TODAY S wireless networks are characterized as a static

Size: px
Start display at page:

Download "TODAY S wireless networks are characterized as a static"

Transcription

1 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY A Spectrum Decson Framework for Cogntve Rado Networks Won-Yeol Lee, Student Member, IEEE, and Ian F. Akyldz, Fellow, IEEE Abstract Cogntve rado networks have been proposed as a soluton to both spectrum neffcency and spectrum scarcty problems. However, they face to a unque challenge based on the fluctuatng nature of heterogeneous spectrum bands as well as the dverse servce requrements of varous applcatons. In ths paper, a spectrum decson framework s proposed to determne a set of spectrum bands by consderng the applcaton requrements as well as the dynamc nature of spectrum bands. To ths end, frst, each spectrum s characterzed by jontly consderng prmary user actvty and spectrum sensng operatons. Based on ths, a mnmum varancebased spectrum decson s proposed for real-tme applcatons, whch mnmzes the capacty varance of the decded spectrum bands subject to the capacty constrants. For best-effort applcatons, a maxmum capacty-based spectrum decson s proposed where spectrum bands are decded to maxmze the total network capacty. Moreover, a dynamc resource management scheme s developed to coordnate the spectrum decson adaptvely dependent on the tme-varyng cogntve rado network capacty. Smulaton results show that the proposed methods provde effcent bandwdth utlzaton whle satsfyng servce requrements. Index Terms Cogntve rado networks, spectrum decson, spectrum characterzaton, real-tme applcaton, best-effort applcaton, mnmum varance-based spectrum decson, maxmum capacty-based spectrum decson, resource management. Ç 1 INTRODUCTION TODAY S wreless networks are characterzed as a statc spectrum assgnment polcy. Recently, because of the ncrease n spectrum demand, ths polcy s faced wth spectrum scarcty at partcular spectrum bands. On the contrary, a large porton of the assgned spectrum s stll used sporadcally leadng to underutlzaton of the sgnfcant amount of spectrum [9]. Hence, dynamc spectrum access technques have recently been proposed to solve these spectrum neffcency problems. The key enablng technology for dynamc spectrum access technques s the cogntve rado technology, whch provdes the capablty to share the wreless channel wth lcensed users (or prmary users) n an opportunstc manner [1]. Cogntve rado (CR) networks are envsoned to provde hgh bandwdth to moble users va heterogeneous wreless archtectures and dynamc spectrum access technques. CR networks, however, mpose unque challenges because of the hgh fluctuaton n the avalable spectrum as well as the dverse qualty-of-servce (QoS) requrements of varous applcatons. To address these challenges, frst, CR networks are requred to determne whch portons of the spectrum are avalable, called spectrum sensng [2], [10]. Furthermore, how to coordnate multple CR users to share the spectrum band, called spectrum sharng, s another mportant ssue n CR networks [7], [16].. The authors are wth the Broadband Wreless Networkng Laboratory, School of Electrcal and Computer Engneerng, Georga Insttute of Technology, Atlanta, GA E-mal: wylee@gatech.edu, an@ece.gatech.edu. Manuscrpt receved 23 July 2008; revsed 28 July 2009; accepted 24 Dec. 2009; publshed onlne 3 Aug For nformaton on obtanng reprnts of ths artcle, please send e-mal to: tmc@computer.org, and reference IEEECS Log Number TMC Dgtal Object Identfer no /TMC Although all these efforts enable CR users to explot spectrum opportuntes effectvely, the heterogenous spectrum envronment ntroduces a new crtcal ssue n CR networks. Generally, CR networks have multple avalable spectrum bands over a wde frequency range that show dfferent channel characterstcs, and need to support applcatons wth dverse servce requrements. Therefore, once avalable spectrum bands are dentfed through spectrum sensng, CR networks need to select the proper spectrum bands accordng to the applcaton requrements. Ths process s referred to as spectrum decson, whch consttutes an mportant but yet unexplored topc n CR networks. To decde on spectrum bands properly, CR networks need to consder the followng ssues:. All avalable spectrum bands show dfferent characterstcs n the CR network. To select the proper spectrum, the CR network needs to characterze avalable spectrum bands by consderng current rado condtons as well as the prmary user (PU) actvty.. The CR network needs to provde a dynamc decson framework to consder all possble events that prevent relable communcatons by closely nteractng wth other CR functonaltes such as spectrum sensng and spectrum sharng.. Accordng to the PU actvtes, total capacty n CR networks vares over tme, whch makes t more dffcult to decde on spectrum bands whle mantanng the servce qualty of other CR users. Thus, the CR network should perform spectrum decson adaptvely dependent on tme-varyng spectrum resources. In ths paper, an adaptve spectrum decson framework s proposed wth the consderaton of all decson events and applcaton types. Frst, a novel capacty model s developed /11/$26.00 ß 2011 IEEE Publshed by the IEEE CS, CASS, ComSoc, IES, & SPS

2 162 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 to descrbe unque characterstcs n CR networks by consderng PU actvty as well as sensng capablty. Accordngly, two dfferent decson schemes are ntroduced. To satsfy the delay constrants n real-tme applcatons, we propose a mnmum varance-based spectrum decson (MVSD) scheme that selects spectrum bands to mnmze capacty varaton. For best-effort applcatons, we propose a maxmum capacty-based spectrum decson (MCSD) scheme to maxmze the total network capacty. Both decson schemes are controlled by a proposed resource management based on the current network condton. The remander of the paper s organzed as follows: Secton 2 presents prevous research and our motvaton. In Secton 3, we propose a novel framework for spectrum decson. In Secton 4, we present a spectrum capacty model used n ths paper. Spectrum decson methods for real-tme and best-effort applcatons are proposed n Sectons 5 and 6, respectvely. Then, we develop a dynamc resource management scheme n Secton 7. Smulaton results are presented n Secton 8. Fnally, conclusons are presented n Secton 9. 2 MOTIVATION 2.1 Related Work The proposed spectrum decson has a smlar objectve to the spectrum sharng n the sense that t performs resource allocaton based on servce requrements. Most of the research on spectrum sharng n CR networks has manly focused on how to effcently allocate ether spectrum or power among CR users subject to nterference constrants. For spectrum allocaton, a global optmzaton scheme s developed based on graph theory [17]. However, whenever the network topology changes accordng to the node moblty, the network needs to completely recompute spectrum assgnment leadng to a hgher computatonal and communcaton overhead. To solve ths problem, a dstrbuted spectrum allocaton based on local barganng s proposed n [4], where CR users negotate spectrum assgnment wthn local self-organzed groups. For the resource-constraned networks such as sensor and ad hoc networks, a rule-based spectrum management s proposed, where CR users access the spectrum ndependently accordng to both local observaton and predetermned rules [5]. In [20], a dynamc channel selecton scheme s developed for delay-senstve applcatons based on a prorty queung analyss and a decentralzed learnng algorthm. Power allocaton among CR users competng the same spectrum s another mportant ssue n spectrum sharng. In [12], an optmal power allocaton scheme s proposed to acheve ergodc and outage capacty of the fadng channel under dfferent types of power constrants and fadng models. In [22], jont beam-formng and power allocaton technques are presented to maxmze the user capacty whle ensurng the QoS of prmary users. Game theory provdes an effcent dstrbuted spectrum sharng scheme by descrbng the conflct and cooperaton among CR users, and hence allowng each user to ratonally decde on ts best acton. Thus, t has been wdely exploted for both channel allocaton [16] and for power allocaton [7]. 2.2 Implementaton Challenge n Spectrum Decson All of the prevous research explaned above has manly addressed spectrum sharng ssues where all operatons are performed wthn the same spectrum band or across contguous channels. Furthermore, to adapt the fast tmevaryng channels, they are generally desgned as a shortterm operaton, such as a packet-based or a tme-slotbased schedulng. However, CR networks necesstate an addtonal resource allocaton capablty when prmary users are detected or CR users newly begn ther sessons, whch are relatvely long-term events. Thus, ths capablty should consder longer-term channel characterstcs, compared to spectrum sharng. In addton, snce avalable spectrum bands are dstrbuted over a wde frequency range, ths functon needs to be mplemented as an nterspectrum operaton. However, ths operaton nevtably ntroduces an addtonal swtchng delay leadng to servce qualty degradaton. Thus, t s not desrable to extend exstng spectrum sharng solutons desgned to adapt to the fast tme-varyng channel to the long-term nterspectrum operaton. Ths unque challenge n CR networks has not been addressed n prevous research. Here our desgn objectve of the spectrum decson framework s to decouple all nterspectrum functonaltes totally from spectrum sharng. 3 THE PROPOSED SPECTRUM DECISION FRAMEWORK 3.1 System Model In ths paper, we consder an nfrastructure-based CR network that has a centralzed network entty, such as a base-staton. The base-staton exerts control over all CR users wthn ts transmsson range. CR users perform the observatons and analyss on rado envronments and feed them to the central base-staton, whch decdes on spectrum avalablty and spectrum allocaton. Each CR user has multple software-defned rado (SDR) transcevers to explot multple spectrum bands over a wde frequency range by reconfgurng the operatng frequency through software operatons. Here, we assume frequency dvson duplex (FDD) systems where uplnk and downlnk channels are separated. Thus, the proposed decson scheme can be appled to each lnk ndependently. When prmary users appear n the spectrum band, CR users need to move to a new avalable band, resultng n a temporary communcaton break. To solve ths problem, we assume that multple noncontguous spectrum bands can be smultaneously used for the transmsson n the CR network. Ths method can create a sgnal that s not only capable of hgh data throughput, but s also mmune to the PU actvty. Even f a prmary user appears n one of the current spectrum bands, the rest of them wll mantan current transmssons [1]. The control channel plays an mportant role n exchangng nformaton regardng sensng and resource allocaton. Several methods are presented n [3], one of whch s assumed to be used as the common control channel n our proposed method.

3 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 163 Fg. 1. The proposed spectrum decson framework. 3.2 Framework Overvew Based on the system model above, we develop a novel framework for spectrum decson. Here, spectrum decson s consdered as an event-based functonalty,.e., the CR network decdes on the proper spectrum bands n the followng events:. CR user appearance: When a new CR user appears n the CR network, t needs to be assgned to new spectrum bands for ts transmsson.. Prmary user appearance: When a prmary user appears n the spectrum band, CR users should move to the new spectrum bands.. Channel qualty degradaton: When channel condton becomes worse, CR users want to swtch to a better spectrum band. To consder all decson events effectvely, the CR network necesstates a unfed framework for spectrum decson. Fg. 1 shows the proposed framework for spectrum decson. A detaled descrpton of ths framework s as follows. By consderng current spectrum condtons, a resource manager determnes f the CR network accepts a new ncomng CR user or not. If a new CR user s allowed to transmt, t s assgned to the proper spectrum bands through spectrum decson. Snce there may be multple CR users competng the same spectrum, spectrum sharng coordnates those multple accesses to prevent collsons, and accordngly to acheve the maxmum capacty. In the event detecton, current spectrum bands and users connectons are montored to detect decson events. The event detecton conssts of two man tasks: spectrum sensng and qualty montorng. When events are detected, the CR network reconfgures ts resource allocaton to mantan the servce qualty. In case of short-term channel varatons such as fast fadng, the CR network reallocates resources wthn the spectrum band through spectrum sharng. If a prmary user s detected or the current spectrum band cannot provde the predetermned servce qualty any longer over a long-term perod, the CR network swtches the spectrum through the resource manager and the spectrum decson. In the proposed framework, CR users perform only event detecton. Based on nformaton gathered from CR users, the base-staton decdes on spectrum avalablty and performs spectrum decson as explaned above. Consequently, the proposed spectrum decson framework provdes a herarchcal QoS guaranteeng scheme: spectrum sharng to allocate the channel and transmsson Fg. 2. Classfcaton of the proposed spectrum decson. power for short-term servce qualtes, and spectrum decson to determne the best spectrum for mantanng the servce qualty over a long term perod. In ths paper, we manly focus on decson functonaltes: spectrum decson and resource management. Spectrum sharng and event detecton functonaltes are out of the scope n ths paper. 3.3 Spectrum Decson Functonaltes In the proposed framework, we consder two types of applcatons: real-tme and best-effort (n ths paper, the terms applcaton and user are nterchangeably used). Accordng to the applcaton type, the proposed spectrum decson can be classfed nto a mnmum varance-based spectrum decson for real-tme applcatons, and a maxmum capacty-based spectrum decson for best-effort applcatons. Decson events manly occur because of ether user actvtes or qualty degradatons. When prmary user appears n the spectrum band or a new CR user begns to transmt, the spectrum decson needs to be ntated. Moreover, the qualty degradaton of ether an entre spectrum band (e.g., ncrease n nterference) or a specfc user connecton (e.g., movng far from the base-staton) can also trgger spectrum decson. If a CR user explots multple spectrum bands, the spectrum decson method becomes more complcated accordng to the event type. When a new CR user appears or the QoS of a certan user becomes worse, multple spectrum bands need to be determned for a sngle user at a tme, called sngle selecton (SS). On the other hand, when a prmary user appears or the qualty of a certan spectrum band becomes worse, multple CR users resdng n that spectrum band lose one of ther current spectrum bands, whch requres multple spectrum decsons for each CR user, called multple selectons (MS). As shown n Fg. 2, accordng to the traffc and event types, spectrum decson can be classfed nto four categores: MVSD-SS, MVSD-MS, MCSD-SS, and MCSD-MS, whch are proposed n Sectons 5.1, 5.2, 6.1, and 6.2, respectvely. For ease of representaton, the mportant notatons used n the subsequent dscusson are summarzed n Table 1. 4 SPECTRUM CHARACTERIZATION To determne the spectrum band properly, t s mportant to dentfy the characterstcs of each spectrum. To ths end, n ths secton, we defne the PU actvty, and accordngly propose a novel CR capacty model.

