Research Letters in Counications Volue 29, Article ID 23276, 4 pages doi:11155/29/23276 Research Letter Chip-Level HARQ Chase Cobining for HSUPA Mohaed Et Tolba, 1, 2 Sair Saoudi, 1, 2 Raphaël Visoz, 3 and Tarik Ait-Idir 4 1 Lab-STICC, Institut Teleco, Teleco Bretagne, UMR CNRS 3192 Lab-STICC, Technopôle Brest-Iroise, CS 83818 29238 Brest Cedex 3, France 2 TELECOM Bretagne institution, Universitéeuropéenne de Bretagne UEB), France 3 Orange Labs, 38-4 rue du generale-leclerc, 92792 Issy Moulineaux Cedex 9, France 4 Counication Systes Departent, Institut National des Postes et Télécounications INPT), Madinat Al Irfane, Rabat 1, Morocco Correspondence should be addressed to Mohaed Et Tolba, ohaedettolba@teleco-bretagneeu Received 19 Noveber 28; Accepted 23 January 29 Recoended by Adnan Kavak Hybrid autoatic repeat request HARQ) is used in high-speed uplink packet access HSUPA) to increase the data rate Chase cobining is the siplest of HARQ algoriths It provides a tie diversity and requires a sall eory In this paper, we propose a technique to increase the chase cobining efficiency In this ethod, HARQ transissions are seen as additional receive antennas which are jointly cobined at chip-level using a space-tie linear iniu square error LMMSE) We deonstrate that the chase cobining diversity gain is considerably increased especially for high-order odulation Copyright 29 Mohaed Et Tolba et al This is an open access article distributed under the Creative Coons Attribution License, which perits unrestricted use, distribution, and reproduction in any ediu, provided the original work is properly cited 1 Introduction High-speed uplink packet access HSUPA) is designed to reach high data rates in universal obile telecounications syste UMTS) uplink Its ipleentation includes new technologies such as fast scheduling as well as hybrid autoatic repeat request HARQ) There are two wellknown HARQ algoriths used for HSUPA: chase cobining and increental redundancy IR) [1] In chase cobining algorith, when a received packet is erroneous, it is saved in the receiver buffer and a negative acknowledgent NACK) is sent to the transitter in order to retransit the packet At the receiver end, the retransitted packet is cobined, before decoding, with the previously saved one [2] This results in a tie diversity gain for ultipath channels In the IR approach, each retransission carries additional redundancy, and a coding gain is provided after cobining different HARQ transissions The conventional rake receiver with receive diversity is used for detecting HARQ transission in HSUPA syste It has been deonstrated that this receiver provides good perforance for high spreading factors with binary phaseshift keying BPSK) odulation However, it exhibits a significant perforance degradation in the case of low spreading factors with high-order odulation The chiplevel LMMSE equalizer has been proposed for HSUPA as a solution to this perforance degradation It was subject to any studies in recent years In [3], authors have shown that the LMMSE chip-level equalizer with receive diversity provides a significant perforance enhanceent for high speed downlink packet access HSDPA) The chip LMMSE has been extended to MIMO ulticode CDMA syste in [4] without taking account of HARQ In [5], the LMMSE is jointly applied with HARQ at sybol-level for spacetie bit interleaved coded odulation STBICM) systes over ultiple input ultiple output MIMO) channel with intersybol interference ISI) Soft bit-level cobining is the siplest way to cobine HARQ transissions before decoding In addition to its ipleentation siplicity, it has been shown that this ethod offers a considerable gain for HSUPA [1] However, its perforance is severely degraded in fading channels when it is used with chase cobining for high-order odulation To overcoe this perforance degradation, we propose a chip-level cobining ethod based on LMMSE chip equalizer which is an interference resistant receiver In this
