AMC-aware QoS proposal for OFDMA-based IEEE WiMAX systems

Transcription

1 AMC-aware QoS proposal for OFDMA-based IEEE82.6 WMAX systems Chad Tarhn, Tjan Chahed GET/Insttut Natonal des Télécommuncatons/UMR CNRS rue C. Fourer - 9 Evry CEDEX - France {chad.tarhn, tjan.chahed}@nt-evry.fr Abstract OFDMA-based IEEE82.6 mplements Adaptve Modulaton and Codng (AMC) whch results n a dfferent bt rate for each user dependng on ts poston n the cell as well as the rado condton t experences; users away from the base staton experence lower throughput. We hence propose n ths paper a new QoS soluton that compensates the degradaton n modulaton by a hgher number of sub-carrers so as to mantan the bt rate of streamng flows at a constant level throughout the whole coverage area. Ths proposal s modeled analytcally n a dynamc user confguraton, where users of both types, streamng and elastc, reman n the system for a fnte duraton. And ths n mult-cell envronment and for dfferent frequency reuse schemes. Index Terms QoS, capacty, OFDMA, IEEE82.6, WMAX, AMC, Reuse parttonng I. INTRODUCTION WMAX systems are based on versons d and e of the IEEE82.6 standard whch defnes a physcal (PHY) and medum access control (MAC) layers for broadband wreless access systems operatng at frequences below GHz. The frst of these standards, publshed n 24, addresses fxed servces whle the second, publshed n 25, s ntended for moble servces. The IEEE 82.6e specfcatons defne three dfferent PHY layers: sngle-carrer (SC) transmsson, Orthogonal Frequency-Dvson Multplexng (OFDM), and OFD Multple Access (OFDMA). The multple access technque used n the frst two of these PHY specfcatons s pure TDMA, whle the thrd technque uses both the tme and frequency dmensons for resource allocaton. From these three PHY technologes, OFDMA has been selected by WMAX Forum [] as the basc technology for portable and moble applcatons. In wreless communcaton systems, random fluctuatons prevent the contnuous use of hghly bandwdth-effcent modulaton and therefore Adaptve Modulaton and Codng (AMC) has become a standard approach n recently developed wreless standards, ncludng WMAX. The dea behnd AMC s to dynamcally adapt the modulaton and codng scheme to the channel condtons so as to acheve the hghest spectral effcency at all tmes [2]. Adaptve modulaton changes the codng scheme and/or modulaton method dependng on channelstate nformaton - choosng t n such a way that t squeezes the most out of what the channel can transmt. In OFDMA, modulaton and/or codng can be chosen dfferently for each sub-carrer, and t can also change wth tme. Indeed, n the IEEE82.6 standard, coherent modulaton schemes are used startng from low effcency modulatons (BPSK wth codng rate /2) to very hgh effcency ones (64-QAM wth codng rate 3/4) dependng on the Sgnal-to-Nose Rato (SNR). Ths results n lower throughput for users further away from the base staton and/or wth worse rado condtons []. Ths may be acceptable for elastc traffc, such as data, but not for streamng, such as voce. For IEEE82.6 to be a QoS-capable multmeda system, t should guarantee streamng applcatons a constant bt rate ndependent of the user poston n the cell as well as rado condtons. To do so, we propose n ths work a new QoS mechansm wheren the degradaton n modulaton for streamng flows s countered by use of a hgher number of sub-carrers. Moreover, n a mult-cell OFDMA system, adjacent cells usng sub-carrers of exactly the same frequency and tme can cause nterference to one another unless nter-cell nterference mtgaton technques are appled. Ths nterference takes the form of collsons, the number of whch ncreases as moble statons get closer to the edge of the cell. To combat ths, a reuse parttonng [3] scheme can be appled and controls the reuse factor n dfferent parts of the cell. II. MODELING PHYSICAL LAYER A. OFDMA sub-carrer allocatons OFDMA s a multple access technque whch dvdes the total Fast Fourer Transform (FFT) space nto a number of sub-channels (set of sub-carrers that are assgned for data exchange) whereas the tme resource s dvded nto tme slots and a frame s constructed by a number of slots. Let N denote the sze of FFT and N the total number of data sub-carrers after reservng the plot and guard sub-carrers whch we dvde nto L groups, each wth K = N /L data sub-carrers. In OFDMA-based WMAX system, resource allocaton s done n the tme-frequency doman: a call may share a subchannel wth other users. The user devce could choose subchannels based on geographcal locaton wth the potental of elmnatng the mpact of deep fades. B. AMC and cell decomposton In ths work, and wthout loss of generalty, we study AMC n the presence of path loss only whch we characterze by a certan value ξ; hgh effcency modulaton s used for users where ξ ξ, correspondng to a large SNR. Ths results 478 Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE GLOBECOM 27 proceedngs. Authorzed lcensed use lmted to: INT-Evry. Downloaded on October 6, 28 at 9:3 from IEEE Xplore. Restrctons apply.