4 164 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 TABLE 1 Symbols Used for the Analytcal Modelng Fg. 3. Expected transmsson tme n mperfect sensng. 4.1 Prmary User Actvty For an effcent spectrum utlzaton, the CR network needs to be aware of the traffc statstcs of prmary networks n each spectrum, called PU actvty. The PU actvty can be modeled as exponentally dstrbuted nterarrvals [21]. In ths model, the PU actvty n spectrum s defned as a two-state brthdeath process wth death rate and brth rate.anon (busy) state represents the perod used by prmary users and an OFF (dle) state represents the unused perod [6], [13]. 4.2 Cogntve Rado Capacty Model In the CR network, the avalable spectrum bands are not contguous and may be dstrbuted over a wde frequency range wth a dfferent bandwdth. For more flexble manpulaton of heterogenous spectrum bands, we employ an orthogonal frequency dvson multplexng (OFDM) as a physcal layer technology, where each spectrum band has a dfferent bandwdth B, consstng of multple subcarrers. Usually, each subcarrer has a dfferent channel gan and a nose level that are tme-varyng. However, when we consder long-term spectrum characterstcs, both fast and frequency selectve fadng effects are mtgated, and hence we can say the channel gan and nose level n the same spectrum are dentcal over a long-term perod. If transmsson power s also dentcal wthn the spectrum, a normalzed channel capacty c ðkþ (bts/sec/hz) of spectrum band can be expressed as c ðkþ ¼r ðkþ=b, where r ðkþ s the capacty of user k when all subcarrers n spectrum are assgned to user k. However, n CR networks, each spectrum cannot provde ts orgnal capacty c ðkþ. Frst, CR users cannot have a relable spectrum permanently and need to move from one spectrum to another accordng to the PU actvty, whch ntroduces the so-called spectrum swtchng delay. Durng the swtchng tme, the transmsson of the CR user s temporarly dsconnected. Here, spectrum swtchng delay ncludes tmes for the spectrum decson process n the base-staton, sgnalng for establshng new channels, and RF front-end reconfguraton. In IEEE Wreless Regonal Area Network (WRAN), swtchng delay s requred to be less than 2 sec [11]. Also conventonal moble broadcastng systems, for example, Qualcomm s MedaFLO, show an average physcal layer channel swtchng delay up to 1.5 sec [18]. Dependng on the development of the hardware technology, we beleve that t wll be much shorter but stll be a sgnfcant factor to nfluence the network performance. Furthermore, CR users are not allowed to transmt durng sensng operatons, leadng to the perodc transmssons wth sensng effcency [13]. These unque features n CR networks show a sgnfcant nfluence on the spectrum capacty C ðkþ. To descrbe all these stochastc actvtes, we defne a new capacty noton, the so-called CR capacty C CR ðkþ, whch s defned as the expected normalzed capacty of user k n spectrum as follows: C CR ðkþ ¼E½C ðkþš ¼ T T þ c ðkþ; ð1þ where represents the spectrum swtchng delay, and T s the expected transmsson tme wthout swtchng n spectrum. Snce CR users face to the spectrum swtchng after the dle perod, the frst term n (1) represents the transmsson effcency when CR users occupy spectrum. If we consder perfect sensng,.e., both false alarm and detecton error probabltes are zero, T s obtaned as 1=, whch s the average dle perod based on the ON-OFF model n Secton 3. On the contrary, n case of mperfect sensng, we should account for the nfluence of sensng capablty. Let t be a sensng perod. Then, the average number of sensng slots n the dle perod n s s d1= =te. From ths, the expected transmsson tme can be obtaned as follows: T ¼ t Xn s 1 ¼ t k¼1 k 1 P f k P f þ 1 1 P f ns 1 P f 1 1 P f ns 1 P f ðn s 1Þ 1 P f ns þ 1 1 P f ns ; where P f represents a false alarm probablty of spectrum at each sensng slot. Here, T can be expressed as the sum of the expected duratons untl when the false alarm s frst detected n each slot. As P f ncreases, T decreases, resultng n decrease n CR capacty, whch s descrbed n Fg. 3. Here, we consder a cooperatve sensng scheme based on OR fuson, where ts detecton error probablty converges to 0 as the number of users ncreases [15]. Thus, the detecton error probablty can be gnored n estmatng CR capacty. 5 SPECTRUM DECISION FOR REAL-TIME APPLICATIONS Real-tme applcatons are senstve to delay and jtter. Moreover, they requre a relable channel to support a sustanable rate durng the sesson tme. Thus, real-tme ð2þ

5 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 165 network capacty, CR user k selects spectrum bands accordng to the followng lnear nteger optmzaton: Maxmze: X C CR ðkþ x ; ð3þ 2A subject to: X x ¼ N; ð4þ 2A Fg. 4. Data loss n real-tme vdeo applcatons. applcatons have strct constrants on the delay bound and the sustanable rate. Generally, real-tme applcatons drop the packets not arrved wthn the delay bound. Even though the network can support sustanable rate R s on average, packets can be delayed and fnally dscarded n the recever due to the varaton of channel capacty, as shown n Fg. 4. Unlke conventonal wreless networks, the CR network has unque delay factors. When CR users ether sense or swtch the spectrum, they need to stop transmsson temporarly, whch prevents the real-tme applcaton from mantanng ts sustanable rate, leadng to delay and jtter. To observe the effect of the delay unquely shown n CR networks, we assume that a bufferng scheme s optmzed to absorb delay factors n conventonal wreless networks, such as applcaton layer, lnk layer, and transmsson delays. Then, the addtonal delay factors unquely ntroduced by CR networks can drectly lead to data losses. For ths reason, we use the data loss rate to evaluate the servce qualty of real-tme applcatons. Also real-tme applcatons are assumed to have a set of dscrete sustanable rates and to adjust ther rates through the negotaton flexbly. Accordng to the decson events, as explaned n Secton 3.3, the proposed spectrum decson for real-tme applcaton can be classfed nto an MVSD-SS and an MVSD-MS. 5.1 Mnmum Varance-Based Spectrum Decson Sngle Selecton (MVSD-SS) Real-tme applcatons need to have more relable and tmenvarant communcaton channels to satsfy strct servce requrements, such as delay constrants and sustanable rates. However, how to maxmze the total network capacty s stll a crucal problem. To address these ssues together, t s essental to guarantee the servce qualty of real-tme applcatons wth mnmum spectrum resources. Thus, the spectrum decson problem can be formulated as an optmzaton to mnmze bandwdth utlzaton subject to the constrants on the sustanable rate, data loss rate, and number of transcevers. However, ths problem s mxed wth the dscrete optmzaton for spectrum selecton and the contnuous optmzaton for bandwdth allocaton, whch s dffcult to solve. Instead, we ntroduce a threestage spectrum decson method as follows: Step 1: Spectrum Selecton From the vew of the date loss rate caused by delay, the network prefers the spectrum bands wth a lower PU actvty. On the other hand, for network capacty, the channel qualty needs to be consdered n spectrum decson. Thus, to mantan servce qualty and acheve the maxmum C CR ðkþw x R sðkþ ð8 2AÞ; ð5þ N where N s the number of the transcevers of the CR user, and W s the currently avalable bandwdth of spectrum that s equal or less than the total bandwdth B, A s the set of currently avalable spectrum bands, and x 2f0; 1g represents the spectrum selecton parameter that equals 1 f spectrum s selected n the bnary nteger optmzaton. Ths optmzaton consders both PU actvty and CR capacty C CR ðkþ smultaneously as shown n (3). The number of the selected bands s restrcted to the number of transcevers N as gven n (4). The last constrant on sustanable rate R s ðkþ n (5) ensures that the selected spectrum bands have enough bandwdth for resource allocaton, whch s explaned n Step 2 (Secton 5.1.2). Snce real-tme applcatons usually requre much strcter servce requrements than best-effort applcatons, they have a hgher prorty for resource allocaton. Thus, avalable bandwdth W ncludes the portons currently occuped by best-effort applcatons as well as the unused porton of the spectrum Step 2: Resource Allocaton Here, the CR network determnes the bandwdth,.e., a set of subcarrers, of the selected spectrum bands to meet the constrants on both sustanable rate R s ðkþ and target data. To allocate the bandwdth properly, frst, we derve total capacty R T ðkþ and data loss rate P loss ðkþ of user k. When bandwdth w ðkþ s allocated to the selected spectrum for user k, the expected total capacty can be obtaned as follows: loss rate P th loss E½R T ðkþš ¼ X 2S C CR ðkþw ðkþ; ð6þ where S s the set of the selected spectrum bands. To satsfy the servce requrement on the sustanable rate, E½R T ðkþš should be equal to R s ðkþ. Unlke total capacty, data loss rate P loss ðkþ, s expressed n a complcated form, as derved n Appendx A. Thus, t cannot be drectly used for the optmzaton. However, snce the varance of the total capacty s proportonal to the data loss rate, as shown n Appendx B, we can use the followng varance for resource allocaton, nstead of the data loss rate. Var½R T ðkþš ¼ X 2S T T þ T T þ c 2 ðkþ 2 w ðkþ 2 : ð7þ Based on the capacty varance obtaned above, the CR network determnes optmal bandwdth w ðkþ of the selected bands to mnmze the varance of the total capacty as follows: Mnmze:Var½R T ðkþš; ð8þ

6 166 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 subject to: XM ¼1 C CR ðkþw ðkþ ¼R s ðkþ; ð9þ w ðkþ <W ð8 2SÞ: ð10þ Equatons (9) and (10) represent the constrants on the sustanable rate and the avalable bandwdth, respectvely. By the Lagrange multpler method, optmal bandwdth w ðkþ can be obtaned as follows: R s ðkþ T þ w ðkþ ¼ c ðkþ T þ T P2S T T þ T : ð11þ Step 3: QoS Checkup Ths optmzaton s based on the mnmum varance, whch guarantees the mnmum data loss rate but may not satsfy target loss rate Ploss th. If the expected data loss rate P lossðkþ gven n (29) s stll hgher than Ploss th after ths optmzaton, we need to perform one of the followng approaches to satsfy the target loss rate:. Aggressve approach: By sacrfcng the bandwdth effcency, the CR network tres to fnd the proper spectrum bands to meet the servce requrements. To ths end, the selected band havng the hghest PU actvty needs to be replaced by the one wth the ðkþ= among the unselected bands that have a lower PU actvty than the orgnal one. If CR network cannot fnd the proper spectrum band n the aggressve approach, t swtches to the conservatve approach as explaned below.. Conservatve approach: Here, real-tme applcatons are assumed to support multple sustanable rates and to adjust ther rates adaptvely. Thus, n ths approach, nstead of ncreasng the bandwdth, the CR network reduces the current sustanable rate to a one-step lower rate through the renegotaton of the servce qualty and repeats the MVSD-SS whle mantanng the bandwdth effcency. Both aggressve and conservatve approaches are appled n spectrum decson combnng wth resource management, whch s explaned n Secton 7. hghest C CR 5.2 Mnmum Varance-Based Spectrum Decson Multple Selectons (MVSD-MS) MVSD-MS s performed when CR users lose one of ther spectrum bands due to ether PU actvty or qualty degradaton on that band. Snce multple users need new spectrum bands at the same tme, frst, they should determne the order of the spectrum decson. Let R lost ðkþ be the lost capacty of user k resultng from spectrum swtchng. Then, the loss rate of user k s obtaned as R lost ðkþ=r s ðkþ. In MVSD-MS, CR users are selected n order from the hghest to the lowest loss rate. After that, they select a sngle spectrum wth the hghest C CR ðkþ= to meet the sustanable rate, and accordngly allocate the bandwdth of all assgned spectrum bands. 6 SPECTRUM DECISION FOR BEST-EFFORT APPLICATIONS The objectve of typcal schedulng methods for best-effort applcatons s to maxmze the network capacty. The spectrum decson for best-effort applcatons has the same objectve, but addtonally needs to explot the PU actvty and long-term channel characterstcs. Smlar to MVSD n Secton 5, spectrum decson for the best-effort applcaton can be classfed nto a maxmum capacty-based spectrum decson sngle selecton and a multple selectons. 6.1 Maxmum Capacty-Based Spectrum Decson Sngle Selecton (MCSD-SS) Optmally, for the maxmum capacty, the CR network has to perform the spectrum decson over all current transmssons at every decson event, whch requres a hgh computatonal complexty. Also, the entre resource reallocaton leads to the spectrum swtchng of the multple users at the same tme, resultng n the abrupt qualty degradaton. Instead, we ntroduce a suboptmal method for besteffort applcatons. If current resource allocaton s optmal, the spectrum decson to maxmze the network capacty can be smplfed as the followng selecton problem to choose spectrum bands so that the decson gan can be maxmzed. Maxmze: X G ; C CR ðkþ;w L ; C CR ðkþ;w x ; ð12þ 2A subject to: X x ¼ N; ð13þ 2A where Gð; C CR ðkþ;w Þ s the expected capacty gan when new user k wth CR capacty C CR ðkþ jons spectrum wth avalable bandwdth W and Lð; C CR ðkþ;w Þ s the expected capacty loss of other users n that spectrum band. A s the set of currently avalable spectrum bands and N s the number of the transcevers of a CR user. x 2f0; 1g represents the spectrum selecton parameter. The decson gan can be defned as the sum of the dfference between capacty gan and capacty loss caused by the addton of a new user. Assume that a spectrum sharng algorthm assgns the bandwdth to the users farly. Then, the capacty of each user competng for the same spectrum can be approxmated as C CR ðkþw =n b; where n b; represents the number of besteffort users currently resdng n spectrum. Based on ths capacty, the decson gan can be derved as follows: G L¼ CCR ðkþw X 1 n b; þ 1 n j2e b; 1 n b; þ 1 C CR ðjþw ; ð14þ where E s the set of the best-effort CR users currently resdng n spectrum band. The frst term represents the capacty gan of new CR user k and the second term descrbes the total capacty loss of other CR users n spectrum. 6.2 Maxmum Capacty-Based Spectrum Decson Multple Selectons (MCSD-MS) Smlar to the MVSD-MS, MCSD-MS enables multple CR users to select a sngle spectrum band. Thus, the CR network frst determnes the order of the spectrum decson, and then chooses a spectrum band for each CR user as follows:. Each CR user who loses ts spectrum band, fnds a canddate spectrum band wth the hghest decson gan.

