Multiresolution MBMS Transmissions for MIMO UTRA LTE Systems

Transcription

1 Multiresolution MBMS Transissions for MIMO UTRA LTE Systes Nuno Souto Aérico Correia Rui Dinis João Carlos Silva Luís Abreu ISCTE / Instituto de Telecounicações/ADETTI Lisboa Portugal nuno.souto@lx.it.pt aerico.correia@iscte.pt joao.carlos.silva@lx.it.pt lusab@gail.co ISR/IST Lisboa Portugal rdinis@ist.utl.pt Abstract Hierarchical constellations constitute a siple technique for achieving ultiresolution and therefore are appealing for MBMS (Multiedia Broadcast and Multicast Service). In this paper we consider the use of M-QAM hierarchical constellations (Quadrature Aplitude Modulation) cobined with MIMO (Multiple Input Multiple Output) for the transission of ulticast and broadcast services in UTRA (Universal Mobile Telecounications Syste Terrestrial Radio Access) Long Ter Evolution (LTE) systes based on Orthogonal Frequency Division Multiplexing (OFDM). Due to the deanding channel estiation requireents and the high sensitivity to interference resulting fro the usage of several antennas and hierarchical constellations an enhanced receiver based on the turbo concept is eployed and its perforance is evaluated. Keywords- MBMS Hierarchical constellations MIMO LTE. I. INTRODUCTION It is widely recognized that OFDM odulations [] are suitable for broadband wireless systes. For this reason they were selected for several digital broadcast systes and wireless networks [] and are also being considered for UTRA LTE [3]. Regarding UTRA LTE special attention is being devoted to the support of MBMS which has already been standardized in 3GPP UTRAN (UMTS Terrestrial Radio Access Network) Release-6 [4] and 7 [5]. The goal is to enable an efficient support of downlink streaing (fro the base station to the obile terinal) and download-and-play type services to large groups of users. Fro the radio perspective MBMS includes point-to-point (PtP) and point-to-ultipoint (PtM) odes. Regarding the PtM ode it sees attractive to eploy hierarchical odulations since it is a siple and flexible enhanceent technique that can increase the transission efficiency due to its ability to provide unequal error protection to different bits and thus provide ultiresolution into a cell. By having several classes of bits with different error protection associated and to which different streas of inforation are apped a given user can attept to deodulate only the ore protected bits or also the bits that carry the additional inforation depending on the propagation conditions. This type of approach is possible whenever the inforation can be scalable like the cases of coded voice or video signals as studied in [6][7]. For this reason hierarchical 6-QAM and 64- QAM constellations have already been incorporated into DVB- T (Digital Video Broadcasting - Terrestrial) standards [8]. MIMO schees have eerged as one of the ost proising ethods for capacity increase in a counication syste [9][0] and are being considered for UTRA LTE [3]. In MIMO systes with coherent detection the channel estiation plays a crucial role since the perforance of the spatial signal processing in the receiver depends on the accuracy of the channel estiates. Furtherore QAM constellations can be severely affected due to inaccurate channel estiates. In this paper we consider the use of QAM hierarchical constellations in a UTRA LTE OFDM based syste eploying ultiple transitting and receiving antennas with the ai of supporting broadcast and ulticast services. To deal with the high sensitivity to channel estiation errors we eploy an iterative receiver capable of perforing joint MIMO detection and channel estiation. This receiver is based on the approach proposed in [] for WCDMA systes. It can apply different MIMO equalization techniques during the iterative process and obtain refined channel estiates by considering the data sybols as extra pilots as proposed in []. The paper is organized as follows. First Section II introduces hierarchical constellations and defines the odel of the MIMO- OFDM syste considered in this study. In Section III the proposed iterative receiver structure and respective channel estiation process are described. Section IV presents soe perforance results obtained with the proposed schee while the conclusions are given on Section V. II. SYSTEM DESCRIPTION A. M-QAM Hierarchical Signal Constellations In hierarchical constellations there are two or ore classes of bits with different error protection and to which different streas of inforation can be apped. By using non uniforly spaced signal points (where the distances along the I or Q axis between adjacent sybols are different) it is possible to odify the different error protection levels. As an exaple a nonunifor 6-QAM constellation can be constructed fro a ain Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.

