Access Analysis of Secondary Users with an Opportunistic perspective in Cognitive Radio Networks

Transcription

1 International Journal of Comutational Intelligence Reearch IN Volume 3, Number 3 07, Reearch India ublication htt:// Acce Analyi of econdary Uer with an Oortunitic erective in Cognitive Radio Network Edwin Marco Kweigabo Reearch cholar, eartment of CE, Achary Nagarjuna Univerity, Guntur, India. K. Gangadhara Rao Aociate rofeor, eartment of CE, Acharya Nagarjuna Univerity, Guntur, India. B. Baavewar Rao 3 rogrammer, Univerity Comuter Centre, Acharya Nagarjuna Univerity, Guntur, India. K. Chandan 4 rofeor, eartment of tatitic, Acharya Nagarjuna Univerity, Guntur, India. Abtract Thi aer analye oortunitic acce of econdary uer in cognitive radio network CRN uing analytical M// queuing model. Through uing ectrum haring, econdary uer oortunitically hare rimary uer ectrum when it emty. Through ectrum haring rimary uer generate revenue and econdary uer maximize it atifaction with ectrum utilization. Number of econdary uer and their acket waiting time in the ytem i calculated. Numerical illutration for the ame i reented uing utility function uer bound and lower bound of rimary arrival wa determined. By uing uer bound and lower bound the otimal econdary

2 446 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan arrival i calculated. Reult indicate that otimal econdary arrival increae and decreae deending on reward and tate of rimary uer arrival. Keyword: Cognitive radio Network, econdary uer, riority baed, determinitic, reward, rofit INTROUCTION ectrum carcity i more growing nowaday due to invention of ortable mobile device like mart hone and Tablet, which have attracted more number of wirele network uer into the ectrum uage. ectrum i limited and carce reource a the available ectrum i already allocated to licened uer, which i not fully utilized in both time and location erective. ectrum allocation and uage haen to be the thrut area for the reearch, thi become more comlex when ectrum i aortioned for both rimary and econdary uer. Thi demand alication of oeration reearch technique ecifically queuing model. The major functionality of the tranceiver in the CRN i to intelligently detecting which communication channel are in ue and which are not, and intantly migrating the econdary uer into a vacant channel while avoiding interference with occuied rimary one. Thi otimize the ue of available reource while maintaining the quality of tranmitted acket. Quality of ervice Qo i an imortant characteritic in deigning the Cognitive Radio a it i chief criterion to elect the bet ectrum band. ectrum management i addreed uing four aect which include ectrum ening, ectrum deciion, ectrum haring and mobility. ectrum management reource are reerved uroely to erve licened uer, uing acce method baed on a fixed ectrum allocation known a Fixed ectrum Acce FA. Thee ectra allocated to licened uer are often underutilized in ace and time a only mall art of it i ued at a certain time intant and location alman A. AlQahtani and Haan Ahmed [5]. In fixed ectrum much of it band remain unued a licened ectrum are excluively owned oerate by their network all the time whether under ue it or not. The band remain idle when not ued by the licened holder on the band. The objective of the ectrum owner in CRN who erve both rimary and econdary uer i to maximize the revenue while erving the regular rimary Uer and attracting maximum number of econdary Uer. Through ectrum management, ectrum owner leae unued ectrum to unlicened wirele network U through oortunitic ectrum acce. Cognitive radio imrove ectrum utilization and econdary market uage of ectrum and imrove income of ectrum owner by triking the right balance between maximizing the ectrum