7 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 167. A CR user wth the hghest decson gan s assgned to the spectrum frst through the optmzaton n (12).. Accordng to the optmzaton result, the CR network updates the current bandwdth allocaton and repeats the MCSD-MS for the remanng CR users who need to be assgned to a new spectrum band. 7 DYNAMIC RESOURCE MANAGEMENT FOR SPECTRUM DECISION Because of the PU actvtes, avalable spectrum bands show tme-varyng characterstcs n the CR network. Thus, wth the only proposed decson schemes, the CR network s not able to explot spectrum resources effcently, and hence results n the volaton of the guaranteed servce qualty. As a result, the CR network necesstates an addtonal resource management scheme to coordnate the proposed spectrum decson methods adaptvely wth bandwdth fluctuatons. The man objectves of the proposed resource management are as follows:. The CR network s capable of determnng the acceptance of a new ncomng CR user wthout any effect on the servce qualty of currently transmttng users.. Durng the transmsson, the CR network needs to mantan the servce qualty of currently transmttng users by consderng the fluctuaton of the avalable bandwdth.. Snce real-tme users usually have a hgher prorty n spectrum access, best-effort users may not have enough resources. Thus, the CR network may be requred to balance the bandwdth between both applcatons. In the followng sectons, we defne the network states to descrbe the current spectrum utlzaton. Based on these states, we present an admsson control scheme, and then propose decson control methods for two dfferent events: CR user and prmary user appearances. 7.1 Spectrum States for Resource Management To explot spectrum resources effcently, the proposed spectrum decson needs to adapt to the tme-varyng network condtons. Thus, we classfy the network condton nto three states accordng to the bandwdth utlzaton. Let W R be the bandwdth currently assgned to real-tme users, and W av be the total avalable bandwdth not occuped by prmary users. W mn represents the mnmum bandwdth to guarantee the servce requrements of current users. W R, W av, and W mn are tme-varyng accordng to the spectrum decson results and PU actvtes. Snce best-effort users do not have strct servce requrements, we consder only the bandwdth assgned to real-tme users n determnng the network state. As shown n Fg. 5, the network states are classfed as follows:. Underloaded state: If the current occupancy of realtme users, W R =W av s less than, the CR network s underloaded. s the predefned overload threshold to determne f the network s overloaded or not.. Overloaded state: When W R =W av >, the CR network s now overloaded. Accordng to the amount of the Fg. 5. The state dagram for resource management. remanng bandwdth, ths state can be classfed nto two substates. If the expected bandwdth requred for the spectrum decson, W req, s less than the currently unused bandwdth W av W R, the CR network s n the begnnng of the overloaded state and stll has enough resources (operatng state). Otherwse, the CR network s almost saturated and does not have enough bandwdth for the current spectrum decson operaton (saturated state). W req s gven n Secton Outage state: If avalable bandwdth W av s below W mn, the CR network cannot provde the guaranteed servce qualty to the currently actve CR users. If becomes hgher, real-tme users can have more stable sustanable rate due to less admsson and rate controls, but the outage probablty wll be hgher. 7.2 Admsson Control The CR network s responsble for guaranteeng the servce requrements of current CR users regardless of bandwdth fluctuatons. Thus, f the CR network cannot mantan the servce requrements, t should reject a new ncomng CR user, referred to as an admsson control. The proposed admsson control method requres the followng procedures:. User characterzaton: Accordng to the rado condton, each CR user requres dfferent bandwdth to acheve the same servce requrements. The rado condton of each user k can be represented as ts normalzed capacty over all spectrum bands CðkÞ as follows: CðkÞ ¼ P M ¼1 CCR ðkþb P M ¼1 B ; ð15þ where M s the number of all spectrum bands and B s the total bandwdth of spectrum.. Bandwdth for guaranteeng the servce qualty: Snce avalable bandwdth W av vares over tme, the CR network cannot always satsfy the servce requrements. Thus, we ntroduce a lower lmt of bandwdth W mn to guarantee the servce requrements of current CR users. Assume that regardless of the bandwdth fluctuaton, the CR network should guarantee an average sustanable rate, R mn ðkþ, over an entre sesson of real-tme user k. Then, the mnmum bandwdth of user k to support R mn ðkþ s expressed as R mn ðkþ=cðkþ. When a new CR user appears, W mn can be expressed as the sum of the

8 168 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 Fg. 6. The flowchart for the proposed decson control: (a) CR user appearance, and (b) prmary user appearance. mnmum bandwdths for all CR users ncludng both current and ncomng users.. Admsson crteron: The proposed spectrum decson s desgned for the states when W av s above W mn. Otherwse, the network s n the outage state, and hence cannot mantan servce requrements of current CR users. However, W mn s tme-varyng accordng to the current users and spectrum avalablty. To mtgate ths temporal resource fluctuaton, we frst determne stable nterval T mn, whch s defned as the average perod where no CR user appears, and accordngly W mn does not change. Assume that the departure rate of CR users s. Then, T mn can be obtaned as 1=ð n r Þ on average, where n r s the number of current real-tme users. To avod resource outage of current CR users, the proposed scheme accepts a new ncomng user only f a resource outage probablty durng ths nterval s greater than the predetermned acceptable outage probablty. Otherwse, t s rejected. The resource outage probablty, P out, s the probablty that W av <W mn, whch s derved n Appendx C. The performance of the admsson control method depends on acceptable outage probablty. If the CR network has a hgher, t can accept more users, resultng n hgher qualty degradaton snce t s hghly probable that W av <W mn. The proposed MVSD method explaned n Secton 5 tres on the assumpton that the network has suffcent avalable bandwdth. Thus, the spectrum decson also needs to consder the addtonal data loss factor resultng from the bandwdth shortage. Snce the network capacty s proportonal to the avalable bandwdth, the data loss rate newly ntroduced by the admsson control can be approxmately estmated as follows: to satsfy target data loss rate P th loss bp loss ¼ W mn E½W av jw av <W mn Š P out : ð16þ W mn Here, the frst term represents the rato of the amount of bandwdth shortage to the bandwdth lmt W mn. Then, the actual data loss rate should be expressed as the sum of Ploss th and bp loss. Assume that real-tme users have maxmum allowable data loss rate P loss. To satsfy ths servce requrement, target rate Ploss th should be decded as follows: P th loss ¼ P loss bp loss : ð17þ The proposed admsson control method s orgnally desgned only for real-tme users. Snce the best-effort users do not have strct servce requrements, they do not need the admsson control scheme. 7.3 Decson Control Here, we propose decson control schemes for both CR and prmary user appearances, whch enable spectrum decson to adapt to the dfferent network states Decson Control n CR User Appearance One of the mportant roles n the decson control s how to allocate spectrum resources wth the mnmum nfluence on current CR users when a new CR user appears. Fg. 6a shows the procedures of the proposed decson control when the new CR user appears. Accordng to the state, the proposed control scheme coordnates the spectrum decson as follows: Underloaded state. Snce the avalable bandwdth s suffcent n the underloaded state, the CR network performs the spectrum decson aggressvely,.e., the aggressve MVSD-SS for real-tme users and the MCSD-SS for best-effort users. Overloaded state. Snce the avalable bandwdth becomes scarce n ths state, the spectrum decson needs to be more spectrum-effcent. Thus, the CR network performs the conservatve MVSD-SS for the real-tme user. However, snce real-tme users occupy much hgher bandwdth through ths operaton, best-effort users may experence bandwdth starvaton n the overloaded state. If the CR network s requred to balance the bandwdth between real-tme and best-effort users, t needs to check the current bandwdth utlzaton of both applcatons before MCSD-SS. Let be a balance coeffcent predetermned by the CR network. If the average bandwdth of current real-tme users, W R =n r s greater than the weghted average bandwdth for best-effort users, ðw av W R Þ=n b, current resource allocaton s consdered to be unbalanced

9 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 169 where n r and n b are the numbers of the current real-tme and the current best-effort users, respectvely. If s greater than 1, real-tme users can occupy more bandwdth, and hence guarantee more stable servce qualty. To solve the resource starvaton problem n best-effort users, we propose a selectve rate control that mantans resource balance n the overloaded state by reducng the sustanable rate of the selected real-tme users. When each real-tme user k reduces ts sustanable rate to a one-step lower rate, the expected bandwdth gan s expressed as RðkÞ=CðkÞ where RðkÞ and CðkÞ s the rate decrement and the normalzed capacty of a real-tme user k, respectvely. Based on the bandwdth gan, the CR network selects real-tme users for the selectve rate control to mnmze total rate reducton subject to the balance constrant, whch can be expressed as the followng lnear nteger optmzaton problem: Mnmze: X k2r RðkÞx k ; ð18þ subject to: W R W W av W R þ W 0; ð19þ n r n b W X RðkÞ CðkÞ x k; x k 2f0; 1g; ð20þ k2r where R s the set of real-tme users currently actve and W s the bandwdth requred for the balance. The real-tme users selected through the above optmzaton reduce ther sustanable rates to the one-step lower rates and then perform the resource allocaton explaned n Secton 5.1. Whenever the best-effort user appears n ths state, the network tres to satsfy the balanced condton. However, to avod the abrupt qualty degradaton of realtme users, a selectve rate control can change the sustanable rate of real-tme users to only a one-step lower rate. Outage state. The servce requrements of CR users cannot be guaranteed because of resource shortage. Thus, all ncomng best-effort users should be rejected n ths state to avod the overall qualty degradaton. New realtme users n ths state are already rejected through the admsson control Decson Control n Prmary User Appearance Once the CR network accepts the users, t should guarantee ther servce requrements durng the transmsson regardless of the bandwdth varaton. Fg. 6b shows the decson control procedure n the prmary user appearance. Accordng to the network states, the proposed scheme can be performed as follows: Underloaded state. Smlar to the CR user appearance, the CR network performs the spectrum decson aggressvely. For a real-tme user, the aggressve MVSD-MS s used whereas the MCSD-MS s executed for a best-effort user. Operatng state. In the overloaded state, the decson control starts to coordnate the bandwdth allocaton to mantan the servce qualty. In the prmary user appearance, the overloaded state can be dvded nto two dfferent substates accordng to the remanng spectrum resources. In the operatng state, the CR network s consdered to be overloaded but stll has enough resources for spectrum decson,.e., avalable bandwdth W av s greater than bandwdth requred for spectrum decson, W req. The expected bandwdth for MVSD-MS, W req can be derved as follows: W req ¼ X R lost ðkþ ðw av W R Þ; ð21þ k2r l P2A CCR ðkþw where W s the avalable bandwdth of spectrum currently unused by both prmary and real-tme CR users and R lost ðkþ s the lost capacty of user k due to the PU actvtes. A s the set of the currently avalable spectrum bands and R l s the set of the real-tme users who lose ther spectrum bands, respectvely. Here, W req s expressed as the sum of the expected bandwdth of user k 2R l requred to support R lost ðkþ. The denomnator n the summaton n (21) represents the total expected capacty of user k over all currently avalable spectrum bands. If the bandwdth of both applcatons s balanced, the CR network performs a conservatve MVSD-MS and an MCSD- MS. Otherwse, t needs a selectve rate control before the spectrum decson smlar to the case n CR user appearance. The only dfference s that a selectve rate control s just appled to the real-tme users losng one of ther spectrum bands to mnmze the nfluence on other real-tme users. Saturated state. The other overloaded state n prmary user appearance s the saturated state where the remanng avalable bandwdth s less than the bandwdth requred for the spectrum decson. In ths case, real-tme CR users cannot fnd new spectrum bands to mantan ther current servce requrements, whch necesstates the renegotaton of ther servce requrements. Let all possble sustanable rates for user k be fr s;1 ðkþ; R s;2 ðkþ;...;r s;nk ðkþg, where n k s the number of all possble sustanable rates. Then, the expected bandwdth of each sustanable rate can be obtaned as R s; ðkþ=cðkþ where CðkÞ s the normalzed capacty of user k gven n (15). Based on the expected bandwdth gans n renegotaton, we propose a full rate control where the sustanable rates of real-tme users currently requestng spectrum decson are optmzed to satsfy both bandwdth and balance constrants. Ths optmzaton problem s expressed as the followng lnear nteger optmzaton, the so-called lockbox problem [8]. Maxmze: X k2r l R s ðkþ ¼ X k2r l X nk ¼1 R s;nk ðkþx ðkþ; ð22þ subject to: W R s þ bw R W av WR s bw R < 0; n r n b ð23þ bw R <W av WR s ; ð24þ bw R ¼ X k2rl X nk ¼1 X nk ¼1 R s; ðkþ CðkÞ x ðkþ; ð25þ x ðkþ ¼1 x ðkþ 2f0; 1g; ð26þ where R l and bw R are the set of the real-tme users who lose ther spectrum bands and ther expected bandwdth. W s R s the bandwdth of the real-tme users not affected by the PU actvtes. Equaton (23) s the constrant on the resource

