Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium QoE Driven Video Sreaming in Cogniive Radio Neworks: The Case of Single Channel Access Zhifeng He, Shiwen Mao Deparmen of Elecrical and Compuer Engineering Auburn Universiy Auburn, AL 36849-5201 USA Email: zzh0008@igermail.auburn.edu, smao@ieee.org Sasry Kompella Informaion Technology Division Naval Research Laboraory Washingon, DC 20375 USA Email: sasry.kompella@nrl.navy.mil Absrac We consider he problem of sreaming muli-user videos over he downlink of a Cogniive Radio Nework (CRN, where each Cogniive User (CU can access one channel a a ime. Moivaed by he prior work ha esablishes a separaion principle for he join design of specrum sensor, sensing, and access polices, we firs model cooperaive specrum sensing as an Ineger Programming problem (IP and develop a Greedy Polymaching scheme o solve i for he opimal sensing sraegies. We hen formulae he problem of CU Qualiy of Experience (QoE maximizaion as a maximum weigh maching problem and solve i wih he Hungarian Mehod for opimal channel assignmens. The proposed specrum sensing and channel assignmen algorihms are compared wih benchmark schemes in simulaions, and are found o ouperform he benchmark schemes in erms of available channels discovered and CU QoE achieved. Index Terms Qualiy of Experience; cogniive radio neworks; maching; muli-user video sreaming; opimizaion. I. INTRODUCTION A recen sudy by Cisco reveals a drasic increase in mobile daa and ha almos 66% of he mobile daa will be video relaed by 2015 [1]. Such dramaic increase in wireless video raffic, coupled wih he depleing specrum resource, poses grea challenges o oday s wireless neworks. I is of grea imporance o improve he wireless nework capaciy by promoing more efficien use of specrum, which can be accomplished by he cogniive radio (CR echnology. CR is an evoluionary echnology for more efficien and flexible access o he radio specrum. In a cogniive radio nework (CRN, Cogniive Users (CUs search for he unoccupied licensed specrum of he Primary User (PU nework and hen opporunisically access deeced specrum holes in an unobrusive manner. CR has been recognized as an effecive approach o suppor bandwidh-demanding mobile services such as wireless video sreaming [2], [3]. In he area of mulimedia communicaions, subjecive assessmen mehods have been sudied inensively [4], which is shown o reflec viewers percepual qualiy more accuraely han he radiional objecive assessmen mehod. The Inernaional Telecommunicaion Union (ITU has proposed sandards on subjecive assessmen mehods for various applicaion scenarios [5]. For video ransmission, qualiy of experience (QoE is an effecive subjecive qualiy assessmen model for he percepual visual qualiy of video sequences. One of he mos widely used QoE meric is he Mean Opinion Score (MOS [6]. In he MOS model, he visual qualiy of a video sequence is no only dependen on he nework environmen such as packe loss rae, nework delay, bu also dependen on he conen ype. For example, under he same nework condiions, he visual qualiy of video conens of fas moions (e.g., spors is generally worse han ha of video conens of slow moions (e.g., news. Since he ulimae goal of mos mulimedia communicaion services is o achieve high percepual qualiy for viewers, i is desirable o incorporae QoE models in such applicaions. In his paper, we address he challenging problem of downlink muli-user video sreaming in CRNs. We consider a CRN consising of one cogniive base saion (CBS and muliple CUs. Wihou loss of generaliy, we assume each CU can sense and access one channel a a ime. The CUs cooperaively sense he PU signals on licensed channels and he CBS infers he licensed channel saes based on he CU sensing resuls wih an OR fusion rule. Once he idle channels are deeced, he CBS hen assigns hem o acive CUs for downlink muli-user video sreaming. We incorporae he video assessmen model proposed in [6], [7], aiming o maximize he CU QoE by opimal designs of specrum sensing and access