2 Research Letters in Counications ethod, HARQ transissions are seen as additional receive antennas Hence, the receive diversity gain is increased The different HARQ transissions receive antennas) are jointly exploited by a single LMMSE filter to provide the equalized chip as the cobining result In this anner, we jointly benefit fro the tie diversity provided by the ultipath channel, and the space diversity offered by the receive antennas It is shown that the proposed ethod outperfors the soft bit-level cobining A considerable gain is provided in the case of high-order odulation 2 HSUPA Syste Overview To reach the required quality of service, HSUPA introduces a new transport channel naed enhanced dedicated channel E-DCH) which supports sophisticated signal processing operations For error detection, it uses a cyclic redundancy check CRC) code which adds 24 bits to each HSUPA transport block The CRC code is followed by a 1/3 turbo code to protect E-DCH data against channel ipairent After channel coding, a rate atching operation is applied to adapt the nuber of bits at the output of turbo code to the desirable data rate One or several E-DCH dedicated physical data channels E-DPDCHs) are then used to carry the E- DCH data [6] The E-DPDCHs are separately interleaved and a bit to sybol apping is applied for each E-DPDCH In the case of BPSK odulation, each bit is apped to a sybol which takes value in { 1, +1} Recently, the 16- quadrature aplitude odulation QAM) is also specified for HSUPA Its constellation is built using two orthogonal 4-pulse aplitude odulation PAM) constellations In 4- PAM, a set of two consecutive bits, b k, k k+1 is converted to a real-valued sybol s k given by s k = 1 5 b k 2+bk+1 ) 1) After bit to sybol apping, each physical channel is spread with a real-valued orthogonal variable spreading factor OVSF) code The spreading factor ay be different fro one physical channel to another The spreading factor SF = 2 is allowed for HSUPA, especially for high data rates After the spreading, each physical channel is weighted by a gain factor in order to copensate the difference between spreading factors After that, the physical channels are IQ in-phase/quadratue-phase) ultiplexed to for a single coplex-valued strea of chips which is scrabled and sent through a ultipath channel A chip-spaced channel odel of length L is assued in this work The signal is received by two antennas at the receiver end A chip LMMSE equalizer of length 2E+1 is jointly applied to the two received signals for channel equalization After despreading, the signal is converted to the log-likelihood ratios LLR) required by the turbo decoder Before the decoding process, the LLR sequence is deinterleaved and inversely rate-atched The decoder is followed by a CRC decoder for residual error detectionwhenapacketisdetectedtobeinerror,itissaved and a NACK is sent to the transitter which retransits the packet At the receiver, the retransitted packet is y 1,1) y 2,1) y 1,Nt) y 2,Nt) LMMSE chip equalizer Despreading Figure 1: Chip-level cobining LLR coputation To turbo decoder cobined with the previously received one This increases the probability of correct decoding 3 Syste Model Consider an observation window of length 2E + 1, and let us denote by x the vector corresponding to the transitted ulticode chip at the instant T c, wheret c is the chip duration When the HARQ is not considered, the received signal at the base station is expressed as ) ) ) y 1 H 1 n 1 y 2 = H 2 x + n 2, 2) where y i = [y i E,, y i,, y i +E] T is the received signal at the ith i = 1, 2) antenna when the vector x = [x E L+1,, x,, x +E ] T is transitted through the ultipath channel represented by the 2E +1) 2E + L) Toeplitz atrix H i between the transit antenna and the ith receive antenna n i = [ni E,, n i,, n i +E] T is a vector of coplex white Gaussian noise saples each of which has the power spectral density N The received signals y 1 and y 2 are jointly delivered to a space-tie LMMSE chip equalizer to reduce the interference effect The equalizer output corresponding to the th chip is expressed as x = w H y, 3) where w is the space-tie LMMSE filter of size 22E +1), ) H denotes the Heritian operator, and y = [ y 1T y 1T is the concatenation of the received signals The space-tie LMMSE filter is expressed as w = HH H + N ) 1 I H E+L, 4) E c where H = [ H 1 H 2] T Its size is 22E +1) 2E + L) HE+L is the E + L)th colun of the atrix H, ande c is the total chip energy After chip equalization, the despreading process is perfored The resulting signal is then used to copute LLR values which are need by turbo decoding When HARQ soft bit-level cobining is considered, chip equalization, despreading, and LLR values coputation are perfored individually for each HARQ transission If a block is erroneous after turbo decoding, its corresponding LLRs are saved to be cobined with those of retransitted block [5] ] T
