PAPER MIMO System with Relative Phase Difference Time-Shift Modulation for Rician Fading Environment

IEICE TRANS. COMMUN., VOL.E91 B, NO.2 FEBRUARY 2008 459 PAPER MIMO System with Relative Phase Difference Time-Shift Modulation for Rician Fading Environment Kenichi KOBAYASHI, Takao SOMEYA, Student Members, Tomoaki OHTSUKI a), Sigit P.W. JAROT, and Tsuyoshi KASHIMA, Members SUMMARY Multiple-Input Multiple-Output (MIMO) systems that realize high-speed data transmission with multiple antennas at both transmitter and receiver are drawing much attention. In line-of sight (LOS) environments, the performance of MIMO systems depends largely on the difference of the phase difference of direct paths from transmit antennas to each receive antenna. When the phase difference of direct paths are close to each other, the spatial division multiplexing (SDM) channels are not orthogonal to each other so signal detection becomes difficult. In this paper, we propose a MIMO system with relative phase difference time-shift modulation (RPDTM) in Rician fading environments. The proposed scheme transmits independent signals from each antenna at each time slot where the relative phase difference between signal constellations used by transmit antennas varies in a pre-determined pattern. This transmission virtually changes the phase difference of direct paths from transmit antennas to each receive antenna without lowering data rate and without knowledge of the channels. In addition, forward error correction coding (ECC) is applied to exploit the time slots where the receiver can detect the signals easily to improve the detection performance. If there are time slots where the receiver can separate the received signal, the receiver can decode the data by using the time slots and the correlation between data. From the results of computer simulation, we show that MIMO system with RPDTM can achieve the better bit error rate (BER) than the conventional MIMO system. We also show that the MIMO system with RPDTM is effective by about Rician factor K = 10 db. key words: MIMO, Rician fading, direct path, phase difference, Turbo code 1. Introduction Multiple-Input Multiple-Output (MIMO) is a technology to realize a large capacity by transmitting independent signals from each transmit antenna at the same time using the same frequency. The potential application of MIMO is wireless LAN and cellular networks. One challenging task is that the receiver has to separate and detect the signals with the same frequency at the same time. Some broadband wireless systems such as Wireless LAN systems are often used in line-of sight (LOS) environments. In [1] [3], the LOS environment is modeled and its capacity has been analyzed. The LOS environments are Rician fading channels where the direct path exists [2] [7]. In Rician fading channels, since the transmission distances Manuscript received November 15, 2006. Manuscript revised June 8, 2007. The authors are with the Graduate School of Electrical Engineering, Tokyo University of Science, Noda-shi, 278-8510 Japan. The author is with the Department of Information and Computer Science, Keio University, Yokohama-shi, 223-8522 Japan. The authors are with NOKIA JAPAN, Nokia Research Center, Tokyo, 153-0064 Japan. a) E-mail: ohtsuki@ics.keio.ac.jp DOI: 10.1093/ietcom/e91 b.2.459 of direct paths are almost equal, all the channel amplitudes of the direct path are almost equal and each phase of one is different. When the phases of the direct paths are close to each other, the spatial division multiplexing (SDM) channel is not orthogonal to each other so that the signal detection becomes difficult. Therefore, the performance of MIMO system in the LOS environments depends largely on the phase difference of direct paths from transmit antennas to each receive antenna. When the phase difference among direct paths from transmit antennas to each receive antenna is 0 rad, for instance, the different sets of transmit signals results in the same received signal, which cannot be separated and detected at the receiver. In addition, when the phase difference of direct paths between receive antennas is 0 rad, the performance of the signal detection is degraded. Thus, several methods for improving the performance of signal detection in LOS environments have been considered. In [7], the transmission technique utilizing linear combination (LC) diversity is proposed for 2 2 MIMO systems to improve the bit error rate (BER) performance in LOS environments. However, the problem of the system is its low data rate owing to transmission of the same signals over some time slots. Also, the performance improvement of LC diversity is not large. In this paper, we propose a MIMO system with relative phase difference time-shift modulation (RPDTM) in Rician fading environments. The proposed scheme transmits independent signals from each antenna at each time slot where relative phase difference between signal constellations used by transmit antennas varies with a pre-determined pattern. This transmission virtually changes the phase difference of direct paths from transmit antennas to each receive antenna without lowering data rate and without