ResearchArticle A Hybrid Approach to Call Admission ...downloads.hindawi.com/journals/afs/2018/2535127.pdf · ResearchArticle A Hybrid Approach to Call Admission Control in 5G Networks

Research ArticleA Hybrid Approach to Call Admission Control in 5G Networks

Mohammed Al-Maitah 1 Olena O Semenova2 Andriy O Semenov2

Pavel I Kulakov3 and Volodymyr Yu Kucheruk3

1Computer Science Department Community College King Saud University Riyadh Saudi Arabia2Faculty of Infocommunications Radioelectronics and Nanosystems Vinnytsia National Technical University Vinnytsia Ukraine3Faculty for Computer Systems and Automation Vinnytsia National Technical University Vinnytsia Ukraine

Correspondence should be addressed to Mohammed Al-Maitah malmaitahksuedusa

Received 27 May 2018 Revised 26 August 2018 Accepted 18 September 2018 Published 8 October 2018

Academic Editor Katsuhiro Honda

Copyright copy 2018 Mohammed Al-Maitah et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Artificial intelligence is employed for solving complex scientific technical and practical problems Such artificial intelligencetechniques as neural networks fuzzy systems and genetic and evolutionary algorithms are widely used for communication systemsmanagement optimization and prediction Artificial intelligence approach provides optimized results in a challenging task ofcall admission control handover routing and traffic prediction in cellular networks 5G mobile communications are designedas heterogeneous networks whose important requirement is accommodating great numbers of users and the quality of servicesatisfaction Call admission control plays a significant role in providing the desired quality of service An effective call admissioncontrol algorithm is needed for optimizing the cellular network system Many call admission control schemes have been proposedThe paper proposes a methodology for developing a genetic neurofuzzy controller for call admission in 5G networks Performanceof the proposed admission control is evaluated through computer simulation

1 Introduction

Nowadays telecommunication operators face new challengesThe fifth generation mobile networks should provide newoutstanding attributes such as higher mobility higher trafficdemands and higher levels of the quality of service

Call admission control offers an effective way to avoidnetwork traffic congestion thus providing a guaranteed qual-ity of service A call admission control algorithm provides adecision whether a call should be accepted into a networkor dropped An efficient call admission control scheme canimprove the call blocking and call dropping probabilities andoverall system utilization [1ndash3]

Employing up-to-date techniques such as artificial intel-ligence in 5G network gives more capability to handletraffic Artificial intelligence can be used in case of networkoverloading and to support decision-making Thus 5G withartificial intelligence can satisfy the expected needs alongwith the new technical challenges [4]

Artificial intelligence-based techniques [5 6] havereplaced conventional techniques in different engineeringapplications and are widely used in telecommunicationnetworks [7ndash9]

Using artificial intelligence technique for call admissioncontrol was proposed in [10] According to [11] applyingthe fuzzy systems provides reducing the call rejection and ahigher quality of service Papers [12ndash14] have examined appli-cation of the fuzzy logic in control systems An algorithmbased on neural networks for decision of call admission hasbeen proposed in [15] Its results have substantiated solvingthe call admission control problem with the neural networkapproach But better results could be obtained by using ahybrid approach proposed in [16] In [17] genetic algorithmshave been proposed to be applied for optimization of the calladmission problem that led to better userrsquos satisfaction [18]

Thus using neural networks fuzzy systems and geneticalgorithms may provide solving the call rejection problemA better method can be derived when combining these

HindawiAdvances in Fuzzy SystemsVolume 2018 Article ID 2535127 7 pageshttpsdoiorg10115520182535127

2 Advances in Fuzzy Systems

three techniques That is because the combination overcomeslimitations of one technique and produces a relevant resultfor improving the quality of service [19] Using artificialintelligence in 5G networks has been investigated in [20ndash22]

The authors propose to apply in 5G mobile networksa genetic neurofuzzy controller with two input linguisticvariables and one output linguistic variable The controllermanages call acceptance or rejection thus allowing conges-tion avoidance The authors have already suggested the fuzzycontroller in [23] and the neurofuzzy controller in [24] Thispaper considers results of a further investigation

The objective of this paper is a genetic neurofuzzycontroller for call admission control in 5G mobile networks

