Docstoc

INTELLIGENT NETWORK SELECTION USING FUZZY LOGIC FOR 4G WIRELESS NETWORKS

Document Sample
INTELLIGENT NETWORK SELECTION USING FUZZY LOGIC FOR 4G WIRELESS NETWORKS Powered By Docstoc
					   INTERNATIONAL JOURNAL OF ELECTRONICS AND
   International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
   0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME
COMMUNICATION ENGINEERING & TECHNOLOGY (IJECET)
ISSN 0976 – 6464(Print)
ISSN 0976 – 6472(Online)
Volume 4, Issue 2, March – April, 2013, pp. 451-461
                                                                           IJECET
© IAEME: www.iaeme.com/ijecet.asp
Journal Impact Factor (2013): 5.8896 (Calculated by GISI)                ©IAEME
www.jifactor.com




    INTELLIGENT NETWORK SELECTION USING FUZZY LOGIC FOR
                   4G WIRELESS NETWORKS

                             C.Amali [1],Bibin Mathew[2], B.Ramachandran[3]
                 [1][2][3]
                         Department of Electronics and Communication Engineering
                              S.R.M University, Chennai, India-603203


   ABSTRACT

           Due to the heterogeneity and the diversity of access networks, various user
   applications with different Quality of Service (QoS) requirements pose new challenges on
   multi-interface Mobile Terminal (MT) in designing optimal network selection algorithm for
   guaranteeing seamless QoS support to the users. Thus, service adaptive QoS metrics of
   mobile users can be improved by sharing the resources of different Radio Access Networks
   (RAN) efficiently. A new intelligent Vertical Handoff (VHO) scheme is presented that
   utilizes Fuzzy Logic based Linguistic Variables to estimate the necessity of handoff and to
   determine a new point of attachment in order to fulfill the end users requirements. As each
   traffic has a different set of QoS requirements, separate Fuzzy Logic Controllers (FLC) are
   used for each traffic to improve the overall performance of proposed system. To maintain
   continuous services while moving in heterogeneous environments, a Fuzzy Multi Attribute
   Decision Making (MADM) access network selection function is used to select a suitable
   network. In order to achieve the service continuity, a vertical handoff decision scheme is
   proposed to enhance the service mobility by selecting the suitable network based on QoS
   requirements of applications and network characteristics.

   Keywords— Fuzzy logic, Handoff Decision, Heterogeneous Networks, NEF, QoS

    1.    INTRODUCTION

          The next generation wireless communication systems are expected to integrate
   multiple Radio Access Technologies (RAT) in terms of services and application
   requirements. Thus, a user will get access to a range of services having different bandwidths
   and QoS requirements with the multimode terminal in "Always Best Connected" (ABC) [1]
   manner.

                                                  451
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

        The strengths of third generation (3G) cellular networks such as UMTS and CDMA
2000 consist of their global coverage where as their weakness lie in bandwidth capacity and
operation costs. With the fast evolution of wireless technologies, a number of wireless access
networks are now emerging thereby creating a heterogeneous network environment for both
mobile users and internet service providers (ISPs). However, any single type of these wireless
networks cannot provide all types of services, e.g., wide coverage and high bandwidth. On
the other hand wireless LAN (WLAN) offers higher bandwidth with low service costs, but it
provides small coverage area [2]. IEEE802.16x or WiMAX systems are optimized to provide
real time high data rate services in WMAN environments. There is no single system that is
good enough to replace all the other technologies.
        The strength of 4G system relays on integrating the existing and newly developed
wireless systems instead of putting efforts into developing new radio interfaces and
technologies to provide seamless mobility and better service quality for mobile users. In this
paper, the coexistence of UMTS, WLAN and WiMAX access networks are considered as a
heterogeneous wireless network. When connections have to switch between heterogeneous
networks for performance and high availability reasons, seamless vertical handoff is
necessary to provide uninterrupted services to the mobile users.
        VHO is the seamless transfer of an ongoing user session between different
heterogeneous radio access technologies. The vertical handover decision process determines
when and where to hand over in a heterogeneous environment when the user is on the move.
Decision criteria include MT speed, user preferences, network conditions and application
requirements. For each network, there is a Received Signal Strength (RSS) threshold value
below which connection break with active base station. Therefore, the signal strength must be
greater than threshold point to maintain the connection with serving network. The signal
becomes weak as mobile moves far away from serving base station and gets stronger signal
towards new base station as it moves closer. There is a need for Handoff if RSS of active
base station decreases below threshold level to maintain the connection.
        The aim of this work is to design and simulate an application specific VHO among
WiMAX, UMTS and WLAN using fuzzy tool. The applications taken into the consideration
are conversational and real time video streaming and each of these applications require a
different QoS. All the three technologies do not support these applications with equal QoS.
Each wireless technology has a limit on mobility support and cost of service offered. Users
will always look for lower cost, keeping QoS intact; hence a VHO algorithm must try to
select the most cost effective network as target network. WLAN has the least cost of service
followed by WiMAX and UMTS. In this work, speed of the vehicle (mobility) is considered
as an important factor to evaluate the networks before ranking the networks based on the
application requirements.
        The remainder of this paper is organized as follows: Related and existing works are
discussed in section 2. Overview of proposed algorithm is presented in section 3. Section 4
comprises various modules of proposed algorithm. Performance assessment is carried out in
section 5 and finally conclusion is given in section 6.