10 170 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 balance explaned n Secton Equaton (24) s the constrant on the avalable bandwdth requred for the spectrum decson. Outage state. Ths state cannot provde a guaranteed servce qualty any longer. Thus, even though the CR network needs the spectrum decson, all CR users who lose ther connectons reduce ther sustanable rate to the mnmum and just wat untl the network condton becomes better. 8 PERFORMANCE EVALUATION 8.1 Smulaton Setup Here, we smulate an nfrastructure-based CR network consstng of one base-staton and multple CR users. Each user s unformly dstrbuted over the network coverage wth the radus of 2 km. The CR network s assumed to operate n 20 lcensed spectrum bands consstng of four VHF/UHF TV, four AMPS, four GSM, four CDMA, and four WCDMA bands. The bandwdth of these bands are 6 MHz (TV), 30 khz (AMPS), 200 khz (GSM), 1.25 MHz (CDMA), and 5 MHz (WCDMA), respectvely. The PU actvtes of each spectrum band, and, are randomly selected over ½0; 1Š. The servce rate of CR traffc s 0.02, and ts arrval rate can be determned accordng to the average number of users. In the smulatons, we assume a lognormal fadng channel model, where the nose power s 115 dbm, the shadowng devaton s 4, and the path loss coeffcent s set to 4 [19]. Transmsson power P k ðfþ s unty over all frequences. Through spectrum sensng, the base-staton s already aware of the spectrum avalablty n ts coverage. Sensng effcency, and false alarm probablty P f are set to 0.9 and 0.99, respectvely. These sensng capabltes are assumed to be dentcal over all spectrum bands. User-based and the band-based qualty degradatons, explaned n Secton 3.3, use the same strateges as prmary user and CR user appearances, respectvely. Thus, we do not consder them n the smulatons. The real-tme applcaton s assumed to support fve dfferent btrates: 64, 128, 256, 512 kbps, and 1.2 Mbps. For the resource management, W mn and R mn are set to 10 MHz and 512 Kbps, respectvely. The overloaded threshold s set to 0.5, the balance coeffcent s 1. The acceptable data loss rate, P loss, and the acceptable outage threshold are set to 0.05 and 0.03, respectvely. To evaluate the performance of our spectrum decson framework, we ntroduce three dfferent cases as follows:. Case 1: CR users explot all functonaltes of the entre spectrum decson framework ncludng MVSD, MCSD, and all resource management functons explaned n Sectons 5, 6, and 7, respectvely.. Case 2: CR users perform the proposed spectrum decson framework wthout the admsson control scheme.. Case 3: CR users use only MVSD and MCSD methods (Case 1 wthout both admsson and decson controls).. Case 4: Instead of the optmzaton schemes n Secton 5.1, the proposed MVSD scheme adopts an exhaustve search to determne proper spectrum bands and ther bandwdth, whch s optmal for real-tme users. Snce there are no prevous work related to spectrum decson, we compare our decson framework wth two straghtforward decson crtera as follows:. Case 5 Capacty-based decson: CR users select the spectrum wth the hghest channel capacty as follows: Maxmze: X 2A C CR ðkþx subject to: X x N x 2f0; 1g 2A X C CR ðkþw x R s ðkþ: 2A ð27þ A s the set of the currently avalable spectrum bands. The last constrant s appled only to the realtme users.. Case 6 PU actvty-based decson: CR users select the spectrum bands wth the lowest PU actvty. Instead of the objectve functon n Case 5, Case 6 uses P 2A 1= x. In the followng sectons, we show the smulaton results n three dfferent scenaros (only real-tme users, only besteffort users, and both of them). 8.2 Real-Tme Applcatons Frst, we consder the scenaro wth only real-tme users to valdate the proposed MVSD-SS and MVSD-MS descrbed n Secton 5. Snce ths scenaro does not requre the decson control for bandwdth balance, Case 3 s not consdered n ths smulaton. The numbers on the graph ndcate the standard devatons of each smulaton, whch show the dstrbuton of the data loss rate over all CR users. Fg. 7a shows how the average number of users nfluences the data loss rate. Here, we assume three spectrum bands and 0.1 sec for the swtchng delay. For ths smulaton, we generate CR user traffc from 10 to 80 on average. When a small number of users are transmttng, each case shows relatvely low data loss rate. However, as the number of users ncreases, other methods (Cases 2, 5, and 6) ncrease the data loss rate. On the contrary, Case 1 stll mantans a certan level of the data loss rate where the admsson control controls the addton of new users adaptvely dependent on current network utlzaton. However, Case 1 shows lttle hgher data loss rate than the acceptable data loss rate. The reason s that durng the transmsson the MVSD-MS scheme mantans all ongong transmssons even though they cannot fnd the spectrum bands to satsfy the acceptable data loss rate, whch causes slght ncrease n the data loss rate. Even though the proposed method does not use admsson control (Case 2), t stll shows a better data loss rate than Cases 5 and 6. In Fg. 7b, we nvestgate the performance of the spectrum decson under four PU actvty scenaros low/ low, low/hgh, hgh/low, and hgh/hgh. Low PU actvty (ether or ) s unformly dstrbuted between 0 and 0.5, and hgh PU actvty s between 0.5 and 1. The average number of users, the number of spectrum bands, and swtchng delay are set to 50, 3, and 0.1 sec, respectvely. In

11 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 171 Fg. 7. Data loss rate n real-tme applcatons: (a) average number of users, (b) PU actvtes, (c) swtchng delay, and (d) number of spectrum bands. all cases, Case 1 shows better performance n data loss than other method (Cases 2, 3, 5, and 6), and a smlar loss rate to Case 4. Also, s s shown that s a more domnant factor to determne the loss rate than snce a hgher ntroduces more frequent swtchng, leadng to a sgnfcant performance degradaton. We also show the relatonshp between the data loss rate and the swtchng delay n Fg. 7c. Here, we assume 50 users and three spectrum bands. The proposed method (Case 1) shows the lower loss rate than other methods by both rejectng users before the transmsson and reducng sustanable rate durng the transmsson. Case 2 stll shows better performance than Cases 5 and 6 because of both MVSD-SS and MVSD-MS. In all cases, a longer swtchng delay results n a hgher data loss rate. As explaned n Secton 3.1, the transmsson wth multple transcevers can mtgate the effect of capacty fluctuatons as well as prevent a temporary dsconnecton of communcaton channels. Ths phenomena are observed n Fg. 7d. Here, we assume 0.1 sec for the swtchng delay and 50 real-tme users. An nterestng pont s that more spectrum bands do not always lead to good performance n the data loss rate. As the number of spectrum bands ncreases, the total amount of PU actvtes over multple spectrum bands ncreases, whch may cause an adverse effect on the data loss rate. In ths smulaton, each does not mprove ts data loss rate sgnfcantly when t has more than two spectrum bands. Consequently, n all smulatons, the proposed method (Case 1) shows almost the same performance as the optmal method (Case 4), but requres less computatonal complexty as explaned n Secton Best-Effort Applcatons To evaluate the performance of MCSD-MS and MCSD-SS descrbed n Secton 6, we compare the proposed method (Case 1) wth Cases 5 and 6. Snce the admsson and decson control functonaltes are not needed n ths scenaro, we do not consder Cases 2, 3, and 4 here. In ths smulaton, we also show how the number of users, PU actvty, swtchng delay, and number of spectrum bands nfluence the total network capacty. As shown n Fg. 8, Case 1 shows hgher capacty compared to the capactybased and PU-actvty-based methods. Fg. 8a ndcates the relatonshp between the number of users and total network capacty where Case 1 shows a better performance over Cases 5 and 6 by explotng the PU actvty and the channel condton at the same tme. In Fg. 8b, we nvestgate how PU actvtes nfluence the performance of the total capacty. Smlarly, Case 1 shows better performance than other cases. Especally, when s lower, Case 1 shows more mprovement due to less frequent swtchng delay. Fg. 8c shows the smulaton results on the total network capacty when 50 best-effort users wth three spectrum bands are assumed. Here, we observe that ncrease n swtchng delay causes an adverse nfluence on network capacty. Fg. 8d nvestgates the smulaton results on the total network capacty when 50 best-effort users wth 0.1 sec swtchng delay are appled. Smlar to the smulaton on real-tme users, Case 1 shows the best performance n two spectrum bands, but less total capacty n more than two spectrum bands, snce t causes a more frequent spectrum swtchng as well as prevents explotng channel dversty. 8.4 Hybrd Scenaro Here, we consder a hybrd scenaro where both real-tme and best-effort users coexst. Smlarly, we assume three Fg. 8. Total network capacty n best-effort applcatons: (a) average number of users, (b) PU actvtes, (c) swtchng delay, and (d) number of spectrum bands.

12 172 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 Fg. 9. Performance n the hybrd scenaro: (a) data loss rate, and (b) user capacty. spectrum bands and 0.1 sec swtchng delay for ths smulaton. Here, we set the total number of actve users to 100 and vary the number of best-effort and real-tme users to nvestgate the performance accordng to the network state. In Fg. 9a, we show the data loss rate of real-tme users on a hybrd scenaro. In the underloaded state,.e., when there less real-tme users n the network, we can see each method shows lower data loss rate. On the other hand, overloaded condtons lead to consderably dfferent performance accordng to the decson methods. Through admsson and decson controls, Case 1 admts real-tme users only when the network can guarantee the servce requrement of current users, and hence mantans the lowest data loss rate. When the proposed method does not use the admsson and decson controls (Cases 2 and 3), the CR network accepts much more real-tme users than t can provde wth the guaranteed servce qualty, leadng to the ncrease n the data loss rate and the decrease n the average capacty of the realtme user as descrbed n Fg. 9b. On the contrary, Cases 5 and 6 show the worst data loss rates. In Fg. 9b, we show the average user capacty of real-tme and best-effort users n the hybrd scenaro. When real-tme users are less than best-effort users, Cases 5 and 6 show the hghest user capacty n real-tme users whle mantanng slghtly hgher data loss rate as that of the proposed methods (Cases 1, 2, and 3). On the contrary, as the number of real-tme users ncreases, Cases 5 and 6 show lower capacty n best-effort users due to the lack of resource management, but the proposed method (Case 1) stll provdes enough capacty to best-effort users. Even though real-tme users occupy most of the bandwdth resources n Cases 5 and 6, they cannot satsfy the servce requrements and show the hghest data loss rate as observed n Fg. 9a. By explotng the admsson control scheme, Cases 1 and 2 show better farness n capacty between both applcaton types whle mantanng the low data loss rate n real-tme users. Though Case 3 does not use both admsson and decson control schemes, t shows slghtly hgher capacty n best-effort users than Cases 5 and 6 snce the MVSD scheme provdes bandwdth-effcent resource allocatons, leadng to ncrease n avalable bandwdth for best-effort users as explaned n Secton 5. Smlarly, the optmal method (Case 4) selects the bandwdth-effcent spectrum for real-tme users, leadng to slghtly hgher capacty n best-effort users than the proposed method (Case 1), whle t acheves almost same data loss rate and user capacty n real-tme users as the proposed method. Fg. 10 shows how the proposed admsson control explots bandwdth resources when 50 real-tme users and 50 best-effort users are transmttng smultaneously. From the smulaton results, we can see the proposed admsson control (Case 1) balances the bandwdth between both applcatons over the entre smulaton tme. On the contrary, n Cases 5 and 6, real-tme applcatons occupy most of the avalable bandwdth to satsfy ther servce requrements, leadng to the bandwdth starvaton of besteffort users. 9 CONCLUSION In ths paper, we ntroduced a framework for spectrum decson to determne a set of spectrum bands by consderng the channel dynamcs n the CR network as well as applcaton requrements. To ths end, frst, a novel spectrum capacty model s proposed that consders unque features n CR networks. Based on ths capacty model, an MVSD s developed for real-tme applcatons, whch determnes the spectrum bands to mnmze the capacty varance. For the best-effort applcatons, an MCSD s proposed where spectrum bands are decded to maxmze the total capacty. Moreover, a dynamc resource management scheme s ntroduced to enable the CR network to Fg. 10. Bandwdth utlzaton n the hybrd scenaro.

