5G Waveform Approaches In Highly Asynchronous Settings Presenter: Gerhard Wunder, gerhard.wunder@hhi.fraunhofer.de EuCNC Workshop Enablers on the road to 5G June 23rd, 2014
What is 5GNOW? 5GNOW (5 th Generation Non-Orthogonal Waveforms for Asynchronous Signalling) is an European collaborative research project supported by the European Commission within FP7 ICT Call 8. Who is in the consortium? Fraunhofer HHI (coordinator), Germany, Dr. Gerhard Wunder Alcatel Lucent (technical coord.), Germany, Thorsten Wild Technische Universität Dresden, Germany, Prof. Gerhard Fettweis CEA-LETI, France, Dr. Dimitri Ktenas IS-Wireless, Poland, Dr. Slawomir Pietrzyk National Instruments, Hungary, Dr. Bertalan Eged www.5gnow.eu, LinkedIn group Vision: 5GNOW is the physical layer evolution of mobile communication network technology such as LTE-Advanced towards emerging application challenges. 28.06.2014 2
(1) 5GNOW Application Challenges (2) 5GNOW Frame Structure Outline (3) 5GNOW Waveform Approaches (4) Conclusions 28.06.2014 3
Gigabit Wireless Connectivity Examples: 3D video streaming, large crowd gatherings 2005 2013 28.06.2014 4
Internet of Things (IoT) Connecting the things of every day life, scalable connectivity for billions of devices Cost below 10$ Coverage (deep indoor) Battery (10 years) Plug&secure, human in the loop 28.06.2014 5
Tactile Internet (TI) Real-time cyber-physical control applications 100ms 10ms 1ms 100 µs latency on physical layer! 28.06.2014 6
Fragmented Spectrum Spectrum paradox: spectrum scarce and expensive but underutilized! EC Digital Agenda forces the systems to deal with fragmented spectrum and white spaces communication (PAPR, 100x better localization) 28.06.2014 7
Application Challenges Wireless Access: flexible scalable fast robust reliable efficient (energy, spectrum) 28.06.2014 8
Frame Structure Concept 28.06.2014 9
Unified Frame Structure Layer Target for 5G integrated air interface: Efficiently combine various types of service and performance classes within a radio frame (from small packet service to high rate bit-pipe ) Type I Type III and Type IV Type II Time Frequency Traffic Type Synch? Access Type Properties I closed-loop scheduled classical high volume data services II open-loop scheduled HetNet and/or cell edge multi-layered high data traffic III open-loop sporadic, contention-based few bits, supporting low latency, e.g. smartphone apps IV open-loop/none* contention-based energy-efficient, high latency, few bits *: none for maximal energy savings at Tx, open-loop for reduced complexity at Rx
Requirements Some arguments against OFDM: Flexibility: Cyclic prefixes reduce spectral efficiency and prohibit flexible handling of frame formats Scalable: Spectral localization is too bad, e.g. in narrowband setting up to 4-6 subcarrier gain by different waveforms Robust and Reliable: OFDM is very sensitive both in time and frequency domain due to FFT Fast: Very difficult to support short symbols with given channel delay spread Efficient (energy, spectrum): OFDM is not robust under incomplete channel state information 28.06.2014 11
Asynchronous Reference Scenario Main observation: OFDM fails in highly asynchronous access scenarios, e.g. for massive MTC communication. MSE [db] -18-20 -22 UFMC, no CFO -24 UFMC, 5% CFO -26 UFMC, 10% CFO CPOFDM, no CFO -28 CPOFDM, 5% CFO CPOFDM, 10% CFO -30-0.5-0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4 0.5 relative delay G. Wunder, P. Jung, M. Kasparick, T. Wild, F. Schaich, S. ten Brink, Y. Chen, I. Gaspar, N. Michailow, A. Festtag, G. Fettweis, N. Cassiau, D. Ktenas, M. Dryjanski, S. Pietrzyk, B. Eged, P. Vago, and F. Wiedmann, 5GNOW: Non-Orthogonal, Asynchronous Waveforms for Future Mobile Applications, IEEE Communications Magazine, 5G Special Issue, Feb. 2014 Outside CP: New waveforms really make a difference! 28.06.2014 12
Waveform Concepts 28.06.2014 13
Filter Bank Multicarrier (FBMC) Features: Non-orthogonal multicarrier modulation (in complex domain) OQAM Prototype filter optimized for time and frequency localization trade-off (PHYDYAS, K=4 optimized for ACLR) Non-adjacent subcarriers are almost perfectly separated Spectral efficiency improved as no cyclic prefix is required Efficient implementation with IFFT/FFT Overlapping of time symbols: ISI solved by OQAM modulation FBMC transmitter with filtering in the frequency domain Nicolas Cassiau, Dimitri Kténas, Jean Baptiste Doré, Time and frequency synchronization for CoMP with FBMC, Tenth International Symposium on Wireless Communication Systems (ISWCS 13), Ilmenau, Germany, August, 2013 15.01.2014 14
High performance receiver for frequencyspreading FBMC A receiver suited for asynchronous uplink multiuser access and fragmented spectrum operation One unique (larger) FFT for all users Relaxed synchro requirements All treatments are realized in the frequency domain BEFORE filtering by prototype filter prototype filter one-tap equalizer Low complexity CFO correction (see on next slide) Three interpolation filters (left, middle and right)
Performance of FBMC Multiple Access with Relaxed Synchronization Due to fair frequency localization e.g. with fragmented spectrum, only the carriers located at the edges of the active spectrum are affected by interference (OFDM: interference is spread over all the active carriers) FBMC waveforms permit a simple way of sharing resources between cell-edge users without strict synchronization between users due to the low level of uplink interference generated by the built-in waveform filter.