knowledge of the channels. In addition, forward error correction coding (ECC) is applied to exploit the time slots where the receiver can detect the signals easily to improve the detection performance. This is because ECC introduces correlation to data and the receiver decodes them by using the correlation. If there are time slots where the receiver can separate the received signal, the receiver can decode the data by using the time slots and the correlation between data. Furthermore, the performance of the proposed scheme can further be improved by using interference cancellation (IC) with log likelihood ratio (LLR) obtained from soft-decision decoding. In this paper, we use Turbo code as typical ECC to use soft-decision decoding. In our proposed system Turbo Copyright c 2008 The Institute of Electronics, Information and Communication Engineers

460 IEICE TRANS. COMMUN., VOL.E91 B, NO.2 FEBRUARY 2008 code is not mandatory. All ECCs that can be used with softdecision decoding can be applied to our proposed system. From the results of computer simulation, we show that the MIMO system with RPDTM can achieve the better BER than the conventional MIMO system. We also show that the MIMO system with RPDTM is effective by about Rician factor K = 10 db. 2. Rician Fading Channel Model [7] We consider the MIMO system with 2 transmit antennas and 2 receive antennas. Let x = [x 1 (t) x 2 (t)] T denote the transmitted signal vector, where [ ] T denotes matrix transposition and t denotes time index. We assume that E[xx H ] = I,where [ ] H denotes conjugate transposition and I denotes identity matrix. Let n = [n 1 (t) n 2 (t)] T denote additive white Gaussian noise (AWGN) vector so that E[nn H ] = N 0 I.Thereceived signal vector r = [r 1 (t) r 2 (t)] T is given by r = Hx + n, (1) where H denotes 2 2 channel matrix as [ ] h11 h H = 12, (2) h 21 h 22 where h ji denotes channel component from transmit antenna i to receive antenna j. The LOS model is composed of direct path and scattered wave. The channel component is expressed as K 1 h ji = K + 1 h ji,d + K + 1 h ji,s, (3) where h ji,d denotes the direct path of the channel component, h ji,s denotes the scattered wave of the channel component, and K denotes Rician factor that is the power ratio of the direct path and the scattered wave. The scattered waves {h ji,s } 2,2 j=1,i=1 are semi-static frequency-flat and uncorrelated complex Gaussian random variables with zero mean and unit variance. Let d ji denote the propagated distance from transmit antenna i to receive antenna j, λ denote the wavelength, and r ji,d denote reception amplitude. The direct path of the channel component h ji,d is expressed as ( ) 2π jd ji h ji,d = r ji,d exp. (4) λ Now, the amplitude of the direct path is considered. r 11,d r 12,d r 21,d r 22,d can be derived, because it is considered that the propagated distance d ji from the transmit antenna i to the receive antenna j is almost all equal [4]. Therefore, h 11,d h 12,d h 21,d h 22,d is derived. Next, the phase of the direct path is considered. Let denote the phases of direct paths from the transmit antennas 1, 2 to the receive antenna j by θ j1,θ j2. The phase difference of direct paths θ j between the channel components of direct paths for receive antenna j is defined by θ j = θ j1 θ j2. (5) Fig. 1 Example when the different sets of transmit signals results in the same received signal. When θ j is 0 rad, for instance, the different sets of transmit signals result in the same received signal, which cannot be separated and detected at the receiver. Figure 1 shows an example when the different sets of transmit signals results in the same received signal. Signal constellations in the left side show the transmit signal constellation from each transmit antenna and that in the right side shows the received one when LOS channel gives no phase shift and no amplitude change. As we can see, the different sets of transmit signals result in the same received signal. For instance, (s11,s22) and (s12,s21) result in the same received signal so that the receiver cannot separate and detect them. The difference of the phase difference of direct paths ϕ between the receive antennas 1, 2isdefinedby ϕ = θ 1 θ 2 ( π ϕ π) (6) The upper bound on BER of uncoded QPSK is determined by ϕ [7]. P (r,ϕ) = 1 2 erfc r ( ) 1 cos ϕ 2 (7) 2 From (7), the worst BER is with ϕ = 0 rad, while the best BER is with ϕ = ±π rad. Therefore, the performance of MIMO system in the LOS environments depends largely on the difference of the phase difference of direct paths from transmit antennas to each receive antenna. In this paper, we focus on improving the worst BER when ϕ = 0 rad without knowledge of the channel at the transmitter. Note that layouts of antennas and combination of the both direction of receiver and transmitter are infinite. In addition each oscillator of transmitter and receiver is not synchronized perfectly. Thus, there may be few cases where the received signal constellations overlap perfectly. However, there may be some cases where the received signal constellations overlap close to perfectly. Therefore, the above assumption results in the worst case performance.