13 LEE AND AKYILDIZ: A SPECTRUM DECISION FRAMEWORK FOR COGNITIVE RADIO NETWORKS 173 coordnate spectrum decson adaptvely dependent on the tme-varyng spectrum resources. Smulaton results show that the proposed spectrum decson framework provdes effcent bandwdth utlzaton whle guaranteeng the servce qualty. APPENDIX A DERIVATION OF THE DATA LOSS RATE IN COGNITIVE RADIO NETWORKS In the CR network, each spectrum band has two dscrete capacty states, 0 and c ðkþw ðkþ accordng to ts PU actvty, as explaned n Secton 4. Here, c ðkþ and w ðkþ are the normalzed capacty and the bandwdth of spectrum for user k, respectvely. Thus, when N spectrum bands are assgned to a CR user k, the total capacty R T ðkþ has 2 N states accordng to the PU actvtes of the selected spectrum bands. Thus, each state m has the followng state probablty: P m ðkþ ¼ Y 2I m T T þ Y 1 T 2B T m þ ; ð28þ where I m and B m are the sets of dle spectrum bands and busy spectrum bands at state m, respectvely. Let the sustanable rate of user k be R s ðkþ and the capacty of each state m be br m ðkþ. From the assumpton that the data loss occurs when channel capacty s below R s ðkþ, the data loss rate can be defned as the rato of the expected capacty loss to the sustanable rate R s ðkþ as follows: P loss ðkþ ¼ R sðkþ P2N m¼1 mnðr sðkþ; br m ðkþþp m ðkþ R s ðkþ P 2 N m¼1 ¼ jr sðkþ br m ðkþjp m ðkþ : 2R s ðkþ APPENDIX B ð29þ DERIVATION OF THE CAPACITY VARIATION IN COGNITIVE RADIO NETWORKS From the capacty state probablty, derved n (28), the varance of the total capacty R T ðkþ can be derved as follows: Var½R T ðkþš ¼ X2N ð br m ðkþ R s ðkþþ 2 P m ðkþ: m¼1 ð30þ By comparng (29) wth (30), we can see that the varance of the total capacty Var½R T ðkþš s proportonal to the data loss rate P loss ðkþ. As a result, we can use the capacty varance for resource allocaton, nstead of the data loss rate. To apply the varance n (30) for the optmzaton, we need another form of the varance expressed n terms of the bandwdth w ðkþ and the normalzed capacty c ðkþ of each spectrum. Snce the spectrum s ndependent wth each other, the varance of the total capacty n the selected spectrum bands can be expressed as follows: X Var½R T ðkþš ¼ Var C ðkþw ðkþ 2S ¼ X 2S Var½C ðkþw ðkþš ¼ X 2S ðe½ðc ðkþw ðkþþ 2 Š E½C ðkþw ðkþš 2 Þ ¼ X c ðkþ 2 w ðkþ 2 T T 2S þ T 2 c ðkþw ðkþ þ ¼ X 2S T T ðt þ T Þ ðt þ Þ 2 c ðkþ 2 w ðkþ 2 ; ð31þ where C ðkþ s the random varable to represent the capacty of spectrum for user k. S s the set of the selected bands. APPENDIX C DERIVATION OF THE RESOURCE OUTAGE PROBABILITY To model PU actvtes n the spectrum, we can use a twostate Markov chan wth the transton probabltes from dle to dle x 00 ¼ 1 e t, from dle to busy x 01 ¼ e t, from busy to dle x 10 ¼ e t, and from busy to busy x 11 ¼ 1 e t, where t s a sensng perod. Then, the dle probablty of spectrum after rt, P dle ðrþ, can be expressed as ether one of the followng probabltes [14]: P 2 ðrþ ¼ P 2b ðrþ ¼ x 10 x 01 þ x 10 þ 1 x 01 x 10 r x 01 x 01 þ x 10 r x 10 x 10 x 01 þ x 10 1 x 01 x 10 x 01 þ x 10 ; ; ð32þ where P 2 ðrþ and P 2b ðrþ are the expected dle probabltes after rt when current spectrum states are dle and busy, respectvely. If a false alarm probablty P f s consdered, the dle probablty of spectrum can be expressed as ether ð1 P f 2 ÞP ðrþ or ð1 P f 2b ÞP ðrþ. Based on these probabltes, we derve the expected resource outage probablty as follows: Snce the network has M spectrum bands, t has 2 M states accordng to the status of each band. Let L be a set of states that experence resource outage,.e., that W av <W mn. I n represents a set of dle spectrum bands at state n. Then, resource outage happens when all spectrum bands n I n ;n2lare dle and the rest of bands 62 I n ;n2lare busy. From ths, the resource outage probablty after rt, P out ðrþ can be derved as follows: P out ðrþ ¼ X n2l Y P dle 2I n ðrþ Y 1 P dle ðrþ : ð33þ 62I n Based on ths probablty, we can obtan the expected resource outage probablty durng rt, P out as P r r 0 ¼1 P outðr 0 Þ=r. ACKNOWLEDGMENTS Ths materal s based upon work supported by the US Natonal Scence Foundaton under Grant No. CNS

14 174 IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 10, NO. 2, FEBRUARY 2011 REFERENCES [1] I.F. Akyldz, W.-Y. Lee, M.C. Vuran, and S. Mohanty, A Survey on Spectrum Management n Cogntve Rado Networks, IEEE Comm. Magazne, vol. 46, no. 4, pp , Apr [2] D. Cabrc, S.M. Mshra, and R.W. Brodersen, Implementaton Issues n Spectrum Sensng for Cogntve Rados, Proc. IEEE Aslomar Conf. Sgnals, Systems and Computers, pp , Nov [3] D. Cabrc, S.M. Mshra, D. Wllkomm, R. Brodersen, and A. Wolsz, A Cogntve Rado Approach for Usage of Vrtual Unlcensed Spectrum, Proc. 14th IST Moble and Wreless Comm. Summt, June [4] L. Cao and H. Zheng, Dstrbuted Spectrum Allocaton va Local Barganng, Proc. IEEE Sensor and Ad Hoc Comm. and Networks (SECON), pp , Sept [5] L. Cao and H. Zheng, Dstrbuted Rule-Regulated Spectrum Sharng, IEEE J. Selected Areas n Comm., vol. 26, no. 1, pp , Jan [6] C. Chou, S. Shankar, H. Km, and K.G. Shn, What and How Much to Gan by Spectrum Aglty? IEEE J. Selected Areas n Comm., vol. 25, no. 3, pp , Apr [7] R. Etkn, A. Parekh, and D. Tse, Spectrum Sharng for Unlcensed Bands, IEEE J. Selected Areas n Comm., vol. 25, no. 3, pp , Apr [8] J.R. Evans and E. Mneka, Optmzaton Algorthms for Networks and Graphs, second ed. CRC Press, [9] FCC, ET Docket No , Spectrum Polcy Task Force Report, Nov [10] M. Gandetto and C. Regazzon, Spectrum Sensng: A Dstrbuted Approach for Cogntve Termnals, IEEE J. Selected Areas n Comm., vol. 25, no. 3, pp , Apr [11] IEEE P802.22/D , IEEE WG, Draft Standard for Wreless Regonal Area Networks Part 22: Cogntve Wreless RAN Medum Access Control (MAC) and Physcal Layer (PHY) Specfcatons: Polces and Procedures for Operaton n the TV Bands, IEEE, Sept [12] X. Kang, Y. Lang, A. Nallanathan, H. Garg, and R. Zhang, Optmal Power Allocaton for Fadng Channels n CR Networks: Ergodc Capacty and Outage Capacty, IEEE Trans. Wreless Comm., vol. 8, no. 2, pp , Feb [13] W.-Y. Lee and I.F. Akyldz, Optmal Spectrum Sensng Framework for Cogntve Rado Networks, IEEE Trans. Wreless Comm., vol. 7, no. 10, pp , Oct [14] W.-Y. Lee and I.F. Akyldz, Spectrum-Aware Moblty Management n Cogntve Rado Cellular Networks, to be publshed. [15] Y.C. Lang, Y. Zeng, E. Peh, and A.T. Hoang, Sensng- Throughput Trade for Cogntve Rado Networks, IEEE Trans. Wreless Comm., vol. 7, no. 4, pp , Apr [16] N. Ne and C. Comancu, Adaptve Channel Allocaton Spectrum Etquette for Cogntve Rado Networks, Proc. Frst IEEE Int l Symp. New Fronters n Dynamc Spectrum Access Networks (DySPAN 05), pp , Nov [17] C. Peng, H. Zheng, and B.Y. Zhao, Utlzaton and Farness n Spectrum Assgnment for Opportunstc Spectrum Access, ACM Moble Networks and Applcatons, vol. 11, no. 4, pp , Aug [18] M.R. Char, F. Lng, A. Mantravad, R. Krshnamoorth, R. Vjayan, G.K. Walker, and R. Chandhok, FLO Physcal Layer: An Overvew, IEEE Trans. Broadcastng, vol. 53, no. 1, pp , Mar [19] T. Rappaport, Wreless Communcatons: Prncples and Practce, second ed. Prentce Hall, [20] H. Shang and M. Schaar, Queung-Based Dynamc Channel Selecton for Heterogeneous Multmeda Applcatons over Cogntve Rado Networks, IEEE Trans. Multmeda, vol. 5, no. 10, pp , Aug [21] K. Srram and W. Whtt, Characterzng Superposton Arrval Processes n Packet Multplexers for Voce and Data, IEEE J. Selected Areas n Comm., vol. 4, no. 6, pp , Sept [22] L. Zhang, Y. Lang, and Y. Xn, Jont Beamformng and Power Allocaton for Multple Access Channels n Cogntve Rado Networks, IEEE J. Selected Areas n Comm., vol. 26, no. 1, pp , Jan Won-Yeol Lee receved the BS and MS degrees from the Department of Electronc Engneerng, Yonse Unversty, Seoul, Korea, n 1997 and 1999, respectvely, and the PhD degree n electrcal and computer engneerng from the Georga Insttute of Technology, Atlanta, n 2009 under the gudance of Professor Ian F. Akyldz. From 1999 to 2004, he was the senor research engneer of the Network R&D Center and Wreless Multmeda Servce Development Dvson at LG Telecom, Seoul, Korea. Currently, he s the deputy drector of the Wreless Technology Department, Central R&D Laboratory, Korea Telecom (KT), Seoul. Hs current research nterests nclude cogntve rado networks, next generaton wreless systems, and wreless sensor networks. He receved the 2008 Researcher of the Year Award n the Broadband Wreless Networkng Laboratory, School of Electrcal and Computer Engneerng, Georga Insttute of Technology. He s a student member of the IEEE. Ian F. Akyldz s the Ken Byers Dstngushed Char professor wth the School of Electrcal and Computer Engneerng, Georga Insttute of Technology (Georga Tech). Snce June 2008, he s an honorary professor wth the School of Electrcal Engneerng at the Unverstat Poltecnco de Catalunya, Barcelona, Span. Also, snce March 2009, he has been an honorary professor wth the Department of Electrcal, Electronc, and Computer Engneerng at the Unversty of Pretora, South Afrca. He s the edtor-n-chef of the Computer Networks (COMNET) journal as well as the foundng edtor-n-chef of the Ad Hoc Networks journal and the Physcal Communcaton journal, all wth Elsever. Hs current research nterests are n cogntve rado networks, wreless sensor networks, and nanocommuncaton networks. He has receved numerous awards, ncludng the 1997 IEEE Leonard G. Abraham Prze Award (IEEE Communcatons Socety) for hs paper enttled Multmeda Group Synchronzaton Protocols for Integrated Servces Archtectures publshed n the IEEE Journal on Selected Areas n Communcatons n January 1996; the 2002 IEEE Harry M. Goode Memoral Award (IEEE Computer Socety) wth the ctaton for sgnfcant and poneerng contrbutons to advanced archtectures and protocols for wreless and satellte networkng ; the 2003 IEEE Best Tutoral Award (IEEE Communcaton Socety) for hs paper enttled A Survey on Sensor Networks, publshed n the IEEE Communcatons magazne n August 2002; the 2003 ACM Sgmoble Outstandng Contrbuton Award wth the ctaton for poneerng contrbutons n the area of moblty and resource management for wreless communcaton networks ; the 2004 Georga Tech Faculty Research Author Award for hs outstandng record of publcatons of papers between ; the 2005 Dstngushed Faculty Achevement Award from School of Electrcal and Computer Engneerng, Georga Tech; the 2009 Georga Tech Outstandng Doctoral Thess Advsor Award for hs 20+ years servce and dedcaton to Georga Tech and producng outstandng PhD students ; and the 2009 ECE Dstngushed Mentor Award from the School of Electrcal and Computer Engneerng, Georga Tech. He has been a fellow of the ACM snce 1996, and he s a fellow of the IEEE.. For more nformaton on ths or any other computng topc, please vst our Dgtal Lbrary at

WIRELESS spectrum is currently regulated by governmental

WIRELESS spectrum is currently regulated by governmental IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 11, NO. 4, APRIL 2012 529 Spectrum-Aware Moblty Management n Cogntve Rado Cellular Networks Won-Yeol Lee, Student Member, IEEE, and Ian F. Akyldz, Fellow, IEEE

More information

Calculation of the received voltage due to the radiation from multiple co-frequency sources

Calculation of the received voltage due to the radiation from multiple co-frequency sources Rec. ITU-R SM.1271-0 1 RECOMMENDATION ITU-R SM.1271-0 * EFFICIENT SPECTRUM UTILIZATION USING PROBABILISTIC METHODS Rec. ITU-R SM.1271 (1997) The ITU Radocommuncaton Assembly, consderng a) that communcatons

More information

Comparative Analysis of Reuse 1 and 3 in Cellular Network Based On SIR Distribution and Rate

Comparative Analysis of Reuse 1 and 3 in Cellular Network Based On SIR Distribution and Rate Comparatve Analyss of Reuse and 3 n ular Network Based On IR Dstrbuton and Rate Chandra Thapa M.Tech. II, DEC V College of Engneerng & Technology R.V.. Nagar, Chttoor-5727, A.P. Inda Emal: chandra2thapa@gmal.com

More information

Queuing-Based Dynamic Channel Selection for Heterogeneous Multimedia Applications over Cognitive Radio Networks

Queuing-Based Dynamic Channel Selection for Heterogeneous Multimedia Applications over Cognitive Radio Networks 1 Queung-Based Dynamc Channel Selecton for Heterogeneous ultmeda Applcatons over Cogntve Rado Networks Hsen-Po Shang and haela van der Schaar Department of Electrcal Engneerng (EE), Unversty of Calforna

More information

Resource Allocation Optimization for Device-to- Device Communication Underlaying Cellular Networks

Resource Allocation Optimization for Device-to- Device Communication Underlaying Cellular Networks Resource Allocaton Optmzaton for Devce-to- Devce Communcaton Underlayng Cellular Networks Bn Wang, L Chen, Xaohang Chen, Xn Zhang, and Dacheng Yang Wreless Theores and Technologes (WT&T) Bejng Unversty

More information

The Spectrum Sharing in Cognitive Radio Networks Based on Competitive Price Game

The Spectrum Sharing in Cognitive Radio Networks Based on Competitive Price Game 8 Y. B. LI, R. YAG, Y. LI, F. YE, THE SPECTRUM SHARIG I COGITIVE RADIO ETWORKS BASED O COMPETITIVE The Spectrum Sharng n Cogntve Rado etworks Based on Compettve Prce Game Y-bng LI, Ru YAG., Yun LI, Fang

More information

Performance Analysis of Multi User MIMO System with Block-Diagonalization Precoding Scheme

Performance Analysis of Multi User MIMO System with Block-Diagonalization Precoding Scheme Performance Analyss of Mult User MIMO System wth Block-Dagonalzaton Precodng Scheme Yoon Hyun m and Jn Young m, wanwoon Unversty, Department of Electroncs Convergence Engneerng, Wolgye-Dong, Nowon-Gu,

More information

The Impact of Spectrum Sensing Frequency and Packet- Loading Scheme on Multimedia Transmission over Cognitive Radio Networks

The Impact of Spectrum Sensing Frequency and Packet- Loading Scheme on Multimedia Transmission over Cognitive Radio Networks Ths artcle has been accepted for publcaton n a future ssue of ths journal, but has not been fully edted. Content may change pror to fnal publcaton. The Impact of Spectrum Sensng Frequency and Pacet- Loadng

More information

A MODIFIED DIRECTIONAL FREQUENCY REUSE PLAN BASED ON CHANNEL ALTERNATION AND ROTATION

A MODIFIED DIRECTIONAL FREQUENCY REUSE PLAN BASED ON CHANNEL ALTERNATION AND ROTATION A MODIFIED DIRECTIONAL FREQUENCY REUSE PLAN BASED ON CHANNEL ALTERNATION AND ROTATION Vncent A. Nguyen Peng-Jun Wan Ophr Freder Computer Scence Department Illnos Insttute of Technology Chcago, Illnos vnguyen@t.edu,

More information

Define Y = # of mobiles from M total mobiles that have an adequate link. Measure of average portion of mobiles allocated a link of adequate quality.