The paper is organized as follows In the Materials andMethods the first subsection introduces linguistic variablesand membership functions for a fuzzy controller the secondsubsection describes structure of the controller and its oper-ation the third subsection considers genetic optimization ofthe controller The fourth subsection presents simulation inMatlab The last section provides conclusions

2 Materials and Methods

21 Fuzzy Controller Fuzzy systems simulate human reason-ing and are applied for characterizing behaviour of nonlinearsystems Fuzzy approach may be employed when there is nomathematical model of the process

Let us consider a mobile system formed by a number ofoperators sharing their Radio Access Networks in order toensure usersrsquo satisfaction Therefore when a user arrives tothe system and his call cannot be admitted it is transferred toanother operator to avoid the rejection

For developing a fuzzy controller its linguistic variablestheir terms and membership functions should be definedInput variables of the fuzzy controller describe all the possiblestates of a process controlled and its output variable describesall possible controlling actions Then a rule base should beassigned consisting of a set of IF-THEN rules to describe con-trolled states To evaluate operability of the fuzzy controller asimulation is performed

Input linguistic variables of the access fuzzy controller arethe effective capacity (E) and offered cell load (O) its outputvariable is the probability for call acceptation rejection ortransfer (P)The architecture of the fuzzy controller is shownin Figure 1

A fuzzy controller transforms input variables to member-ship function values evaluates a fuzzy output according to arule base and then converts the fuzzy output to a crisp one

The proposed fuzzy controller is to be converted into anadaptive neurofuzzy inference system (ANFIS) suitable foroperating under uncertain conditions For updating the rulebase a genetic algorithm is to be usedThe architecture of thegenetic neurofuzzy controller is shown in Figure 2

The linguistic variable of the effective capacity E is definedwith terms ldquolowrdquo ldquomediumrdquo and ldquohighrdquo

119879 (119864) = 119871119900119908 (119871) 119872119890119889119894119906119898 (119872) 119867119894119892ℎ (119867) (1)

Rule base

Fuzzification DefuzzificationInference EO

P

Figure 1 Architecture of the fuzzy controller

Neural Network

Fuzzification

Genetic Algorithm

Defuzzification

Rule Base

EO

P

Figure 2 Architecture of the genetic neurofuzzy controller

The linguistic variable of the offered load O is defined withterms ldquolowrdquo ldquomediumrdquo and ldquohighrdquo

119879 (119874) = 119871119900119908 (119871) 119872119890119889119894119906119898 (119872) 119867119894119892ℎ (119867) (2)

The linguistic variable of the call acceptation P is defined withterms ldquotransferrdquo ldquorejectrdquo and ldquoacceptrdquo

119879 (119875) = 119879119903119886119899119904119891119890119903 (119879) 119877119890119895119890119888119905 (119877) 119860119888119888119890119901119905 (119860) (3)

ldquoRejectrdquo means the call will be blocked ldquoAcceptrdquo means thecall will be admitted to the system ldquoTransferrdquo means the callwill be transferred to the other operator

Figures 3 and 4 present input variable membershipfunctions which are of trapezoid type and defined as

120583 (119909) =

0 119894119891 119909 le 1198861 119900119903 119909 ge 1198864(119909 minus 1198861)(1198862 minus 1198861) 119894119891 119909 isin (1198861 1198862] 1 119891 119909 isin (1198862 1198863) (1198864 minus 119909)(1198864 minus 1198863) 119894119891 119909 isin (1198863 1198864)

(4)

Figure 5 illustrates output variable membership functionswhich are singletons and defined as

120583 (119909) = 0 119894119891 119909 = 1198981 119894119891 119909 = 119898 (5)

The suggested fuzzy controller operates according to the rulebase consisting of nine rules Input and output values arespecified in Table 1

The fuzzy controller operates as followsEach input variable gets a corresponding fuzzy value

E 997888rarr ELEMEHO 997888rarr OLOMOH (6)

Advances in Fuzzy Systems 3

450 800

(E)

E

L HM

600200 350

Figure 3 Membership functions for the effective capacity Defini-tion of the membership functions L low M medium H High

M

4515 60

O

(O) HL

200 40

Figure 4Membership functions for the offered cell load Definitionof the membership functions L low M medium H High

Table 1 Fuzzy rule table denoting the output P values according tothe input E and O values

