NONHOMOGENEOUS NETWORK TRAFFIC CONTROL SYSTEM USING QUEUEING THEORY

Document Sample
NONHOMOGENEOUS NETWORK TRAFFIC CONTROL SYSTEM USING QUEUEING THEORY Powered By Docstoc
					 International Journal of
                              JOURNAL 3, Issueand Technology (IJCET), ISSN 0976
                                       Engineering
 INTERNATIONALComputer VolumeOF COMPUTER ENGINEERING –
 6367(Print), ISSN 0976 – 6375(Online)             3, October-December (2012), © IAEME
                            & TECHNOLOGY (IJCET)
ISSN 0976 – 6367(Print)
ISSN 0976 – 6375(Online)
Volume 3, Issue 3, October - December (2012), pp. 394-405
                                                                            IJCET
© IAEME: www.iaeme.com/ijcet.asp
Journal Impact Factor (2012): 3.9580 (Calculated by GISI)               ©IAEME
www.jifactor.com




     NONHOMOGENEOUS NETWORK TRAFFIC CONTROL SYSTEM
                USING QUEUEING THEORY
                        1
                      Dr.K.PRASADH, 2 Mr.R.SENTHILKUMAR,
   1
     PRINCIPAL, MOOKAMBIKA TECHNICAL CAMPUS, MUVATTUPUZHA, KERALA,
                                        INDIA.
        2,
           RESEARCH SCHOLOR, SINGHANIA UNIVERSITY, RAJASTHAN, INDIA.
                       E-Mail: {ksprasaadh, trskme }@gmail.com.


 ABSTRACT

              In computer networking, network traffic control is the progression of running,
 prioritizing, calculating or minimizing the network traffic across different network environment.
 It is essential to compute the network traffic for an efficient communication to establish the
 sources of network congestion and harass those problems particularly. To make the network
 traffic flow and communication as an effective one, the previous work used an optimal set of
 distributed traffic control laws (DCLs) for growing demand of heavy internet application in
 heterogeneous network environment. The optimality of the traffic control is achieved through
 multi-path based rate adaptation and load balancing schemes. But there is a great extent of
 optimal values to be misspecified and the rate adaptation has less congestion control and fairness
 for network traffic control. To make the network traffic control more defined, in this work, we
 are going to present a queuing theory for controlling the network traffic in non-homogeneous
 environment. This work discovers how to construct the vital model of network traffic study
 based on Queuing Theory. Using this, the network traffic forecasting ways and the firm
 congestion rate formula are obtained. By integrating the general network traffic monitor
 parameters, the inference and monitor process for the non-homogeneous network traffic
 reasonably computed. The performance of the proposed non-homogeneous network traffic
 control scheme using queuing theory is estimated with different set of nodes contrast to an
 existing optimal set of distributed traffic control laws.


 Key words: Traffic control, Heterogeneous network, queuing theory, congestion rate.



                                                394
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
1. INTRODUCTION

         Nowadays, it is mere significant to maintain computers up to date with safety
measures. Worms and viruses can utilize a set of network usage, and are usually banned or
predetermined by concerning the proper operating system patches. It is necessary to keep the
network traffic more precise from network jamming characteristics. Network access is an
integral and often-critical part of day-to-day business for most computed users because of
network traffic. Network traffic monitoring is an imperative method for network presentation
analysis and observe. The present study seeks out to discover how to construct the necessary
form of network traffic analysis based on Queuing Theory. Therefore we can apprehend the
inference and tracing process for the network traffic judiciously.
            Queuing Theory, besides called random service theory evaluate the arbitrary
instruction of queuing occurrence, and constructs up the arithmetical model by examining the
date of the network. During the calculation of the system, we can disclose the directive about the
filing probability and decide the finest scheme for the system. Assuming Queuing Theory to
guess the network traffic, it turn out to be the significant ways of network performance
calculation, examination and inference and, through this way, we can reproduce the true system,
it is practical and consistent for organizing, monitoring and defending the network.
            The quickly promising multimedia applications advertise in today’s active bureau has
established severe confronts to network bandwidth on confined area networks. An approach to
normalize and organize the network resource proficiently is mission decisive for endeavor
networks. Organizing the network proficiently regularly adjourns the requirement to promote the
network and reduce costs.
                The environment of relevance traffic might be exemplified by constant or variable
bit rate, permanent or burst distribution of bandwidth, movable or permanent instant
relationships among the end points and delay understanding. Local area network covers after
current broadband networks in eminence of service and group of service technologies. The
network strategy must state suitable network access rights and resources to examine the
discriminated types of traffic. Application precise computation is imperative for rational traffic
load and bandwidth organization. User-specific categorization is practical for substantiation,
security and obligation of privileged behavior to definite users.