2.     RELATED WORK

       To improve the performance of the handoff scheme, a network discovery algorithm
based on the fuzzy logic multiple objective decision making system is presented in [3]. In [4]
& [5], an adaptive fuzzy based handoff decision system is developed which consider
parameters such as data rate, cost and RSSI to obtain training elements. With the training

                                             452
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

element of ANFIS from fuzzy based system, the rules and the membership function can be
properly tuned to optimize the handoff performance. In [6], a Modified Weight Function
based Network Selection Algorithm (MWF-NSA) that considers user preference and
application profile is proposed in deciding the weight functions of the networks. In [7] fuzzy
multiple attribute decision is used for vertical handoff decision. The application requirements
for both conversational and streaming are given in [8].
        Although, there have been various vertical handoff algorithms available in the
literature, our proposed work aims to incorporate the user speed, characteristics of networks,
user preference, cost and QoS requirements of different traffic in fuzzy logic based network
selection algorithm to select the optimal network to provide efficient vertical handoff for
heterogeneous wireless networks. In the proposed algorithm, the network related, terminal
related, user related and service related attributes are included, which improves the accuracy
of the VHO.

3.     OVERVIEW OF PROPOSED ALGORITHM

        Due to the heterogeneity and the diversity of access networks, various user
applications with different QoS requirements pose new challenges in designing optimal
network selection algorithm for guaranteeing seamless QoS support to the users. Thus, VHO
is necessary to provide uninterrupted services to mobile users anywhere and anytime in 4G
Networks. The various techniques used for executing handoff can be classified into Mobile-
Controlled Handoff (MCHO), Network-Controlled Handoff (NCHO) and Mobile-Assisted
Handoff (MAHO). In MCHO, the mobile node continuously monitors the signal strength of
access points and initiates the handoff procedure when certain handoff decision is triggered.
Fuzzy logic based VHO algorithms are best suitable for MCHO in integrated networks. The
fuzzy logic controllers are designed for the different modules of proposed algorithm using
fuzzy rules base. The fuzzification comprises the process of transforming the crisp inputs into
the fuzzy sets via the membership functions. The fuzzy rules base consists of a collection of
fuzzy if-then rules to represent the human knowledge about the problem.
        In fuzzification process, if inputs X={x1, x2, …, xm} are memberships of fuzzy set
     ~ ~         ~
Y={ A , B ,…, M }, respectively, then the degree of membership of X={ x1, x2, …, xm} is
given by

                             X    X     X
                         (µ L , µ M , µ H )     where µY ( X ) ∈ [0,1]

The reason for using Fuzzy MADM is that

        i) multiple parameters can be processed simultaneously and provides the best solution
for VHO decision when the input exhibits uncertainty.
        ii) the traditional fuzzy based Vertical Handoff Decision Algorithm (VHDA) needs
defuzzification, which may increase the handoff delay. But the proposed VHDA selects the
best suitable network based on the values of network evaluation function.