2 n the dvson of the cell nto r regons, =...r, whch we may consder to be concentrc crcles of radus R for smplcty, but mght be of dfferent topology f we take nto account other phenomena, such as fast-fadng. In each regon, users have the same modulaton scheme and experence thus a correspondng bt rate whch decreases as users get further from the base staton. To calculate the area covered by each modulaton scheme, we must know the maxmal dstance between Base Staton (BS) and users usng a correspondng modulaton. Ths dstance s determned usng the maxmal SNR a user should receve wthout data loss. In [5], swtchng ponts between modulaton/codng schemes are proposed dependng on receved SNR and allow us to calculate the maxmal dstance a user should have. The path loss for the free space model s gven by [8]: PL[dB] = log[g E G R ( λ 4πR )2 ] = logg E logg R +2log( 4πR λ ) () where G E s the emtter antenna gan, G R s the recever antenna gan, R s the dstance between the emtter and the recever and λ s the wavelength. Ths path loss s also equal to: PL[dB] = P E [dbm] SNR[dB] N[dBm] (2) where P E s the emtted power and N s the thermal nose whch s equal to: N[dBm] =log(τtw) (3) τ = watt/k Hz s the Boltzmann constant, T s the temperature n Kelvn (T = 29) and W s the transmsson bandwdth n Hz. Usng the above equatons, we can calculate the relatonshp between the dstance and the SNR as follows: R = λ PE [dbm]+log(ge )[db]+log(gr)[db] SNR[dB] N[dBm] 2 4π (4) The area of each regon S s gven by: S = π (R 2 R 2 ) where R =. For the sake of llustraton, let us consder the followng example based on the lcensed band for WMAX to outdoor use n France whch starts at a frequency of 3.4GHz and whch has system bandwdth equal to 2MHz. At ths bandwdth, the thermal nose s equal to.97dbm. Accordng to the maxmum allowed Effectve Isotropc Radated Power (EIRP) of W, where the emtters are assumed to have an emsson power P E of W for users. We consder wthout loss of generalty the case of antennas n BS and user equpment wthout gan,.e., G E = G R =. In Fgure, we represent the dstance assgned to SNR for swtchng ponts. The proporton of each surface area per PHY assumpton s determned and shownntablei. Recever SNR[dB] Fg BPSK /2 QPSK 3/4 6 QAM /2 6 QAM 3/4 64 QAM 2/3 64 QAM 3/ cell radus [m] Receved SNR functon of the dstance Modulaton Codng rate Recever SNR(dB) Surface [%] BPSK / / QAM / / QAM 2/ / TABLE I IEEE82.6 PHY ASSUMPTIONS C. Throughput of the cell The nstantaneous physcal bt rate elastc users n regon S s gven by: R s,e of streamng or R s,e = Ls,e K C log 2 (M) ( BLER) T s S c (5) = L s,e K B E ( BLER) where L s,e s the number of sub-channels to be assgned to streamng/elastc users n regon S, K s the number of data sub-carrers assgned to each sub-channel, C s the codng rate of the M-ary modulaton, T s s the OFDMA symbol duraton gven by: T s = T b + T g wth T b the useful symbol perod (n unts of mcroseconds) N gven by W n and T g the guard perod equal to G T b, W s the bandwdth (MHz), n s the samplng factor, G s the rato of cyclc prefx (CP) to useful tme, S c s the sector coeffcent, B s the baud rate (symbols/sec), E s the effcency of the modulaton (bts/symbol) n each regon S and BLER s the perceved Block Error Rate. Although streamng flows wll have a dfferent number of sub-channels L s per regon, ths amounts to one streamng class only wth rate R s. As of (TCP-based) elastc calls, snce they tolerate reducton n ther throughput, we shall not do anythng : they wll smply share the left over capacty. On Note that for each SINR value, we can determne a couple of values (E,BLER) and these values are determned by lnk level curves E = f(sinr) and BLER = g(sinr) 478 Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE GLOBECOM 27 proceedngs. Authorzed lcensed use lmted to: INT-Evry. Downloaded on October 6, 28 at 9:3 from IEEE Xplore. Restrctons apply.