2. LITERATURE REVIEW

                To control the network traffic in the network environment, several approaches have been
presented earlier for controlling the traffic schemes at different forms. The existing approaches utilized
algorithms centered on TCP kinds of traffic counting both experiential algorithms supported on control
theory [1]. The widespread strategy for network traffic control is top-down control by concerning so-
called development. A network can be splitted into numerous sub-networks and each sub-network can be
splitted into some structure blocks. By using this approach [5], the scenario-based top-down control turns
out to be more active, stretchy and more adaptive to the present traffic pattern [3].
For the network traffic organization concern, [6] proposed an improved ant algorithm with response to
purpose conservatory and active pheromone devised the course variety behavior of each ant will be
subjective according to their distinction. Privacy threat is one of the serious concerns in multi-hop
wireless networks, where attacks such as transfer study [4] and stream tracing can be simply started, [7]
proposed a new network coding supported privacy-preserving strategy against traffic study in multi-hop
wireless systems.


                                                  395
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
                  The difficulty underneath multicast transfer in heterogeneous wireless networks
[9] with system conventions is determined by MLMR algorithm [8] in order to attain the best
coding sub-graph. The supervising of the network traffic based on queuing theory in
heterogeneous environment [10] the examining of network traffic is required for estimating the
effectiveness and self-reliance from steady operations of network [11] which we converse the
presentation and calculation of network traffic organization and will give a proposal for manage
the presentation of effort traffic based on queuing theory [12].
              To improve the network traffic control scheme, in this work, the queuing theory is
used for controlling the network traffic in the non-homogeneous network environment.

3. HETEROGENEOUS NETWORK TRAFFIC CONTROL SCHEME USING QUEUING
THEORY
               The proposed work used queuing theory for efficiently controlling the network
traffic schemes raised in the non-homogeneous network environment. Using queuing theory, the
non-homogeneous network traffic is controlled by following the queue based the packet data
arrival time and sending rate to the destination. The process of the queuing theory for network
traffic control scheme in a non-homogeneous environment is explained briefly under this section.
The architecture diagram of the proposed non-homogeneous network traffic control scheme
using queuing theory is shown in fig 3.1.



                                                 Set of source &
                        Heteroge                 destination nodes
                        neous
                        network




                                                  Network
             Apply queuing                        traffic occurs
             theory




                 N/w traffic controlled by
                 computing arrival time
                 and service time of nodes


 Fig 3.1 Architecture Diagram of the proposed non-homogeneous network traffic control
                             scheme using queuing theory


                                              396
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
3.1 Model of queuing theory

        In network communication, the sending and receiving of packet data and the arrangement of the
data policy, interpret and distributing to the superior layer, in all these procedure, we can discover an easy
queuing model. According to the Queuing Theory, this communicate practice can be distracted as
Queuing theory model like fig. 3.2. In view of this type of easy data broadcasting system suits the queue
model.


                             λ'
                                                              TJ      TD         TC




                                            Fig 3.2 Queuing theory model

From the above fig. 3.2,

WhereTs=TJ+TD+TC..… (eqn 1)

          Parameter                                 Description
              λ'                       Packet distributing rate of the sender.
             TN                              Transportation delay time
              λ                             Packet data incoming speed
             Nq             Amount of data packets accumulated in the buffer (temporary
                                                      storage).
                γ                   Packets error rate in sending from receiver
               Ts                            Packet data servicing time
               TJ                                 Decoding time
               TD                                Dispatching time
               TC                                Calculating time.
                a                                inter-arrival time
               µ                                    Service rate

Table 1 Constraints Description

         The arrival rate, λ, is the average rate new nodes arrive measured in arrivals per time period.
Common units are access/second. The inter-arrival time, a, is the average time between nodes arrivals. It
is measured in time per nodes. A common unit would be seconds/access.
a = 1 / λ ………………………….. (eqn 2)

Queuing systems are usually described by three values separated by slashes
                Arrival distribution / service distribution / # of servers
Where
• M = Markovian or exponentially distributed
• D = Deterministic or constant.
• G = General or binomial distribution
The pseudo code below described the queuing model for different types of services.