Performance of FBMC Multiple Access with Relaxed Synchronization In case of QPSK, without guard carrier, the capacity is close to synchronous transmission and the level of interference is much lower than the required SNR to allow the decoding of QPSK. For 16-QAM modulation, FBMC gives a significantly better capacity, particularly in the range of [10-20]dB of SNR. Due to the better frequency localization, only the carriers located at the border of the user spectrum are affected by interference. For 64-QAM, the FBMC waveform clearly outperforms the OFDM waveform. Interference dominates the SINR for OFDM waveform and consequently for a given capacity of 5 bits/s/carrier, the SNR loss is of around 5dB.
Universal Filtered Multicarrier (UFMC) s 1k s 2k IDFT spreader V 1k + P/S IDFT spreader V 2k + P/S Filter F 1k with length L Filter F 2k with length L x 1k x 2k + x k Generalization of Filtered OFDM and FBMC (FMT) UFMC complexity similar to OFDM Huge knowledge base of OFDM processing can be re-applied to UFMC Baseband to RF s Bk IDFT spreader V Bk + P/S Filter F Bk with length L x Bk channel Frequency domain symbol processing (e.g. per subcarrier equalization) 2N point- FFT 0 0 Time domain preprocessing (e.g. windowing) + S/P zeropadding RF to Baseband noise n + other users 0 V. Vakilian, T. Wild, F. Schaich, S.t. Brink, J.-F. Frigon, "Universal-Filtered Multi-Carrier Technique for Wireless Systems Beyond LTE", IEEE Globecom'13, Atlanta, December 2013 28.06.2014 18
Uplink CoMP: UFMC vs OFDM UFMC adds increased robustness in CoMP against timefrequency misalignments CFO is estimated and compensated. CFO estimation error Δε UFMC OFDM Parameters CFO 10% of subcarrier spacing QPSK FFT size 128 12 subc. per PRB 6 PRBs allocated Filter length / CP length 16 UFMC: Dolph- Chebychev filters with 120 db att. Frequency-selective fading channel (16 taps)
Generalized Frequency Division Multiplexing (GFDM) Waveform Properties Multidimensional block structure with cyclic prefix: Circular sub-carrier pulse shape: Overlapping (non-orthogonal) sub-carriers: I. Gaspar, N. Michailow, A. Navarro Caldevilla, E. Ohlmer, S. Krone and G. Fettweis, Low Complexity GFDM Receiver Based On Sparse Frequency Domain Processing, 77th IEEE Vehicular Technology Conference (VTC Spring'13), Dresden, Germany, June 2013 28.06.2014 20
GDFM Successive Interference Cancellation (SIC) 64QAM, RRC, a=0.2 64QAM, RRC, a=0.4 256QAM, RRC, a=0.4 AWGN Rayleigh multipath Theoretical BER of orthogonal system can be reached with SIC 28.06.2014 21 19. September 2013 Nicola Michailow Slide 21
Conclusions 28.06.2014 22
Conclusions 5G visions like the IoT have very specific application demands and require highly asynchronous access in time and frequency New waveforms such as FBMC, UFMC, GFDM have very desirable properties and are significantly more robust to temporal and spectral fragmentation of traffic Two major upcoming things: System simulation to show benefit for fragmented traffic Unifying theory to explore in terms of Gabor signaling Demonstration of multiuser uplink 28.06.2014 23
Thank you for your attention! www.5gnow.eu Contact Dr. Gerhard Wunder gerhard.wunder@hhi.fraunhofer,de www.hhi.fraunhofer.de/wn Fraunhofer Heinrich Hertz Institute Berlin, Germany 28.06.2014 24