3 the bass of Processor Sharng (PS) [9]. In dong so, we are actually gvng prorty to streamng calls over data ones. The former are subject to admsson control as we cannot accept more users than the (dedcated) resources, sub-carrers n ths case, assgned to them. The number of sub-channels L e allocated to an elastc call n regon S s thus gven by: L e = L r = Ls ns r = ne (6) x ndcates the largest nteger that s less than or equal to x. n s and ne are the number of streamng and elastc flows n regon S. In total, we obtan one (hgh prorty) class for streamng applcatons wth constant bt rate R s and r best-effort classes for elastc ones correspondng to the r regons of our system. III. MODELING AT THE PHYSICAL LAYER We assume that streamng calls arrve to regon S accordng to a Posson process wth ntensty λ s and use l s subcarrers for an exponentally dstrbuted tme wth mean /µ s ndependent of the share of the resources they get. We also assume that elastc flows arrve to the system accordng to a Posson process wth ntensty λ e and assume for tractablty that ther servce rate s exponentally dstrbuted wth mean µ e = Re E[Z] where E[Z] s the mean fle sze 2. We now model our new, QoS-capable system as a Contnuous Tme Markov Chan (CTMC) by takng nto account the proposed prortes for the ntegraton of streamng and elastc flows as well as the way they share resources. The state s characterzed by the followng row vector: n := (n s,n s 2,..., n s r,n e,n e 2,..., n e r) where n s and n e, for =...r, represent the number of streamng and elastc calls n regon S, respectvely. The state space of the system s gven by: S := { r n N 2r (l s n s + l e n e ) N} (7) = where l s and l e denote the number of sub-carrers allocated to streamng and elastc calls n regon S respectvely and N s the maxmum number of sub-carrers n the cell. A. Analyss We now determne the steady-state probablty vector bdπ = {π( n ) n S }. Note that the correspondng system s non homogeneous as the departure rate of elastc calls depends on the overall number of calls n the system whereas streamng calls do not. The soluton of the steady-state dstrbuton s obtaned by solvng the set of lnearly ndependent equatons gven by: { π Q = n S π( (8) n )= 2 In fact, the total length of an elastc flow n unts of packets s found to follow a log normal dstrbuton, accordng to the measurement-based modellng [7] To construct the transton matrx Q, we must consder all possble transtons between neghborng states. Let q( n n ) denote the transton probablty from state n to neghborng states n. Note that when we accept a new call n regon S, r the state s noted by n s,e and when a call + termnates the servce the next state s n s,e. We thus have the followng transton rates: q( n n s )=λ s + q( n n s )=n s µs q( n (9) n e )=λ e + q( n n e )=n e µe ( n ) and the values q( n n ) must be obtaned as the sum of all terms n each lne n matrx Q s equal to zero for r. B. Performance measures Based on the steady-state probabltes of our CTMC, we now determne the performance measures relatve to our model. The call blockng probabltes of both types of flows, B s and B e n regon S, are obtaned by summng up the steady state probablty of saturaton states: B s,e = π( n ) () n S s,e where S s,e s the subset of states n S for whch any new call of class- s blocked when arrvng to the system due to lack of resources. Formally, S s,e := { r n S (lkn s s k + lkn e e k)+l s,e >N} k= The mean transfer tme of class- elastc flows can be calculated from the Lttle s formula by: n T e = λ e ( Be ) () where n e s the mean number of elastc flows n regon S. IV. REUSE PARTITIONING IN OFDMA WIMAX In wreless communcatons systems, users who are close to the base staton (BS) experence a hgher sgnal qualty compared to those who are located near the cell edge. Users wt hgh sgnal qualty can tolerate a hgher level of nterference, whch makes a denser reuse of channels possble. Reuse parttonng has been proposed n [6] as a means to ncrease the total system capacty. A. Basc concept Under the reuse parttonng allocaton polcy, the coverage area of a sngle cell s dvded nto several regons, each regon havng ts own reuse factor. Consderng two regons only, ths amounts to the fact that users n the nner regon have access to the total frequency band L whereas those n the outer regon are restrcted to only L sub-channels n the target cell and L sub-channels n the neghborng cell, wth L <L and L <Lthe least overlappng. In dong so, users at the edge shall experence a lower nterference fgure n terms of collsons, as quantfed next Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE GLOBECOM 27 proceedngs. Authorzed lcensed use lmted to: INT-Evry. Downloaded on October 6, 28 at 9:3 from IEEE Xplore. Restrctons apply.