E OL M H

L R T TM A R TH A A R

Theminimum operations are performed

w1 = min [ELOL] w2 = min [ELOM] w3 = min [ELOH] w4 = min [EMOL] w5 = min [EMOM] w6 = min [EMOH] w7 = min [EHOL] w8 = min [EHOM] w9 = min [EHOH]

(7)

The crisp value is obtained

119875 = sum9119894=1 119908119894 sdot 119875119894sum9119894=1119908119894 (8)

R

0 100

P

(P) AT

minus100

Figure 5 Membership functions for the call acceptance Definitionof the membership functions T transfer R reject A accept

22 Neural Network Neural networks consist of intercon-nected artificial neurons They are able to acquire store andemploy expert knowledge and can learn new patterns Neuralnetworks are applied for such problems as optimizationclassification patternmatching function approximation anddata clustering

For improving the operability of the fuzzy controller itis converted into a neurofuzzy controller For this aim anadaptive neurofuzzy inference system (ANFIS) was chosenIt is a neural network operating like a fuzzy system underuncertain conditions combining both fuzzy logic and neuralnetworks principles ANFIS is capable of approximatingfunctions Its advantage is combining fuzzy reasoning withlearning capabilities when solving a problem

Figure 6 presents a block diagram of the neurofuzzycontroller

The access neurofuzzy controller has fuzzy rules of theform

If a = Am

and b = Bn

then c = Cpm = 1 2 3 n = 1 2 3 p = 1 2 3

(9)

The output of each node in layer 1 is the membership degreeof input

Y1j = 120583Aj (a) for j = 1 2 3Y1j = 120583Bjminus3 (b) for j = 4 5 6 (10)

The output of each node in layer 2 is the weighting factor ofthe fuzzy rule

Y2i = wi = 120583Am (a) sdot 120583Bn (b) i = 1 9 (11)

The output of each node in layer 3 is the ratio of the firingstrength of the fuzzy rule to the total value of all firingstrengths

Y3i = ni = wi(w1 + w2 + + w9) (12)

The output of each node in layer 4 is multiplication of thenormalized output with the node function

Y4i = nifi (13)


L

M

H

L

M

H

П

П

П

N

N

N

1

2

9

+P

f

O

E

Figure 6 Block diagram of the neurofuzzy controller Each neuron in first layer is used to fuzzify received crisp inputs Nine neurons insecond layer correspond to a fuzzy rule Neurons in the third layer are used for the normalization of the values The fourth layer is thedefuzzification one The single neuron in fifth layer represents an output of the neurofuzzy system

Table 2 Encoding values

E O PL 00 00 00 TM 01 01 01 RH 11 11 11 A

The output of layer 5 gives the output of the controller

Y5 = n1c1 + n2c2 + + n9c9 (14)

23 Genetic Algorithm Genetic algorithms provide an accu-rate and fast solution for optimization problems They actaccording to processes of selection mutation crossover andreproductionThey are able to optimize both continuous anddiscrete variables So a genetic algorithm is employed forupdating the rule base

A chromosome represents a solution encoded into genesHere a gene corresponds to a linguistic variable value Table 2shows encoding the fuzzy rules into chromosomes The termldquoLowrdquo of the input linguistic variables is specified as ldquo00rdquoTheterm ldquoMediumrdquo of the input linguistic variables is specifiedas ldquo01rdquo The term ldquoHighrdquo of the input linguistic variables isspecified as ldquo11rdquoThe term ldquoTransferrdquo of the output variable isspecified as ldquo00rdquo The term ldquoRejectrdquo of the output variable isspecified as ldquo01rdquo The term ldquoAcceptrdquo of the output linguisticvariable is specified as ldquo11rdquo

Here each fuzzy rule is defined by a set of 5 genes thatidentify fuzzy sets for each of two input and output variables

Chromosome for the 1-st fuzzy rule looks so

0 0 0 0 0 1 (15)

Chromosome for the 9-th fuzzy rule looks so

1 1 1 1 0 1 (16)

The fitness function is

119865 = 1120576 + 119877119872119878119864 (17)

where RMSE (root mean square error) is an objective func-tion showing the error value between the actual output andthe predicted output