Define Y = # of mobiles from M total mobiles that have an adequate link. Measure of average portion of mobiles allocated a link of adequate quality. Wreless Communcatons Technologes 6::559 (Advanced Topcs n Communcatons) Lecture 5 (Aprl th ) and Lecture 6 (May st ) Instructor: Professor Narayan Mandayam Summarzed by: Steve Leung (leungs@ece.rutgers.edu)

More information

Distributed Uplink Scheduling in EV-DO Rev. A Networks

Distributed Uplink Scheduling in EV-DO Rev. A Networks Dstrbuted Uplnk Schedulng n EV-DO ev. A Networks Ashwn Srdharan (Sprnt Nextel) amesh Subbaraman, och Guérn (ESE, Unversty of Pennsylvana) Overvew of Problem Most modern wreless systems Delver hgh performance

More information

Dynamic Optimization. Assignment 1. Sasanka Nagavalli January 29, 2013 Robotics Institute Carnegie Mellon University

Dynamic Optimization. Assignment 1. Sasanka Nagavalli January 29, 2013 Robotics Institute Carnegie Mellon University Dynamc Optmzaton Assgnment 1 Sasanka Nagavall snagaval@andrew.cmu.edu 16-745 January 29, 213 Robotcs Insttute Carnege Mellon Unversty Table of Contents 1. Problem and Approach... 1 2. Optmzaton wthout

More information

Distributed Channel Allocation Algorithm with Power Control

Distributed Channel Allocation Algorithm with Power Control Dstrbuted Channel Allocaton Algorthm wth Power Control Shaoj N Helsnk Unversty of Technology, Insttute of Rado Communcatons, Communcatons Laboratory, Otakaar 5, 0150 Espoo, Fnland. E-mal: n@tltu.hut.f

More information

NATIONAL RADIO ASTRONOMY OBSERVATORY Green Bank, West Virginia SPECTRAL PROCESSOR MEMO NO. 25. MEMORANDUM February 13, 1985

NATIONAL RADIO ASTRONOMY OBSERVATORY Green Bank, West Virginia SPECTRAL PROCESSOR MEMO NO. 25. MEMORANDUM February 13, 1985 NATONAL RADO ASTRONOMY OBSERVATORY Green Bank, West Vrgna SPECTRAL PROCESSOR MEMO NO. 25 MEMORANDUM February 13, 1985 To: Spectral Processor Group From: R. Fsher Subj: Some Experments wth an nteger FFT

More information

Performance Evaluation of QoS Parameters in Dynamic Spectrum Sharing for Heterogeneous Wireless Communication Networks

Performance Evaluation of QoS Parameters in Dynamic Spectrum Sharing for Heterogeneous Wireless Communication Networks IJCSI Internatonal Journal of Computer Scence Issues, Vol. 9, Issue 1, No 2, January 2012 ISSN (Onlne): 1694-0814 www.ijcsi.org 81 Performance Evaluaton of QoS Parameters n Dynamc Spectrum Sharng for Heterogeneous

More information

To: Professor Avitabile Date: February 4, 2003 From: Mechanical Student Subject: Experiment #1 Numerical Methods Using Excel

To: Professor Avitabile Date: February 4, 2003 From: Mechanical Student Subject: Experiment #1 Numerical Methods Using Excel To: Professor Avtable Date: February 4, 3 From: Mechancal Student Subject:.3 Experment # Numercal Methods Usng Excel Introducton Mcrosoft Excel s a spreadsheet program that can be used for data analyss,

More information

Resource Control for Elastic Traffic in CDMA Networks

Resource Control for Elastic Traffic in CDMA Networks Resource Control for Elastc Traffc n CDMA Networks Vaslos A. Srs Insttute of Computer Scence, FORTH Crete, Greece vsrs@cs.forth.gr ACM MobCom 2002 Sep. 23-28, 2002, Atlanta, U.S.A. Funded n part by BTexact

More information

HUAWEI TECHNOLOGIES CO., LTD. Huawei Proprietary Page 1

HUAWEI TECHNOLOGIES CO., LTD. Huawei Proprietary Page 1 Project Ttle Date Submtted IEEE 802.16 Broadband Wreless Access Workng Group Double-Stage DL MU-MIMO Scheme 2008-05-05 Source(s) Yang Tang, Young Hoon Kwon, Yajun Kou, Shahab Sanaye,

More information

Selective Sensing and Transmission for Multi-Channel Cognitive Radio Networks

Selective Sensing and Transmission for Multi-Channel Cognitive Radio Networks IEEE INFOCOM 2 Workshop On Cogntve & Cooperatve Networks Selectve Sensng and Transmsson for Mult-Channel Cogntve Rado Networks You Xu, Yunzhou L, Yfe Zhao, Hongxng Zou and Athanasos V. Vaslakos Insttute

More information

Keywords LTE, Uplink, Power Control, Fractional Power Control.

Keywords LTE, Uplink, Power Control, Fractional Power Control. Volume 3, Issue 6, June 2013 ISSN: 2277 128X Internatonal Journal of Advanced Research n Computer Scence and Software Engneerng Research Paper Avalable onlne at: www.jarcsse.com Uplnk Power Control Schemes

More information

Throughput Maximization by Adaptive Threshold Adjustment for AMC Systems

Throughput Maximization by Adaptive Threshold Adjustment for AMC Systems APSIPA ASC 2011 X an Throughput Maxmzaton by Adaptve Threshold Adjustment for AMC Systems We-Shun Lao and Hsuan-Jung Su Graduate Insttute of Communcaton Engneerng Department of Electrcal Engneerng Natonal

More information

Research of Dispatching Method in Elevator Group Control System Based on Fuzzy Neural Network. Yufeng Dai a, Yun Du b

Research of Dispatching Method in Elevator Group Control System Based on Fuzzy Neural Network. Yufeng Dai a, Yun Du b 2nd Internatonal Conference on Computer Engneerng, Informaton Scence & Applcaton Technology (ICCIA 207) Research of Dspatchng Method n Elevator Group Control System Based on Fuzzy Neural Network Yufeng

More information

A Fuzzy-based Routing Strategy for Multihop Cognitive Radio Networks

A Fuzzy-based Routing Strategy for Multihop Cognitive Radio Networks 74 Internatonal Journal of Communcaton Networks and Informaton Securty (IJCNIS) Vol. 3, No., Aprl 0 A Fuzzy-based Routng Strategy for Multhop Cogntve Rado Networks Al El Masr, Naceur Malouch and Hcham

More information

Uplink User Selection Scheme for Multiuser MIMO Systems in a Multicell Environment

Uplink User Selection Scheme for Multiuser MIMO Systems in a Multicell Environment Uplnk User Selecton Scheme for Multuser MIMO Systems n a Multcell Envronment Byong Ok Lee School of Electrcal Engneerng and Computer Scence and INMC Seoul Natonal Unversty leebo@moble.snu.ac.kr Oh-Soon

More information

Cooperative Multicast Scheduling Scheme for IPTV Service over IEEE Networks

Cooperative Multicast Scheduling Scheme for IPTV Service over IEEE Networks Cooperatve Multcast Schedulng Scheme for IPTV Servce over IEEE 802.16 Networks Fen Hou 1, Ln X. Ca 1, James She 1, Pn-Han Ho 1, Xuemn (Sherman Shen 1, and Junshan Zhang 2 Unversty of Waterloo, Waterloo,

More information

Study of Downlink Radio Resource Allocation Scheme with Interference Coordination in LTE A Network

Study of Downlink Radio Resource Allocation Scheme with Interference Coordination in LTE A Network Internatonal Journal of Future Computer and Communcaton, Vol. 6, o. 3, September 2017 Study of Downln Rado Resource Allocaton Scheme wth Interference Coordnaton n LTE A etwor Yen-Wen Chen and Chen-Ju Chen

More information

Approximating User Distributions in WCDMA Networks Using 2-D Gaussian

Approximating User Distributions in WCDMA Networks Using 2-D Gaussian CCCT 05: INTERNATIONAL CONFERENCE ON COMPUTING, COMMUNICATIONS, AND CONTROL TECHNOLOGIES 1 Approxmatng User Dstrbutons n CDMA Networks Usng 2-D Gaussan Son NGUYEN and Robert AKL Department of Computer

More information

Priority based Dynamic Multiple Robot Path Planning

Priority based Dynamic Multiple Robot Path Planning 2nd Internatonal Conference on Autonomous obots and Agents Prorty based Dynamc Multple obot Path Plannng Abstract Taxong Zheng Department of Automaton Chongqng Unversty of Post and Telecommuncaton, Chna

More information

Enhanced Uplink Scheduling for Continuous Connectivity in High Speed Packet Access Systems

Enhanced Uplink Scheduling for Continuous Connectivity in High Speed Packet Access Systems Int. J. Communcatons, Network and System Scences, 212, 5, 446-453 http://dx.do.org/1.4236/jcns.212.5855 Publshed Onlne August 212 (http://www.scrp.org/journal/jcns) Enhanced Uplnk Schedulng for Contnuous

More information

EE360: Lecture 7 Outline Cellular System Capacity and ASE Announcements Summary due next week

EE360: Lecture 7 Outline Cellular System Capacity and ASE Announcements Summary due next week EE360: Lecture 7 Outlne Cellular System Capacty and ASE Announcements Summary due next week Capacty Area Spectral Effcency Dynamc Resource Allocaton Revew of Cellular Lecture Desgn consderatons: Spectral

More information

King s Research Portal

King s Research Portal Kng s Research Portal DOI: 10.1109/TWC.2015.2460254 Document Verson Peer revewed verson Lnk to publcaton record n Kng's Research Portal Ctaton for publshed verson (APA): Shrvanmoghaddam, M., L, Y., Dohler,

More information

Ergodic Capacity of Block-Fading Gaussian Broadcast and Multi-access Channels for Single-User-Selection and Constant-Power

Ergodic Capacity of Block-Fading Gaussian Broadcast and Multi-access Channels for Single-User-Selection and Constant-Power 7th European Sgnal Processng Conference EUSIPCO 29 Glasgow, Scotland, August 24-28, 29 Ergodc Capacty of Block-Fadng Gaussan Broadcast and Mult-access Channels for Sngle-User-Selecton and Constant-Power

More information

Rejection of PSK Interference in DS-SS/PSK System Using Adaptive Transversal Filter with Conditional Response Recalculation

Rejection of PSK Interference in DS-SS/PSK System Using Adaptive Transversal Filter with Conditional Response Recalculation SERBIAN JOURNAL OF ELECTRICAL ENGINEERING Vol., No., November 23, 3-9 Rejecton of PSK Interference n DS-SS/PSK System Usng Adaptve Transversal Flter wth Condtonal Response Recalculaton Zorca Nkolć, Bojan

More information

Adaptive Modulation for Multiple Antenna Channels

Adaptive Modulation for Multiple Antenna Channels Adaptve Modulaton for Multple Antenna Channels June Chul Roh and Bhaskar D. Rao Department of Electrcal and Computer Engneerng Unversty of Calforna, San Dego La Jolla, CA 993-7 E-mal: jroh@ece.ucsd.edu,

More information

Topology Control for C-RAN Architecture Based on Complex Network

Topology Control for C-RAN Architecture Based on Complex Network Topology Control for C-RAN Archtecture Based on Complex Network Zhanun Lu, Yung He, Yunpeng L, Zhaoy L, Ka Dng Chongqng key laboratory of moble communcatons technology Chongqng unversty of post and telecommuncaton

More information

QoS Provisioning in Wireless Data Networks under Non-Continuously Backlogged Users

QoS Provisioning in Wireless Data Networks under Non-Continuously Backlogged Users os Provsonng n Wreless Data Networks under Non-Contnuously Backlogged Users Tmotheos Kastrnoganns, and Symeon Papavasslou, Member, IEEE School of Electrcal and Computer Engneerng Natonal Techncal Unversty

