Tokyo Wireless Technology Summit ~Wireless Technologies Enabling Breakthrough Towards The Future~, 7 March, 2014, Waseda University, Tokyo, Japan

Transcription

1 Tokyo Wireless Technology Summit ~Wireless Technologies Enabling Breakthrough Towards The Future~, 7 March, 2014, Waseda University, Tokyo, Japan Toward Spectrum-Energy Efficiency of Wireless Networks Fumiyuki Adachi (), Distinguished Professor Wireless Signal Processing & Networking (WSP&N) Lab., Dept. of Communications Engineering, Graduate School of Engineering, Tohoku University, Japan adachi@ecei.tohoku.ac.jp Acknowledgment: Special thanks to members of Wireless Signal Processing & Networking (WSP&N) Lab. The secret to humor is surprise@facebook OUTLINE Wireless Evolution Challenges for 5G Wireless Toward Green Wireless Concluding Remarks 2014/03/07 FA/Tohoku University 1 Wireless Evolution 2014/03/07 FA/Tohoku University 2

2 Wireless Has Been Continuously Evolving In early 1980 s, communications systems changed from fixed point-to-point to wireless anytime, anywhere communication. Cellular systems have evolved from narrowband network of around 10kbps to wideband networks of around 10Mbps. Now on the way to broadband networks of 100Mbps (LTE). Service type Voice Multimedia point -topoint 0G Voice only 1G Narrowband Era 2G ~2.4kbps ~64kbps Wideband Era LTE-A Year We are here 2014/03/07 FA/Tohoku University 3 3G 4G 3.5G 3.9G ~2Mbps 100M~1Gbps 50~100Mbps ~14Mbps W-CDMA CDMA2000 TD-SCDMA HSDPA (W-CDMA) 3G LTE Broadband Era 5G >1Gbps Gigabit wireless Wireless Has Been Continuously Evolving There was a big technical leap from 2G to 3G systems. Voice+data Voice+Data 1G Analog (FDMA) ~2.4kbps 2G Digital (TDMA) ~64kbps Big leap 3G/3.5G Digital (CDMA) ~2Mbps ~14Mbps 3.9G (3G LTE) 4G (LTE-A) Improved frequency utilization Narrowband Increased no. of voice-band channels Broadband Increased peak rate (increased throughput) 3.5G (HSPA,5MHz) 3.9G (LTE,~20MHz) 4G (LTE-A, ~100MHz) Up Down Up Down Up Down 2014/03/07 FA/Tohoku University Mbps 14.4 Mbps 75 Mbps 300 Mbps 15bps /Hz 30bps /Hz

3 Networked Society People are always connected to networks Society is relying on communications networks Unlimited desire for data rate, but limited wireless resources and a wide range of mobility High mobility user Internet Very high mobility user Stationary user Low mobility user 2014/03/07 FA/Tohoku University 5 Future Wireless Services Problems Limited bandwidth Limited power Cloud Computing Network A variety of data services through Internet Wireless Access Network Gigabit wireless pipe (>1Gbps/user) User terminals with high quality display and gigabit wireless processing 2014/03/07 FA/Tohoku University 6

4 Explosive Growth of Mobile Traffic (x1,000 in 10 years) Traffic volume Voice>Data 3G/HSPA (14Mbps) 1 VoiceData Technology Evolution 3.9G/LTE (300Mbps) Voice<<Data, Video 4G/LTE-A (3Gbps) /03/07 FA/Tohoku University 7 LTE-Advanced May Not Be Sufficient LTE-advanced (4G) networks are expected to provide broadband packet data services of up to 1Gbps/BS In December 2007, ITU allocated 3.4~3.6GHz band for 4G services. Only 200MHz is available for global use. Although one-cell reuse of 100MHz is possible, an effective bandwidth (around 25% of total) which can be used at each BS is only around 12.5MHz/link. 1Gbps/12.5MHz is equivalent to 80bps/Hz/BS!! 5G networks may require >>1Gbps/BS capability. Development of advanced wireless techniques that achieve a spectrum efficiency of >>80bps/Hz/BS is demanded. 3.5G (HSPA,5MHz) 3.9G (LTE,~20MHz) 4G (LTE-A, ~100MHz) Up Down Up Down Up Down 14.4 Mbps 14.4 Mbps 75 Mbps 300 Mbps 15bps/ Hz 30bps/ Hz 2014/03/07 FA/Tohoku University 8