119877119872119878119864 = radic 1119899119899sum119894=1

(119875119903119890119886119897 minus 119875119901119903119890119889)2 (18)

where n is a number of observationsThe genetic algorithm is as follows

(1) Generate random population of chromosomes

(2) Evaluate the fitness values of every chromosome

(3) Select two parent chromosomes according to theirfitness value

(4) Crossover the parents chromosomes to form childchromosomes

(5) Mutate the child chromosomes

(6) If the end condition is satisfied finish else go to step(2)

24 Simulation The authors have used the Matlab programto confirm the operability of the proposed neurofuzzy con-troller Figure 7 illustrates the simulated in Matlab fuzzycontroller Two input variables and one output variable werespecifiedThe rule base was assigned To check the operabilityof the fuzzy controller we set the input values and run thesimulation to obtain the output value

Let the effective capacity E=253 and offered load O=15According to Figure 8 we get the probability for call accepta-tion P=355


Figure 7 Fuzzy controller in Matlab The fuzzy controller is ofSugeno-type for converting into ANFIS

Figure 8 Simulation results Probability for call acceptation is355 It means the call is likely to be passed to the system

Let the effective capacity E=143 and offered load O=18According to Figure 9 we get the probability for call accepta-tion P=-60

Figure 10 presents the structure of the neurofuzzy con-troller

Figure 11 illustrates data applied for training the neuro-fuzzy controller

Figure 12 shows the training errorThe simulation employed the hybrid learning rule com-

bining the gradient rule and the least squares estimate Errortolerance was 0 The number of iterations was 10 The inputvalue membership functions after the training process wereof a trapezoidal type

Figure 9 Simulation results Probability for call acceptation is -60It means the call is likely to be transferred to another operator

Figure 10 Structure of the proposed neurofuzzy controller inMatlab software all nodes in five layers and links between them areshown

3 Conclusions

Thenext generation wireless networks are more complex anddynamic possessing many design challenges for the networkplanning and management

This paper considers the artificial intelligence basedmethodology for designing the controller for call admissionin 5G networks to support their quality of service

To give a better solution to the optimization problemthe suggested controller is built combining a fuzzy systemwith an artificial neural network and genetic algorithmArtificial neural networks learn the network behaviour andmake accurate predictions estimating when a call to beadmitted in a changing situation Fuzzy systems deal withuncertainty at the call admission schemes they can reducethe number of rejected calls thus increasing the quality of


Figure 11 Training data Nine desired output values and corre-sponding input values were specified as an array The x-axis showsnumber of the arrayrsquos rowThe y-axis shows desired output value

Figure 12 Training error It equals 0 meaning the training wassuccessful

service Genetic algorithms can optimize given values andprovide the resource reservation and guaranteed services forcalls

The proposed fuzzy controller was converted into theneurofuzzy system The genetic algorithm was proposedfor improving the rule base The simulation results sustainthat the considered genetic neurofuzzy controller can beemployed in 5G mobile networks The proposed approachcan increase the quality of service by decreasing the callblocking probability of the new incoming calls in a network

Data Availability

No data were used to support this study

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

References

[1] S B Deshmukh and V V Deshmukh ldquoCall admission controlin cellular networkrdquo International Journal of Advanced Researchin Electrical Electronics and Instrumentation Engineering vol 2no 4 pp 1388ndash1391 2013

[2] D E Asuquo E E Williams and E O Nwachukwu ldquoA surveyof call admission control schemes in wireless cellular networksrdquoInternational Journal of Scientific amp Engineering Research vol 5no 2 pp 111ndash120 2014

[3] V S Kolate G I Patil and A S Bhide ldquoCall admission con-trol schemes and handoff prioritization in 3g wireless mobilenetworkrdquo International Journal of Engineering and InnovativeTechnology vol 1 no 3 pp 92ndash97 2012

[4] M Mahith and S V Sheela ldquoSurvey on artificial intelligence in5Grdquo International Journal of Advance Engineering and ResearchDevelopment vol 5 no 03 pp 1277ndash1281 2018

[5] R Fuller ldquoOn fuzzy reasoning schemesrdquo The State of the Artof Information Systems Applications in 2007 Turku Centre forComputer Science Abo vol 16 pp 85ndash112 1999

[6] R Fuller ldquoFuzzy logic and neural nets in intelligent systemsrdquoInformation Systems Day Turku Centre for Computer ScienceAbo vol 17 pp 74ndash94 1999