                                                    397
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME


                  Step 1: Compute λ, µ, d-packet data
                  Step 2: Evaluate ρ,
                                   ρ = λ/µ
                  Step 3: If incoming of d is random with
                  deterministic service, then
                          The avg queue length = (2 ρ – ρ2) / 2 (1 -
                  ρ)
                  Step 4: Else If incoming of d is random with
                  random service, then
                         The avg queue length = 2 – ρ / 2 µ (1 - ρ)
                  Step 5: End If

                  Fig 3.3 Pseudo code for Queuing theory model

The above figure (Fig 3.3) described the process of computing the queue length based on
incoming speed, service rate of the packet data enters into the network. Based on the deputation
of services, the incoming data packet is defined. By computing the service rate, incoming rate of
data packet in the non-homogeneous network environment, the proposed network traffic control
scheme in non-homogeneous environment using queuing theory is done by using differential
equations. The table (table 1) described the parameters used in the queuing theory concepts of
non-homogeneous environment.

3.2 Queuing theory and non-homogeneous network traffic control

                 The network traffic is very common in the non-homogeneous environment. The
structure will be in inferior form, when the non-homogeneous network traffic becomes under
tremendous state, in which guides to the network congestion. There are huge contracts of study
about tracing the congestion at present, besides, the credentials which make utilize of Queuing
Theory to explore the non-homogeneous network traffic rate emerge more and more. For
forecasting the traffic rate, we often test the data disposal function of the router used in the
network. Considering a router’s arrival rate of data flow in groups is λ, and the average time
which the routers use to dispose each group is 1/µ, the buffer of the routers is B, if a certain
group arrives, the waiting length of the queue in groups has already reached, so the group has to
be lost. When the arriving time of group timeouts, the group has to resend the packet data.
Suppose, the group’s average waiting time is 1/µ, we identify Pi (l) to be the arrival probability
of the queue length for the routers group at the moment of t, supposing the queue length is i,
P (t) = (P0 (l), P1 (l), . , Pi (l) ), i = 0,1, . . . B+1 ………….. (eqn 3)
The table below described the parameters used in the queuing theory and non-homogeneous
network traffic control schemes.

                                                398
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
            Parameter                                    Description
                B                                 Buffer space of the routers
                 i                                      Queue Length
               P (t)                Probability of incoming speed of packet data at time t
               1/µ                                  Average Waiting Time
              AC (t)                                    Jamming rate
               Pn(t)               Probability of n incoming speed of packet data at time t
             Pn(t+ t)            Probability of n incoming speed of packet data at time t+ t
              o(∆t)                           No data packets arrived at time t
                           Table 2 Constraints Description

Evaluation of non-homogeneous network congestion rate

        Network congestion rate is varying all the time. The instant jamming rate and the steady
jamming rate are frequently used to examine the network traffic in non-homogeneous network
monitor. The instant rate AC (t) is the jamming rate at the instant of t. The AC (t) can be attained
by explaining the system length of the queue’s prospect distributing, which is called Pn+1(t).
        Let Pn (t) (n=0,1,. . .,i+1) to be the incoming probability of the queue time-span for the
routers set at the instant of t by allowing for the queue time-span is n . Then the queuing system
of the router’s date sets suits simple Markov Process. In proportion to Markov Process, Pn(t)
satisfies the subsequent system of discrepancy difference equations. Let,

                 Pn(t) = prob { n data packets in the system in time t }………….. (eqn 4)
Pn(t+ t) = prob {n data packets present in the system in time (t + t)} ………… (eqn 5)

Case 1:
For n ≥ 1
Pn(t+∆t) = Prob { n no. of data packets present in the system at time t } × prob { no data
          Packets coming in time (∆t)} × prob {no data packet leaving in time ∆t}
          + Prob { ( n -1) no. of data packets present in the system at time t } × prob { 1
          data packet coming in time (∆t)} × prob { no data packet leaving in time ∆t }
          + Prob {(n +1) no. Of data packets present in the system at time t} × prob {
           No data packets coming in time (∆t)} × prob {1 data packet leaving in time
           ∆t}+. . .
⇒ Pn (t+∆t) = Pn (t) × {1- λ n ∆t + o (∆t)} × {1- µ n ∆t + o (∆t)} + Pn-1 (t) {λ n-1 ∆t + o (∆t) }
× {1- µ n-1 ∆t + o(∆t) }
               + Pn+1 (t) {1- λ n+1 ∆t + o (∆t)} × {µ n+1 ∆t + o(∆t) } + o(∆t)
                                                                      …………. (eqn 6)
Case 2:
For n =0,
P0 (t+∆t) = prob {no data packet present in the system in time (t+∆t) } = prob {no data
           packet present in time t } × prob { no data packet coming in time ∆t }
           + prob {one data packet present in time t} × prob {no data packet coming In
           time ∆t } × prob { one data packet leaving in time ∆t } .
⇒ P0 (t+∆t) = P0 (t) × {1- λ 0 ∆t + o(∆t)} + P1 (t) × {1- λ 1 ∆t + o(∆t)} …….. (eqn 7)