3 Acce Analyi of econdary Uer with an Oortunitic erective allocation to U and rotecting interet of the U alman A. AlQahtani and Haan Ahmed [5]. Thi aer mainly focue on maximizing the ectrum allocation for the U which in turn maximize the rofit. To arrive thi objective the CRN i conidered a M// queuing model [] where U and U are arrived with a oion roce and their arrival rate are and. The common determinitic ervice rate for both i µ. The objective of the aer i twofold. i to derive the utility function for finding the otimum number of U for given, utilization factor Ɵ, reward for the comletion of ervice of each U i α and oerational cot of U i β iito find the oortunitic for variou value of, along with uitable numerical illutration. The remaining aer i organized a follow. In ection II the review of the literature i reented. The aumed CRN model i deicted in ection III. The derived exreion for the utility function and otimum are reented in ection IV. The numerical illutration i given in ection V. Finally concluion are drawn in ection VI and alo given future coe of reearch. REVIEW OF THE LITERATURE everal tudie have been conducted on oortunitic ectrum acce in cognitive radio network, on how the U uer oortunitically acce free channel to tranmit their acket. However no tudy conducted on econdary uer otimization in the given uer and lower bound of rimary arrival. Iameldin uliman and Janne Lehtomaki [] Analyzed oortunitic ectrum acce in cognitive radio network. In their analyi rimary uer i conidered to have higher riority uer over econdary uer and it doe not care about the tatu of econdary uer in it tranmiion. Through uing time lotted ytem, econdary uer ene a free time lot and tart tranmiion and when it ene the reence of U, it remove it acket form the channel. Waiting time in a queue for U and U i determined. Average waiting time wa comared under different cenario to determine which ha minimum delay. However how ectrum acceed oortunitically in given bound i not hown. Oortunitic ectrum acce channel determined according to targeted Qo before making deciion on which channel to ue at the beginning of each lot on a channel. A CR uer deire to requet information regarding available channel characteritic from the bae tation baed on information ened on channel. The ectrum i elected according to the Qo requirement and the ectrum characteritic. ata rate and bandwidth are determined according to deciion rule and aroriate ectrum band choen Aylin Turhan, Murat Alanyali, and avid tarobinki [6].

4 448 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan rimary uer have higher riority ectrum acce and they don t care about the tate of U and interrution can occur at any time even a time U tart tranmiion. U leae a channel when it i free however it i reemted for a certain eriod on U arrival and reume when U finihe tranmiion. But during the channel handoff between U and U, the rimary Uer U may face interference for a certain time of delay on rejoining the channel Iameldin uliman and Janne Lehtomaki []. How the econdary uer otimize the channel in cae it ha almot to finih tranmiion, rimary uer interfere tranmiion, intead of tolerating it hort until U finih tranmiion. Narvada Khedun and Vandana Baoo [] modelled the ytem baed on the M// ytem with riority queuing. re-emtive and non re-emtive aroach to determine which ue minimum time. The M// wa extended to M// model, and then the erformance of M// non re-emtive and re-emtive riority queuing model wa acceed. Waiting time uing M// modal wa comared to waiting time of M// un Zhang, Tao Jiang, Lei Zhang, aiming Qu, and We eng [3] emloyed the reemtive reume riority R M/M/ queuing model to analyze the tranmiion delay of riority baed econdary uer U in the oortunitic ectrum acce OA baed cognitive radio network. The interruted acket tranmiion time wa reduced by aigning of interruted uer to higher riority than the newly arrived U. The exreion of tranmiion delay wa derived baing on the R M/M/ queuing modal. An analytical reult wa comared with imulation reult to determine U minimum delay. Mina Fahima and Abdoraoul Ghaemi [4] formulated R M/G/ model to invetigate delivery time of U in cognitive radio network. The rate of channel handoff wa invetigated uing re-emtive reume riority M/G/ model on multile handoff and data delivery time for U. tability region for arrival rate of econdary uer wa determined. alman A. AlQahtani and Haan Ahmed [5] rooed and invetigated efficient admiion with eviction control for CRN to imrove ectrum utilization efficiency. They conidered a tolerance waiting time for the U and enhanced eviction olice. They rooed cheme which rovide better balance between U Qo and rimary uer rotection. They comared call admiion control CAC and delayed rimary uer U to determine the imact on tranmiion and rofit maximization on admiion control. Aylin Turhan, Murat Alanyali, and avid tarobinki [6] determined otimal admiion control of econdary uer U in cognitive radio CR conidering a rimary uer arrive to the ytem and find all channel occuied. U uer i reemted excet if all channel are occuied by U. They have alied admiion