[7] A A Atayero and M K Luka ldquoApplications of soft computingin mobile and wireless communicationsrdquo International Journalof Computer Applications vol 45 no 22 pp 48ndash54 2012

[8] M Alkhawlani and A Ayesh ldquoAccess network selection basedon fuzzy logic and genetic algorithmsrdquo Advances in ArtificialIntelligence vol 2008 Article ID 793058 12 pages 2008

[9] D D N Pon Kumar and K Murugesan ldquoPerformance analysisof AI based QoS scheduler for mobile WiMAXrdquo ICTACTJournal on Communication Technology vol 03 no 03 pp 572ndash579 2012

[10] X Yang and J Bigham ldquoA call admission control scheme usingneuroevolution algorithm in cellular networksrdquo in Proceedingsof the 20th International Joint Conference on Artificial Intelli-gence (IJCAI rsquo07) pp 186ndash191 Hyderabad India January 2007

[11] T Ovengalt K Djouani and A Kurien ldquoA fuzzy approach forcall admission control in LTE networksrdquo Procedia ComputerScience vol 32 pp 237ndash244 2014

[12] P Kejik and H Stanislav ldquoFuzzy logic based call admissioncontrol in UMTS systemrdquo in Proceedings of the 1st InternationalConference on Wireless Communication Vehicular TechnologyInformation Theory and Aerospace amp Electronic System Technol-ogy pp 375ndash378 Aalborg Denmark May 2009

[13] P Kejik and S Hanus ldquoComparison of fuzzy logic and geneticalgorithm based admission control strategies for UMTS sys-temrdquo Radioengineering vol 19 no 1 pp 6ndash10 2010

[14] C M Anecia and M S Pushpalatha ldquoFACM fuzzy basedresource admission control for multipath routing in MANETrdquoInternational Journal of Advanced Research in Computer Scienceamp Technology vol 2 no 1 pp 114ndash119 2014

[15] BNaeem RNgah S ZMHashimWMaqbool andM B AlildquoA neural network based approach for call admission control inheterogeneous networksrdquo Life Science Journal vol 11 no 11 pp238ndash242 2014


[16] C-J Chang L-C Kuo Y-S Chen and S Shen ldquoNeural fuzzycall admission and rate controller forWCDMAcellular systemsproviding multirate servicesrdquo in Proceedings of the InternationalWireless Communications and Mobile Computing Conference(IWCMC rsquo06) pp 383ndash388 Vancouver Canada July 2006

[17] R Pratyusha D Pratyusha A Konar and A K Nagar ldquoCalladmission control using artificial bee colony optimizationrdquoin Proceedings of the International Conference on Genetic andEvolutionary Methods (GEM rsquo12) pp 1ndash7 2012

[18] B Ramesh H S Gowrishankar and P S Satyanarayana ldquoAqos provisioning recurrentneural network based call admissioncontrol for beyond 3G networksrdquo International Journal ofComputer Science Issues vol 7 no 2 pp 7ndash15 2010

[19] GMahesh S Yeshwanth andUVManikantan ldquoSurvey on softcomputing based call admission control in wireless networksrdquoInternational Journal of Computer Science and InformationTechnologies vol 5 no 3 pp 3176ndash3180 2014

[20] R Gbegbe O Asseu K E Ali G L Diety and S HamoudaldquoCall admission control algorithm for energy saving in 5G H-cran networksrdquo Asian Journal of Applied Sciences vol 10 no 4pp 179ndash185 2017

[21] K Heena and M H Mirza ldquoSurvey of a new call admissioncontrol scheme based on pattern prediction for 5G mobilewireless cellular networkrdquo International Journal of InnovativeEngineering Research vol 7 no 1 pp 005ndash010 2017

[22] T Sigwele P Pillai A S Alam and Y F Hu ldquoFuzzy logic-basedcall admission control in 5G cloud radio access networks withpreemptionrdquo EURASIP Journal on Wireless Communicationsand Networking vol 2017 no 1 2017

[23] O Semenova A Semenov K Koval A Rudyk and V ChuhovldquoAccess fuzzy controller for CDMA networksrdquo in Proceedings ofthe International Siberian Conference on Control and Communi-cations (SIBCON rsquo13) pp 1-2 Krasnoyarsk Russia September2013