                                                399
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
Case 3:
For n =C+1,
PB+1(t+∆t) = prob {(B+1) no. of data packet present in the system in time (t+∆t )} =
              Prob {B no. Of data packet present in time t} × prob {1 data packet coming
              In time ∆t} × prob {no data packet leaving in time ∆t}
             + Prob {(B+1) no of data packets present in time t} × prob {no data
               packet leaving in time ∆t }
⇒ PB+1(t+∆t) = PB (t) × {λB ∆t + o (∆t)} × {1- µB ∆t + o (∆t)} + PB+1 (t) × {1- µB+1 ∆t + o(∆t)}
        …………… (eqn 8)
The instant jamming rate of the non-homogeneous network can not be utilized to estimate the steady
operating state of the system, so it is necessary to acquire the constant jamming rate of the system. The
steady jamming rate means it will not modify with the time varying, when the system mechanism in a
steady operating situation. The classification of the steady jamming rate is
AC (t) = lim AC (t) ………. (eqn 9)
            t− ∞
In view of dispensing of the steady extent of the queue and C as the defense of the router, the steady
jamming rate can be attained in two ways: initially, we acquire the instant jamming rate, and then build its
bound out. According to its classification, it can be attained with the dispensing of the extent of the queue.
The second method is distribution is done through steady state equations. The pseudo code below (Fig
3.4) described the process of queuing theory with non-homogeneous network environment.

                        Step 1: Identify the incoming packet data d
                        Step 2: If more number of d arrives in non-homogeneous
                        network environment                network traffic occurs
                        Step 3: Apply Queuing theory
                        Step 4: Compute traffic rate
                                          To test the data packet lost
                        Step 5: Find the incoming speed of data packet λ
                        Step 6: Find the average time of packet data left the group
                                          1/µ
                        Step 7: If λ exceeds the time t,
                        Step 8: resend the packet data
                        Step 9: Compute P (t)
                        Step 10: End If
                        Step 11: If network congestion occurs
                                          Compute AC (t)
                        Step 12: End if
                        Step 13: Based on queue length, the probability of non-
                        homogeneous network
                                 Congestion rate evaluated


                                                     400
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME

                      Step 14: If n>=1
                                        Use eqn 6
                      Step 15: else If n = 0
                                        Use eqn 7
                      Step 16: Else
                                        Use eqn 8
                      Step 17: End If
                      Step 18: End if
                      Step 19: Control the network traffic by computing the traffic
                      rate and congestion rate
                      Step 20: End


Fig 3.4 Pseudo code for Queuing theory model in non-homogeneous network environment

         The above fig (Fig 3.4) described the entire process of controlling the non-homogeneous
network environment using queuing theory model with different set of equations. Based on
incoming speed of the data packet, the service rate of the network is evaluated and the traffic rate
is also being identified to test the data packet lost. If network congestion occurs, the congestion
rate is computed based on the queue length and the probability of the non-homogeneous network
traffic is computed based on the packet data entering and leaving the group into the non-
homogeneous network environment. Based on the queuing system described, the non-
homogeneous network traffic is controlled efficiently.

4. EXPERIMENTAL EVALUATION

       The proposed non-homogeneous network traffic control method is efficiently done
through queuing theory. Based on queuing theory, the incoming data packets of the non-
homogeneous network environment followed the queue based on incoming and service rate. The
experimental results on the non-homogeneous network of different internet application are
achieved from the test data sets of traffic streams used from Internet Service Providers from their
universal connectivity servers. The diverse environment of universal connectivity servers
presents dynamically unreliable traffic streams as per the user insist and the variable character of
the non-homogeneous metrics vital for the direct schemes to be accepted. The performance of
the proposed non-homogeneous network traffic control method using queuing theory is
measured in terms of

i) Packet data traffic
ii) Queuing efficiency
iii) Network congestion



                                                    401
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
5. RESULTS AND DISCUSSION
         From this work, it is being observed that the proposed non-homogeneous network traffic scheme
efficiently controlled by queuing theory (NHNTC by QT) and the network traffic is keenly observed by
maintaining queue and the network congestion are also been eradicated by eomputign the traffic rate.
Compared to an existing optimal set of distributed traffic control laws (OSDTC), the proposed network
traffic scheme using queuing theory performs well. The below table and graph described the performance
of the proposed non-homogeneous network traffic scheme efficiently controlled by queuing theory.