More information

Fuzzy Logic Power Control in Cognitive Radio

Fuzzy Logic Power Control in Cognitive Radio Fuzzy Logc ower Control n Cogntve Rado Zeljko Tabakovc Croatan ost and Electronc Communcatons Agency Jursceva 3 Zagreb, Croata zeljkotabakovc@telekomhr Sonja Grgc, Mslav Grgc Unversty of Zagreb, Faculty

More information

Test 2. ECON3161, Game Theory. Tuesday, November 6 th

Test 2. ECON3161, Game Theory. Tuesday, November 6 th Test 2 ECON36, Game Theory Tuesday, November 6 th Drectons: Answer each queston completely. If you cannot determne the answer, explanng how you would arrve at the answer may earn you some ponts.. (20 ponts)

More information

A NSGA-II algorithm to solve a bi-objective optimization of the redundancy allocation problem for series-parallel systems

A NSGA-II algorithm to solve a bi-objective optimization of the redundancy allocation problem for series-parallel systems 0 nd Internatonal Conference on Industral Technology and Management (ICITM 0) IPCSIT vol. 49 (0) (0) IACSIT Press, Sngapore DOI: 0.776/IPCSIT.0.V49.8 A NSGA-II algorthm to solve a b-obectve optmzaton of

More information

A Novel Optimization of the Distance Source Routing (DSR) Protocol for the Mobile Ad Hoc Networks (MANET)

A Novel Optimization of the Distance Source Routing (DSR) Protocol for the Mobile Ad Hoc Networks (MANET) A Novel Optmzaton of the Dstance Source Routng (DSR) Protocol for the Moble Ad Hoc Networs (MANET) Syed S. Rzv 1, Majd A. Jafr, and Khaled Ellethy Computer Scence and Engneerng Department Unversty of Brdgeport

More information

A Novel DSA-Driven MAC Protocol for Cognitive Radio Networks

A Novel DSA-Driven MAC Protocol for Cognitive Radio Networks Wreless Sensor Networ, 29, 2, 6-2 do:.4236/wsn.29.2 7 Publshed Onlne July 29 (http://www.scrp.org/journal/wsn/). A Novel DSA-Drven MAC Protocol for Cogntve Rado Networs Hua SONG, Xaola LIN School of Informaton

More information

PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION. Evgeny Artyomov and Orly Yadid-Pecht

PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION. Evgeny Artyomov and Orly Yadid-Pecht 68 Internatonal Journal "Informaton Theores & Applcatons" Vol.11 PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION Evgeny Artyomov and Orly

More information

Opportunistic Beamforming for Finite Horizon Multicast

Opportunistic Beamforming for Finite Horizon Multicast Opportunstc Beamformng for Fnte Horzon Multcast Gek Hong Sm, Joerg Wdmer, and Balaj Rengarajan allyson.sm@mdea.org, joerg.wdmer@mdea.org, and balaj.rengarajan@gmal.com Insttute IMDEA Networks, Madrd, Span

More information

The Synthesis of Dependable Communication Networks for Automotive Systems

The Synthesis of Dependable Communication Networks for Automotive Systems 06AE-258 The Synthess of Dependable Communcaton Networks for Automotve Systems Copyrght 2005 SAE Internatonal Nagarajan Kandasamy Drexel Unversty, Phladelpha, USA Fad Aloul Amercan Unversty of Sharjah,

More information

RESOURCE CONTROL FOR HYBRID CODE AND TIME DIVISION SCHEDULING

RESOURCE CONTROL FOR HYBRID CODE AND TIME DIVISION SCHEDULING RESOURCE CONTROL FOR HYBRID CODE AND TIME DIVISION SCHEDULING Vaslos A. Srs Insttute of Computer Scence (ICS), FORTH and Department of Computer Scence, Unversty of Crete P.O. Box 385, GR 7 Heraklon, Crete,

More information

antenna antenna (4.139)

antenna antenna (4.139) .6.6 The Lmts of Usable Input Levels for LNAs The sgnal voltage level delvered to the nput of an LNA from the antenna may vary n a very wde nterval, from very weak sgnals comparable to the nose level,

More information

Improved Detection Performance of Cognitive Radio Networks in AWGN and Rayleigh Fading Environments

Improved Detection Performance of Cognitive Radio Networks in AWGN and Rayleigh Fading Environments Improved Detecton Performance of Cogntve Rado Networks n AWGN and Raylegh Fadng Envronments Yng Loong Lee 1, Wasan Kadhm Saad, Ayman Abd El-Saleh *1,, Mahamod Ismal 1 Faculty of Engneerng Multmeda Unversty

More information

Spectrum Sharing For Delay-Sensitive Applications With Continuing QoS Guarantees

Spectrum Sharing For Delay-Sensitive Applications With Continuing QoS Guarantees Spectrum Sharng For Delay-Senstve Applcatons Wth Contnung QoS Guarantees Yuanzhang Xao, Kartk Ahuja, and Mhaela van der Schaar Department of Electrcal Engneerng, UCLA Emals: yxao@seas.ucla.edu, ahujak@ucla.edu,

More information

Clustering Based Fractional Frequency Reuse and Fair Resource Allocation in Multi-cell Networks

Clustering Based Fractional Frequency Reuse and Fair Resource Allocation in Multi-cell Networks Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE ICC 21 proceedngs Clusterng Based Fractonal Frequency Reuse and Far Resource

More information

Joint Adaptive Modulation and Power Allocation in Cognitive Radio Networks

Joint Adaptive Modulation and Power Allocation in Cognitive Radio Networks I. J. Communcatons, etwork and System Scences, 8, 3, 7-83 Publshed Onlne August 8 n ScRes (http://www.scrp.org/journal/jcns/). Jont Adaptve Modulaton and Power Allocaton n Cogntve Rado etworks Dong LI,

More information

A study of turbo codes for multilevel modulations in Gaussian and mobile channels

A study of turbo codes for multilevel modulations in Gaussian and mobile channels A study of turbo codes for multlevel modulatons n Gaussan and moble channels Lamne Sylla and Paul Forter (sylla, forter)@gel.ulaval.ca Department of Electrcal and Computer Engneerng Laval Unversty, Ste-Foy,

More information

High Speed ADC Sampling Transients

High Speed ADC Sampling Transients Hgh Speed ADC Samplng Transents Doug Stuetzle Hgh speed analog to dgtal converters (ADCs) are, at the analog sgnal nterface, track and hold devces. As such, they nclude samplng capactors and samplng swtches.

More information

Traffic Modeling and Performance Evaluation in GSM/GPRS Networks

Traffic Modeling and Performance Evaluation in GSM/GPRS Networks Proceedngs of the 3th WSEAS Internatonal Conference on COMMUNICATIONS Traffc Modelng and Performance Evaluaton n GSM/ Networks Cornel Balnt, Georgeta Budura, Marza Eugen Poltehnca Unversty of Tmsoara Bd..

More information

A Spreading Sequence Allocation Procedure for MC-CDMA Transmission Systems

A Spreading Sequence Allocation Procedure for MC-CDMA Transmission Systems A Spreadng Sequence Allocaton Procedure for MC-CDMA Transmsson Systems Davd Motter, Damen Castelan Mtsubsh Electrc ITE 80, Avenue des Buttes de Coësmes, 35700 Rennes FRAE e-mal: {motter,castelan}@tcl.te.mee.com

More information

Robust Power and Subcarrier Allocation for OFDM-Based Cognitive Radio Networks Considering Spectrum Sensing Uncertainties

Robust Power and Subcarrier Allocation for OFDM-Based Cognitive Radio Networks Considering Spectrum Sensing Uncertainties 8 H. FATHI, S. M. S. SADOUGH, ROBUST POWER AD SUBCARRIER ALLOCATIO FOR OFDM-BASED COGITIVE RADIO... Robust Power and Subcarrer Allocaton for OFDM-Based Cogntve Rado etworks Consderng Spectrum Sensng Uncertantes

More information

Redes de Comunicação em Ambientes Industriais Aula 8

Redes de Comunicação em Ambientes Industriais Aula 8 Redes de Comuncação em Ambentes Industras Aula 8 Luís Almeda lda@det.ua.pt Electronc Systems Lab-IEETA / DET Unversdade de Avero Avero, Portugal RCAI 2005/2006 1 In the prevous epsode... Cooperaton models:

More information

Decision aid methodologies in transportation

Decision aid methodologies in transportation Decson ad methodologes n transportaton Lecture 7: More Applcatons Prem Kumar prem.vswanathan@epfl.ch Transport and Moblty Laboratory Summary We learnt about the dfferent schedulng models We also learnt

More information

Energy Efficiency Analysis of a Multichannel Wireless Access Protocol

Energy Efficiency Analysis of a Multichannel Wireless Access Protocol Energy Effcency Analyss of a Multchannel Wreless Access Protocol A. Chockalngam y, Wepng u, Mchele Zorz, and Laurence B. Mlsten Department of Electrcal and Computer Engneerng, Unversty of Calforna, San

More information

NOVEL ITERATIVE TECHNIQUES FOR RADAR TARGET DISCRIMINATION

NOVEL ITERATIVE TECHNIQUES FOR RADAR TARGET DISCRIMINATION NOVEL ITERATIVE TECHNIQUES FOR RADAR TARGET DISCRIMINATION Phaneendra R.Venkata, Nathan A. Goodman Department of Electrcal and Computer Engneerng, Unversty of Arzona, 30 E. Speedway Blvd, Tucson, Arzona

More information

Power Minimization Under Constant Throughput Constraint in Wireless Networks with Beamforming

Power Minimization Under Constant Throughput Constraint in Wireless Networks with Beamforming Power Mnmzaton Under Constant Throughput Constrant n Wreless etworks wth Beamformng Zhu Han and K.J. Ray Lu, Electrcal and Computer Engneer Department, Unversty of Maryland, College Park. Abstract In mult-access

More information

Traffic balancing over licensed and unlicensed bands in heterogeneous networks

Traffic balancing over licensed and unlicensed bands in heterogeneous networks Correspondence letter Traffc balancng over lcensed and unlcensed bands n heterogeneous networks LI Zhen, CUI Qme, CUI Zhyan, ZHENG We Natonal Engneerng Laboratory for Moble Network Securty, Bejng Unversty

More information

Parameter Free Iterative Decoding Metrics for Non-Coherent Orthogonal Modulation

Parameter Free Iterative Decoding Metrics for Non-Coherent Orthogonal Modulation 1 Parameter Free Iteratve Decodng Metrcs for Non-Coherent Orthogonal Modulaton Albert Gullén Fàbregas and Alex Grant Abstract We study decoder metrcs suted for teratve decodng of non-coherently detected

More information

Performance Study of OFDMA vs. OFDM/SDMA

Performance Study of OFDMA vs. OFDM/SDMA Performance Study of OFDA vs. OFD/SDA Zhua Guo and Wenwu Zhu crosoft Research, Asa 3F, Beng Sgma Center, No. 49, Zhchun Road adan Dstrct, Beng 00080, P. R. Chna {zhguo, wwzhu}@mcrosoft.com Abstract: In

More information

Impact of Secondary MAC Cooperation on Spectrum Sharing in Cognitive Radio Networks

Impact of Secondary MAC Cooperation on Spectrum Sharing in Cognitive Radio Networks Impact of Secondary MAC Cooperaton on Spectrum Sharng n Cogntve ado Networks Tarq Elkourd and Osvaldo Smeone CWCSP, ECE Dept. New Jersey Insttute of Technology Unversty Heghts, Newark, New Jersey 0702

More information

Optimal State Prediction for Feedback-Based QoS Adaptations

Optimal State Prediction for Feedback-Based QoS Adaptations Optmal State Predcton for Feedback-Based QoS Adaptatons Baochun L, Dongyan Xu, Klara Nahrstedt Department of Computer Scence Unversty of Illnos at Urbana-Champagn b-l, d-xu, klara @cs.uuc.edu Abstract

More information

AN ALGORITHM TO COMBINE LINK ADAPTATION AND TRANSMIT POWER CONTROL IN HIPERLAN TYPE 2

AN ALGORITHM TO COMBINE LINK ADAPTATION AND TRANSMIT POWER CONTROL IN HIPERLAN TYPE 2 AN ALGORITHM TO COMBINE LINK ADAPTATION AND TRANSMIT POWER CONTROL IN HIPERLAN TYPE 2 Markus Radmrsch Inst. f. Allgem. Nachrchtentechnk, Unv. Hannover, Appelstr. 9a, 3167 Hannover, Germany Tel.: +49-511-762

More information

AN IMPROVED BIT LOADING TECHNIQUE FOR ENHANCED ENERGY EFFICIENCY IN NEXT GENERATION VOICE/VIDEO APPLICATIONS

AN IMPROVED BIT LOADING TECHNIQUE FOR ENHANCED ENERGY EFFICIENCY IN NEXT GENERATION VOICE/VIDEO APPLICATIONS Journal of Engneerng Scence and Technology Vol., o. 4 (6) 476-495 School of Engneerng, Taylor s Unversty A IMPROVED BIT LOADIG TECHIQUE FOR EHACED EERGY EFFICIECY I EXT GEERATIO VOICE/VIDEO APPLICATIOS

More information

Low Complexity Duty Cycle Control with Joint Delay and Energy Efficiency for Beacon-enabled IEEE Wireless Sensor Networks

Low Complexity Duty Cycle Control with Joint Delay and Energy Efficiency for Beacon-enabled IEEE Wireless Sensor Networks Low Complexty Duty Cycle Control wth Jont Delay and Energy Effcency for Beacon-enabled IEEE 8254 Wreless Sensor Networks Yun L Kok Keong Cha Yue Chen Jonathan Loo School of Electronc Engneerng and Computer

More information

Control Chart. Control Chart - history. Process in control. Developed in 1920 s. By Dr. Walter A. Shewhart