4 B. Collsons There are several approaches as how to dstrbute avalable sub-carrers to sub-channels. The smplest one conssts of pckng randomly sub-carrers from avalable ones such that any avalable sub-carrer has the same probablty to get allocated to an arrvng user. We refer to ths method as random allocaton of sub-carrers whch does not requre any coordnaton between cells. In the case of two cells, the followng lemma establshes the mean number of collsons when the number of occuped sub-carrers s (L K) and (L K2) n cell and cell 2 respectvely. Lemma : The mean number of collsons n L subchannels s gven by: E[C K,K2] = L E [C K,K2] (2) where E [C K,K2] s the mean number of collsons n one group of sub-carrers and s n turn gven by: c max E [C K,K2] = c Pr(c K,K2) (3) c=c mn Pr(c K,K2) beng the probablty of havng c collsons n ths group and s equal to: K c ( N/L )( N/L c )( N/L c nc ) nc= c nc K2 c cmax K c ) (4) c=c mn nc= ( N/L c )( N/L c nc )( N/L c nc K2 c c s the number of colldng sub-carrers. It belongs to {c mn : c max } wth: and c mn = max(,k+k2 N/L) c max = mn(k,k2,n/l) and nc s the number of non-colldng sub-carrers n cell ; nc {:K c}. The proof s contaned n Reference []. Lemma 2: The mean number of collsons n the system s equal to []: E[C] = 2 ( Π (K ))E[C K,K2] (5) K,K2 = where E[C K,K2] s gven from Equaton (2) and Π (K ) = n S π( n )Pr(K n ) s the probablty of havng LK sub-carrers n cell. V. NUMERICAL RESULTS We now present some numercal results consderng an OFDMA system wth an FFT sze of 248 sub-carrers and 2 regons wth AMC respectvely 6-QAM 3/4 (E = 3 bt/symbol) and (E 2 = bts/symbol). These effcency parameters correspond to the transmsson modes wth convolutonally-coded modulaton. Moreover, let L =6 sub-channels; l s = sub-channel; l2 s = 3 sub-channels; λ s =. users/sec; λ e =. users/sec; µ s =.2; BLER =; mean fle sze E[Z] = 5Kb. These values are delberately small to ease numercal burden; a scalng up s straght-forward. Based on Equaton (5) and usng a baud value B = 2666symbol/sec and K =48, R s = R s 2 = 28Kbps. Fgure 2 shows the blockng probablty for streamng flows wth and wthout QoS as a functon of an ncreasng arrval rate of streamng flows. Blockng probablty of stream flows QAM 3/4 wth QoS 6 QAM 3/4 wthout QoS QPSK /4 wth QoS wthout QoS Arrval rate of stream flows Fg. 2. Blockng probablty for streamng flows n two regons, wth and wthout QoS We observe that wthout QoS, the blockng rate n the nner regon s the same as n the outer regon. Ths s due to the fact that n ths case, streamng flows are assgned the same number of sub-carrers n both regons. Wth our QoS proposal however, the blockng rate ncreases n both regon and s hgher n the outer one, as dstant users are gven more resources. In terms of throughput, recall that our strategy s completely seamless for both types of users whch get the same bt rate. The dstncton between nner and outer regon s dfferent for data users who ndeed belong to two dfferent classes n terms of throughput, one per each regon, but have an equal blockng probablty, both when QoS s enabled or not. Consequently, the mean transfer tmes are dfferent n each regon, as shown n Fgure 3, and get larger when QoS s enabled. We next turn to the case where reuse parttonng s enabled (L =3sub-channels). Fgure 4 shows the blockng probablty for streamng flows for the two regons, nner and outer, when both QoS and reuse parttonng are enabled. We observe that wth respect to Fgure 2, the blockng probablty remans the same n the nner regon but gets (unacceptably) hgher n the outer one. Fgure 5 shows the mean transfer tme n the two regons when agan QoS and reuse parttonng are both enabled. We observe that wth respect to Fgure 3, both values get hgher, wth a bas n the outer regon. These ncreases n the blockng probabltes and mean transfer tmes are nevertheless counter-balanced by an ncrease n the total cell throughput, as show n Fgure 6 for the case of two cells. Ths means that the lower number of sub-carrers n the edge of the cells s compensated for by a lower number 4783 Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE GLOBECOM 27 proceedngs. Authorzed lcensed use lmted to: INT-Evry. Downloaded on October 6, 28 at 9:3 from IEEE Xplore. Restrctons apply.