5 Important Technical Issues for 5G Spectrum issue MIMO antenna multiplexing to increase bps/hz But, more importantly Frequency reuse to increase bps/hz/bs or bps/hz/km 2 Signal Energy issue Single-carrier waveform with reduced peak power But, this is not enough Doubly-selective channel issue Frequency-selective channel: frequency-domain equalization (FDE) Time-selective channel: Higher frequency band will be used Quite high tracking ability against high Doppler shifts is necessary 2014/03/07 FA/Tohoku University 9 Challenges For 5G Wireless Spectrum Issue Energy Issue Channel Issue 2014/03/07 FA/Tohoku University 10

6 Spectrum Issue 2014/03/07 FA/Tohoku University 11 Frequency Reuse Because of limited available bandwidth, the same frequency must be reused From the spectrum efficiency (bps/hz/km 2 )pointofview, the same frequency needs to be reused at locations as close as possible Co-channel interference is a limiting factor on frequency efficiency Co-channel interference management becomes a crucial issue to realize spectrum efficient broadband networks 2014/03/07 FA/Tohoku University 12

7 Peak data rate per BS B S However, the available bandwidth B may be around 100MHz only How to improve SE? MIMO may be a savior bps Hz km 2014/03/07 FA/Tohoku University 13 Is Super-high Level Modulation Helpful? To achieve 1Gbps/BS using 12.5MHz bandwidth, the required spectrum efficiency is 1Gbps/12.5MHz=80bps/Hz/BS! We need to adopt 2 80 QAM. However, use of extremely high level of modulation is impractical due to strong cochannel interference. Signal constellation 2014/03/07 16QAM (4bps/Hz) 256QAM (8bps/Hz) 1024QAM (10bps/Hz) FA/Tohoku University 14

8 MIMO May Be A Savior Independent data streams are transmitted simultaneously from transmit antennas using the same carrier frequency. Spatial multiplexing is to increase achievable data rate with the limited bandwidth, i.e., the channel capacity in bps/hz. N t antennas N r antennas Multipath channel Coding S/P Signal detection Decod. 2014/03/07 FA/Tohoku University 15 However, MIMO Cannot Solve Energy Issue Transmit power is another important issue N t F N t F d t d bsmt N bsmt N Distance between BS and MT r r N N N t r t Received SINR N N N t r r N N r t n n r N S t t h n n r t N r S 2014/03/07 FA/Tohoku University 16

9 Signal Energy Issue 2014/03/07 FA/Tohoku University 17 Communication Range Shrinks For broadband communications, communication range shrinks significantly because of the transmit power limitation. Fundamental change is necessary in wireless access network structure. Core Network Radio control station Base station Core Network Base station 2014/03/07 FA/Tohoku University 18

10 Uniform Quality Is The Target Uniform quality over an BS area Uniform quality over an BS area Throughput Present cellular systems Cell edge BS Distance from BS 2014/03/07 FA/Tohoku University 19 Green Wireless Until LTE, much effort has been paid to improving the SE Recently, strong attention has been paid not only to SE but also to EE Future broadband wireless networks require significant improvement of both SE and EE How to simultaneously improve SE and EE? 2014/03/07 FA/Tohoku University 20

11 Small-cell Network Simultaneous improvement of SE and EE Reducing the cell radius by a factor of 30 (1,000m 30m) X1,000 capacity increase Reduced transmit power by a factor of 150,000 1km 30m x 1,000 capacity increase 2014/03/07 FA/Tohoku University 21 Another Important Issue: Doubly-selective Channel Signal waveform design: singlecarrier or multi-carrier? Simple one-tap frequency-domain equalization (FDE) 2014/03/07 FA/Tohoku University 22

12 Doubly-selective Channel Transmitted radio waves are reflected or diffracted by some large buildings, creating resolvable paths having time delays of multiple of (signal bandwidth) -1 Each resolvable path is the sum of irresolvable paths created by local scatterers surrounding a mobile The path gain h l (t) varies in time according to the movement of mobile terminal since resolvable paths are added constructively at one time and destructively at another time d -4 Large obstacles Local scatterers Transmitter Reflection/ diffraction Receiver 2014/03/07 FA/Tohoku University 23 Frequency-selective Channel The transfer function H(f, t) of broadband channel at time t is not constant and varies over the signal bandwidth This channel is called the frequencyselective channel Advanced equalization technique is necessary H f L l h jf l L=16 uniform power delay profile with l-th path time delay =100l + [-50,50)ns l 2014/03/07 FA/Tohoku University 24

