International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 A High Capacity and Low Decoding Complexity of Multiple Active Spatial Modulation System over Correlated Channels Uma.S Student M.E-communication systems, Loyola Institute of Technology, Chennai-600 123. Suganthi.K Student M.E- Communication Systems, Loyola Institute of Technology, Chennai-600 123. Abstract: A generalized spatial modulation (SM) scheme with multiple active transmit antennas, named as Multiple ActiveSpatial Modulation (MA-SM), is alternative to the STBC system. It work on the principle of multiple transmission of data stream from different transmit antennas in multiple time slots. By allowing multiple transmit antennas in the spatial modulation system (SM) and high multiplexing gain of Vertical Bell Lab Layered (V-BLAST) systems to transmit different symbol at the same time instants. Design of MA-SM for arbitrary number of transmits antennas and the modulation scheme was presented and mapped in to the high dimensional constellation including spatial dimension. Finally to compare the MA-SM with different MIMO schemes, and implemented in to the correlated channel (Rayleigh and Nakagami) fading channels. And by using the spatial correlation to improve the channel capacity and compare the capacity along with antenna mutual coupling and also compare the low complexity detection of MA-SM. Index terms-Spatial Modulation(SM), vertical-Bell lab layered Space-Time (V-BLAST), maximum likelihood (ML) detection, multiple-input multiple-output (MIMO) systems. I. INTRODUCTION Multiple Input Multiple Output (MIMO) multiplexing is a promising technology that could greatly increase the channel capacity without additional spectral resources. Many recent research results have concluded that the multiple-input multiple-output (MIMO) wireless communication architecture is a promising approach to achieve high bandwidth efficiencies. MIMO wireless channels can be simply defined as a link for which both the transmitting and receiving ends are equipped with multiple antenna elements. The capacity of a MIMO system not only depends on the number of channels (N. M), but also depends on the correlation between the channels. The channel correlation of a MIMO system is mainly due to two components (spatial correlation, antenna mutual coupling). This advanced communication technology has the potential to resolve the bottleneck in traffic capacity for future wireless networks. ISSN: 2231-5381 MIMO is an effective way to improve the capacity and reliability, comparing with single antenna wireless systems[1],[2].several MIMO techniques have been comprehensively studied recently studied among which the space time block code (STBC) for two transmit antennas. Offers a low- complexity maximum likelihood (ML) decoding due to its orthogonal structure. Based on this property of orthogonality, orthogonal space time block codes (OSTBCs) was presented in [3],[4].OSTBCs are special class of space time codes which exploits the spatial diversity and offer low complexity ML decoding. However ,rate one OSTBC exists for two transmit antennas only to increase the data rate a new class of semi-orthogonal codes was proposed in [5],[6] known as quasi orthogonal space time block codes (QOSTBCs) they are all full rate codes with pair wise decoding complexity .however ,the QOSTBCs of[5],[6] cannot achieve full diversity. To achieve full diversity, QOSTBC in [7],[8] was proposed by talking half of the symbols from rotated constellation. To further reduce the decoding complexity without compromising on the data rates, a new and distinct class of codes were designed using the concept of co-ordinate interleaving .these codes are popularly known as co-ordinate interleaved orthogonal designs(CIODs) [9],[10].the CIODs are full rate codes which achieve single-symbol decidability. In [9], [10] CIODs for PAM and QAM constellation are discussed. The existing STBCs retransmit each symbol in space and time which reduce the capacity of the system. This reduction in capacity can be improved by using a mapping function for 16QAM constellation in Alamouti STBC [2], in this M-PAM constellation and extended it to square QAM constellations. Using this mapping function I proposed an STBC for four