KOBAYASHI et al.: MIMO SYSTEM WITH RELATIVE PHASE DIFFERENCE TIME-SHIFT MODULATION 461 3. MIMO System with Relative Phase Difference Time- Shift Modulation 3.1 Relative Phase Difference Time-Shift Modulation (RPDTM) The proposed scheme transmits independent signals from each antenna where relative phase difference between signal constellations used by transmit antennas varies with predetermined pattern. A pre-determined phase shift p i (t) for the transmitted signal x i (t) from the transmit antenna i at time t is defined by (t mod T) p i (t) = P i T 1, t = 0,, L f 1 (8) where P i denotes the maximum phase shift for the transmit antenna i, L f denotes the frame length, and T denotes a period of change of relative phase difference between signal constellations. For instance, we consider MIMO with 2 transmit and 2 receive antennas where each transmit antenna uses QPSK, and the period of change of relative phase difference between signal constellations is T = 2. From transmit antenna 1, signals are transmitted in π/4-shift QPSK constellation, while different signals are transmitted in QPSK constellation from transmit antenna 2. The relative phase difference between signal constellations changes periodically, like 0,π/4, 0,π/4,. Thus, the receiver can detect the signals easier at least in the time slots with relative phase difference either 0 or π/4. As mentioned above, this transmission virtually changes the phase difference of direct paths from transmit antennas to each receive antenna without lowering data rate and without knowledge of the channels. 3.2 Transmitter A transmitter model of the proposed scheme is shown in Fig. 2. ECC is applied to exploit the above-mentioned time slots where the receiver can detect the signals easily to improve the detection performance. First, the transmitter encodes the information sequence independently for each antenna. Then, the encoded sequence is passed through the interleaver and the modulator. Finally, signals with a predetermined phase shift are transmitted from transmit antenna 1, while signals without a pre-determined phase shift are transmitted from transmit antenna 2. 3.3 Receiver A receiver model of the proposed scheme is shown in Fig. 3. The receiver has the iterative configuration based on IC. In the initial detection, the received signals are detected by maximum likelihood detection (MLD). MLD is known as an optimal symbol detection algorithm in terms of detection performance and chooses the symbol combination that minimizes the squared Euclidean distance so that ˆx = arg min r Ha 2, (9) a where a denotes all the possible symbol combinations. and the received signal r = [r 1 (t) r 2 (t)] T with RPDTM is expressed as [ ] r1 (t) r 2 (t) [ ][ h11 h = 12 x1 (t)exp h 21 h 22 x 2 (t) ( Pi (t mod T) T 1 ) ]. (10) Then, the signal streams are passed through de-interleaver and Turbo decoder independently for each antenna, and the information sequence is decoded. We set the number of iterations for the initial detection to I = 0. In the iterative detection, the detection order is determined based on the sum of LLR for bit of each signal stream. In the iterative process, the receiver first detects one signal stream with high reliability and decodes it. The replica is generated from the decoded signals. The replica of x i (n) from transmit antenna i at time n is given by x i (n) = E[x i (n)] = ˆxP (x i (n) = ˆx) ˆx A = ˆx A [1 + exp( ( 1) { ˆx} j c i, j )] 1, (11) log 2 A ˆx j=1 where A denotes the constellation, A shows total number of signal points in the constellation, ˆx denotes one of the symbols, { ˆx} m is the mth bit in the symbol ˆx,and c i, j denotes LLR obtained by Turbo decoder. Then, the replica is subtracted from the received signals. The other signal stream is detected and decoded, and its replica is generated. The replica is subtracted from the received signal, and the signal Fig. 2 Transmitter model of MIMO system with RPDTM. Fig. 3 Receiver model of MIMO system with RPDTM.

