power of research
management personnel's supervision over outsourcing projects, improving
the observance of security systems and professional ethics from the one
giving outsourcing services, promoting education (to other personnel in
outsourcing, research managers and finally training the working results to
them and being sure of knowledge attraction), promoting the analysis of
performance costs, making decision to increase or decrease inefficient
organizational units that create no added value to the organization's
outsourcing activity (changing structure), initiating or changing the
combination of supervision council, promoting or changing attraction
strategy and knowledge transfer, reviewing the performances in the working
value cycle in various sectors (group, faculty, viceroys, … ) in order to
prevent tasks projection and the like.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 37
6. RESULTS
As it was mentioned, using over a decade experience of outsourcing in a
national center and presenting models in this area resulted to integrated and
unprecedented scientific models. This contribution mainly is based on a
knowledge management module. Using this module and integrating it with
previous modules would result in a sort of intelligence which prevents
occurrence of previous mistakes automatically. This article aimed to present
a pattern for outsourcing research projects. Although much studies have
been conducted concerning outsourcing and the models related to it, it has
less been taken into consideration the organizational memory and creating a
continuous improvement in the field of outsourcing. The effective role of
this unit in the access to the past and organizational learning that used
through it and more paid attention in this article and a model based on it is
presented. In this article, the experiences gained by Public center of Iranian
ICT studies were used as a case study for this issue. These experiences also
require organizational memory unit in the outsourcing company and shows
the devastating impacts of lack of attention to.Outsourcing in each case,
especially for the research projects can be managed using the model above
and improve it through the time. The benefits in using this model for the
outsourcing organizations are as follows:
- Following a specific framework and making a discipline in
outsourcing
- Being aware of the infrastructures required for applying and
managing the outsourcing process
- The optimal designation of organizational units and their
relationship with a successful outsourcing application
- Reducing the costs and increasing the physical and spiritual revenues
caused by a proper selection of outsourcing shareholder
- Learning in the field of outsourcing and satisfaction of domestic and
foreign customers caused by progressing of the organization
onwards
- The ability for counseling and guiding other organizations in the
field of outsourcing and sharing the knowledge and experience with
them
- Using organizational memory for the cases other than outsourcing
in the organization (improving other units and organizational
dimensions)
- Other benefits caused by being the organization as a learner in the
field of outsourcing