13 Why Single-carrier (SC) Transmission for Uplink? Nyquist-filtered SC signal has lower PAPR than OFDM No ISI at the transmitted waveform due to Nyquist filtering SC is suitable for the uplink transmission OFDM Less expensive power amplifier is required 256 subcarriers st fct stf t st f t c c 2014/03/07 FA/Tohoku University 25 SC SC-FDE Block transmission of N c symbols Insertion of cyclic prefix (CP) at the transmitter FFT/IFFT at the receiver Simple one-tap FDE Receiver Transmitter Transmit data block Nc-point IFFT Data demod. W(0) W(k) Received data block W(N c -1) Data mod. FDE Nc-point FFT +CP -CP *H. Sari, G. Karam, and I. Jeanclaude, "Transmission Techniques for Digital Terrestrial TV Broadcasting," IEEE Commun. Mag., vol. 33, pp , February *D. Falconer, S. Ariyavisitakul, A, Benyamin-Seeyar and B. Eidson, Frequency Domain Equalization for Single-Carrier Broadband Wireless Systems, IEEE Communications Magazine, Vol. 40, No. 4, pp , April *F. Adachi, D. Garg, S. Takaoka, and K. Takeda, Broadband CDMA techniques, IEEE Wireless Commun. Mag., Vol. 12, No. 2, pp. 8-18, April /03/07 FA/Tohoku University 26

14 Towards Green Wireless 2014/03/07 FA/Tohoku University 27 Green Wireless Until LTE-A, much effort has been paid to improve to the spectrum efficiency. However, signal energy efficiency is becoming more and more important Unfortunately, spectrum and energy efficiencies are in a tradeoff relationship Improving both spectrum and energy efficiencies at the same time is an important technical issue Spectrum efficiency (bps/hz/km 2 ) Interference limited 5G Noise limited SE-EE tradeoff curve Energy efficiency (bit/j) 2014/03/07 FA/Tohoku University 28

15 Small-cell Network Simultaneous improvement of SE and EE Reducing the cell radius by a factor of 30 (1,000m 30m) X1,000 capacity increase Reduced transmit power by a factor of 150,000 Small-cell network 1km 30m x 1,000 capacity increase Near single-user access for increasing the bandwidth/user and hence, user data rate Millimeter wave bands can also be used Near single user per BS 2014/03/07 FA/Tohoku University 29 Green Wireless High spectrum efficiency Single-cell frequency reuse to boost bps/hz/km 2 High energy efficiency A few mw for a few 10Mbps How to achieve the above simultaneously while avoiding wireless control signaling problem? Restructuring cellular networks is necesaary 2014/03/07 FA/Tohoku University 30

16 Distributed Antenna Network Distributed MIMO technology Hybrid waveform with reduced peak power (extreme case is the single-carrier waveform) Frequency-domain signal processing (equalization) 2014/03/07 FA/Tohoku University 31 Distributed Antenna Network Distributed antenna network (DAN) is designed to realize a nano-cell network with simultaneously increased spectrum and energy efficiencies. Many antennas belonging to a base station (SPC: signal processing center) are distributed around SPC. Each distributed antenna forms a cell Resource allocation control (frequency, time, power) is carried out by SPC. F. Adachi, K. Takeda, T. Obara, T. Yamamoto, and H. Matsuda,"Recent Advances in Singlecarrier Frequency-domain Equalization and Distributed Antenna Network," IEICE Trans. Fundamentals, Vol.E93-A, No.11, pp , Nov F. Adachi, K. Takeda, T. Yamamoto, R. Matsukawa, and S. Kumagai,"Recent Advances in Single-carrier Distributed Antenna Network,," Wireless Communications and Mobile Computing, Volume 11, Issue 12, pp , Dec. 2011, doi: /wcm F. Adachi, W. Peng, T. Obara, T. Yamamoto, R. Matsukawa, and M. Nakada, Distributed Antenna Network for Gigabit Wireless Access, International Journal of Electronics and Communications (AEUE), Elsevier, Vol. 66, Issue 6, pp , 2012, DOI: /j.aeue /03/07 FA/Tohoku University 32

17 Distributed Antenna Network Ubiquitous antennas as an entrance to core network Co-located antennas Path loss Shadowing loss Multipath fading Multi-access or Multi-user detection Distributed antennas 2014/03/07 FA/Tohoku University 33 Path loss Shadowing loss Multipath fading Single-user access Conceptual Structure of DAN Short range communication combined with single-user access is the crucial requirement! Many antennas are spatially distributed around a signal processing center (SPC), which is a gateway to the network Distributed MIMO multiplexing/di versity/relay Virtual wireless transceiver Antennas are connected with a SPC by optical links With a high probability, some antennas can always be visible from MT Distributed antenna layer to form a user centric cell Coherent optical link SPC SPC Signal processing layer SPC SPC 2014/03/07 FA/Tohoku University 34