462 IEICE TRANS. COMMUN., VOL.E91 B, NO.2 FEBRUARY 2008 stream detected first is detected and decoded again. When all the signal streams are decoded, the number of iterations for the iterative detection is updated as I I + 1. This process is performed iteratively until the performance converges. 4. Simulation Results We evaluate the BER performance of the proposed MIMO system with RPDTM in Rician fading environment by computer simulation. Table 1 lists the simulation parameters. The phase difference of direct paths for receive antenna 1 is equal to that of direct paths for receive antenna 2 because we assume that the difference of the phase difference of direct paths between receive antennas is ϕ = 0 rad. Thus, we omit the antenna index of the phase difference of direct paths and denote it by θ. The phase difference of direct paths is set to θ = 0,π/6rad, where θ = 0 rad is the case when it is most difficult to detect the signals for ϕ = 0rad, and θ = π/6 rad is the case when it is easiest to detect the signals for ϕ = 0 rad [7]. The phase difference of direct paths is set to be uniformly distributed, where the phase difference of direct paths is assumed to be unknown at the transmitter. 4.1 Performance of the Uncoded MIMO System with RPDTM Figure 4 shows BER versus E b /N 0 for the uncoded MIMO system with RPDTM for K = 10 db. In Fig. 4, we set the maximum phase shift for the transmit antenna 1 to P 1 = π/4 rad and the period of change of relative phase difference between signal constellations to T = 2. From Fig. 4, for θ = 0 rad that is the worst environment to detect the signals, the BER of the MIMO system with RPDTM is improved compared to that of the conventional MIMO system. On the other hand, for θ = π/6 rad that is the easiest environment Table 1 Simulation parameters. Number of Tx Antennas N t 2 Number of Rx Antennas N r 2 Modulation QPSK Phase Difference of Direct Paths θ 0,π/6rad, uniform distribution ( π rad θ π rad) Difference of Phase Difference of Direct Paths between 0rad Receive Antennas ϕ Rician Factor K 3, 10, 20 db Period of Change of Relative Phase Difference T 2, 4 slots Error Correcting Codes Turbo Code (constraint length = 3) Code Rate 1/2 Code length 1000 bits Interleaver S-Random Decoding Algorithm MAP Algorithm Number of Decoding Iterations 8 Frame Length L f 500 symbols Channel Estimation Ideal to detect the signals, the relative phase difference of direct paths from transmit antennas at the time slot with phase shift becomes P θ = π/12 rad that is not the best environments with θ = π/6 to detect the signals. Thus, the BER of the MIMO system with RPDTM is degraded compared to that of the conventional MIMO system. For the phase difference of direct paths that is uniformly distributed, the phase difference of direct paths at all the time slots changes independently. We consider that the occurrence probability of the time slot where it is easy to detect the signals is statistically equal to that of the time slot where it is difficult to detect in terms of the phase difference of direct paths. Thus, the BER of the MIMO system with RPDTM is almost equal to that of the conventional MIMO system when the phase difference of direct paths that is uniformly distributed. 4.2 Performance Evaluation of the Proposed MIMO System with RPDTM for Rician Factor K = 10 db In this subsection, we evaluate the performance of the proposed MIMO system with RPDTM for Rician factor K = 10 db. Figure 5 shows BER versus the number of iterations I of IC for MIMO system with RPDTM (θ = 0rad, P 1 = π/4rad, T = 2). I = 0 denotes the initial detection without IC. From Fig. 5, we can see that BERs of MIMO system with RPDTM are improved by using IC. This is because IC can remove the interference from other transmit antennas remaining in the decoder output. We can also see that two iterations is sufficient for the BER performance to converge. Thus, in the following we use the number of iterations I = 2. Figure 6 shows BER versus maximum phase shift for the transmit antenna 1 for MIMO system with RPDTM (θ = 0 rad). From Fig. 6, we can see that maximum phase shift P 1 = π/4 rad is optimal for T = 2. This is because for P 1 = π/4 rad, there is at least one time slot where it is easy to detect the signals. We can also see that maximum phase Fig. 4 BER versus E b /N 0 for the uncoded MIMO system with RPDTM (K = 10 db, P 1 = π/4rad, T = 2).