5 Acce Analyi of econdary Uer with an Oortunitic erective control on U to determine the number of admitted U. Through uing dynamic rogramming, otimal admiion control olicy maximized long run average rofit. They found otimal admiion control olice, and they comared otimal admiion olicy with total number of uer in the ytem and they determined the effect otimal admiion control olicy of U to the dynamic rogramming to determine which i more efficient. TEM MOEL The ytem model i adoted from M// reemtive riority model a uggeted [], wherein U will have higher riority with an ability to reemt U. The queuing model i reented in the following ection. M// Queuing Model Thi aer conider the Model a uggeted [] and oortunitic ectrum acce in which there i one rimary uer and one econdary uer on the channel. The econdary uer accee the channel at the beginning and tranmit acket when channel i free. It i aumed that the channel ha erfect ening noie and error free available for tranmiion. The ectrum allocation i baed on the FCF cheduling mechanim, giving the higher riority to U. Time ent by the acket in the ytem i defined uing M// riority queuing cheme, in which acket are tranmitted according to oion ditribution roce and ervice time i determinitic. The waiting time of each acket in the ytem ha three comonent, the time until the beginning of the next lot, time ent in a queue waiting for the ervice to begin followed by the average ervice time tranmiion time. Reward i collected from each ucceful U tranmiion, the ucceful tranmiion time of the U acket deend on the three factor the acket arrival time and the number of acket both rimary and econdary waiting in the ytem for tranmiion and the time each econdary acket end in the queue waiting for free channel i determined. By uing utility function otimum econdary arrival i determined in which oortunitic lower bound for rimary uer when otimum acce of econdary uer for given rimary uer earning and U oeration cot found and through uing uer bound for rimary uer when oortunitic acce for U given U reward and oeration cot. Aumtion The arrival of the acket follow the oion roce where arrival occur at the rate

6 450 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan Notation μ The ervice rate i determinitic contant for all arrival rimary and econdary uer, the channel i divided into fixed time lot of the equal ize, in which when a acket arrive it occuie one lot. The queue ervice diciline i FCF The ytem i aumed to have an infinite buffer ize Error free rimary uer arrival rate econdary uer arrival rate ervice time E[ ] Average ervice time N N Average number of acket in the queue of rimary uer Average acket in the queue for the econdary uer W Waiting time in a queue for rimary uer W Waiting time in a queue for econdary uer Q T Total time ent by the rimary uer in the ytem Total time ent by econdary uer in the ytem Waiting time till the beginning of a new lot Utilization factor for the rimary queue Utilization factor for econdary queue Reward collected on ucceful econdary Tranmiion Oerational Cot ertaining to the channel when the ame i offered to the U. Utilization i a quantified meaure of the channel uage over a eriod of time.

7 Acce Analyi of econdary Uer with an Oortunitic erective.. 45 In thi model the otimal number of econdary uer i calculated analytically baed on the M// model [] a follow. The ervice time of U and U are ame i.e. The waiting time for any new rimary uer, include T exiting rimary uer. W = T + N and tranmiion time of the W = T + W = T W ince N W T W Waiting time for U in queue i um of waiting time until the beginning of a lot, the time which rimary uer end in the queue, the time during which the exiting econdary uer tranmit their reective acket, and the ubequent arrived U ervice time. W T N N W 3 Aly little formula and ubtitute into 3 then waiting time of econdary uer in a queue i W T T W W W W W W 4 W T W 5 ubtitute into 5 then W T 6 Average delay time/acket of the rimary uer.

8 45 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan T 7 Average delay time/acket of the econdary uer. 8 T From little formula, the exected number of acket of the econdary uer in the ytem i given a N T W 9 Otimal number of econdary Uer In thi ection rofit to conider for U i ued to derive the otimum number of aid by U econdary uer on the channel. The utility function deend on reward for uing the ectrum and oeration cot for maintaining U acket in the ytem. N W W 0 N, i total reward charged on ucceful econdary tranmiion. W, i total oeration cot for acket waiting in the ytem. The rofit of ectrum owner on ucceful econdary uer tranmiion i total revenue minu oeration cot a hown below ubtitute 9 into 0 and alying little formula then rofit function a follow T Analyi by obtain otimum number of econdary uer i given in index I u T 3 By imlification to get the boundary equation T