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 38
This method has no limitation or disadvantage relative to current models
and methods and the further organizational infrastructures that requiresto be
implemented is that organizational memory unit.
7. REFERENCES
[1] Carbone, J., EMS profits get squeezed. Purchasing 134, 29–33, 2005.
[2] Pei, Z. ,Zhen-xiang, Z. and Chun-ping, H.," Study on Critical Success Factors for IT
Outsourcing lifecycle", International Conference on Wireless Communications,
Networking and Mobile Computing, 2007. WiCom2007.Volume , Issue , 21-25 Sept.
2007, Pp4379 - 4382
[3] Reyes Gonzalez, Jose Gasco and Juan Llopis, “Information System Outsourcing: A
Literature Analysis”, Information & Management, Vol. 43, 2006, (7), pp.821-834.
[4] Outsourcing the strategic decisions of successful enterprises, a meeting with the
presence of management authorities, Tadbir magazine, Issue 166, Esfand 2005
[5] Executive guidelines of research management, Iran telecommunications research center,
Mordad 2006.
[6] Higgins, M., "The outsourcing of R&D through acquisitions in the pharmaceutical
industry," Journal of Financial Economics 80 (2006) 351–383
[7] Ulset, S. , "R&D outsourcing and contractual governance: An empirical study of
commercial R&D projects, " Journal of Economic Behavior &Organisation Vol. 30
(1996) 63-82
[8] Gilley, K.M., Rasheed, A., "Making more by doing less: ananalysis of outsourcing and
its effects on firm performance,"Journal of Management 26 (4), 2000, pp.763–790.
[9] Billi, J.E., Pai, C.W., Spahlinger, D.A., "Strategic outsourcing ofclinical services: a
model for volume-stressed academic medicalcenters," Health Care Management Review
29 (4), 2004,pp. 291–297.
[10] Roberts, V., "Managing strategic outsourcing in the healthcareindustry," Journal of
Healthcare Management 46 (4), 2001,pp.239–249.
[11] Chen, I.J., Paulraj, A., "Towards a theory of supply chain management: the constructs
and measurements," Journal of Operations Management 22 (2), 2004,pp.119–150.
[12] Shy, O., Stenbacka, R., "Strategic outsourcing," Journal of Economic Behavior and
Organization 50 (2), 2003, pp.203–224.
[13] Fine, C.H., Whitney, D.E., Is the make-buy decision a core competence? In: Muffato,
M., Pawar, K. (Eds.), Logistics in the Information Age, ServiziGraficiEditoriali. Pandova,
Italy, 1999, pp.31-63.
[14] Quinn, J.B., Hilmer, F.G., "Strategic outsourcing," Sloan Management Review 35
(4),2004, pp. 43–55.
[15] Quinn, J.B., "Strategic outsourcing: leveraging knowledge capabilities," Sloan
Management Review 40 (4), 1999, pp.9–21.
[16] Grover, V., Malhotra, M.K., "Transaction cost framework in operations and supply
chain management research: theory and measurement," Journal of Operations
Management 21 (4), 2003, pp. 457–473.
[17] Holcomb, H. and Hitt, M., "Toward a model of strategic outsourcing," Journal of
Operations Management 25 (2007) 464–481
[18] McIvor, R., "What is the right outsourcing strategy for your process?," European
Management Journal (2008) 26,24
[19] Lei, D., Hitt, M.A., "Strategic restructuring and outsourcing: the effect of mergers and
acquisitions and LBOs on building firm skills and capabilities," Journal of Management
21 (5), 1995, pp.835–859.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 39
[20] Brandes, H., Lilliecreutz, J., Brege, S., "Outsourcing success or failure?," European
Journal of Purchasing & Supply Management 3 (2), 1997,pp. 63–75.
[21] Lonsdale, C., "Effectively managing vertical supply relationships: a risk management
model for outsourcing," Supply Chain Management: An International Journal 4 (4),
1999,pp. 176–183.
[22] Bragg, S.M., Outsourcing: A Guide to… Selecting the Correct Business Unit…
Negotiating the Contract Maintaining Control of the Process. Wiley, New York, USA,
1998.
[23] Wasner, R., "The outsourcing process -strategic and operational realities," Ph.D.
Thesis, Department of Management and Economics, Divisions of Industrial Marketing,
Linkoping University, Sweden, 1999.
[24] Mommea, J. and Hvolby, H. , "An outsourcing framework: action research in the
heavy industry sector," European Journal of Purchasing & Supply Management No. 8
,2002, pp.185–196.
[25] Laudon, K.C., Laudon, J.P., Management Information Systems FNew Approaches to
Organization and Technology, 5th Edition. Prentice-Hall International, Englewood Cliffs,
NJ., 1998.
[26] Momme, J., "Framework for Outsourcing: based on theoretical review and empirical
findings from Danish heavy industry," In: Hvolby, H.H. (Ed.), The Fourth SMESME
International Conference, Department of Production. Aalborg University, Denmark, pp.
265–274, 2001.
[27] Momme, J. and Hvolby, H. , "How Core Competence Thinking and Outsourcing
Interrelate," Proceedings of the 13th IPS Fugls0 Research Seminar, Department of
Production, Aalborg University, Denmark, 1998, pp.233 .260.
This paper may be cited:
Ahmadabadi, M. N., Bagheri, A. and Abolghasemi, F., 2014. Optimization
of Outsourcing ICT Projects in Public Organizations; Case Study: Public
Center of Iranian ICT Studies. International Journal of Computer Science
and Business Informatics, Vol. 11, No. 1, pp. 20-39.