2 QPSK constellation where each sybol is in fact another QPSK constellation as shown in Figure. Figure. Non-unifor 6-QAM constellation. The basic idea is that the constellation can be viewed as a 6- QAM constellation if the channel conditions are good enough or as a QPSK constellation otherwise. In the latter situation the received bit rate is reduced to half. These constellations can be characterized by the paraeter k D /D (0<k 0.5) as shown in Figure. If k 0.5 the resulting constellation corresponds to a unifor 6-QAM. This approach can be naturally extended to any QAM constellation size M where the nuber of possible classes of bits with different error protection is / log M. B. Transitted Signals In Figure we show a transitter chain that incorporates QAM hierarchical constellations into a UTRA LTE based MIMO- OFDM transission. In the proposed schee there are / log M parallel chains for the different input bit streas that will have unequal error protection. For 6-QAM we can use two parallel chains while for 64-QAM we can use three chains. Each strea is encoded interleaved and apped into the constellation sybols in the odulation appers according to the iportance attributed to the chain. Pilot sybols are inserted into the odulated data sequence which is then converted to the tie doain using an IDFT (Inverse Discrete Fourier Transfor). The resulting strea is then split into several saller streas which are transitted siultaneously by M tx transitting antennas. Note that in the proposed schee the coding is not perfored independently for each different antenna. Instead each data sequence is encoded and divided equally aong the transitting antennas by the Serial to Parallel block. The objective is to try to obtain soe diversity for the sae encoded block. In this paper we consider the frae structure of Figure 3 for a MIMO-OFDM syste with N carriers. According to this structure the pilot sybols are ultiplexed with the data sybols using a spacing of NT sybols in the tie doain and NF OFDM blocks in the frequency doain. To avoid interference between pilots of different transitting antennas FDM (Frequency Division Multiplexing) is eployed for the pilots which eans that pilot sybols cannot be transitted over the sae subcarrier in different antennas. Data sybols are not transitted on subcarriers reserved for pilots in any antenna therefore the iniu allowed spacing in the N M +. frequency doain is ( ) in F Before being transitted the sequences of sybols are converted to the tie doain through { x il i 0... N } IDFT { S k 0... N } where S is the sybol transitted by the k th subcarrier of the l th OFDM block using antenna. The transitted OFDM signals are then expressed as tx N x () t xil ht ( t i Ts) () l i NG with T s denoting the sybol duration N G the nuber of saples at the cyclic prefix ( x il xn il i N G ) and h T (t) the adopted pulse shaping filter. Figure 3. Frae structure for a OFDM transission (P pilot sybol D data sybol T s sybol duration). log M Figure. Transitter chain. Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.

3 R log M R N rx kl ( q) {( ˆ n H ) } { S ˆ } kl Figure 4. Iterative receiver structure. III. ITERATIVE RECEIVER A. Receiver Structure To achieve reliable channel estiation and data detection we eploy a receiver capable of jointly perforing these tasks through iterative processing. The structure of the iterative receiver is shown in Figure 4 where N rx receiving antennas are eployed. According to the figure the signal which is considered to be sapled and with the cyclic prefix reoved is converted to the frequency doain after an appropriate size- N DFT operation. If the cyclic prefix is longer than the overall channel ipulse response the resulting sequence received in antenna n can be expressed as M tx n n n + R S H N () n with Hkl denoting the overall channel frequency response between transit antenna and receiving antenna n for the kth frequency of the lth tie block and Nkl denoting the corresponding channel noise. The sequences of saples () enter the MIMO equalizer (Spatial Deultiplexer block) which separates the siultaneous transitted streas. This can be accoplished with an MMSE (Miniu Mean Squared Error) equalizer [3] a ZF (Zero Forcing) equalizer [3] a Maxiu Likelihood Soft Output criterion (MLSO) or an interference canceller (IC) []. It is possible to perfor soe of the receiver iterations using one spatial deultiplexing technique like the MMSE and the others using a different one like the IC as was studied in []. In any case after MIMO equalization the deultiplexed sybol sequences are serialized and pass through the deodulator de-interleaver and channel decoder blocks. This channel decoder has two outputs. One is the estiated inforation sequence and the other is the sequence of log-likelihood ratio (LLR) estiates of the code sybols. These LLRs go through the Decision Device which outputs either soft-decision or hard decision estiates of the code sybols and enter the Transitted Signal Rebuilder which perfors the sae operations of the