`
`doc. IEEE 802.11-02-708r1 1 / 23
`
`Alexei Gorokhov, Paul Mattheijssen, Manel Collados,
`Bertrand Vandewiele, Gunnar Wetzker
`
`Philips Research
`
`PHY options for high throughput wireless LANs
`
`Right time and place for MIMO?
`
`Performance enhancement via antenna selection
`
`Experimental set-up and channel measurements
`
`MIMO architectures for high data rates
`
`Experimental results
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000001
`
`
`
`doc. IEEE 802.11-02-708r1 2 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
` Higher network capacity
`
` MIMO
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000002
`
`
`
`doc. IEEE 802.11-02-708r1 3 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
` Higher network capacity
`
` MIMO
`
` Currently 12 channels
`2 x rate 6 channels
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000003
`
`
`
`doc. IEEE 802.11-02-708r1 4 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
` Higher network capacity
`
` MIMO
`
` Over 64-QAM: severe requirements
`to analogue & mixed signal circuits
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000004
`
`
`
`doc. IEEE 802.11-02-708r1 5 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
` Higher network capacity
`
` MIMO
`
`
`
`
`Iterative demodulation with “anti-Gray” maps
`Turbo coded CODEC
`
` Higher complexity: power & area
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000005
`
`
`
`doc. IEEE 802.11-02-708r1 6 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
` Higher network capacity
`
` MIMO
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000006
`
`
`
`doc. IEEE 802.11-02-708r1 7 / 23
`
`Requirements
`
`Possible solutions
`
` Higher data rates (> 100Mbps)
`
` Increase bandwidth per link
`
` Increased throughput
`
` Higher order modulation
`
` Extended range
`
` Better coverage
`
` More powerful CODEC
`
` TX / RX diversity
`
`
`
`Higher network capacity
`
` MIMO
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000007
`
`
`
`doc. IEEE 802.11-02-708r1 8 / 23
`
`RX
`
`Motivation
`
` Theoretical throughput scales linearly
`w.r.t. the # of antennas:
`
` Increased range / coverage in NLOS
`environments
`
` Cheap RF-CMOS technology:
`fractional cost per RF front-end (5.x GHz)
`
`TX
`
`Constraint
` Keep DSP complexity limited
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000008
`
`
`
`doc. IEEE 802.11-02-708r1 9 / 23
`
`RX
`
` Need for extra degrees of freedom at
`RX to ensure enough diversity ( )
`
` High incremental cost of adding RF
`front-end versus the cost of antenna
`
` Use antennas and front-ends at
`RX, select adaptively a subset of
`antennas,
`
` optimal selection is rather complex
` simple sub-optimal selection possible
`
`select out of
`antennas
`
`This image cannot currently be displayed.
`
`TX
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000009
`
`
`
`doc. IEEE 802.11-02-708r1 10 / 23
`
`
`
`
`
`
`
`
`
`
`
`
`
`4 TX chains
`4 RX chains
`f ~ 5.8GHz
`BW20MHz
`
`14 bit ADC
`35 dB AGC
`
`Receiver
`
`Transmitter
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000010
`
`
`
`doc. IEEE 802.11-02-708r1 11 / 23
`
`LNA
`
`mixer
`
`DAC/ADC
`
`AGC+LPF
`
`Monopoles
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000011
`
`
`
`Total TX power 14dBm
`
`doc. IEEE 802.11-02-708r1 12 / 23
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9 10 11 12
`
`
`
`
`
`
`“soft” walls, much glass
`few heavy metallic constructions
`lots of furniture
`few concrete walls / blocks
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000012
`
`
`
`doc. IEEE 802.11-02-708r1 13 / 23
`
`80% quantil
`50% quantil
`10% quantil
` 1% quantil
`
`2.5
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`RMS delay spread [x symbol rate]
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`
`RMS delay spread versus range
`
`80% quantil
`50% quantil
`10% quantil
` 1% quantil
`
`50
`
`40
`
`30
`
`20
`
`10
`
`00
`
`SNR per RX antenna [dB]
`
`-10
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`
`Signal-to-noise ratio per RX
`antenna versus range
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000013
`
`
`
`doc. IEEE 802.11-02-708r1 14 / 23
`
`TX signal path
`
`DEMUX
`
`FEC
`encoder
`
`FEC
`encoder
`
`interleaver
`
`mapper
`
`interleaver
`
`mapper
`
`stream
`cycling
`
`IFFT
`
`IFFT
`
`cyclic
`extension
`
`pulse
`shaping
`
`cyclic
`extension
`
`pulse
`shaping
`
`RX signal path
`
`FFT
`
`sampling
`
`FFT
`
`sampling
`
`pulse
`shaping
`
`pulse
`shaping
`
`latency of ~ one TX/RX cycle
`
`+ +
`
`-
`
`-
`
`deinterleave
`demap
`
`MMSE
`filter
`
`deinterleave
`demap
`
`MRC
`
`interleaver
`
`mapper
`
`FEC
`encoder
`
`FEC
`encoder
`
`FEC
`encoder
`
`MUX
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000014
`
`
`
`doc. IEEE 802.11-02-708r1 15 / 23
`
`TX signal path