policies. I is obviously a challenging problem o joinly design he specrum sensing and access polices for QoE-aware muli-user video sreaming, due o he large number of design facors and he complex ineracions ha should be modeled in a crosslayer opimizaion framework. In [8], a separaion principle is esablished for he join design problem, which decouples he design of sensing from ha of sensor and access policy. Moivaed by his ineresing work, we decouple he problem ino wo sub-problems: (i discovering a sufficien amoun of licensed channels reliably and quickly o mee he bandwidh demand of he CUs; and (ii allocaion of he available channels o he CUs according o heir respecive QoE requiremens and nework saus. We propose a disribued Greedy Polymaching algorihm ha can compue opimal soluions o he channel sensing sub-problem, as well as a Hungarian mehodbased approach o achieve opimal soluions o he channel assignmen sub-problem. Simulaion resuls demonsrae he superior performance of he proposed mehods in erms of he MOS ha CUs can achieve under various nework scenarios, 978-1-4799-3512-3/14/$31.00 2014 IEEE 1388
Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium wih comparisons o he benchmark schemes. The remainder of his paper is organized as follows. Secion II reviews relaed work. The sysem model and problem formulaion are presened in Secion III. The opimal soluion algorihms are proposed in Secion IV and validaed by simulaions in Secion V. Secion VI concludes he paper. II. RELATED WORK This paper is relaed o he prior work on Qualiy of Service (QoS and QoE provisioning and video sreaming over CRNs. We briefly review he relaed work in he following. In [4], he qualiy of mulimedia and audio-visual services is evaluaed using a model ha is based on subjecive opinions, such as aesheic feeling and aciviy feeling, and nework parameers such as delay, packe loss rae, and frame rae. However, conen ype of he mulimedia is no considered in his work. The video qualiy models proposed in [9] and [10] consider encoder based disorions only. The auhors in [6] cluser video sequences ino differen conen ypes according o he feaures of hese sequences such as emporal (movemen and spaial (edges, brighness characerisics by cluser analysis, and hen a percepual video qualiy meric in erms of MOS is derived based on he conen ype of he video sequence and some nework parameers such as packe error rae, Sender BiRae (SBR, and frame rae. In [7], he auhors develop a MOS predicion model as a funcion of he conen ype of he esed video sequence and some oher parameers such as SBR, frame rae, and Block Error Rae, for video sequences sreamed over UMTS (Universal Mobile Telecommunicaions Sysem nework. As in [6], a conen ype classificaion framework is used in his paper. Afer all hese, he auhor proposes an QoE Driven adapaion scheme which, for a given conen ype of esed video sequence, aking he disorion of he video sequence a he receiver side and he channel loss rae as he inpu, dynamically adjuss he SBR, which is he oupu, in order o mainain a cerain level of MOS under changing nework environmen. However, he adapaion scheme may no be adequae for video sreaming in CRN where channel availabiliy is uncerain. The problem of video sreaming over CRNs has also been sudied in a few prior works. The ransmission of mulimedia over CRN is firs proposed by J. Miola in [11]. Auhors of [12] develop an aucion game model o deliver conen-aware mulimedia. In [3] he qualiy opimizaion problem is formulaed as an mixed ineger nonlinear programming (MINLP problem and solved wih effecive algorihms. The auhors in [13] consider he scenario where mulimedia ransmission is scheduled in CRN and a QoE Driven channel allocaion scheme is proposed o opimize he mulimedia ransmission of prioriy-based CUs, where he MOS model proposed in [7] is used. Specifically, each CU has differen QoE requiremens and hus has differen prioriy in uilizing he idle channels of he PU sysem. Upon he re-appearance of an acive PU on he