transitter (interleaving odulation conversion of serial to parallel streas). The reconstructed sybol sequences are then used for a refineent of the channel estiates and also for possible iproveent of the spatial deultiplexing task (in case of eploying an IC) for the subsequent iteration. The possible MIMO equalization techniques are now going to be briefly described. Using atrix notation the MMSE estiates of the transitted sybols in subcarrier k and OFDM block l is given by H ( σ ) Sˆ Hˆ Hˆ Hˆ + I R (3) H where S ˆ is the M tx estiated transitted signal vector with one different transit antenna in each position H ˆ is the N rx M tx channel atrix estiate with each colun representing a different transit antenna and each line representing a different receive antenna R is the N rx received signal vector with one different receive antenna in each position and σ is the noise variance. The ZF estiate can be siply obtained by setting σ to 0 in (3). In the MLSO criterion we use the following estiate for each sybol ˆ E S R S si Λ ( s ) p ( Rkl ) P S s p S s ( R ) kl i i k l k l i (4) where s i corresponds to a constellation sybol fro the E is the expected value P () odulation alphabet Λ [ ] Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.

4 represents a probability and p () a probability density function (PDF). Considering equiprobable sybols P( Skl si) M where M is the constellation size. The PDF values required in (4) can be coputed as interf ( i) ( i M ) p R S s p R S s S (5) tx M with interf M tx Skl Λ interf p( R S s S ) ( πσ ) (6) Hˆ ( n:) s N n rx i exp Nrx n σ interf where Skl is a (M tx -) vector representing a possible cobination of sybols transitted siultaneously by all antennas except antenna s is a M tx vector coposed by interf S and s i and H ˆ ( :) n is the n th line of channel atrix H ˆ. An IC can also be used inside the Spatial Deultiplexer block but usually is only recoendable after the first receiver iteration []. In this case in iteration q for each transit antenna and receive antenna n the IC subtracts the interference caused by all the other antennas. This can be represented as where ( ) ( q ˆ ) S M tx n n ˆ n ( q) ( ) ( ) ( q ˆ ) R R S H (7) represents the data sybols estiates of the previous iteration for transit antenna subcarrier k and OFDM block l. B. Channel Estiation To obtain the frequency channel response estiates for each transitting/receiving antenna pair the receiver applies the following steps in each iteration: () The channel estiate between transit antenna and receive antenna n for each pilot sybol position is siply coputed as Pilot ( S ) * H R. (8) n n Pilot Skl Pilot where S corresponds to a pilot sybol transitted in the k th subcarrier of the l th OFDM block using antenna. Obviously not all indexes k an l will correspond to a pilot sybol since > and >. N T () Channel estiates for the sae subcarrier k transit antenna and receive antenna n but in tie doain positions (index l) that do not carry a pilot sybol can be N F R obtained through interpolation using a finite ipulse response (FIR) filter with length W as follows: Hˆ W n j n + t k l+ j N ht H (9) T j ( W ) where t is the OFDM block index relative to the last one j carrying a pilot (which is block with index l) and h t are the interpolation coefficients of the estiation filter which depend on the channel estiation algorith eployed. There are several proposed algoriths in the literature like the optial Wiener filter interpolator [4] or the low pass sinc interpolator [5]. (3) After the first iteration the data estiates can also be used as pilots for channel estiation refineent. IV. NUMERICAL RESULTS To study the behaviour of the proposed MIMO-OFDM schee and respective iterative receiver several siulations were perfored for a 6-QAM (k 0.4) hierarchical constellation. Two classes of bits with different error protection were used. Each individual inforation strea was encoded with a block size chosen so that the final encoded and odulated strea fitted a sub-frae coposed of 7 OFDM blocks (corresponding to a 0.5s duration). All the paraeters used for these siulations were based on UTRA LTE 3GPP docuents [6] and [7] for a 0MHz bandwidth. Table shows the respective paraeters. Transission BW 0 MHz CP length 7 FFT size 04 Nuber of occupied sub-carriers 600 Sub-frae duration (s) 0.5 Sub-carrier spacing (khz) 5 sybol duration (ns) 30 pilot power (db) 0 OFDM sybols per subfrae Table. Siulation paraeters for 0MHz bandwidth. The channel ipulse response is based on Vehicular A environent [9] with Rayleigh fading assued for the different paths. A velocity of 30 k/h was eployed unless otherwise stated. The channel encoders were rate-/ turbo codes based on two identical recursive convolutional codes characterized by G(D) [ (+D +D 3 )/(+D+D 3 )] [8]. A rando interleaver was used within the turbo encoders. Table shows the different cobinations of receiver and turbo decoder iterations applied as well as the respective MIMO decoding ethods used. 7 Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.