9 Acce Analyi of econdary Uer with an Oortunitic erective From the above equation, only i varying and the remain are contant. o the bound for i a follow: The uer bound for i. 0 4 The lower bound of the atify the following condition. 0 The tability region for the rimary arrival THE NUMERICAL ILLUTRATION For fixed value of,, to tudy boundarie in IV and to calculate for variou Four cenario are being conidered for admitting econdary uer.. increaing. and increaing 3. increaing 4. decreaing The ervice rate i 4, Maximum bound reward i 5 and oeration cot i. ervice rate and oeration cot are contant. The lower and uer bound for rimary uer arrival a in the following range , 5

10 454 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan Value er cenario are a hown in table below increaing and increaing increaing decreaing Table Table Table3 Table4 cenario : increaing U Arrival increaing Otimal econdary arrival acket/econd rimary arrival U Figure : for given value,,, Td are 4, 5,, and reectively for variou value of,which i increaing

11 Acce Analyi of econdary Uer with an Oortunitic erective For given value of,,, T d a decreae, decreaing of free ectrum caue increae from lower bound, free ectrum looking for free ectrum to increae. Thi idea i the ame a decreaing free ectrum ace o a to increae ectrum demand. From economic theory of demand and uly when the uly decreae the rice increae and when uly increae rice decreae. cenario : and increaing Otimal econdary arrival U Arrival and Reward Increaing rimary Arrival U Figure : for given value,, Td are 4,, and reectively for variou and increaing Increaing decreae free ectrum. The decrement of ectrum force the market to increae, thee enable ectrum owner to get high rofit due to increaed for the ame ervice. However increaing reduce demand a the econdary uer are not willing to ay extra for the ame ervice. cenario 3: i increaing Otimal econdary Arrival Reward increaing Reward U Figure3: for given value,,, Td are 3.9, 4,, and reectively for variation i increaing

12 456 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan For given value,,, Td a increae from low to uer bound, decreae. Thi i the ame a the economic theory in which when rice i increaed roduct demand decreae. Low increae demand however ectrum owner get low rofit. High decreae demand a uer are not willing to ay extra for ame ervice but ectrum owner rofit increae a extra i charged for ame ervice. cenario 4: i decreaing Otimal econdary Arrival Reward decreaing Reward U Figure4: for given value,,, Td are 3.9, 4,, and reectively for variou decreaing For given,,, T d, a decreae from uer bound to lower bound the increae. From the economic theory when the rice decreae the demand increae. Increaing decreae the ectrum owner rofit. If few i ay high which enable ectrum owner to get uer rofit, a few ay more for ame and before and after increaing, until reache at oint of maximum atifaction where few afford rooed. Thi factor force ectrum owner to reduce o a to attract more. CONCLUION Thi aer ha determined which maximize ingle channel utilization through uing utility function. Number of acket admitted into the ytem and their waiting time wa calculated uing M// model in which ucceful tranmiion of acket, the ectrum owner get rewarded. Through utility function, uer and lower bound of in which the otimal number of econdary uer admitted i calculated. and have effect on admitted. Increaing with contant increae. Increaing

