04.02.2016 @ 5G System Concept Seminar RF towards 5G Researchers: Tommi Tuovinen, Nuutti Tervo & Aarno Pärssinen
5.2.2016 2 Outline 5G challenges for RF Key RF system assumptions Channel SNR and related key RF parameters Minimum SNR definitions and data rates TX EVM & RX SNR models Link adaptation & maximum channel loss RF link budget analysis with MIMO channel model Data rates and link output power requirements 5G RF implementation aspects What s next
5.2.2016 3 5G challenges for RF High data rate cmw and mmw Antenna array gain to compensate for higher losses RF parallelism & high frequency power & cost Smaller antenna area only RFIC viable mmw systems E2E RF specifications for dense cellular Spatial multiplexing New tool in 5G RF
5.2.2016 6 Key RF system assumptions 2015 TDD OFDM system fc = 10.1 GHz System bandwidth of 500 MHz Modulations BPSK, QPSK, 16QAM, 64QAM & 256QAM Target peak data rates 10 Gbit/s peak rate and 1 Gbit/s cell edge ---------------------- After WRC-15 decisions target frequencies get higher (many bands will be studied in the range 23.5-86GHz) Other changes to be expected in BW s etc.
5.2.2016 7 Channel SNR and Related Key RF Parameters Total link SNR budget Calculated for different coding rates using repetitive coding scheme SNR specifications distributed between TX & RX TX: EVM-specific modulation requirements Constant for different coding rates RX: Benefit of coding utilized in RX sens. requirements Link margin can be defined for MIMO beam-specific path gains TX output power and RX input levels treated as pairs
5.2.2016 8 Minimum SNR Definitions and Maximum Data Rate DSP is abstracted to minimum SNR definitions Fading margin or beam-specific path gains can be used instead of LOS channel path loss Any simulation based model with SNR vs. BER can be used in analysis
5.2.2016 9 SNR Distribution between TX & RX Total link SNR distributed between TX & RX TX EVM-specific modulation requirements Benefit of coding utilized in RX sensitivity requirements
5.2.2016 10 TX EVM & RX SNR Models SNR requirements are linked to practical implementation TX: Based on commercial class A PA at 10 GHz RX: 9 db NF for UE with peak SNR limitations
5.2.2016 11 Link Adaptation & Maximum Channel Loss TX maximum and RX minimum power levels Can be tied to achievable data rates values! Maximum channel loss can be defined Range can be increased by providing antenna array gain
5.2.2016 13 Example: RF link budget Analysis at 10 GHz MIMO channel model included Note: co-channel interference not taken into account To be cancelled in BB or included in SNR requirements Practical antenna limitations at 10 GHz 4 antennas to UE (optimistically 8.4 dbi) 16 antennas to BS per link (17.5 dbi)
5.2.2016 14 Example: RF link budget Analysis at 10 GHz Indoor rush hour Mobility small (< 3 km/h) 8 active users per 100 m2 Experimental-based MIMO channel Channel simulated for each UE independently MIMO beams with eigen value decomposition Each eigen presents one distinguishable MIMO path A. Roivainen, C. F. Dias, N. Tervo, V. Hovinen, M. Sonkki, and M. Latva-aho, Indoor geometry based stochastic channel model at 10 GHz, Submitted to IEEE Trans. Antennas and Prop., Nov. 2015.
5.2.2016 15 MIMO Beam-Specific Path Gains at 10 GHz Observations from MIMO channel simulations LOS dynamic range some 30 db on average Some 9-10 db per each weaker beam NLOS dynamic range some 10 db on average Some 3-4 db per each weaker beam
5.2.2016 16 Data Rates and TX Output Power at 10 GHz NLOS example Sum results and per MIMO-beam
5.2.2016 17 RF architecture: Digital MIMO, RF beamforming or hybrid? Digital MIMO for data rate (RF) beamforming for range and interference mitigation Beam centric propagation environment vs.
5.2.2016 18 Massive MIMO & 5G RF Implementation Antenna aspects Base station beamforming examples at 10 GHz Some 30-36 dbi beamforming gain Some 24-27 dbi beamforming gain
5.2.2016 19 Massive MIMO & 5G RF Implementation Antenna aspects User equipment beamforming examples at 10 GHz Some 0-9 dbi beamforming gain Some 6-15 dbi beamforming gain
5.2.2016 20 EVM ~ 1.5 % with 256 QAM In-band distortion? Spectral efficient waveforms suffer high PAPR OFDM still the most potential candidate Tradeoffs and challenges in mmw PA research Efficiency in cmw/mmw? Tx beamforming architecture PA array concept Integration with antennas Size Massive MIMO & 5G RF Implementation Some PA Aspects
5.2.2016 27 What s next? Refine analysis with some practical radiation limits Look co-channel interference issues against practical implementation Go deeper to practical implementation aspects Antennas, external PA s and other FE, RFIC Extend system design to new bands from WRC-15 Several bands 23.5 86GHz