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 40
An Optimized CBIR Using Particle
Swarm Optimization Algorithm
Subhakala S.
Sri Krishna college of technology,
Coimabatore,India.
Bhuvana S.
Sri Krishna college of technology,
Coimabatore,India.
Radhakrishnan R.
Sri shakthi institute of Engineering and technology,
Coimabatore,India.
ABSTRACT
Storage and retrieval of images over a large database is an important issue. Content Based
Image Retrieval system provides solution for this issue. In Content Based Image
Retrieval(CBIR) similar images are retrieved using low level features such as color,
texture, edge, etc that are extracted both from the query image and the database. In CBIR
less amount of retrieval time with high accuracy is desired property. The proposed system
achieves this property by using Particle Swarm Optimization algorithm. The proposed
system consists of the following phases (i) Color feature extraction using (luminance(y),
blue chrominance (u), red chrominance (v)) method (ii) Texture feature extraction using
Grey Level Co-occurrence Matrix (iii) Edge feature extraction using Edge Histogram
Descriptor (iv) Measurement of Similarity between Query image and the Database image
using Euclidean Distance. (v) Optimization of retrieved result using Particle Swarm
Optimization. In comparison with the existing approach, the proposed approach
significantly improves the precision and recall of the retrieval system.
Keywords
Accuracy, Particle Swarm Optimization, Luminance, Chrominance, Edge Histogram
Descriptor.
1. INTRODUCTION
The Content Based Image Retrieval is a method which uses visual contents
to search images from large image repositories. According to user’s interest,
has been an active research area over the last few years. Users are exploiting
the opportunity [1] to access remotely stored images in all kinds of new and
exciting ways. However this has the problem of locating a desired image in
a large and varied collection. This leads to the rise of new research and
development area CBIR, the retrieval of images in the basis of features
automatically extracted from the image themselves.
The increase in computing power and electronic storage capacity has lead
to an exponential increase in the amount of digital content available to users

IJCSBI.ORG
ISSN: 1694-2108 | Vol. 11, No. 1. MARCH 2014 41
in the form of images and video, which form the bases of much
entertainment, educational and commercial applications. Consequently, the
search for the relevant information in the large space of image and video
databases has become more challenging. How to achieve accurate retrieval
results is still an unsolved problem and an active research area.
The currently available CBIR technique retrieves stored images from a
collection of given images by comparing features [6] [7] and automatically
extracted from the image themselves. The most commonly used features are
color, shape, texture etc.
The proposed system uses color, texture [11] and edge feature extraction.
From the extracted features similarity is measured using Euclidean distance.
The results are optimized using Particle Swarm Optimization. This system
will achieve the better retrieval accuracy.
1.1 Feature Extraction
Feature extraction is a form of dimensionality reduction. It reduces the
input size. This approach is helpful when images are in larger size. Reduced
feature representation is required for tasks such as query matching and
similarity retrieval. Feature extraction [3] is very different from feature
selection. Feature extraction consists in transforming arbitrary data, such as
text or images, into numerical features usable for machine learning.
1.2 Similarity Measurement
To compare the similarity between images, distance between that images
are measured. Example for a similarity measures such as Euclidean distance,
histogram intersection etc.
1.3 Optimization
Optimization [5] is defined as a set of methods and techniques for the
design and use of technical systems as fully as possible within the
parameters. Optimization can be classified into two categories: local and
global. The basic difference between local and global optimization is the
size of region where optimization conditions hold. A local optimum has an
extreme function value as compared to the points contained in the small
neighbourhood. However, the global optimum has the extreme function
value amongst all the points in the whole design space. Even though
clustering algorithms are simple and effective, they are sensitive to
initialization and easily trapped in local optima. There are many
optimization algorithms like Particle Swarm Optimization to overcome the
drawbacks of clustering techniques.

Vol 11 No 1 - March 2014

Recommended

Recommended

More Related Content

What's hot

What's hot (15)

Similar to Vol 11 No 1 - March 2014

Similar to Vol 11 No 1 - March 2014 (20)

More from ijcsbi

More from ijcsbi (20)

Recently uploaded

Recently uploaded (20)

Vol 11 No 1 - March 2014