transmit antennas which achieves high coding gain and full diversity. In this STBC renamed as multiple antenna space modulation (MASM). a novel scheme approaching even higher capacity by combining the amplitude/phase modulation techniques with antenna index modulation, named Spatial Modulation (SM), is proposed to extend the constellation into a three dimension one (both the complex plane and the spatial dimension are involved) [1]. http://www.ijettjournal.org Page 2003 International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 Symbols are emitted from a selected antenna after being mapped through a traditional modulator. Therefore, the information is conveyed not only by the amplitude/phase modulation techniques, but also by the antenna indices. The ICI and IAI in SM system will also be avoided if only one antenna is active all over the transmission. Hence the low complexity decoder is capable of prominent performance. The Space Time Block Coding-Spatial Modulation (STBC-SM) scheme, in which SM is combined with Space Time Block Code (STBC) to exploit high spectral efficiency from SM and enjoy coding gains from STBC. At the transmitter side, mapped symbols are emitted from several chosen antennas after being coded with STBC encoder. At the receiver side, a demodulator combining ML algorithm along with the linear STBC decoder is shown to be optimal. Similar to STBC, STBC-SM suffers from either low multiplexing gain or high computational complexity when the block code size extends to more than two. Since the ML decoder for STBC-SM employs an exhaustive search of antenna sets, the decoder complexity increases exponentially as the antenna subset expands. Both the low multiplexing gain and computational complexity, a multiple active –spatial modulation (MA-SM) scheme and near optimal decoder with linear complexity are proposed. The main contribution of this paper: 1) A novel scheme of multi-antenna transmission for SM, named MA-SM, is proposed; in which several transmit antennas carrying different information symbols are active during each time slot. Similar to traditional SM, information bits in MA-SM are mapped into both spatial dimension and traditional complex dimension. As a new approach, we consider the antenna sets with arbitrary number of active antennas rather than a single antenna index in spatial dimension to further explore multiplexing gains. Antenna set deytector Spatial Modulator Demappe r-1 Mod-1 S/P H P/S Demappe r- Mod- Fig.1, System model. II. MULTI -ACTIVE SPATIAL MODULATION A. Proposed MA-SM scheme The general system model consists of a MIMO wireless link with transmit antennas and receive antennas, which is illustrated in Figure 1. The source information bits are transmitted from of the transmit antennas after being mapped through an M order Quadrature Amplitude Modulation (M-QAM). Through the × wireless channel H and the -dim additive white Gaussian noise (AWGN) = [ , …. ] the received signal is given by (1) where ρ is the average signal to noise ratio (SNR) at each receive antenna, H and are independent and identically distributed (i.i.d) entries according to C (0, 1) (complex Gaussian zero mean distribution with variance 1) and S is the constellation set of M -QAM. The transmitted symbol X is comprised with QAM symbols emitted from the antennas ….. , respectively for denoting convenience, the antenna group ( ….. ) will be written as =(0,1,…..1….),j {1,…..[ ]2}where corresponds stand for the states of transmit antennas and each 0 and 1 represents the off and on of the corresponding antenna respectively for example ,in a system where =4, =2,the possible antenna groups could be denoted as =(1,1,0,0,), =(1,0,1,0), =(1,0,0,1), =(0,1,1,0), =(0,1 ,0,1), =(0,0,1,1). = ≜[0 Where ….. S, ISSN: 2231-5381 + ….. ….. (1) 0] {1.... . }. 2) A general principle for designing the MA-SM code is given. By carefully designing over the antenna sets and the rotation angle applied to symbols, more diversity gains are available. 3) A near-optimal decoder with low complexity is derived. In order to reduce the computational complexity, it separates the antenna set detection from the traditional demodulator. Compared to the ML detection, the proposed detection algorithm reduces the computational complexity prominently. 