idle licensed channel, each CU uilizing he idle licensed channels will evacuae from he curren channel i is using o avoid conflic wih he acive PU. Alhough some ineresing works have been done on video sreaming over CRNs, an imporan par of he challenging problem is how o discover a sufficien number of idle channels reliably and quickly, which is no well addressed in prior work bu is essenial for supporing he demand of large bandwidh for video applicaions. A holisic approach is necessary for supporing QoE-aware video sreaming in CRNs, which inegraes effecive soluions o boh he specrum sensing problem and resource allocaion problem for video sessions, considering heir differen characerisics relaed o MOS. We aim o develop such an approach in his paper. III. SYSTEM MODEL AND PROBLEM STATEMENT A. Sysem Model We consider a primary nework operaing on N 1 orhogonal licensed channels. There is a CR nework co-locaed wih he primary nework, consising of a CBS supporing M 1 CUs. The CUs sense he PUs usage of he licensed channels and access he licensed channels in an opporunisic manner. As in priori work [14], [15], we assume he CUs, when hey are no receiving daa, measure he SNRs of he PU ransmissions over all he licensed channels and repor he measured SNRs o he CBS hrough some feedback mechanism. Based on such feedback, he CBS hen assigns hose CUs wih good channel condiions o sense each licensed channel, so as o improve he sensing performance. We consider he downlink muli-user video sreaming scenario, where he CBS sreams a video o each acive CU using he license channels ha are deeced idle. We assume ime is divided ino a series of non-overlap GOP (Group of Picures windows, each consising of T ime slos. Each ime slo can be furher divided ino he following four phases as in Algorihm 1 for specrum sensing and access for muli-user video sreaming. Algorihm 1: Specrum Sensing and Access for QoEdriven Muli-user Video Sreaming 1 Phase 1: The CBS deermines for each CU which channel o sense based on SNR feedback, and broadcass he sensing schedule o he CUs ; 2 Phase 2: Each CU follows he sensing schedule o sense he channel, and hen repors he sensing resul o he CBS ; 3 Phase 3: The CBS makes wo decisions: (i he channel availabiliy on he curren ime slo, based on he sensing resuls and fusion rule; and (ii channel assignmen o he CUs for muli-user video ransmission on he curren ime slo, based on channel availabiliy, channel condiion, he Conen Type (denoed as CT of each CU, and oher informaion. Then he CBS broadcass he channel access schedule o he CUs ; 4 Phase 4: The CBS uses he assigned channels o ransmi video daa and each CU follows he channel access schedule o receive video daa. 1389
Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium Noe ha a he very beginning of he firs GOP window, he SNR informaion used in Phase 1 may no be available. However, such informaion can be obained via esimaion or learning echniques, or by leing CUs probing he channels when CUs are idle [16]. B. Problem Saemen 1 Opimal Assignmen Problem for Specrum Sensing (OAPSS : In a pracical wireless nework scenario, CUs are locaed a differen geographical posiions wih differen performance on deecing he primary signal on a paricular licensed channel [14]. I has been proved ha by selecing a suiable subse of CUs o sense a paricular channel can improve he performance on deecion of PU signal, hus reducing he probabiliy of causing inerference o he PU ransmissions [15], [17]. A useful meric o evaluae he performance of deecing a PU signal is probabiliy of miss deecion, which is he probabiliy ha a CU fails o deec he exisence of an exising PU signal. Usually cooperaive sensing is used o improve he deecion performance by fusing he sensing resuls from muliple CUs, and a cerain fusion rule is required o combine hese resuls. In his paper he OR fusion rule is used a he CBS o decide he presence or absence of he PU signal on a paricular channel. Wih he OR rule, if any of he CUs repors he presence