5 Receiver ethod Turbo decoder iterations per receiver loop Iterative Receiver Iterations with MMSE 3 with MMSE+3 with IC 3 with MLSO Table. Different receiver ethod applied for the siulations. Most of the BER (Bit Error Rate) results presented next will be shown as a function of E S /N 0 where E S is the average sybol energy and N 0 is the single sided noise power spectral density. For channel estiation purposes pilot sybols were distributed using a spacing of NF 6 + Mtx and N T 4 or 7 (the two possible configurations proposed in [6]) and a sinc filter interpolation with length W was used at the receiver. In the graphs legends MPB designates ost protected bits IPB eans interediate protected bits and LPB corresponds to least protected bits. Figure 5 copares the perforance of the different receiver ethods of Table for a MIMO x transission eploying a 6-QAM (with k 0.4) hierarchical constellation. It is visible that although the receiver with the MMSE equalizer alone perfors worse than when using the MLSO equalizer the perforance can be substantially iproved and achieve lower BLERs than with the MLSO when applying also an IC in the last receiver iterations. For the reainder of the paper the receiver configuration considered will be ethod (MMSE+IC). Figure 6 shows the behaviour of the receiver for different velocities. According to the results the perforance is alost insensitive to velocity until 0 k/h being visible only a sall degradation in the perforance of the LPB. For higher velocities the perforance quickly degrades for the LPB but does not change significantly for the MPB even at 300 k/h. Figure 7 copares the perforance of a MIMO x transission eploying a 6-QAM (with k 0.4) hierarchical constellation with the two possible pilot spacings in the tie doain. It is visible that both cases have siilar perforances (close to the perfect estiation curves) which eans that it is possible to adopt the larger pilot spacing and therefore increase the transission efficiency if one or two antennas (we verified that the conclusion is also true even for four transit antennas). It is iportant to reeber however that reducing the nuber of pilot sybols will sacrifice the syste robustness for higher velocities as will be shown further ahead. V. CONCLUSIONS In this paper we have studied the use of QAM hierarchical constellations with the ai of supporting ulticast and broadcast transissions in a MIMO-OFDM syste siilar to the one being considered for UTRA LTE. It was verified through siulations that the iterative receivers schees studied are able to achieve good perforances for all the bit streas including those with lower error protection levels even for very high velocities. Therefore the proposed transitter/receiver schee can provide unequal error protection which is adequate for supporting MBMS transissions in UTRA LTE. It was also observed that if a axiu of four transit antennas are being used with 6- QAM hierarchical odulations the option with only half of the pilots sybols proposed for UTRA LTE ( N T 7 ) is adequate unless a higher robustness is desired for high velocities for all inforation streas. VI. ACNKOWLEDGMENTS This paper was supported in part by the C-MOBILE project IST and FCT (pluriannual funding and project U- BOAT PTDC/EEA-TEL/67066/006). BLER Method _ Method Method Es/N0 (db) o x LPB MPB Figure 5. 6-QAM (k 0.4) hierarchical constellation v30 K/h x MIMO transission using several receiver ethods. 4. BLER v3k/h _ v30k/h v0k/h v300k/h Es/N0 (db) N T o x LPB MPB Figure 6. 6-QAM (k 0.4) hierarchical constellation. x MIMO transission using receiver ethod for several velocities. 4. N T Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.