4) Theoretical analysis and computer simulations to substantiate the efficiency of MA-SM. A closed form expression for the upper bound on the bit error probability (BEP) is de-rived for the selection of transmission as well as the detection algorithm. Simulation results demonstrate the superior performance of MA-SM when applied to several communication systems by comparing with several widely used algorithms. B.System Design and Optimization In MA-SM system, the information bits are conveyed by both the complex symbols and the indices of the active antennas from which those symbols are transmitted. At the transmitter side, NP antennas are chosen to carry different symbols during the transmission, which results in the increase of multiplexing gain. Theoretically, there is no limitation on , which implies that could be allowed to be any number no larger than to benefit the available Multiplexing gain. However, this will lead to the exponentially increasing complexity at the receiver side and the ICI and IAI would degrade the performance seriously. http://www.ijettjournal.org Page 2004 International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 The constellation set of MA-SM could be denoted as the Cartesian product of the complex set including both the real and imaginary parts of the transmit symbols and the antenna groups with a discrete topology. Besides, the available antenna groups composed of the antennas are always more than2[ ] which means that we can carefully select the active antenna groups to minimize interference. Transmitter Spatial Modulation Symbol Channel QAM mapping Serial to parrel AWGN QAM demap ping As to the complex symbol optimization, it could be observed from (4) that the minimum distance between symbols could be maximized by properly choosing the rotation angles. The 3-dim constellation could be treated as one constituted by different constellation planes. Each of them is a standard QAM constellation and planes are distinguished by the antenna groups. Since the Euclidean distance between symbols locating on the same complex plane is maximized with the QAM modulation, rotation angles for them on the same plane should be exactly the same. C.Complexity In this section, the complexity of both the transmitter and receiver are taken into consideration. The complexity of the decoder for MA-SM is compared to the complexity of the ML decoder and some other optimal decoders for SM and GSM. The number of operations needed is used to estimate the receiver complexity. Receiver Parrel to serial means = 2 selection procedure could be skipped. Otherwise, the distance definition indicates that we should try to avoid overlapping antenna indices between different groups since groups sharing the same antenna indices will lead to the increase of the linear dependence probability of channel space, which is the main cause of detection error. Spatial demodu lation M R C Error rate III. T HEORETICAL A NALYSIS Figure.2. Block Diagram of MA-SM Since the minimum distance between codewords dominates the BEP, transmission scheme can be optimized by maximizing the minimum distance involved. Unlike the traditional complex space where Euclidean distance could be applied, the three dimensional space here contains a discrete dimension that confuses the definition of distance. Referring to the most widely used distance definition in discrete metric space as in a similar definition could be derived. δ(a, b) = 0, 1, = ≠ We now derive the bit error probability (BEP) for the proposed decoder in system to estimate its performance. In this mainly focus on system with N T transmit antennas and NP active .antennas employing bit phase shift keying (BPSK).The analysis could easily be extended to other cases. For our convenience, we assume that the power of the transmit symbols is normalized and the Gaussian noises added on all the receive antennas are with same variance thus the system model could be rewritten as Y=HX+ (2) to find the hybrid distance d in which the Frobinenious Norm equals to the Euclidean distance and , denotes the rotation angle applied to QAM symbol Si∈S emmited from antenna group . is defined as the number of different indices between two antenna groups, thus the minimum distance between