of a PU signal hen he CBS decides ha he channel is busy; Oherwise, he CBS decides ha he channel is idle. Here we use an M 1 N 1 marix X o denoe he sensing ask assignmen a ime slo, while he enry locaed in he i-h row and j-h column is defined as { 1, CU i senses channel j in ime slo x = (1 0, oherwise. Le P denoe he probabiliy of deecion on channel j by d CU i a ime slo. According o he OR fusion rule, he probabiliy of deecion on channel j a ime slo is M 1 ( x Pd j =1 1 Pd. (2 For an energy deecor, we have [17] ( (λ Pd = 1 2 erfc σn 2 ς 1 K 2 ( 2ς +1, (3 where λ is he hreshold of energy deecion on channel j by CU i a ime slo, σn 2 is he power of he i.i.d. Addiive Whie Gaussian Noise (AWGN a he CU, ς is he SNR of PU s signal on channel j a CU i, K is he number of samples on channel j by energy deecion. In (3, erfc(z = 2 π e u2 z du is he complemenary error funcion, and le erfc 1 ( denoe he inverse funcion of erfc (. Inordero guaranee he proecion of he PUs, we se Pd = P d by uning λ for all i, j. Thus he probabiliy of deecion of he aciviy of a PU will be greaer han P d if he channel is sensed by some CUs. In he case ha a channel is no sensed by any of he CUs, i will no be used for video sreaming. Under he assumpions ha he PU signal is complex valued phase-shif keying (PSK and he noise is circularly symmeric complex Gaussian (CSCG, hen CU i s probabiliy of false alarm on channel j, denoed by P, can be expressed as [17] f P f = 1 2 erfc ( 2ς +1erfc 1 ( 2 P d + K 2 ς. (4 The objecive of sensing ask assignmen is o maximize he probabiliy of deecing all he idle channels a ime slo, while mainaining fairness among he probabiliies of deecion of he N 1 channels. I has been shown ha proporional fairness can be achieved by maximizing he sum of logarihmic funcions. The opimal sensing ask assignmen problem is o maximize he following objecive funcion. log (1 P fj j=1 N 1 = M 1 log N 1 M 1 = log (1 P f x (1 P f x M 1 = ϕ x, (5 where Pf j is he probabiliy of false alarm on channel j and ϕ. =log (1 P f. We assume ha each CU can sense one channel a each ime slo, and he number of CUs ha can be assigned o sense a channel i a each ime slo is unresriced. Therefore, he opimal sensing ask assignmen problem is formulaed as max : s.. M 1 j=1 x ϕ x (6 x =1, for all i (7 {0, 1}, for all i, j. (8 2 Opimal Assignmen Problem for Video Transmission (OAPVT : We consider he QoE model named Mean Score Opinion (MOS proposed in [7]. The MOS of CU i during ime slo, denoed by Ψ, can be expressed as Ψ = α + CT i γ +(β + CT i δln ( SBR (9 = α + CT i γ +(β + CT i δln ( ( B j log 2 1+SNR, where α = 3.9860, β = 0.0919, γ = 5.8497, andδ = 0.9844 are consans, CT i is he Conen Type of he video sequences required by CU i, B j is he bandwidh of channel j in kbps, and SNR is he SNR of he video signal using channel j measured a CU i a ime slo [7]. We assume ha N 2 channels are deermined o be idle afer he sensing phase, where N 2 N 1. We consider a general 1390
Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium case where no all he CUs have daa o receive a all imes. Insead, he probabiliy of a CU has daa o receive a each GOP window is 0 ξ 1. The number of CUs ha have daa o receive in a GOP window (called acive CUs is denoed as M 2,whereM 2 M 1.AnM 2 N 2 marix Y is used o represen channel access assignmen on ime slo, while he enry locaed in he i-h row and j-h column is { 1, assign channel j o CU i in ime slo y = 0, oherwise. (10 We consider he case where each CU can use a mos one channel a each ime slo due o hardware consrains, and each channel can be used by a mos one CU a each ime slo. We aim o maximize he expeced average MOS of all he CUs during a GOP window by assigning he available channels. max : [ 1 E T T Ψ i =1 ] = 1 T T E [ Ψ i]. =1 The above objecive funcion can be maximized if we maximize he expeced MOS incremen of he M 2 CUs during each ime slo [3], which can be wrien as E [ Ψ ] i M 2 N 2 = E [ Ψ ] y j=1 N 2 [ ( = Pr H 0j s j =1 φ +Pr( H1j s j =1 θ ] y, j=1 where