[24] O Semenova A Semenov and P Kulakov ldquoAccess Neuro-Fuzzy Controller for W-CDMA Networksrdquo in Proceedings ofthe 4th International Scientific-Practical Conference Problems ofInfocommunications Science and Technology (PIC SampT rsquo17) pp38ndash41 Kharkiv Ukraine October 2017

Computer Games Technology

International Journal of

Hindawiwwwhindawicom Volume 2018

Hindawiwwwhindawicom

Journal ofEngineeringVolume 2018

Advances in

FuzzySystems


Volume 2018


ReconfigurableComputing



Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

thinspArtificial Intelligence

Hindawiwwwhindawicom Volumethinsp2018


Civil EngineeringAdvances in


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia


Biomedical Imaging



Engineering Mathematics


RoboticsJournal of



Computational Intelligence and Neuroscience


Mathematical Problems in Engineering

Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018


Human-ComputerInteraction

Advances in


Scientic Programming

Submit your manuscripts atwwwhindawicom


three techniques That is because the combination overcomeslimitations of one technique and produces a relevant resultfor improving the quality of service [19] Using artificialintelligence in 5G networks has been investigated in [20ndash22]

The authors propose to apply in 5G mobile networksa genetic neurofuzzy controller with two input linguisticvariables and one output linguistic variable The controllermanages call acceptance or rejection thus allowing conges-tion avoidance The authors have already suggested the fuzzycontroller in [23] and the neurofuzzy controller in [24] Thispaper considers results of a further investigation

The objective of this paper is a genetic neurofuzzycontroller for call admission control in 5G mobile networks

The paper is organized as follows In the Materials andMethods the first subsection introduces linguistic variablesand membership functions for a fuzzy controller the secondsubsection describes structure of the controller and its oper-ation the third subsection considers genetic optimization ofthe controller The fourth subsection presents simulation inMatlab The last section provides conclusions

2 Materials and Methods

21 Fuzzy Controller Fuzzy systems simulate human reason-ing and are applied for characterizing behaviour of nonlinearsystems Fuzzy approach may be employed when there is nomathematical model of the process

Let us consider a mobile system formed by a number ofoperators sharing their Radio Access Networks in order toensure usersrsquo satisfaction Therefore when a user arrives tothe system and his call cannot be admitted it is transferred toanother operator to avoid the rejection

For developing a fuzzy controller its linguistic variablestheir terms and membership functions should be definedInput variables of the fuzzy controller describe all the possiblestates of a process controlled and its output variable describesall possible controlling actions Then a rule base should beassigned consisting of a set of IF-THEN rules to describe con-trolled states To evaluate operability of the fuzzy controller asimulation is performed

Input linguistic variables of the access fuzzy controller arethe effective capacity (E) and offered cell load (O) its outputvariable is the probability for call acceptation rejection ortransfer (P)The architecture of the fuzzy controller is shownin Figure 1

A fuzzy controller transforms input variables to member-ship function values evaluates a fuzzy output according to arule base and then converts the fuzzy output to a crisp one

The proposed fuzzy controller is to be converted into anadaptive neurofuzzy inference system (ANFIS) suitable foroperating under uncertain conditions For updating the rulebase a genetic algorithm is to be usedThe architecture of thegenetic neurofuzzy controller is shown in Figure 2

The linguistic variable of the effective capacity E is definedwith terms ldquolowrdquo ldquomediumrdquo and ldquohighrdquo

119879 (119864) = 119871119900119908 (119871) 119872119890119889119894119906119898 (119872) 119867119894119892ℎ (119867) (1)

Rule base

Fuzzification DefuzzificationInference EO

P

Figure 1 Architecture of the fuzzy controller

Neural Network

Fuzzification

Genetic Algorithm

Defuzzification

Rule Base

EO

P

Figure 2 Architecture of the genetic neurofuzzy controller

The linguistic variable of the offered load O is defined withterms ldquolowrdquo ldquomediumrdquo and ldquohighrdquo

119879 (119874) = 119871119900119908 (119871) 119872119890119889119894119906119898 (119872) 119867119894119892ℎ (119867) (2)

The linguistic variable of the call acceptation P is defined withterms ldquotransferrdquo ldquorejectrdquo and ldquoacceptrdquo

119879 (119875) = 119879119903119886119899119904119891119890119903 (119879) 119877119890119895119890119888119905 (119877) 119860119888119888119890119901119905 (119860) (3)

ldquoRejectrdquo means the call will be blocked ldquoAcceptrdquo means thecall will be admitted to the system ldquoTransferrdquo means the callwill be transferred to the other operator