                                              453
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME




                     :
                Fig.1: Functional Block Diagram of Proposed Mechanism

       Fuzzy logic based proposed algorithm is divided into four modules. The functional
                           sed                           fig.
block diagram of proposed mechanism is shown in fig.1. The first module determines
whether handover is necessary when the mobile terminal is moving across heterogeneous
     ess                       mod                                              higher
wireless networks. The next module is executed only if handoff factor is high than the
                         .                                         D
handoff threshold value. In the second module, User Satisfaction Degree (USD) is evaluated
                                                                                   network
based on the speed, network load and cost of service in order to select the netwo that
supports. In the third module, QoS requirements of conversational and real time video
streaming traffic are considered to determine QoS factor for the available networks to provide
                                                                  performance       networ
uninterrupted services to mobile users. In fourth module, the performance of networks is
                                               Function
evaluated by calculating Network Evaluation Function (NEF) based on the va      values of USD
                                                                             2.
and QoS factor of ongoing service. The simulation scenario is shown in fig.2. The scenario is
common to both urban and suburban environments where all the networks will be present.
The simulation is based on locations where all the three networks are available. WiMAX
           ea
coverage area is larger than UMTS and WLAN for simulation in the proposed scheme.scheme




                :
           Fig.2: Simulation Scenario for Urban and Suburban Environments

4.     FORMULATION OF PROPOSED ALGORITHM

       The proposed FUZZY based VHO algorithm is formulated using 4 modules as shown
                                                                             netw
in fig.3. In this algorithm, MT speed, service cost, application profile and network
characteristics are considered for the evaluation of available networks. The detailed
explanation of the modules is described in the following sub sections.

                                             454
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME




                       Fig.3: Flow Diagram of Proposed Algorithm

4.1     HANDOFF FACTOR ESTIMATION
        This module examines the conditions of serving network and estimates the handoff
factor. By estimating the handoff factor to initiate the handoff at the right time, end user
satisfaction can be maximized which in turn improves the overall QoS in heterogeneous
wireless networks. RSS of serving network is estimated using suitable path loss models based
on the environments in which mobile terminal is moving. If RSS of serving network keep on
decreasing, there is a need for handoff to another network to provide uninterrupted services to
mobile users.
 For this case, FLC is deployed using fuzzy logic toolbox in MATLAB. The Fuzzy Inference
System (FIS) accepts the fuzzified values and interprets the fuzzified values to the outputs
based on the user defined rules. Whenever the RSS drops below the threshold value, the
probability of handoff becomes high. Thus, handoff factor is estimated to select the best
suitable network. The next stage of network evaluation process is evaluated only when
handoff factor is high. The corresponding FIS model for handoff Factor estimation is shown
in fig.4. Mamdani system is used here because of its wide spread acceptance and it is well
suited to human input.




                          Fig.4: RSS Handoff Decision controller


4.2     USER SATISFACTION DEGREE (USD) CALCULATION
        Our proposed scheme chooses only a few parameters that are critical to maximize the
end users satisfaction while performing VHO in heterogeneous environment. In this stage, the
performance of networks are evaluated based on the parameters like MT speed, service cost
and network load. The QoS requirements of ongoing service can be provided easily by
maximizing end user satisfaction based on their preference, location and application
requirements. Thus available networks are evaluated to select the network capable of
satisfying the user request during VHO operation. For example, networks with less coverage
area cannot support the users with high speed. It requires large number of handoff to
complete the ongoing service in MT.

                                             455
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

        Before evaluating the networks based on given application, it is necessary to evaluate
the networks based on speed, user preference (cost) and network conditions (load) to reduce
the outage probability and also to avoid unnecessary handovers in heterogeneous wireless
networks. Thus, the end users can specify their needs and preferences by assigning
membership functions to each system parameter. FLC is designed by assigning membership
functions to speed, cost and network load to evaluate the networks according to their
characteristics. FIS accepts the input fuzzy sets and maps with outputs based on the fuzzy
rules. For example, if the MT speed is low and user preference is low cost, then WLAN will
be selected. The corresponding membership functions are given in fig.5




                  Fig.5: Membership Functions of Speed, Load and Cost

4.3     QoS FACTOR EVALUATION
        After estimating the user satisfaction degree, the characteristics of different traffic
must be taken into account in designing optimal network selection algorithm for guaranteeing
seamless QoS support to the users. In this module, two fuzzy logic controllers are deployed
for conversational and video streaming traffic. Service sensitivity parameters such as RSS,
data rate and connection delay are considered to achieve guaranteed QoS to mobile users.
The membership functions are assigned as low, medium and high to the input variables and
fuzzy rules are made as per the requirements of 3GPP QoS classes.
        For conversational traffic, RSS is an important factor to maintain a good link between
MT and BS with low delay and minimum bandwidth requirements. But, real time video
streaming application requires large bandwidth and RSS with tolerable delay. The application
requirements considered for the simulation is shown in Table1.