Research Letters in Counications 3 Table 1: MCSs configurations MCS Nuber of codes Min SF Coding rate Modulation Max bit rate kbps) 1 4 2 73 BPSK 45 2 4 2 74 16-QAM 819 4 Chip-Level Cobining In the chip-level cobining approach, each HARQ transission is seen as two additional receive antennas The transissions are jointly exploited by the LMMSE chip equalizer as shown in Figure 1Lety p,q) denote the received vector of the th chip on the pth antenna at qth HARQ transission, and n p,q) denote the noise vector corresponding to the th chip on the pth antenna at qth HARQ transission The syste odel after jth transission is expressed as see [7]) Y j) = H j) x + n j), 5) where Y j) = [y 1,1) ) T,y 2,1) ) T,,y 1,j) ) T,y 2,j) ) T ], n j) = [n 1,1) ) T,n 2,1) ) T,,n 1,j) ) T,n 2,j) ) T ], and H j) is the channel atrix after jth transission It is written as follows: H 1,1) H 2,1) H j) =, 6) H 1,j) H 2,j) where H p,q) p = 1, 2; q = 1,, j) is a Toeplitz atrix of size 2E +1) 2E + L)givenby h p,q) L 1 h p,q) H p,q) = h p,q) L 1 h p,q), 7) h p,q) L 1 h p,q) h p,q) l l L 1) in 7) are the channel taps associated to the pth receive antenna of the qth HARQ transission The tie delay between two successive HARQ transissions, round-trip delay RTD), is taken into account in the propagation channel generation HARQ transissions are separately received in tie by two receive antennas At the receiver side, with the aid of a buffer and concatenation procedure expressed in 6), each HARQ transission can be viewed as a source of two ore virtual antennas In other words, the delay diversity will translate into space diversity which is exploited by a chip-level LMMSE equalizer The resulting chip after LMMSE filter is expressed by x j) = w j)) H Y j), 8) BLER 1 1 1 1 2 1 3 5 5 1 Received E c /N per antenna db) Single trans Method12trans) Method22trans) Method 1 3 trans) Method 2 3 trans) Figure 2: Siulation results for MCS1 where w j) is written as ) 1 w j) = H j) H j)) H N + I H j) E c = A 1 H j) E+L, where H j) E+L is the E + L)th colun of the atrix H j) The coputation of the atrix inversion in 9) is done using the QR decoposition of the atrix A [8] Chip-level cobining ethod can be ipleented using a recursive technique This allows to get siilar coputationnal coplexity and a slightly increased eory requireent copared with soft bit-level cobining where HARQ transissions are separately equalized at the chip-level 5 Siulation and Results Coputer siulations are done for HSUPA syste The transitter is ipleented according to 3GPP technical specifications [9] The predefined odulation and coding schees MCSs) siulated in this work are shown in Table 1 For both MCSs two physical channels are spread with SF = 2, and the two others are spread with SF = 4 The transission is done over a ultipath channel which has a profile siilar to that of ITU-Pedestrian B which is considered to be block fading Because of the low obility of HSUPA obile, its speed is fixed at 3 K/h LMMSE chip filter of length 15 is applied for channel equalization Perfect channel knowledge is assued at the receiver SISO E+L 9)