KOBAYASHI et al.: MIMO SYSTEM WITH RELATIVE PHASE DIFFERENCE TIME-SHIFT MODULATION 463 Fig. 5 BER versus the number of Iterations of IC for MIMO system with RPDTM (K = 10 db, θ = 0rad, P 1 = π/4rad, T = 2). Fig. 7 BER versus E b /N 0 for MIMO system with RPDTM (K = 10 db, θ = 0rad). Fig. 6 BER versus maximum phase shift for the transmit antenna 1 for MIMO system with RPDTM (K = 10 db, θ = 0rad). shift θ = π/3 rad is optimal for T = 4. This is because for P 1 = π/3 rad, there are more time slots where it is easy to detect the signals than the others. That is, for P 1 = π/3rad and the pre-determined phase shift (0, π/9, 2π/9, π/3) rad, the phase shifts of the time slots 1 and 2 of π/9, 2π/9rad are close to θ = π/6 rad that is the case when it is easy to detect the signals. On the other hand, for example, for P 1 = π/2 rad and the pre-determined phase shift pattern (0,π/6,π/3,π/2rad), there is only one time slot where it is easy to detect the signals with the phase shift of π/6rad. Figures 7 9 show BER versus E b /N 0 for MIMO system with RPDTM. To benchmark MIMO system with RPDTM, we consider 1 2 Turbo coded MRC scheme. The modulation scheme of MRC scheme uses 16QAM to make the transmission rate the same as 2 2MIMOsystem with RPDTM. From Fig. 7, we can see that BER of Fig. 8 BER versus E b /N 0 for MIMO system with RPDTM (K = 10 db, θ = π/6rad). MIMO system with RPDTM can be improved compared to that of MRC by using IC with the soft replica generated from soft output of Turbo decoder. We can also see that for θ = 0 rad that is most difficult case to detect the signals, BER of MIMO system with RPDTM can be improved compared to that of the conventional MIMO system. This is because by Turbo code, MIMO system with RPDTM can exploit the time slots where the receiver can detect the signals easily to improve the detection performance. From Fig. 8, we can see that for θ = π/6 rad that is the easiest case to detect the signals, BER of MIMO system with RPDTM can be improved compared to that of the conventional MIMO system, though BER of the uncoded MIMO system with RPDTM is degraded compared to that of the uncoded conventional

464 IEICE TRANS. COMMUN., VOL.E91 B, NO.2 FEBRUARY 2008 Fig. 9 BER versus E b /N 0 for MIMO system with RPDTM (K = 10 db, θ = uniform distribution). Fig. 10 BER versus E b /N 0 for MIMO system with RPDTM (K = 3dB, θ = 0rad). MIMO system in Fig. 4. This is because coding gain can be obtained because of virtual fluctuation of channels by phase shift. From Fig. 9, we can see that BER of MIMO system with RPDTM can be improved compared to that of the conventional MIMO system when the phase difference of direct paths that is uniformly distributed. It is clear that for the phase difference of direct paths that is uniformly distributed, MIMO system with RPDTM can achieve better BER than conventional MIMO system, because MIMO system with RPDTM can achieve better BER than conventional MIMO system irrespective of the phase difference of the direct paths. From Figs. 7 9, we can see that BERs of MIMO system with RPDTM with T = 4 are slightly improved compared to those of MIMO system with RPDTM with T = 2. This is because with T = 4, there exists more slots where it is easy to detect the signals than with T = 2. 