          Table 1: Conversational and Streaming Application Requirements
          Application     Symmetry      Data rate      Delay        RSS

         Conversational       Two Way        4-64 kbit/s   <150 ms             High
                                                           <400 ms
                                                           (max limit)

           Streaming          One Way          16-384      < 10 s              High
                                                kbit/s



                                             456
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

4.4    NETWORK EVALUATION FUNCTION ESTIMATION
       After filtering the available networks using user satisfaction degree and QoS factor,
NEF is calculated. NEF is determined to provide optimization between user satisfaction
degree and QoS factor. The outputs of the previous modules are combined in parallel to
include the mobility, QoS requirements of applications and network characteristics in order to
provide best solution to the network selection problem.
                                                                                ne
       By selecting the network with maximum USD and QoS factor as target network, the
                         quired
number of handoffs required for MT to complete its connection can be reduced with
                       ,
guaranteed QoS. Thus, the MT is always connected to the best suitable network for service
continuity and also to guarantee QoS requirements of ongoing service.

5.     PERFORMANCE ASSESSMENT

         The performance of the VHDA is tested within the framework of chosen simulation
scenario as shown in fig.2. The NEF is calculated using SIMULINK model of MATLAB.
The user satisfaction degree is plotted with respect to velocity for different percentage of
                                 .    correspon                                6.
traffic load and cost of service. The corresponding figures are shown in fig.6 The fig.6.1
                      ntage
shows at low percentage of cost and traffic load, WLAN is the preferred network up to 20
Km/hr thereafter WiMAX USD value is increasing. In fig.6.2 after 20km/hr both UMTS and
WiMAX are having same value up to 50km/hr. After that WiMAX USD is increasing since
                            .                    low
it supports higher velocity. For medium cost and low traffic load WiMAX USD is performing
                               6.3.            service
better which is shown in fig.6.3. The cost of service is high in UMTS and it supports high
                                                fig.
traffic load, that’s why UMTS USD is higher in fig.6.4.




                       n
Fig.6: User Satisfaction Degree of WLAN, UMTS and WiMAX (6.1) for 10% Cost and
10% Traffic Load.(6.2) for 10% Cost and 50% Traffic Load. (6.3) for 50% Cost and
             10% Traffic Load. (6.4) for 70% Cost and 70% Traffic Load

                                             457
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

                                      propo                         7.
       The SIMULINK model of the proposed algorithm is shown in fig.7. The model works
                          l
according to the functional block diagram given in Fig.1.




                       :
                  Fig.7: SIMULINK model of the proposed mechanism




                                             458
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

                                                  example,
        The NEF table is shown in table 2. As an example four cases of USD are taken and in
each USD case, four different QoS cases are considered. The examples considered here a      are
(a) 15 % speed and 10% cost of service and traffic load. (b) 30% speed, 50% traffic load and
10% cost of service. (c) 50% speed and cost of service and 10% traffic load. (d) 70% of
                                                               examples     ur
speed, cost of service and traffic load. For all the above examples, four cases of QoS
                                                      ,
parameters are considered. QoS parameters are RSS, data rate and delay and are represented
        ,
by high, medium and low in table 2. This table gives the final selection of network based on
NEF value calculated from USD and QoS values. Delay is considered to be an important
factor in conversational application whereas data rate is more important in video s streaming,
                               uirement
which means bandwidth requirement is more in video streaming applications. WLAN and
WiMAX is having higher bandwidth when compared to UMTS. UMTS and WiMAX is
having higher coverage area and supports more load than WLAN. The corresponding bar
diagrams of the examples shown in table 2 are given in fig.8. From the fig.8, it is understood
that for low value of speed, traffic load and cost, WLAN is the preferred network for both
                                     applications. As the USD & QoS values changes, there is
conversational and video streaming applicat
                                                                   eed,
a corresponding change in the final NEF. For medium values of speed, load and cost, UMTS
                                                                             applications.
is the preferred network for conversational and WiMAX for video streaming applications As
                              ,
the cost of service increases, UMTS is selected because of its higher cost than WLAN and
                           rs
WiMAX. If QoS parameters like RSS and data rate have a low value and when delay is high  high,
then all the networks have a low NEF for all the cases of USD. In this case, no networks are
selected for incoming traffic.