two codewords X and x is maximized, the performance gain is achieved. (4) An error that can be occurred in the demodulator could be categories into two scenarios due to the separated steps in demodulator. First is the error occur in the active antenna detection (denoted by PAntErr) and the second one is that the error occurs in traditional demapping when antenna detection is corrected (denoted by PModErr) thus the overall bit error probability could be bounded as Perror=1-(1-PAntErr) (1-PModErr) d , = { +|| , - , || } ⎧ It shows that the optimization could be executed in the selection of antenna groups and complex symbol respectively we will first consider the antenna group optimization. Denoting [ ]= q the number of illegal sets is The active antennas are detected with a projection operator that is linear.an explicit formula for the projection for the k-th antenna could be denoted as written as -2 which provides redundancy for antenna set selection. When no illegal set available which ISSN: 2231-5381 (5) (3) http://www.ijettjournal.org = = HX+ (6) Page 2005 International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 IV. N UMERICAL R ESULTS In this section some simulation results for the MASM system with different numbers of transmit antennas and make comparisons with other MIMO systems, such as SM,VBLAST,Alamouti’s scheme and STBC-SM.the bit error rate(BER) performance of these systems was evaluated by Monte carlo simulations for various spectral efficiencies as a function of a the average SNR per receive antenna and in cases the independence of channels is assumed unless otherwise specified. In order to be convincing, comparisons are realized under the same transmission rate without restraint on the constellations. The SM system uses the optimal decoder derived in the GSM employs the detection algorithm in the V-BLAST system uses linear decorrelator detection and the STBC-SM uses the optimal detector introduced in Rotation parameters and antenna groups for different transmission rates in MA-SM are selected. Fig.4 Transmit vs Receive diversity C.BER performance qam modulation with masm A.M-QAM constellation diagram The transmitted signals must traverse a potentially difficult environment with scattering reflection and so. In MIMO transmits the multiple data at the same time so that the data’s are occurs aliasing, to avoid this aliasing the signal are mapped and the transmitted to the channel it’s very useful to find the distance between the data’s and avoid interference. Fig.5 BER performance of MASM wit QAM modulation D.Comparison with Alamouti’s STBC, STBC-SM and VBLAST schemes Fig .3M-QAM constellation diagram B.Transmit vs. receive diversity STBC uses the diversity techniques have they provide the reliable communication in multi antenna system. In Fig.1 different diversity techniques are used at the transmit vs receiver (no diversity –one transmit antenna, one receive antenna,.Alamouti two transmit antenna , one receive antenna Maximal ratio combiner one transmit antenna, two receive antennas and theoretical second order diversity).in that the Xaxis SNR and Y-axis BER, the SNR is increased and BER are reduced. The maximal ratio combiner provides the better SNR performance compare to other techniques. ISSN: 2231-5381 Fig.6.comparison of different MIMO schemes Comparison with some traditional MIMO schemes are presented here over different transmission rates.fig.6 http://www.ijettjournal.org Page 2006 International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 capacity of rayleigh fading channel using spatial correlation 1 0.9 0.8 0.7 0.6 CDF (C ) gives comparison at 6 bit/s/Hz of MASM with =4, =2with 16-QAM modulation,Alamouti’s STBC uses 1024 QAM modulation,STBC-SM employs 256 QAM modulation and 8 transmit antennas and V-BLAST uses =2 and 32 QAM modulation.it shows that MA-SM provides better SNR gains over V-BKAST,STBC-SM,and Alamouti’s STBC at BER value of 10 ,respectively.it shows that MA-SM becomes more efficient at high transmission rate. D.Decoding complexity comparison 0.5 0.4 0.3 0.2 Comparison of complex ity among sc hemes 0.1 0.045 r1 r2 1.2 r3 r4 0 0.04 6 8 10 12 C(Bits/Sec/(Hz )) 14 16 18 0.035 1 0.03 Fig.8.capacity of Rayleigh fading channel using spatial correlation 2. Mutual