`
`FEC
`encoder
`
`space
`frequency
`
`interleaver
`
`mapper
`
`IFFT
`
`mapper
`
`IFFT
`
`cyclic
`extension
`
`cyclic
`extension
`
`pulse
`shaping
`
`pulse
`shaping
`
`RX signal path
`
`FEC
`decoder
`
`space
`frequency
`de-
`interleaver
`
`demapper
`
`demapper
`
`2 x 2
`MMSE
`filter
`
`FFT
`
`sampling
`
`FFT
`
`sampling
`
`pulse
`shaping
`
`pulse
`shaping
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000015
`
`
`
`doc. IEEE 802.11-02-708r1 16 / 23
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`9
`
`3
`
`6
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`S-F modulation / layered RX
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`
`MIMO channel
`
`Outage capacities versus range
`outage rate 1%
`optimal RX antenna selection
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000016
`
`
`
`doc. IEEE 802.11-02-708r1 17 / 23
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`9
`
`3
`
`6
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`S-F interleaving / MMSE
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`
`MIMO channel
`
`Outage capacities versus range
`outage rate 1%
`optimal RX antenna selection
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000017
`
`
`
`doc. IEEE 802.11-02-708r1 18 / 23
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`9
`
`3
`
`6
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`S-F interleaving / MMSE
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`1 x 1 RX 1
`1 x 1 RX 4
`2 x 2 RX 2
`2 x 2 RX 4
`3 x 3 RX 3
`3 x 3 RX 4
`4 x 4 RX 4
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`
`S-F modulation / layered RX
`
`Outage capacities versus range
`outage rate 1%
`optimal RX antenna selection
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Maximum rate [Mbps]
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000018
`
`
`
`doc. IEEE 802.11-02-708r1 19 / 23
`
`MIMO capacities
`
` MIMO capacity scales almost linearly w.r.t. the number of TX/RX antennas
`space-frequency modulation with layered RX : ~90% of theoretical limit
`
`space-frequency interleaving with MMSE:
`~60% of theoretical limit
`
` …………… with adaptive RX selection: ~80% of theoretical limit
`
`Feasibility aspects
`
`
`
`
`
`
`
`channel processing beyond 2 x 2 system is hardly feasible (baseband)
`layered reception yields higher complexity & processing latency,
`seems prohibitive beyond 2 x 2 systems
`sub-optimal RX antenna selection looks attractive
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000019
`
`
`
`doc. IEEE 802.11-02-708r1 20 / 23
`
`2 x 2 RX 2
`2 x 2 RX 4 (opt)
`2 x 2 RX 4 (sub)
`2 x 4 RX 4 (MRC)
`1 x 1 RX 1
`
`200
`180
`160
`140
`120
`
`100
`80
`60
`40
`20
`0
`
`Maximum rate [Mbps]
`
`2 x 2 RX 2
`2 x 2 RX 4 (opt)
`2 x 2 RX 4 (sub)
`2 x 4 RX 4 (MRC)
`1 x 1 RX 1
`
`200
`180
`160
`140
`120
`
`100
`80
`60
`40
`20
`0
`
`Maximum rate [Mbps]
`
`3
`
`6
`
`9
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`S-F modulation / layered RX
`Outage capacities versus range
`outage rate 1%
`
`9
`
`3
`
`6
`
`12 15 18 21 24 27 30 33 36
`Range (m)
`S-F interleaving / MMSE
`
` S-F modulation with MMSE receiver & sub-optimal RX selection looks attractive
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000020
`
`
`
`doc. IEEE 802.11-02-708r1 21 / 23
`
`Candidate FEC structures
`
`Standard convolutional code
`rate (1/2) 64-state code [133,171]8
`
`puncture to achieve mandatory / supplementary rate modes
`soft-input Viterbi decoding at RX
`easy to implement, IEEE 802.11 acceptance
`expected to be sensitive to SINR discrepancy
`
`
`
`
`
`
`
`
`
`Turbo- CODEC similar to that of UMTS
`rate (1/3) PCCC with 8-state components [13,15]8
`
`puncture to achieve desired rates
`iterative SISO decoding (Max-Log-MAP)
`reduced SINR margin, less sensitive to SINR discrepancy
`rather high complexity
`
`
`
`
`
`
`
`
`
`
`
`
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000021
`
`
`
`doc. IEEE 802.11-02-708r1 22 / 23
`
` 108 Mbps
` 72 Mbps
` 36 Mbps
` 24 Mbps
`
` 96 Mbps
` 64 Mbps
` 32 Mbps
`
`3
`
`6
`
`9
`
`15
`12
`Range (m)
`
`18
`
`21
`
`24
`
`Signalling
`108Mbps 64QAM, rate 3/4
`96Mbps 64QAM, rate 2/3
`72Mbps 16QAM, rate 3/4
`64Mbps 16QAM, rate 2/3
`36Mbps 16QAM, rate 3/4
`32Mbps QPSK, rate 2/3
`24Mbps QPSK, rate 1/2
`
`100.0%
`
`10.0%
`
`1.0%
`
`0.1%
`
`Outage rate [%]
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000022
`
`
`
`doc. IEEE 802.11-02-708r1 23 / 23
`
`Observations
` Maximum data rate scales linearly w.r.t. to the
`number of antennas
`
` Receive antenna selection improves substantially
`maximum data rates (limited number of TX/RX chains)
`
`
`
`2 x 2 space division multiplexing with selection 2 of 4 RX
`antennas 200%-300% of single-antenna rates
`
`Philips
`Research
`
`HUAWEI EXHIBIT 1020
`HUAWEI VS. SPH
`
`000023