Figures 3 and 4 present input variable membershipfunctions which are of trapezoid type and defined as

120583 (119909) =

0 119894119891 119909 le 1198861 119900119903 119909 ge 1198864(119909 minus 1198861)(1198862 minus 1198861) 119894119891 119909 isin (1198861 1198862] 1 119891 119909 isin (1198862 1198863) (1198864 minus 119909)(1198864 minus 1198863) 119894119891 119909 isin (1198863 1198864)

(4)

Figure 5 illustrates output variable membership functionswhich are singletons and defined as

120583 (119909) = 0 119894119891 119909 = 1198981 119894119891 119909 = 119898 (5)

The suggested fuzzy controller operates according to the rulebase consisting of nine rules Input and output values arespecified in Table 1

The fuzzy controller operates as followsEach input variable gets a corresponding fuzzy value

E 997888rarr ELEMEHO 997888rarr OLOMOH (6)


450 800

(E)

E

L HM

600200 350


M

4515 60

O

(O) HL

200 40



E OL M H




(7)


119875 = sum9119894=1 119908119894 sdot 119875119894sum9119894=1119908119894 (8)

R

0 100

P

(P) AT

minus100






If a = Am

and b = Bn

then c = Cpm = 1 2 3 n = 1 2 3 p = 1 2 3

(9)






Y3i = ni = wi(w1 + w2 + + w9) (12)


Y4i = nifi (13)


L

M

H

L

M

H

П

П

П

N

N

N

1

2

9

+P

f

O

E



E O PL 00 00 00 TM 01 01 01 RH 11 11 11 A


Y5 = n1c1 + n2c2 + + n9c9 (14)





0 0 0 0 0 1 (15)


1 1 1 1 0 1 (16)


119865 = 1120576 + 119877119872119878119864 (17)


119877119872119878119864 = radic 1119899119899sum119894=1

(119875119903119890119886119897 minus 119875119901119903119890119889)2 (18)




















3 Conclusions









Data Availability




References































Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in





450 800

(E)

E

L HM

600200 350


M

4515 60

O

(O) HL

200 40



E OL M H




(7)


119875 = sum9119894=1 119908119894 sdot 119875119894sum9119894=1119908119894 (8)

R

0 100

P

(P) AT

minus100






If a = Am

and b = Bn

then c = Cpm = 1 2 3 n = 1 2 3 p = 1 2 3

(9)






Y3i = ni = wi(w1 + w2 + + w9) (12)


Y4i = nifi (13)


L

M

H

L

M

H

П

П

П

N

N

N

1

2

9

+P

f

O

E



E O PL 00 00 00 TM 01 01 01 RH 11 11 11 A


Y5 = n1c1 + n2c2 + + n9c9 (14)





0 0 0 0 0 1 (15)


1 1 1 1 0 1 (16)


119865 = 1120576 + 119877119872119878119864 (17)


119877119872119878119864 = radic 1119899119899sum119894=1

(119875119903119890119886119897 minus 119875119901119903119890119889)2 (18)




















3 Conclusions









Data Availability




References































Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in





L

M

H

L

M

H

П

П

П

N

N

N

1

2

9

+P

f

O

E



E O PL 00 00 00 TM 01 01 01 RH 11 11 11 A


Y5 = n1c1 + n2c2 + + n9c9 (14)





0 0 0 0 0 1 (15)


1 1 1 1 0 1 (16)


119865 = 1120576 + 119877119872119878119864 (17)


119877119872119878119864 = radic 1119899119899sum119894=1

(119875119903119890119886119897 minus 119875119901119903119890119889)2 (18)




















3 Conclusions









Data Availability




References































Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in














3 Conclusions









Data Availability




References































Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in









Data Availability




References































Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in



















Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in









Advances in

FuzzySystems


Volume 2018













Journal of

Journal of



Hindawi


Advances in

Multimedia


Biomedical Imaging





RoboticsJournal of









Volume 2018



Advances in




Documents

ResearchArticle A Hybrid Approach to Call Admission ...downloads.hindawi.com/journals/afs/2018/2535127.pdf · ResearchArticle A Hybrid Approach to Call Admission Control in 5G Networks