18 TDD Allows Introduction of Transmit Equalization TDD can exploit the channel reciprocity to introduce the transmit equalization without the CSI feedback TDD Frame Uplink Down link time Transmit equalization SPC Channel estimation Receive equalization Uplink 2014/03/07 FA/Tohoku University 35 Downlink Cooperative Diversity (STBC-JTRD) Space-Time Block Coded Joint Transmit/Receive Diversity (STBC-JTRD) is suitable for downlink application since it allows an arbitrary number of transmit antennas. Transmit FDE (channel state information (CSI) is necessary at TX) to obtain frequency-diversity gain. Only simple addition/subtraction and complex conjugation operations are required at the receiver. #N dan -1 Data mod. FFT STBC-JTRD encoder SPC H.Tomeba,K.TakedaandF.Adachi, Space- Time Block Coded Joint Transmit/Receive Diversity in a Frequency-Nonselective Rayleigh Fading Channel, IEICE Trans. #N mt -1 Commun., Vol.E89-B, No.8, pp , #0 Aug N mt receive 2014/03/07 antennas FA/Tohoku University IFFT +CP CP #1 #0 FFT N dan distributed antennas Mobile Terminal STBC- JTRD decoder IFFT Data demod. 36

19 Uplink Cooperative Diversity (FD-STTD) Estimated symbol-blocks s 0 (t)^ s 1 (t)^ #N dan -1 Data demod. FFT + STTD decoding with FDE FFT -CP #0 #1 N dan distributed antennas N mt =2 st st s s DAN SPC N c tnc N c tnc #1 #0 +CP +CP 2 symbol-blocks STTD encoding S.M. Alamouti, A simple transmit diversity technique for wireless communications, IEEE Journal on Selected Areas in Communications, Vol. 16, No. 8, pp , October Mobile Terminal 2014/03/07 FA/Tohoku University 37 s 0 (t) Spatial Distribution of Downlink Throughput DAN can achieve much higher throughput s 1 (t) throughput at the cell edge is about 2.6 (bps/hz) in DAN while it is about 1.2 (bps/hz) in CN. Type II S-P2 (Incremental Redundancy) using rate-1/3 turbo coding. Data mod. Throughput (bps/hz) QAM E s /N 0 =0 (db) (N dan,n mt )=(4, 2) MMSE weight 1.0 Spatial distribution of throughput when (N dan, N mt ) = (4, 2) for DAN using 7 distributed antennas, 16QAM, and E s /N 0 =0 (db) 2014/03/07 FA/Tohoku University 38

20 Cooperative Multiplexing Distributed antennas can be used for spatial multiplexing to significantly increase the data rate. DAN SPC Distributed antenna 2014/03/07 FA/Tohoku University 39 Cooperative Multiplexing TS-SC MIMO multiplexing using QRM-MLBD N dan N c N g N c N g Ndan H T. Yamamoto, K. Takeda, and F. N mt Adachi, Training sequence-aided QRM- MLD block signal detection for singlecarrier MIMO spatial multiplexing, Nmt Nmt Proc. IEEE International Conference on Communications (ICC 2011), Kyoto, Japan, 5-9 June, /03/07 FA/Tohoku University 40

21 Spatial Distribution of Uplink Throughput DAN achieves higher throughput than CN (note that CN can achieve high throughput only near SPC). (a) DAN (b) CN *TS-SC MIMO multiplexing (N c =64 and N g =16) *QRM-MLBD(M=16), 16QAM *Turbo-coded HARQ type-ii S-P4 strategy *Packet size=2048. *(N dan, N mt )=(2,2), E s /N 0 =5dB *An L=16-path frequency-selective block Rayleigh fading channel with uniform power delay profile 2014/03/07 FA/Tohoku University 41 DAN Is Not Almighty DAN formulate user-centric personal cells Short range communication link with almost single user access Frequency reuse at short distance Role of signal processing center (SPC Density of SPCs is similar to traditional cellular network Wireless signal processing such as space-time coding, equalization, etc Resource allocation (frequency, time, power) to distributed antennas Coherent optical SPC-antenna links Wireless and optical convergence on signal processing and communications However, traffic density is not uniform and hence, DAN is not almighty 2014/03/07 FA/Tohoku University 42

22 Heterogeneous Network Heterogeneous network is a realistic approach Small-cell network (e.g., DAN) to cover hot-spot area Large-cell network (3G, LTE) to cover wide area Access control High mobility users Core network Hot spot area SPC SPC 2014/03/07 FA/Tohoku University 43 Concluding Remarks 5G requires energy & spectrum efficient network Heterogeneous network is a realistic approach Small-cell network (e.g. DAN) to provide short range communications High speed data services Significantly reduced signal energy New frequency band e.g., millimeter wave bands Large-cell network is still necessary Call control signaling High mobility users M2M trafficlow data rate but mbillions of devices Improved dependability Simultaneous operation of different types of networks Acknowledgment Special thanks to members of Wireless Signal Processing & Networking (WSP&N) Lab. 2014/03/07 FA/Tohoku University 44