Positioning with 5G mmWave Massi sive-MIMO Systems Henk Wymeersch - - PowerPoint PPT Presentation

Positioning with 5G mmWave Massi sive-MIMO Systems

Henk Wymeersch Department of Electrical Engineering Chalmers University of Technology Gothenburg, Sweden https://goo.gl/KbrQEF email: henkw@chalmers.se

Summer School on 5G V2X Communications, June 11th, 2018 1

With help from Arash Shahmansoori, Zohair Abu Shaban, Nil Garcia, Gabriel Garcia, Mike Koivisto, and others.

Gonzalo Seco-Granados Department of Telecommunications Universitat Autonoma de Barcelona Bellaterra, Spain https://spcomnav.uab.cat/~gseco email: gonzalo.seco@uab.cat

Main idea a of radio-base ased posi sitioning

• Waveform conveys information about geometry
• Different measurements can be taken

– Signal strength: path loss vs fingerprinting – Time: TOA, TDOA, RTT – Angle: AOA, AOD – Frequency (Doppler): FOA

• Main resources: delay resolution (bandwidth

th), angle resolution (ante tenna elements ts), frequency resolution (time), interference mitigation (SNR)

2 Summer School on 5G V2X Communications, June 11th, 2018

Outline

• 5G positioning: 5 selling points
• 5G positioning: performance bounds
• 5G positioning: algorithms
• Conclusions
• References

3 Summer School on 5G V2X Communications, June 11th, 2018

Radio-base sed positioning

4

accuracy cost 100m 10m 1m 1 cm 10 cm ?? 5G potential 3G 4G GPS UWB WiFi uplink TDOA downlink TOA 2G Uplink Cell ID fingerprinting Two-way ranging and uplink TDOA

Summer School on 5G V2X Communications, June 11th, 2018

Radio-base sed positioning

5 Summer School on 5G V2X Communications, June 11th, 2018

What do we mean by 5G?

6

Large antenna arrays Directional transmission Large bandwidths Higher carrier frequencies Device-to-device communication Network densification

image: Qualcomm image: Ericsson image: University

• f Bristol

image: YouTube

Summer School on 5G V2X Communications, June 11th, 2018

5 Selling points s for 5G positioning

1. High carrier frequencies 2. Large bandwidths 3. Large number of antennas 4. D2D communication 5. Network densification

7

5G

Summer School on 5G V2X Communications, June 11th, 2018

High carrier er frequenc encies es

• Received power due to path loss, shadowing, multipath fading
• Path loss: countered by array gains
• Shadowing: severe penetration loss so no shadowing
• Multipath fading: no diffraction, limited scattering and little reflection
• Communication channel is dominated by LOS and a few location-

dependent clusters

8

Sparse communication channel, related to the physical environment

Below 6 GHz: full matrix (i.i.d., Gaussian) Above 28 GHz: low-rank matrix Each “effective path” corresponds to cluster

Summer School on 5G V2X Communications, June 11th, 2018

Large e bandwidths

1. From Fisher information: large bandwidth leads to better delay (distance) estimation accuracy 2. More resolvable multipath components: two paths are resolvable when

9

High degree of resolvability of multipath

Meissner, Paul, et al. "Accurate and robust indoor localization systems using ultra- wideband signals." arXiv preprint arXiv:1304.7928 (2013).

2 GHz bandwidth

30 meter

Summer School on 5G V2X Communications, June 11th, 2018

Large e number er of antenn ennas as

• From Fisher information

– Large number of RX antennas: better AOA resolvability – Large number of TX antennas: smaller beamwidth, better AOD resolvability

10

High degree of resolvability of angles

Summer School on 5G V2X Communications, June 11th, 2018

D2D communicat ation

• 5G will have sidelinks
• Measurements between devices
• Can improve location accuracy and coverage

11

Cooperative positioning based on D2D measurements

Wymeersch, Henk, Jaime Lien, and Moe Z. Win. "Cooperative localization in wireless networks." Proceedings of the IEEE 97.2 (2009): 427-450.