5 QAM 3/4 wth QoS 6 QAM 3/4 wthout QoS QPSK /4 wth QoS wthout QoS.7.6 Mean transfer tme for elastc calls (s) Mean transfer tme for elastc calls (s) QAM 3/ Arrval rate of stream flows.8 Arrval rate of stream flows Fg. 3. QoS Mean transfer tme for elastc flows n two regons, wth and wthout Fg. 5. Mean transfer tme for elastc flows n two regons wth reuse parttonng.45 6 x Blockng probablty of streamng flows QAM 3/4 Overall cell throughput Wthout reuse parttonng Wth reuse parttonng Arrval rate of streamng flows Arrval rate of streamng flows Fg. 4. Blockng probablty for streamng flows wth reuse parttonng Fg. 6. Overall cell throughput of collsons on those sub-carrers. And hence t s worth the mplementaton. VI. CONCLUSION We proposed n ths paper a new QoS paradgm that makes OFDMA-based IEEE82.6 capable of offerng streamng applcatons a sustanable rate throughout the whole coverage area whle offerng the leftover capacty to elastc ones. We specfcally proposed to compensate for the degraded performance of streamng flows that are away from the base staton and/or experencng bad rado condtons, due to path loss and fadng, by assgnng them a hgher number of sub-carrers. Usng reuse parttonng mproves the overall system throughput albet hgher ndvdual blockng rates and transfer tmes. In our next step, we shall consder the ntegraton of a moblty pattern n WMAX, n the framework of IEEE82.6e. In ths case, our present QoS proposal s even more essental as moble users need to be guaranteed a sustanable performance as they move around the system. REFERENCES [] [2] A. Molsch, Wreless Communcatons, IEEE Press, 25. [3] M. Johansson, Dynamc Reuse Parttonng Wthn Cells Based on Local Channel and Arrval Rate Fluctuatons, Techncal Report, Uppsala Unversty, Sweden, [4] IEEE , Part 6: Ar Interface for Fxed and Moble Broadband Wreless Access Systems, IEEE Standard for local and Metropoltan Area Networks, February 26. [5] IEEE , Part 6: Ar Interface for Fxed Broadband Wreless Access Systems, IEEE Standard for local and Metropoltan Area Networks, October 24. [6] S.W. Halpern, Reuse Parttonng n Cellular systems, Proc. VTC, pp , 983. [7] A. B. Downey, The structural cause of fle sze dstrbutons, ACM SIGMETRICS Performance Eval. Rev., vol. 29, pp , June 2. [8] G. L. Stuber, Prncples of Moble Communcaton, 2nd ed. Norwell, MA:Kluwer, 2. [9] N. Benameur, S. Ben Fredj, F. Delcogne, S. Oueslat-Boulaha and J.W. Roberts, Integrated Admsson Control for Streamng and Elastc Traffc, QofIS 2, Combra, September 2. [] C. Tarhn, T. Chahed, On capacty of OFDMA-based IEEE82.6 WMAX ncludng Adaptve Modulaton and Codng (AMC) and ntercell nterference, LANMAN 27, Prnceton NJ, June 27. [] S-E. Elayoub, B. Foureste and X. Auffret, On the capacty of OFDMA 82.6 systems, ICC 26, Istanbul, June Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE GLOBECOM 27 proceedngs. Authorzed lcensed use lmted to: INT-Evry. Downloaded on October 6, 28 at 9:3 from IEEE Xplore. Restrctons apply.