4 Research Letters in Counications 1 References BLER 1 1 1 2 1 3 2 4 6 8 1 12 14 16 18 Received E c /N per antenna db) Single trans Method 1 2 trans) Method 2 2 trans) Method 1 3 trans) Method 2 3 trans) Figure 3: Siulation results for MCS2 decoding is perfored using the Max-Log-MAP algorith with eight iterations In the following, we label the soft bitlevel cobining ethod, Method 1 The proposed cobining ethod is labeled Method 2 Siulations are done for chase cobining using Method 1 and Method 2 The two ethods are copared in ters of block error rate BLER) which is coputed as function of received E c /N per antenna Note that received E c /N is averaged over HARQ transissions Figure 2 shows the BLER results for MCS1 After two and three HARQ transissions, we observe that Method 2 outperfors Method 1 A gain of 3 db is provided after three transissions for BLER= 1 2 Figure 3 plots the BLER results, obtained with two and three HARQ transissions for MCS2 It is seen that Method 2 provides a significant gain over Method 1 for two and three HARQ transissions The proposed technique Method 2) offers an additional gain of 3 db at 1 2 with three HARQ transissions It is also observed that the perforance obtained by Method 1 after three transissions is reached by Method 2 with only two transissions This significant increase of chase cobining efficiency with Method 2 is due to the space diversity offered by the receive antennas, and the tie diversity provided by the chase algorith Moreover, the different transissions are jointly exploited by the LMMSE chip equalizer which is an interference-resistant detector [1] M Bertinelli and E Malkaäki, HARQ for WCDMA enhanced uplink: link level perforance in SHO, in Proceedings of the 15th IEEE International Syposiu on Personal, Indoor and Mobile Radio Counications PIMRC 4), vol 4, pp 2856 286, Barcelona, Spain, Septeber 24 [2] D Chase, Code cobining a axiu-likelihood decoding approach for cobining an arbitrary nuber of noisy packets, IEEE Transactions on Counications, vol 33, no 5, pp 385 393, 1985 [3] J-B Landre and A Saadani, Receive diversity and LMMSE equalization benefits for HSDPA: realistic network throughputs, in Proceedings of the 18th IEEE International Syposiu on Personal, Indoor and Mobile Radio Counications PIMRC 7), pp 1 5, Athens, Greece, Septeber 27 [4] B-H Ki, X Zhang, and M Flury, Linear MMSE spacetie equalizer for MIMO ulticode CDMA systes, IEEE Transactions on Counications, vol 54, no 1, pp 171 1714, 26 [5] T Ait-Idir, H Chafnaji, and S Saoudi, Joint hybrid ARQ and iterative space-tie equalization for coded transission over the MIMO-ISI channel, in Proceedings of the IEEE Wireless Counications and Networking Conference WCNC 8), pp 622 627, Las Vegas, Nev, USA, March-April 28 [6] 3GPP TS 25213 v75, Spreading and odulation FDD), Release 7, May 28 [7] ENOnggosanusi,AGDabak,YHui,andGJeong, Hybrid ARQ transission and cobining for MIMO systes, in Proceedings of the IEEE International Conference on Counications ICC 3), vol 5, pp 325 329, Anchorage, Alaska, USA, May 23 [8] W H Press, S A Teukolsky, W T Vetterling, and B P Flannery, Nuerical Recipes in C: The Art of Scientific Coputing, Cabridge University Press, Cabridge, UK, 1992 [9] 3GPP TS 25212 v78, Multiplexing and channel coding FDD), Release 7, May 28 6 Conclusion In this paper, we have proposed a new technique for HARQ chase cobining based on the LMMSE chip equalizer in HSUPA syste In this ethod, each transission is seen as two additional receive antennas at the input of the LMMSE filter The proposed ethod has been copared with the soft bit-level cobining for two and three transissions Siulation results have deonstrated a significant increase in chase cobining efficiency when using the proposed ethod especially for 16-QAM odulation
Rotating Machinery http://wwwhindawico Volue 214 The Scientific World Journal http://wwwhindawico Volue 214 Distributed Sensor Networks Sensors http://wwwhindawico Volue 214 http://wwwhindawico Volue 214 http://wwwhindawico Volue 214 Control Science and Civil http://wwwhindawico http://wwwhindawico Volue 214 Volue 214 Subit your anuscripts at http://wwwhindawico Electrical and Coputer Robotics http://wwwhindawico http://wwwhindawico Volue 214 Volue 214 VLSI Design OptoElectronics Navigation and Observation http://wwwhindawico Volue 214 http://wwwhindawico http://wwwhindawico Cheical http://wwwhindawico Volue 214 Volue 214 Active and Passive Electronic Coponents Antennas and Propagation http://wwwhindawico Aerospace http://wwwhindawico Volue 214 http://wwwhindawico Volue 21 Volue 214 Modelling & Siulation in Volue 214 http://wwwhindawico Volue 214 Shock and Vibration http://wwwhindawico Volue 214 Acoustics and Vibration http://wwwhindawico Volue 214