s j =1indicaes he channel is sensed idle; Pr(H 0j and Pr(H1j are he probabiliy of channel j o be idle or busy a ime slo, respecively; Pr(H0j s j =1and Pr(H 1j s j =1 are he condiional probabiliy for channel j o be idle or busy condiioned on he sensing resul, respecively; μ and ν are he received SNR a CU i using channel j which is indeed idle or busy a ime slo, respecively; and Pr ( H 0j s j =1 = (1 P fj Pr(H0j (1 P dj Pr(H1j+ (1 P fj Pr(H0j Pr ( H 1j s j =1 =1 Pr ( H 0j s j =1 φ = α + CT iγ +(β + CT i δln ( B j log 2 ( 1+μ θ = α + CT iγ +(β + CT i δln ( B j log 2 ( 1+ν, Define w as w =Pr( H 0j s j =1 φ +Pr( H 1j s j =1 θ. (11 Algorihm 2: Greedy Poly-Maching Algorihm 1 for i =1 M 1 do 2 for j =1 N 1 do 3 x =0; 4 end 5 j =argmax j {1,,N1} {ϕ } ; 6 x =1; 7 end The opimal channel access problem is formulaed as max : s.. N 2 ω y (12 j=1 N 2 j=1 y 1, i {1,,M 2 }. (13 y 1, j {1,,N 2} (14 y {0, 1}, for all i, j. (15 IV. SOLUTION ALGORITHMS AND ANALYSIS A. Poly-maching Based Soluion o OAPSS As can be seen from he above, afer suiable subsiuions, boh he OAPSS and OAPVT problems are formulaed as he well-known General Assignmen Problem (GAP, in he form of Ineger Programming (IP problems, which has been proved o be NP-hard o solve. However, we find ha here OAPSS is a special case of he GAP. Since here is no consrain on how many CUs can be assigned o a channel, he problem is acually a Maximum Weigh Poly-Maching (MWPM problem on a biparie graph ha maches CUs o licensed channels wih edge weighs defined as ϕ, while a channel can be mached o muliple CUs [18]. I can be solved by he following greedy sraegy in Algorihm 2 [18]. This greedy sraegy has a ime complexiy of O(M 1 N 1.In fac his is a disribued algorihm, since each CU can choose is bes channels o sense and here is no need o involve he CBS in his phase. Since he CUs can iniialize heir x s and launch heir searching procedures in Line 5 in parallel, his disribued sraegy has a ime complexiy of O(N 1.Wealso show ha he Greedy Poly-maching Algorihm is opimal. Theorem 1. The Greedy Poly-maching Algorihm 2 achieves he opimal soluion o he OAPSS problem. Proof: Exchanging he summaion order, he objecive funcion of he OAPSS problem (6 becomes ( N1 M1 j=1 ϕ x,where N 1 ϕ x is he uiliy ha CU i can achieve under he wo consrains. Since each CU can have a mos { one channel, he maximum uiliy CU i can achieve is max j ϕ x}, which is accomplished by in Line 5 of Algorihm 2. Since he opimal sraegies of he CUs do 1391
Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium Parameers TABLE I SIMULATION PARAMETERS Value M 1 30 N 1 30 K 10 4 f s 10 6 H z T 100 B j 10 6 H z μ 25dB o 15dB ν 80dB o 60dB ς 30dB o 10dB P d ( } 0.95 max j {P H0j 0.9 # of channels 35 30 25 20 15 10 5 # of channels sensed idle (proposed # of channels sensed idle (random # of channels acually idle # of channels missed deecion (proposed # of channels missed deecion (random no conflic wih each oher and hus are independen wih oher, he maximum uiliy of he CUs are also independen of each oher. I follows ha max ( M N1 1 j=1 ϕ x = M1 ( { maxj ϕ x}, and Algorihm 2 is opimal. B. Hungarian Mehod Based Soluion o OAPVT In he OAPVT problem, since each CU can use a mos one channel (see (13 and each channel can be used by a mos one CU (see (14, i can be seen ha he OAPVT problem becomes a maximum weigh maching problem on a biparie graph ha maches acive CUs o available channels, while only one edge is allowed for any CU and channel and he edge weighs are defined as ω. This maximum weigh maching problem can be effecively solved in polynomial ime using he Hungarian mehod, and he soluion is opimal. In our case, he ime complexiy of using Hungarian mehod o solve he OAPVT problem is O((M 2 +N 2 (M 2 N 2, where M 2 +N 2 is he oal number of verices and M 2 N 2 is he oal number of possible edges in he biparie graph represening he OAPVT problem. V. SIMULATION RESULTS In his secion, he performance of he proposed algorihms is validaed wih Malab simulaions. We