Control Chart. Control Chart - history. Process in control. Developed in 1920 s. By Dr. Walter A. Shewhart Control Chart - hstory Control Chart Developed n 920 s By Dr. Walter A. Shewhart 2 Process n control A phenomenon s sad to be controlled when, through the use of past experence, we can predct, at least

More information

Multiagent Jamming-Resilient Control Channel Game for Cognitive Radio Ad Hoc Networks

Multiagent Jamming-Resilient Control Channel Game for Cognitive Radio Ad Hoc Networks Multagent Jammng-Reslent Control Channel Game for Cogntve Rado Ad Hoc Networks Brandon F. Lo and Ian F. Akyldz Broadband Wreless Networkng Laboratory, School of Electrcal and Computer Engneerng Georga

More information

Exploiting Dynamic Workload Variation in Low Energy Preemptive Task Scheduling

Exploiting Dynamic Workload Variation in Low Energy Preemptive Task Scheduling Explotng Dynamc Worload Varaton n Low Energy Preemptve Tas Schedulng Lap-Fa Leung, Ch-Yng Tsu Department of Electrcal and Electronc Engneerng Hong Kong Unversty of Scence and Technology Clear Water Bay,

More information

Subcarrier allocation for OFDMA wireless channels using lagrangian relaxation methods

Subcarrier allocation for OFDMA wireless channels using lagrangian relaxation methods Unversty of Wollongong Research Onlne Faculty of Informatcs - Papers (Archve) Faculty of Engneerng and Informaton Scences 2006 Subcarrer allocaton for OFDMA wreless channels usng lagrangan relaxaton methods

More information

Fair Scheduling for Real-Time Multimedia Support in IEEE Wireless Access Networks

Fair Scheduling for Real-Time Multimedia Support in IEEE Wireless Access Networks Far Schedulng for Real-Tme Multmeda Support n IEEE 802.16 Wreless Access Networs Yaser P. Fallah Inst. of Transportaton Studes (Depts EECS & CEE), Unversty of Calforna at Bereley, USA e-mal: yaserpf@eecs.bereley.edu

More information

Multiband Jamming Strategies with Minimum Rate Constraints

Multiband Jamming Strategies with Minimum Rate Constraints Multband Jammng Strateges wth Mnmum Rate Constrants Karm Banawan, Sennur Ulukus, Peng Wang, and Bran Henz Department of Electrcal and Computer Engneerng, Unversty of Maryland, College Park, MD 7 US Army

More information

Distributed Resource Allocation and Scheduling in OFDMA Wireless Networks

Distributed Resource Allocation and Scheduling in OFDMA Wireless Networks Southern Illnos Unversty Carbondale OpenSIUC Conference Proceedngs Department of Electrcal and Computer Engneerng 11-2006 Dstrbuted Resource Allocaton and Schedulng n OFDMA Wreless Networks Xangpng Qn

More information

A Mathematical Model for Restoration Problem in Smart Grids Incorporating Load Shedding Concept

A Mathematical Model for Restoration Problem in Smart Grids Incorporating Load Shedding Concept J. Appl. Envron. Bol. Sc., 5(1)20-27, 2015 2015, TextRoad Publcaton ISSN: 2090-4274 Journal of Appled Envronmental and Bologcal Scences www.textroad.com A Mathematcal Model for Restoraton Problem n Smart

More information

The Application of Interpolation Algorithms in OFDM Channel Estimation

The Application of Interpolation Algorithms in OFDM Channel Estimation The Applcaton of Interpolaton Algorthms n OFDM Estmaton Xjun ZHANG 1,, Zhantng YUAN 1, 1 School of Electrcal and Informaton Engneerng, Lanzhou Unversty of Technology, Lanzhou, Gansu 730050, Chna School

More information

PERFORMANCE EVALUATION OF A SPECTRUM-SENSING TECHNIQUE FOR LDACS AND JTIDS COEXISTENCE IN L-BAND

PERFORMANCE EVALUATION OF A SPECTRUM-SENSING TECHNIQUE FOR LDACS AND JTIDS COEXISTENCE IN L-BAND Proceedngs of SDR'1-WInnComm-Europe, 7-9 June 01 PERFORMANCE EVALUATION OF A SPECTRUM-SENSING TECHNIQUE FOR LDACS AND JTIDS COEXISTENCE IN L-BAND Gulo Bartol (Unverstà degl Stud d Frenze, Florence, Italy;

More information

Optimal Placement of PMU and RTU by Hybrid Genetic Algorithm and Simulated Annealing for Multiarea Power System State Estimation

Optimal Placement of PMU and RTU by Hybrid Genetic Algorithm and Simulated Annealing for Multiarea Power System State Estimation T. Kerdchuen and W. Ongsakul / GMSARN Internatonal Journal (09) - Optmal Placement of and by Hybrd Genetc Algorthm and Smulated Annealng for Multarea Power System State Estmaton Thawatch Kerdchuen and

More information

QoS-Aware Parallel Sensing/Probing Architecture and Adaptive Cross-Layer Protocol Design for Opportunistic Networks

QoS-Aware Parallel Sensing/Probing Architecture and Adaptive Cross-Layer Protocol Design for Opportunistic Networks IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 1 QoS-Aware Parallel Sensng/Probng Archtecture and Adaptve Cross-Layer Protocol Desgn for Opportunstc Networks Mohammad J. Abdel-Rahman, Member, IEEE, HarshKumarShankar,andMarwanKrunz,Fellow,

More information

Medium Access Control for Multi-Channel Parallel Transmission in Cognitive Radio Networks

Medium Access Control for Multi-Channel Parallel Transmission in Cognitive Radio Networks Medum ccess Control for Mult-Channel Parallel Transmsson n Cogntve Rado Networs Tao Shu, Shuguang Cu, and Marwan Krunz Department of Electrcal and Computer Engneerng Unversty of rzona Tucson, Z 85721 {tshu,

More information

Efficient Large Integers Arithmetic by Adopting Squaring and Complement Recoding Techniques

Efficient Large Integers Arithmetic by Adopting Squaring and Complement Recoding Techniques The th Worshop on Combnatoral Mathematcs and Computaton Theory Effcent Large Integers Arthmetc by Adoptng Squarng and Complement Recodng Technques Cha-Long Wu*, Der-Chyuan Lou, and Te-Jen Chang *Department

More information

Network Reconfiguration in Distribution Systems Using a Modified TS Algorithm

Network Reconfiguration in Distribution Systems Using a Modified TS Algorithm Network Reconfguraton n Dstrbuton Systems Usng a Modfed TS Algorthm ZHANG DONG,FU ZHENGCAI,ZHANG LIUCHUN,SONG ZHENGQIANG School of Electroncs, Informaton and Electrcal Engneerng Shangha Jaotong Unversty

More information

Performance Analysis of Scheduling Policies for Delay-Tolerant Applications in Centralized Wireless Networks

Performance Analysis of Scheduling Policies for Delay-Tolerant Applications in Centralized Wireless Networks Performance Analyss of Schedulng Polces for Delay-Tolerant Applcatons n Centralzed Wreless Networks Mohamed Shaqfeh and Norbert Goertz Insttute for Dgtal Communcatons Jont Research Insttute for Sgnal &

More information

An Adaptive Over-current Protection Scheme for MV Distribution Networks Including DG

An Adaptive Over-current Protection Scheme for MV Distribution Networks Including DG An Adaptve Over-current Protecton Scheme for MV Dstrbuton Networks Includng DG S.A.M. Javadan Islamc Azad Unversty s.a.m.javadan@gmal.com M.-R. Haghfam Tarbat Modares Unversty haghfam@modares.ac.r P. Barazandeh

More information

An Efficient Scheduling For Diverse QoS Requirements in WiMAX

An Efficient Scheduling For Diverse QoS Requirements in WiMAX An Effcent Schedulng For Dverse QoS Requrements n WMAX by Xaojng Meng A thess presented to the Unversty of Waterloo n fulfllment of the thess requrement for the degree of Master of Appled Scence n Electrcal

More information

Uncertainty in measurements of power and energy on power networks

Uncertainty in measurements of power and energy on power networks Uncertanty n measurements of power and energy on power networks E. Manov, N. Kolev Department of Measurement and Instrumentaton, Techncal Unversty Sofa, bul. Klment Ohrdsk No8, bl., 000 Sofa, Bulgara Tel./fax:

More information

Impact of Interference Model on Capacity in CDMA Cellular Networks. Robert Akl, D.Sc. Asad Parvez University of North Texas

Impact of Interference Model on Capacity in CDMA Cellular Networks. Robert Akl, D.Sc. Asad Parvez University of North Texas Impact of Interference Model on Capacty n CDMA Cellular Networks Robert Akl, D.Sc. Asad Parvez Unversty of North Texas Outlne Introducton to CDMA networks Average nterference model Actual nterference model

More information

High Speed, Low Power And Area Efficient Carry-Select Adder

High Speed, Low Power And Area Efficient Carry-Select Adder Internatonal Journal of Scence, Engneerng and Technology Research (IJSETR), Volume 5, Issue 3, March 2016 Hgh Speed, Low Power And Area Effcent Carry-Select Adder Nelant Harsh M.tech.VLSI Desgn Electroncs

More information

Full-duplex Relaying for D2D Communication in mmwave based 5G Networks

Full-duplex Relaying for D2D Communication in mmwave based 5G Networks Full-duplex Relayng for D2D Communcaton n mmwave based 5G Networks Boang Ma Hamed Shah-Mansour Member IEEE and Vncent W.S. Wong Fellow IEEE Abstract Devce-to-devce D2D communcaton whch can offload data

More information

Decomposition Principles and Online Learning in Cross-Layer Optimization for Delay-Sensitive Applications

Decomposition Principles and Online Learning in Cross-Layer Optimization for Delay-Sensitive Applications Techncal Report Decomposton Prncples and Onlne Learnng n Cross-Layer Optmzaton for Delay-Senstve Applcatons Abstract In ths report, we propose a general cross-layer optmzaton framework n whch we explctly

More information

A Benchmark for D2D in Cellular Networks: The Importance of Information

A Benchmark for D2D in Cellular Networks: The Importance of Information A Benchmark for D2D n Cellular Networks: The Importance of Informaton Yğt Özcan, Catherne Rosenberg Unversty of Waterloo {yozcan,cath}@uwaterloo.ca Fabrce Gullemn Orange Labs, France fabrce.gullemn@orange.com

More information

Joint Subcarrier and CPU Time Allocation for Mobile Edge Computing

Joint Subcarrier and CPU Time Allocation for Mobile Edge Computing Jont Subcarrer and CPU Tme Allocaton for Moble Edge Computng Ynghao Yu, Jun Zhang, and Khaled B. Letaef, Fellow, IEEE Dept. of ECE, The Hong Kong Unversty of Scence and Technology Hamad Bn Khalfa Unversty,

More information

REAL-TIME SCHEDULING IN LTE FOR SMART GRIDS. Yuzhe Xu, Carlo Fischione

REAL-TIME SCHEDULING IN LTE FOR SMART GRIDS. Yuzhe Xu, Carlo Fischione REAL-TIME SCHEDULING IN LTE FOR SMART GRIDS Yuzhe Xu, Carlo Fschone Automatc Control Lab KTH, Royal Insttue of Technology 1-44, Stockholm, Sweden Emal: yuzhe@kth.se, carlof@kth.se ABSTRACT The latest wreless

More information

Weighted Penalty Model for Content Balancing in CATS

Weighted Penalty Model for Content Balancing in CATS Weghted Penalty Model for Content Balancng n CATS Chngwe Davd Shn Yuehme Chen Walter Denny Way Len Swanson Aprl 2009 Usng assessment and research to promote learnng WPM for CAT Content Balancng 2 Abstract

More information

Joint Power Control and Scheduling for Two-Cell Energy Efficient Broadcasting with Network Coding

Joint Power Control and Scheduling for Two-Cell Energy Efficient Broadcasting with Network Coding Communcatons and Network, 2013, 5, 312-318 http://dx.do.org/10.4236/cn.2013.53b2058 Publshed Onlne September 2013 (http://www.scrp.org/journal/cn) Jont Power Control and Schedulng for Two-Cell Energy Effcent

More information

Enhancing Throughput in Wireless Multi-Hop Network with Multiple Packet Reception

Enhancing Throughput in Wireless Multi-Hop Network with Multiple Packet Reception Enhancng Throughput n Wreless Mult-Hop Network wth Multple Packet Recepton Ja-lang Lu, Paulne Vandenhove, We Shu, Mn-You Wu Dept. of Computer Scence & Engneerng, Shangha JaoTong Unversty, Shangha, Chna

More information

Combined Beamforming and Scheduling for High Speed Downlink Packet Access

Combined Beamforming and Scheduling for High Speed Downlink Packet Access Combned Beamformng and Schedulng for Hgh Speed Downlnk Packet Access Alexander Seeger Informaton and Communcaton Moble Semens A.G. Hofmannstr.5, 8359 Munch, Germany Alexander.Seeger@semens.com Marcn Skora

More information

Dynamic Resource Allocation Algorithm of UAS by Network Environment and Data Requirement

Dynamic Resource Allocation Algorithm of UAS by Network Environment and Data Requirement Dynamc Resource Allocaton Algorthm of UAS by Network Envronment and Data Requrement Hye-Rm Cheon, Jun-Woo Cho, and Jae-Hyun Km Department of Electrcal and Computer Engneerng Ajou Unversty Suwon, Republc

More information

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 13, NO. 12, DECEMBER

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 13, NO. 12, DECEMBER IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 3, NO. 2, DECEMBER 204 695 On Spatal Capacty of Wreless Ad Hoc Networks wth Threshold Based Schedulng Yue Lng Che, Student Member, IEEE, Ru Zhang, Member,

More information