13 Acce Analyi of econdary Uer with an Oortunitic erective and decreae. Keeing contant and decreaing the reward increae.the tudy i imortant for maximizing the reward of rimary uer and alo the maximum utilization of ectrum. The future coe of thi work lie in determining the equilibrium oint where ectrum uly and demand get atifaction for both the ide. REFERENCE [] Iamalden. uliman,jane Lehtomaki, Queueing Analyi of Oortunitic Acce in Cognitive Radio, Centre for wirele communication CWC, IEEE comuter ociety, , 009 [] Narvada Khadun, Vandana Baoo, Analyi of riority Queuing with Multichannel in Cognitive Radio Network, IEEE Journal v, 8569-,05. [3] un Zhang, Tao Jiang, Lei Zhang, aiming Qu, Wei eng, Analyi on the Tranmiion elay of riority Baed econdary Uer in Cognitive Radio Network, IEEE Journal vol , 03. [4] Mina Fahami, Abdoraoul Ghaemi, Analyi of the R M/G/ Queuing ytem for Cognitive Radio Network with Handoff Management, IEEE Journal vol , 04. [5] alman A. AlQahtani, Haan Ahmed An Admiion Control cheme for econdary Uer in Cognitive Radio Network, IEEE Journal vol , 04. [6] Aylin Turhan, Murat Alanyali, avid tarobinki, Otimal Admiion Control of econdary Uer in reemtive Cognitive Radio Network, Boton, 0 [7] K.L. Muruganantha raad, B.Uh, A comarion between M/M/ and M// queuing model to vehicular traffic at Kanyakumari ditrict, IOR-JM vol 3 5 [8] hunfu Jin, uan Zhao, Wuyi ue, Zolt affer, erformance Analyi and Otimization of an Adative Admiion Control cheme in Cognitive Radio Network, Hindawi ublihing Cororation, Volume 03. [9] hung Li, Zizhan Zheng, Eylem Ekici, Ne hroff, Maximizing ytem Throughut by Cooerative ening in Cognitive Radio Network, [0] Rahul Urgaonkar, Michael J. Neely, Oortunitic cheduling with Releiability Gurantee in Cognitive Radio Network, IEEE INFOCOM, 008. [] Li Chun Wang, Chung Wei Wang, ectrum Handoff for Cognitive Radio Network: Reactive ening or roactive ening?, IEEE vol , 008.

14 458 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan [] Li Chun Wang, Chung Wei Wang, Chung Ju Chang, Modelling aand Analyi for ectrum Handoff in Cognitive Radio Network, IEEE Tranaction on mobile comuting, vol, , 0 [3] Li Chun Wang, Chung Wei Wang, Fumiyuki Adachi, Load Balancing ectrum eciion for Cognitive radio Network, IEEE elected area in communication, vol9 No. 4, , 0. [4] Li Chun Wang, Chung Wei Wang, Fumiyuki Adachi, Modelling and Analyi for reactive deciion ectrum Handoff in Cognitive Radio Network, IEEE Globecom, , 00. [5] Li Chun Wang, Chung Wei Wang, Fumiyuki Adachi, Modeling and Analyi for roactive deciion ectrum Handoff in Cognitive Radio network, IEEE, 09. [6] Ekendeke Ndiomo Maurice, Bin Guo, Reearch on Lifetime Etimation of econdary connection Uing the R M/G/ Queering Network Model, ICCA, 03 [7] Long Bao Le, Ekram Hoain, Reource Allocation for ectrum Underlay in in Cognitive Radio Network, IEEE Tranaction on wirele communication, vol7, 76 84, 008. [8] in Lin Huang, iang Wu, Jun Wu, uan u, Min Wang, Queuing Theory Baed ectrum Allocation in Cognitive Radio Network, International Conference on Wirele Communication, , 04. [9] aid Lakhal, Abdellah Idrii, Queue Management of econdary Uer in a Cognitive Radio Network, Fifth International on NGN,04. [0] iego acheco aramo, Vicent la and Jorge Martinez Bauet, Otimal Admiion Control in Cognitive Radio Network, 00. [] Ahmed M. Alaa, table Throughut Region of Cognitive Relay Network with Imerfect ening and Finate Relaying Buffer, conference aer 04. [] Gaofei un, iaohua Tian, ouyun u, inbing Wang, rofit maximization for econdary uer in dynamic ectrum auction of cognitive radio network, Journal of wirele communication and mobile comuting, 33 34, 05.

15 Acce Analyi of econdary Uer with an Oortunitic erective Index. For eay imlification we let and Exand the term To determine otimal rofit gain on econdary admiion we differentiate with reect to Aly different ion by art From V V U U V UV After imlification 0 6 econdary Arrival U derivation olve for uing quadratic formula and U arrival rate mut be oitive and negative value i neglected.

16 460 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan From a ac b b 4 7 c b a ubtitute in the above formula ubtitute the value of and in the equation above u Index econd derivative roof 8

17 Acce Analyi of econdary Uer with an Oortunitic erective.. 46 Uing by art differentiation, we get 4 9 When we imlify the above equation we get 3 30 ubtitute the value of and in the above equation, we get equation 3.

18 46 K. Gangadhara Rao, Edwin Marco Kweigabo, B. Baavewar Rao & K. Chandan