consider a scenario in which he PUs and CUs are randomly disribued around a CBS wihin he service radius of he CBS. Table II liss he values of he parameers used in he simulaions, where f s is he sampling frequency a he CUs wih energy deecion. The QoE video model relaed parameers are from [7]. We compare he proposed scheme wih a benchmark scheme presened in [18], called Daa Rae (DR Driven scheme in he simulaions, in which channels are assigned o end users o maximize he overall daa rae of all he end users. The proposed channel sensing algorihm Algorihm 2 is also inegraed wih he DR Driven scheme for a fair comparison. The effeciveness of he proposed sensing algorihm is presened in Fig. 1. { We ( increase } he minimum channel idle probabiliy min i,j Pr H 0j from 0.1 o 0.47 and plo he number of channels ha are really idle and he number of channels ha are sensed idle. Here we compare wih he random sensing scheme as in [19], where each CU randomly 0 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 min P(H j 0j Fig. 1. Performance of channel sensing vs. he minimum channel idle probabiliy (95% confidence inervals ploed as error bars. and independenly selecs one of he N 1 channels o sense wih equal probabiliy. As uilizaion of he channels decreases, he number of idle channels increases. The proposed sensing algorihm can discover more idle channels for CUs o use. Moreover, he number of channels ha miss deecion is less han 0.5 on average, which is less han N 1 (1 P =1.5. d j Recall ha N 1 is he oal number of channels, P is he d j probabiliy of deecion, so N 1 (1 P is he expeced d j number of channels missed deeced. So he sensing algorihm offers an accepable level of proecion o he PUs, and is effecive in deecing idle channels for he CUs. We nex compare he expeced MOS of all he CUs a each ime slo (denoed as Ψ during an enire GOP window. In Fig. 2, we plo he achieved MOS sum of all he CUs achieved by he { proposed ( } scheme and he DR Driven scheme. We se min j Pr H 0j = 0.5 and raffic load ξ = 1 in his simulaion. Fig. 2 shows ha he proposed QoE-aware scheme achieves a consisenly high QoE sum han DR Driven does during he enire GOP window. This is because any soluion derived by solving he daa rae maximizaion problem in [18] yields a lower bound on he opimal objecive value of he OAPVT problem, which aims a achieving he maximum expeced average MOS by channel assignmen. Fig. 3 demonsraes how he CU video qualiy is affeced by he raffic load of he CUs (i.e., ξ. Le avg Ψ denoe he average MOS sum of he CUs during an enire GOP window. The achieved average MOS sums by he proposed scheme and he DR Driven scheme are ploed, where 95% confidence inervals are ploed as error bars. As he CU raffic load is increased, more CUs need channels for video ransmission. We can see ha while he number of he really idle channels is greaer han he number of acive CUs, he average MOS sum of boh schemes increases wih ξ, and he performance gap beween he wo schemes grows larger. While he number of really idle channels is no greaer han he number of acive CUs, he average MOS sum of boh schemes remain he same, 1392
Globecom 2014 - Communicaions Sofware, Services and Mulimedia Symposium 70 65 60 QoE driven DR driven Cener (BWAC sie a Auburn Universiy. Any opinions, findings, and conclusions or recommendaions expressed in his maerial are hose of he auhor(s and do no necessarily reflec he views of he NSF. Ψ 55 50 45 40 35 30 0 10 20 30 40 50 60 70 80 90 100 Time slos in a GOP window Fig. 2. avg_ψ 52 50 48 46 44 42 40 38 36 34 32 30 MOS sum of all he CUs over ime in an enire GOP window. QoE driven DR driven 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 ξ Fig. 3. Average MOS sum of he CUs over an enire GOP window, avg Ψ, for differen CU raffic loads ξ (wih 95% confidence inervals. since no more channel resource is available o saisfy he need of he exra CUs. VI. CONCLUSION In his paper, we invesigaed he problem of QoE-aware video sreaming over CRNs. An IP problem on