6 BLER perf. esti realistic esti. N T 4 _ realistic esti. N T Es/N0 (db) o x LPB MPB Figure 7. 6-QAM (k 0.4) hierarchical constellation v30 K/h x MIMO transission using receiver ethod for several pilot spacings. VII. REFERENCES [] L.Ciini Analysis and Siulation of a Digital Mobile Channel using Orthogonal Frequency Division Multiplexing IEEE Trans. on Co. Vol. 33 No. 7 July 985. [] R.van Nee and R.Prasad OFDM for Wireless Multiedia Counications Artech House Publ [3] 3GPPP TR 5.84-v7..0 3rd Generation Partenership Project: Technical Specification Group Radio Access Network; Physical Layers Aspects for Evolved UTRA Septeber 006. [4] 3GPP v Introduction of Multiedia Broadcast/Multicast Service (MBMS) in RAN 3GPP June 006. [5] 3GPP v.7..0 Introduction of Multiedia Broadcast/Multicast Service (MBMS) in RAN 3GPP Septeber 006. [6] T. Cover Broadcast channels IEEE Trans. On Infor. Theory vol.it-8 pp.-4 January 97. [7] K. Rachandran A. Ortega K. M. Uz and M. Vetterli Multiresolution broadcast for digital HDTV using joint source/channel coding IEEE J. Select. Areas Coun. vol. January 993. [8] Digital video broadcasting (DVB) fraing structure channel coding and odulation for digital terrestrial television (DVB-T) Mar ETSI European Telecounication Standard ETS [9] E. Telatar Capacity of ultiantenna Gaussian channels AT&T Bell Laboratories Technical Meorandu June 995. [0] G.J. Foschini and M.J. Gans On liits of wireless counications in a fading environent when using ultiple antennas Wireless Pers.Coun. vol. 6 pp Mar [] N. Souto R. Dinis F. Cercas J. Silva A. Correia An Iterative Receiver for WCDMA Systes with MIMO Transissions and Hierarchical Constellations Proc. IEEE International Syposiu on Spread Spectru Techniques and Applications - ISSSTA 006 August 006 Manaus Brasil. [] M. C. Valenti Iterative Channel Estiation and Decoding of Pilot Sybol Assisted Turbo Codes Over Flat-Fading Channels IEEE Journal on Selected Areas in Co. vol. 9 no. 9 pp Septeber 00. [3] S.M. Kay Fundaentals of Statistical Signal Processing: Estiation Theory. Englewood Cliffs NJ: Prentice-Hall 993. [4] J. K. Cavers An analysis of Pilot Sybol Assisted Modulation for Rayleigh Fading Channels IEEE Trans. On Veh. Tech. vol. 40 no. 4 pp Nov. 99. [5] Y.-S. Ki C.-J. Ki G.-Y. Jeong Y.-J. Bang H.-K. Park and S. S. Choi New Rayleigh fading channel estiator based on PSAM channel sounding technique in Proc. IEEE International Conf. on Co. pp Montreal Canada June 997. [6] 3GPP Feasibility study for evolved Universal Terrestrial Radio Access (UTRA) and Universal Terrestrial Radio Access Network (UTRAN) Tech rep. 5.9 v [7] 3GPP Feasibility Study for Orthogonal Frequency Division Multiplexing (OFDM) for UTRAN enhanceent Tech rep v [8] 3GPP 5.-v6..0 Multiplexing and Channel Coding (FDD). [9] Selection procedures for the choice of radio transission technologies of UMTS ETSI TR 0 v3..0 Sophia Antipolis France 998. Authorized licensed use liited to: SOUTHWEST JIAOTONG UNIVERSITY. Downloaded on April at 06:30 fro IEEE Xplore. Restrictions apply.