          :                           Conversational               Applications
   Table 2: Example of NEF Values for Conversational and Streaming Applicatio




                                             459
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME




             :                              (
       Fig.8: Graph showing NEF values (from Table 2) of different networks for
                                 applications. .1)
conversational and streaming applications (8.1) NEF for 15% speed and 10% of cost of
                               .2)NEF for 30% speed, 50% traffic load and10% cost of
   service and traffic load. (8.2)
           .3)NEF                                                         .
service. (8.3)NEF for 50% speed, 10% traffic load and 50% cost of service. (8.4)NEF for
                70% speed and 70% traffic load and 70% cost of service


                                             460
International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN
0976 – 6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 2, March – April (2013), © IAEME

6.     CONCLUSION

        This paper proposes an intelligent fuzzy logic based network selection for 4G wireless
networks. The developed algorithm considers the USD as well as QoS parameters for both the
conversational and video streaming applications.The USD, QoS and NEF are evaluated using
separate fuzzy logic controllers and tested using SIMULINK .Whenever there is a need for
handoff, the NEF is calculated using USD and QoS of parameters from the available list of
networks present at that time. The network with high NEF is selected as the preferred network
and is going to serve the user. In the proposed algorithm, the network related, terminal related,
user related and service related attributes are included to minimize the number of VHO. Thus, the
best suitable network is selected for the ongoing traffic by providing optimization between the
complexity and improved QoS. The proposed scheme can benefits both users and networks by
handling the uncertainity and time varying information using fuzzy logic rules.
        For further research, fuzzy logic technique can be integrated with other MADM methods
to provide more efficient network selection and also to analyze the effect of complexity in the
proposed scheme.

REFERENCES

[1] E.Gustaffson and A.Jonsson, “Always Best Connected”, IEEE wireless Communication ,
vol.10, no.1, Feb.2003, pp.49-55.
[2] Q. Song and A. Jamilipour , “Network Selection in an Integrated Wireless LAN and UMTS
Environment Using Mathematical Modeling and Computing Techniques”, IEEE Wireless
Communications, June 2005; DOI: 10.1109/MWC.2005.1452853.
[3] M. Lahby, C. Leghris. and A. Adib. “A Hybrid Approach for Network Selection in
Heterogeneous Multi-Access Environments”,In the Proceedings of the 4th IFIP International
Conference on New Technologies, Mobility and Security (NTMS), pp.1-5, Paris France, Feb
2011.
[4] Xu Haibo, Tian Hui, Zhang Ping, “A novel terminal-controlled handover scheme in
heterogeneous wireless networks”, Computers and Electrical Engineering 36 (2010) pp 269–279.
 [5] Ali Çalhan & Celal Çeken, “An Adaptive Neuro-Fuzzy Based Vertical Handoff Decision
Algorithm for Wireless Heterogeneous Networks” proceedings of 21st Annual IEEE International
symposium on Personal, Indoor and Mobile Radio Communications, 2010, pp 2271-2276.
 [6] C.Amali and B.Ramachandran, “Modified Weight Function Based Network Selection
Algorithm for 4G Wireless Networks”, ACM International Conference on Advances in
Computing, Communication and Informatics(ICACCI 2012) Chennai, India, pp. 292-299, 3rd –
5th August 2012.
[7] L. Xia, LG. Jiang, and C. He, "A Novel Fuzzy Logic Vertical Handoff Algorithm with Aid of
Differential Prediction and Pre-Decision Method", Proc. of IEEE ICC'06, 2006, pp. 5665-5670.
[8] Carlos Ramirez-Perez and Victor M. Ramos R. “A QoS hierarchical decision scheme for
vertical handoff”, 2012 8th International Caribbean Conference on Devices, Circuts and Systems
(ICCDCS).
[9] Prof. J. R. Pathan, Prof. A. R. Teke, Prof. M. A. Parjane and Prof. P.S. Togrikar, “Dropping
Based Contention Resolution for Service Differentiation to Provide Qos in Wdm Obs Networks”,
International Journal of Computer Engineering & Technology (IJCET), Volume 4, Issue 1, 2013,
pp. 218 - 228, ISSN Print: 0976 – 6367, ISSN Online: 0976 – 6375.
[10] Jayashree Agrakhed, G. S. Biradar and V. D. Mytri, “Optimal Qos Routing With Prioritized
Region Scheduling Over Wmsn”, International Journal of Computer Engineering & Technology
(IJCET), Volume 3, Issue 1, 2013, pp. 289 - 304, ISSN Print: 0976 – 6367, ISSN Online: 0976 –
6375.

                                              461

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:5/12/2013
language:Latin
pages:11