     No. of packets (p)                         Packet data traffic (kbps)
                                 Proposed NHNTC by QT              Existing OSDTC
              25                             75                            100
              50                             90                            135
              75                            110                            210
             100                            124                            280
             125                            150                            350
                          Table 5.1 No. of packets vs. Packet data traffic

The above table (table 5.1) described the packet data traffic arised when more number of packets entered
into the network for packet data communication. The outcome of the proposed non-homogeneous
network traffic scheme efficiently controlled by queuing theory (NHNTC by QT) is compared with an
existing optimal set of distributed traffic control laws (OSDTC).


                                                   350
                                                   300
                                                   250
                                    Packet data 200
                                    traffic (kbps) 150
                                                    100
                                                     50
                                                      0
                                                          25    50    75   100 125
                                                          No. of packets


                                            Proposed NHNTC by QT Existing OSDTC

                           Fig 5.1 No. of packets vs. Packet data traffic

Fig 5.1 described the process of packet data traffic arised in the non-homogeneous network environment.
The proposed non-homogeneous network traffic scheme efficiently controlled by queuing theory used
queuing models for organizing the data packets according to the servicing and incoming time. The packet
data traffic management is being efficient in the proposed non-homogeneous network traffic scheme
efficiently controlled by queuing theory compared to an existing optimal set of distributed traffic control
laws only distribute and formed the packet data based on control laws. In the proposed NHNTC by QT,
the queuing model is efficiently formed by evaluating the traffic rate and the network congestion rate. The
variance in efficiency of packet data traffic is 40-60% low in the proposed NHNTC by QT.

                                                               402
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME


     No. of packets (p)                                  Queuing Efficiency (%)
                                Proposed NHNTC by QT                              Existing OSDTC
             25                           20                                             14
             50                                   38                                    25
             75                                   62                                    54
            100                                   84                                    76
            125                                  100                                    87
                       Table 5.2 No. Of packets vs. Queuing efficiency

The above table (table 5.2) described the queuing efficiency when more number of packets
entered into the network for packet data communication. The outcome of the proposed non-
homogeneous network traffic scheme efficiently controlled by queuing theory (NHNTC by QT)
is compared with an existing optimal set of distributed traffic control laws (OSDTC).


                                                100
                                                 80
                                                  60
                                      Queuing
                                   Efficiency (%) 40

                                                 20
                                                   0
                                                       25    50    75   100 125
                                                       No. of packets


                                          Proposed NHNTC by QT Existing OSDTC

                             Fig 5.2 No. Of packets vs. Queuing efficiency

Fig 5.2 described the queuing efficiency of the non-homogeneous network environment. The
proposed non-homogeneous network traffic scheme efficiently controlled by identifying the
packet data lost in the network. The packet data traffic management is being efficient, since it
followed the queue based on the probability of the nodes in the network in the proposed non-
homogeneous network traffic scheme efficiently controlled by queuing theory compared to an
existing optimal set of distributed traffic control laws only distribute and formed the packet data
based on control laws. In the proposed NHNTC by QT, the queuing model is efficiently formed
by evaluating the service ti. The variance in efficiency of packet data traffic is 20-25% high in
the proposed NHNTC by QT.


                                                       403
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME

                          No. of        Network Congestion Rate
                         nodes (n)                 (%)
                                        Proposed        Existing
                                       NHNTC by         OSDTC
                                           QT
                            10              5              9
                            20             13             16
                            30             21             27
                            40             28             36
                            50             34             49
                     Table 5.3 No. Of nodes vs. Network Congestion rate

The above table (table 5.3) described the network congestion rate when more number of packets
entered into the network for packet data communication. The outcome of the proposed non-
homogeneous network traffic scheme efficiently controlled by queuing theory (NHNTC by QT)
is compared with an existing optimal set of distributed traffic control laws (OSDTC).