specrum sensing was formulaed and solved wih a Greedy Polymaching Algorihm. Then a channel assignmen problem was formulaed as an IP and solved wih he Hungarian Mehod o derive he opimal channel assignmen, where QoE is used as he performance meric. We showed ha boh proposed algorihms achieve opimal soluions for specrum sensing and channel access, respecively. The proposed schemes were validaed wih simulaions. REFERENCES [1] Cisco, Visual Neworking Index (VNI, Feb. 2014. [Online]. Available: hp://www.cisco.com/. [2] D. Hu, S. Mao, Y. T. Hou, and J. H. Reed, Scalable video mulicas in cogniive radio neworks, IEEE J. Sel. Areas Commun., Special Issue on Wireless Video Transmission, vol.29, no.3, pp.334 344, Apr. 2010. [3] D. Hu, and S. Mao, Sreaming scalable videos over muli-hop cogniive radio neworks, IEEE Trans. Wireless. Commun., vol.11, no.9, pp.3501 3511, Nov. 2011. [4] K. Yamagishi and T. Hayashi, Opinion model using psychological facors for ineracive mulimodal services, IEICE Trans. Communicaion., E89-B(2:281 288, Feb. 2006. [5] J. You, U. Reier, M. Hannuksela, M, Gabbouj, and A. Perkis, Percepual-based qualiy assessmen for audio-visual services: A survey, Signal Processing: Image Communicaion., vol.25, no.7, pp.482 501, Aug. 2010. [6] A. Khan, L. Sun, and E. Ifeachor, Conen clusering based video qualiy predicion model for MPEG4 video sreaming over wireless neworks, in Proc. IEEE ICC 09., Dresden, Germany, June 2009, pp.1 5. [7] A. Khan, L. Sun, and E. Ifeachor, QoE predicion model and is applicaion in video qualiy adapaion over UMTS neworks, IEEE Trans. Mulimedia, vol.14, no.2, pp.431 442, Apr. 2012. [8] Y. Chen, Q. Zhao, and A. Swami, Join design and separaion principle for opporunisic specrum access in he presence of sensing errors, IEEE Trans. Inf. Theory, vol.54, no.5, pp.2053 2071, May 2008. [9] A. Eden, No-reference image qualiy analysis for compressed video sequences, IEEE Trans. Broadcas., vol.54, no.3, pp.691 697, Sep. 2008. [10] Q. Huynh-Thu and M. Ghanbari, Temporal aspec of perceived qualiy in mobile video broadcasing, IEEE Trans. Broadcas., vol.54, no.3, pp.641 651, Sep. 2008. [11] J. Miola III, Cogniive radio for flexible mobile mulimedia communicaions, in Mobile New. Appl. J. (MONET, vol.6, no.5, pp.435 441, Sep. 2001. [12] Y. Chen, Y. Wu, B. Wang, and K. J. R. Liu, Specrum aucion games for mulimedia sreaming over cogniive radio neworks, IEEE Trans. Commun., vol.58, no.8, pp.2381 2390, Aug. 2010. [13] T. Jiang, H. Wang, and A. Vasilakos, QoE-Driven channel allocaion schemes for mulimedia ransmission of prioriy-based secondary users over cogniive radio neworks, IEEE J. Sel. Areas Commun., vol.30, no.7, pp.1215 1224, Aug. 2012. [14] W. Wang, B. Kasiri, J. Cai, and A.S. Alfa. Channel Assignmen of Cooperaive Specrum Sensing in Muli-Channel Cogniive Radio Neworks, Proc. IEEE ICC 11, kyoo, Japan, June. 2011, pp.1 5. [15] W. Wang, B. Kasiri, J. Cai, and A.S. Alfa. Disribued cooperaive muli-channel specrum sensing based on dynamic coaliional game. Proc. IEEE GLOBECOM 10, Miami, FL, Dec. 2010, pp.1 5. [16] Alexander W. Min, Kang G. Shin, Join Opimal Sensor Selecion and Scheduling in Dynamic Specrum Access Neworks, IEEE Trans. Mobile Compuing, vol.12, no.8, pp.1532 1545, Aug. 2013. [17] B. Wang, K.J. Ray Liu, and T.Charles. Clancy. Evoluionary cooperaive specrum sensing game: How o collaborae? IEEE Trans. Commun., vol.58, no,3, pp.890 900, Mar. 2010. [18] K. Kar, L. Xiang, and S. Sarkar, Throughpu-opimal scheduling in mulichannel access poin neworks under infrequen channel measuremens, IEEE Trans. Wireless. Commun., vol.7, no.7, pp.2619 2629, July 2008. [19] H. Su, and X. Zhang, Cross-layer based opporunisic MAC proocols for QoS provisionings over cogniive radio wireless neworks, IEEE J. Sel. Areas Commun., vol.26, no.1, pp.118 129, Jan. 2008. ACKNOWLEDGMENT This work is suppored in par by he US Naional Science Foundaion (NSF under Gran CNS-0953513, and hrough he NSF I/UCRC Broadband Wireless Access & Applicaions 1393