4.3 Performance Evaluation of the Proposed MIMO System with RPDTM for Rician Factor K = 3, 20 db In this subsection, we evaluate the performance of the proposed MIMO system with RPDTM for Rician factor K = 3, 20 db. Figures 10 and 11 show BER versus E b /N 0 for MIMO system with RPDTM for Rician factor K = 3, 20 db, respectively. From Fig. 10, we can see that BER of MIMO system with RPDTM can be improved compared to that of MRC. However, BERs of MIMO system with RPDTM are not improved compared to that of the conventional MIMO system. This is because time slots where it is easy to detect the signals are not produced by RPDTM when the power ratio of the direct path and the scattered wave is low like K = 3 db. On the other hands, from Fig. 11 we can see that BERs of MIMO system with RPDTM are largely improved compared to that of the conventional MIMO system when K = 20 db. However, we can see that BER of MIMO system Fig. 11 BER versus E b /N 0 for MIMO system with RPDTM (K = 20 db, θ = 0rad). with RPDTM is degraded compared to that of MRC. From the results, it can be said that the our proposed scheme is effective by about K = 10 db. 5. Conclusions In this paper, we proposed a MIMO system with RPDTM that offers improved signal detection performance in Rician fading environments. The proposed scheme transmits independent signals from each antenna at each time slot where relative phase difference between signal constellations used by transmit antennas varies with a pre-determined pattern. This transmission virtually changes the phase difference of direct paths from transmit antennas to each receive antenna without lowering data rate and without knowledge of the channels. In addition, ECC was applied to exploit the time slots where the receiver can detect the signals easily to improve the detection performance. If there

KOBAYASHI et al.: MIMO SYSTEM WITH RELATIVE PHASE DIFFERENCE TIME-SHIFT MODULATION 465 are time slots where the receiver can separate the received signal, the receiver can decode the data by using the time slots and the correlation between data. From the results of computer simulation, we showed that the MIMO system with RPDTM can achieve the better BER than the conventional MIMO system for the phase difference of direct paths θ = 0,π/6rad, and the phase difference of direct paths that is uniformly distributed. We also showed that the MIMO system with RPDTM is effective by about Rician factor K = 10 db. Note that [8] has similar effects as our proposed system. However, the system in [8] needs to design interleaver and de-interleaver suitable for the channels, such as interleaver depth, while our proposed system does not need those. References [1] P.F. Driessen and G.J. Foschini, On the capacity formula for multiple input multiple output wireless channels, IEEE Trans. Commun., vol.47, no.2, pp.173 176, Feb. 1999. [2] D. Gesbert, H. Bolcskei, D.A. Gore, and A.J. Paulraj, Outdoor MIMO wireless channel: Models and performance prediction, IEEE Trans. Commun., vol.50, no.12, pp.1926 1934, Dec. 2002. [3] D. Gesbert, H. Bolcskei, D.A. Gore, and A.J. Paulraj, MIMO wireless channel: Capacity and performance prediction, Proc. IEEE PIMRC 2004, vol.2, pp.1126 1130, Sept. 2004. [4] Y. Murakami, K. Kobayashi, M. Orihashi, and T. Matsuoka, Analysis of BER performances using channel matrix eigenvalue in MIMO systems under Rayleigh and Rician fading channels, Proc. IEEE PIMRC 2004, vol.2, pp.1126 1130, Sept. 2004. [5] Y. Li, P.H.W. Fung, Y. Wu, and S. Sun, Bit-interleaved coded modulation in linear dispersion coded MIMO system over spatially correlated Rician fading channel, Global Telecommunications Conference, 2004. GLOBECOM 04. IEEE, vol.1, pp.5 9, Dec. 2004. [6] Y. Murakami, K. Kobayashi, T. Matsuoka, and K. Takahashi, Performance analysis based on channel matrix eigenvalue for MIMO