                                                50

                                                40
                                    Network      30
                                 Congestion rate
                                      (%)        20
                                                 10
                                                  0
                                                      10   20    30   40   50
                                                      No. of nodes


                                         Proposed NHNTC by QT Existing OSDTC

                          Fig 5.3 No. Of nodes vs. Network Congestion rate

Fig 5.3 described the process of network congestion rate in the non-homogeneous network
environment. Since the proposed non-homogeneous network traffic scheme efficiently used the
queuing models for organizing the data packets according to the servicing and incoming time,
the network congestion rate is low. The packet data traffic management is being efficient in the
proposed non-homogeneous network traffic scheme efficiently controlled by queuing theory
compared to an existing optimal set of distributed traffic control laws only distribute and formed
the packet data based on control laws. In the proposed NHNTC by QT, the network congestion is
efficiently cleared and the variance in efficiency of network congestion rate is 20-30% low in the
proposed NHNTC by QT.
            Finally, it is considered that the proposed NHNTC by QT efficiently followed the
queuing model by distributing the packet data based on the servicing time and arrival rate of the
packet data in the non-homogeneous network environment. The network traffic also be
controlled in a reliable manner.

                                                       404
International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 –
6367(Print), ISSN 0976 – 6375(Online) Volume 3, Issue 3, October-December (2012), © IAEME
6. CONCLUSION

         In existing optimal set of distributed traffic control laws, multi-path based rate adaptation
and load-balancing schemes are used for traffic control schemes. The issues raised over existing
optimal set of distributed traffic control laws are well handled by the proposed non-
homogeneous network traffic scheme efficiently controlled by queuing theory. The proposed
non-homogeneous network traffic scheme, at first, formed the queue based on the packet data
arrival. If more number of packets arrived, the proposed used queuing model by maintaining and
servicing the packet data based on arrival rate. After that, if network congestion occurs, the
proposed NHNTC by QT efficiently controlled the traffic by evaluating the traffic rate. The
experimental results showed that the proposed non-homogeneous network traffic scheme
efficiently controlled by queuing theory is efficient in terms of network congestion rate, queuing
efficiency and the proposed one outperforms well in the network traffic control process in the
non-homogeneous network environment.

REFERENCES:

[1] Bernardo A. Movsichoff, Constantino M. Lagoa, and Hao Che “ End-to-End Optimal
Algorithms for Integrated QoS, Traffic Engineering, and Failure Recovery “ IEEE /ACM
Transactions on networking, Vol. 15, No. 4, August 2007.
[2] B. A. Movsichoff, C. M. Lagoa, and H. Che, “Decentralized optimal traffic engineering in
connectionless networks,” IEEE J. Sel. Areas Commun., vol. 23, pp. 293–303, Feb. 2005.
[3] F. Paganini, Z. Wang, J. C. Doyle, and S. H. Low, “Congestion control for high performance,
stability, and fairness in general networks,” IEEE/ACM Trans. Netw., vol. 13, no. 1, pp. 43–56,
2005.
[4] Ping Luo, Hui Xiong, et. Al., ‘ Information-Theoretic Distance Measures for Clustering
Validation: Generalizationand Normalization ‘, IEEE TRANSACTIONS ON KNOWLEDGE
AND DATA ENGINEERING, VOL. 21, NO. 9, SEPTEMBER 2009.
[5] YubinWang VranckenJ. Et.Al., “Implementing scenarios coordination for road network
traffic control”, 2010 IEEE International Conference on Systems Man and Cybernetics (SMC).
[6] Qi Bing Jun Lu et. Al., ‘An Improved Ant Algorithm for Network Traffic Control’, 4th
International Conference on Wireless Communications, Networking and Mobile Computing,
2008. WiCOM '08.
[7] Yanfei Fan Yixin Jiang et.al., “Network Coding Based Privacy Preservation against Traffic
Analysis in Multi-Hop Wireless Networks”, IEEE Transactions on Wireless Communications,
2011.
[8] Shah-Mansouri, V. et. Al., “Lifetime-resource tradeoff for multicast traffic in wireless
sensornetworks”, IEEE Transactions on Wireless Communications, 2010.
[9] Duk Kyung Kim Griffith, D. et. Al., „A New Call Admission Control Scheme for
Heterogeneous Wireless Networks“, IEEE Transactions on Wireless Communications, 2010.
[10] Kamali,S.H. Hedayati,M.et.Al.,“The Monitoring of the Network Traffic Based On Queuing
Theory and Simulation in Heterogeneous NetworkEnvironment”, ICCTD '09. International
Conference on Computer Technology and Development, 2009.
[11] Jong-hwan Kim et. Al., “Active Queue Management for Flow Fairness and Stable Queue
ength”, IEEE Transactions on Parallel and Distributed Systems, 2011.



                                                 405

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:41
posted:12/29/2012
language:
pages:12