coupling 0.8 ra tio ra tio 0.025 0.6 0.02 0.015 0.4 0.01 capacity of rayleigh fading channel using mutual coupling 1 0.2 0.005 0.9 0 6 7 8 9 10 11 R(bits/s/Hz) 12 13 14 15 0 6 7 8 9 10 11 R(bits/s/Hz) 12 13 14 15 0.8 0.7 Fig.7.complexity comparison The complexity of both the transmitter and receiver are taken into consideration.the complexity of the decoder MA-SM is compared to the complexity of the ML decoder and some other optimal decoders for SM and GSM. The number of transmit antennas are needed for target transmission has enormous implication on the on the complexity of the transmitter .because there are multiple antennas being active simultaneously conveying different symbols, the number of antennas needed decreases prominently for a given size of constellation. In fig 5.1 r1 denotes the complexity ratio of optimal MA-SM receiver to optimal GSM decoder under the same target rate R=10, r2 denotes the ratio of optimal MASM decoder to the optimal SM decoder, r3 denotes the ratio of proposed low complexity MA-SM decoder to optimal GSM decoder and r4 denotes the ratio of proposed MA-SM decoder to optimal SM decoder. In this at least 8 transmit antennas are needed in a GSM scheme and 64 antennas are needed in SM system while only 4 transmit antennas are essential in an MA-SM system, so that the complexity is less compared with SM and GSM. E. MA-SM over correlated channel 1. Spatial correlation In multipath wireless communication environment, the wireless channels are not independent from each other but due to scatterings in the propagation paths, the channels are related to each other with different degrees. This kind of correlation is called spatial correlation. ISSN: 2231-5381 C DF (C) 0.6 0.5 0.4 0.3 0.2 0.1 0 4 5 6 7 8 9 C(Bits/Sec/(Hz)) 10 11 12 13 14 Fig.9.Capacity of Rayleigh fading channel using antenna mutual coupling Comparison between spatial correlation and mutual coupling Spatial correlation Mutual coupling dt=0.2500 dt=0.2000 dr=0.1500 dr=0.1500 Capacity=12.7605 Capacity=9.200 V. CONCLUSION AND DISCUSSIONS In this a generalized mapping rules for Quadrature Amplitude Modulation (QAM) constellation .using this mapping function we propose an STBC for four transmit antennas and MASM for 4*2 antennas.MA-SM offers significant improvements of system performance compared with SM ,STBC and V-BLAST systems and it provides the better capacity in the spatial correlation compare to antenna mutual coupling. It concludes that the MA-SM scheme can be useful for high rate wireless communication systems http://www.ijettjournal.org Page 2007 International Journal of Engineering Trends and Technology (IJETT) - Volume4Issue5- May 2013 REFERENCES [1] S. M. Alamouti, “A simple transmit diversity technique for wireless communications,” IEEE J. Sel. Areas Commun., vol. 16, pp. 1451–1458, Oct. 1988. [2] P. Wolniansky, G. Foschini, G. Golden, and R. Valenzuela, “V-blast: an architecture for realizing very high data rates over the rich-scattering wireless channel,” in Proc. 1998 International Symp. Signals, Syst.,Electron., pp. 295– 300. [3] H. Jafarkhani, Space-Time Coding, Theory and Practive. Cambridg University Press, 2005. [4] V. Tarokh, H. Jafarkhani, and A. R. Calderbank, “Spacetime block codes from orthogonal designs,” IEEE Trans. Inf. Theory, vol. 45, no. 5,pp. 1456–1467, July 1999. [5] E.Biglieri, Y. Hong, and E. Viterbo, “On fast-decodable space-time block codes,” IEEE Trans. Inf. Theory, vol. 55, no. 2, pp. 524–530, Feb. 2009. [6] E. Ba¸ar and Ümit Aygölü, “High-rate full-diveristy space-time blocks codes for three and four transmit antennas,” IET Commun., vol. 3, no. 8 pp. 1371–1378, Aug. 2009. [7] “Full-rate full-diversity STBCS for three and four transmit antennas,” Electron. Lett. vol. 44, no. 18, pp. 1076– 1077, Aug. 2008. [8] D. Tse and P. Viswanath, Fundamentals of Wireless Communication Cambridge University Press, 2005. [9] J. Jeganathan, A. Ghrayeb, L. Szczecinski, and A. Ceron, “Space shift keying modulation for MIMO channels,” IEEE Trans. Wireless Commun., vol. 12, pp. 3692–3703, July 2009. [10] R. Mesleh, H. Haas, S. Sinaovic, C. W. Ahn, and S. Yun, “Spatial modulation,” IEEE Trans. Veh. Technol., vol. 57, no. 4, pp. 2228–2241,July 2008. ISSN: 2231-5381 http://www.ijettjournal.org Page 2008