systems in LOS environments, IEICE Trans. Fundamantals, vol.e88-a, no.10, pp.2926 2936, Oct. 2005. [7] Y. Murakami, K. Kobayashi, T. Matsuoka, and K. Takahashi, Design of transmission technique utilizing linear combination diversity in consideration of LOS environments, IEICE Trans. Fundamantals, vol.e88-a, no.11, pp.3127 3133, Nov. 2005. [8] K. Kobayashi, Y. Murakami, M. Orihashi, and T. Matsuoka, Varying interleave patterns with iterative decoding for improved performance in MIMO systems, Personal Indoor and Mobile Radio Communications (PIMRC2004), vol.2, pp.1429 1433, Sept. 2004. Takao Someya received the B.S. and M.E. degrees in electrical engineering from Tokyo University of Science, Noda, Japan, in 2004 and 2006, respectively. Tomoaki Ohtsuki received the B.E., M.E., and Ph.D. degrees in Electrical Engineering from Keio University, Yokohama, Japan in 1990, 1992, and 1994, respectively. From 1994 to 1995 he was a Post Doctoral Fellow and a Visiting Researcher in Electrical Engineering at Keio University. From 1993 to 1995 he was a Special Researcher of Fellowships of the Japan Society for the Promotion of Science for Japanese Junior Scientists. From 1995 to 2005 he was with Science University of Tokyo. In 2005 he joined Keio University. He is now an Associate Professor at Keio University. From 1998 to 1999 he was with the department of electrical engineering and computer sciences, University of California, Berkeley. He is engaged in research on wireless communications, optical communications, signal processing, and information theory. Dr. Ohtsuki is a recipient of the 1997 Inoue Research Award for Young Scientist, the 1997 Hiroshi Ando Memorial Young Engineering Award, Ericsson Young Scientist Award 2000, 2002 Funai Information and Science Award for Young Scientist, IEEE the 1st Asia-Pacific Young Researcher Award 2001, and the 5th International Communication Foundation (ICF) Research Award. He is an Editor of IEICE Transactions on Communications and a Technical Editor of IEEE Wireless Communications Magazine. He is a senior member of the IEEE and a member of the SITA. Sigit P.W. Jarot received the B.E., M.E. and PhD.Eng. degrees from Keio University, Yokohama, Japan, in 1998, 2000 and 2003 respectively. From 2000 to 2003, he was a research associate in the Department of Information and Communication Science, Keio University. In 2003, he joined Nokia Research Center, Nokia Japan Co. Ltd. Currently, he is a member of research staff. His research interests include mobile communication systems particularly 3GPP/LTE (long term evolution), and visible light communication. Kenichi Kobayashi received the B.S. and M.E. degrees in electrical engineering from Tokyo University of Science, Noda, Japan, in 2005 and 2007, respectively. Tsuyoshi Kashima received the B.S. and M.S. degrees in theoretical physics from Tokyo University, Tokyo, Japan, in 1998, and 2000 respectively. Since joining Nokia Research Center, Tokyo in 2000, he was engaged in system level research on mobile radio communications including cellular and multi-hop networks, and is now engaged mainly in MAC layer development work for 3GPP LTE standardization. Since 2004, he has also started Ph.D. study in Tokyo Institute of Technology, Tokyo, Japan with continuing his work in Nokia. His Ph.D. research topics are on the optimal receiver architecture utilizing iterative joint processing. Messagepassing algorithm and Monte Carlo method are especially in his interest. Mr. Kashima is a member of IEEE. He received the Paper Award from IEICE in 2007.