Error [m] CCDF

Summer School on 5G V2X Communications, June 11th, 2018

Network densi sificat ation

• Many access nodes, which could be location references
• High chance of LOS at short distances
• LOS link most useful for positioning

12 http://www.5gworkshops.com/5G_Channel_Mod el_for_bands_up_to100_GHz(2015-12-6).pdf

LOS link generally available for positioning

Summer School on 5G V2X Communications, June 11th, 2018

Outline

• 5G positioning: 5 selling points
• 5G positioning: performance bounds
• 5G positioning: algorithms
• Conclusions
• References

13 Summer School on 5G V2X Communications, June 11th, 2018

5G mmWave positioning: model with L clust sters

14

Estimate position (and orientation) in the presence of unknown scatterer locations Limited number of RF

• chains. Precoding matrix F

and combining matrix W.

Summer School on 5G V2X Communications, June 11th, 2018

Some e geomet etric intuition

• Abstraction
• Downlink
• Known BS position
• Consider downlink

transmission.

15

AOD = direction from BS TOA = range from BS estimate position

• rientation

AOA, AOD no yes AOA, TOA no* no AOD,TOA yes no AOA,AOD,TOA yes yes

*unless orientation is known

Summer School on 5G V2X Communications, June 11th, 2018

Performanc ance bounds

• Received waveform
• Channel parameters and location parameters
• Virtual anchors have no physical meaning for scatterers
• Parametrization can also be in terms of scatterers and points of incidence

16 Summer School on 5G V2X Communications, June 11th, 2018

Fisher er informat ation matrix of chan annel el param amet eter ers

• Unknown parameter
• Noise-free observation
• FIM has entries

17

This part was available in delay domain 7L real parameters L

Summer School on 5G V2X Communications, June 11th, 2018

Fisher er informat ation matrix of chan annel el param amet eter ers

• Each sub-block (e.g.,

, ) is Hadamard product

• Conclusion: each sub-block will be almost diagonal

18

Receiver component Transmitter component Signal & gain component ⦿ ⦿

Tends to diagonal when paths have distinct AOA (for large number of receive antenna) Tends to diagonal when paths have distinct delays (for large signal bandwidth) Tends to diagonal when paths have distinct AOD (for large number of transmit antennas), under full MIMO

Summer School on 5G V2X Communications, June 11th, 2018

Fisher er informat ation matrix of chan annel el param amet eter ers

• Original FIM
• Rearrange parameters

19

Each path provides independent information

Summer School on 5G V2X Communications, June 11th, 2018

Fisher er informat ation matrix in position space

• Remove channel amplitude with Schur complement
• Introduce parameter of interest
• Determine FIM
• Compute EFIM
• f position and orientation
• From EFIM we compute PEB and OEB
• EFIM can be expressed as sum over the paths

20

invert FIM can be nonsingular (Potential for SLAM) invert

Summer School on 5G V2X Communications, June 11th, 2018

Uplink vs downlink

Downlink Beam in known direction FIM in channel space Position relates to delay, AOD FIM in location space Leads to different scaling: Uplink Beam in unknown direction FIM in channel space Position relates to delay, AOA FIM in location space

21

Same Different!

Unknown

• rientation

Unknown

• rientation

Summer School on 5G V2X Communications, June 11th, 2018

Results

• 3D scenario (unknown [position, azimuth, elevation])
• 12 x 12 arrays at TX and RX
• 38 GHz carrier, 125 MHz bandwidth (beam squint ignored)
• 1 mW transmit power, 16 training symbols
• Single path and 4 path channel
• No combining (W = I)

22

• Z. Abu-Shaban, Xiangyun Zhou, T. Abhayapala, G. Seco-Granados, H. Wymeersch, "Error Bounds for Uplink and

Downlink 3D Localization in 5G mmWave Systems", IEEE Transactions on Wireless Communications, May 2018.

downlink uplink

Summer School on 5G V2X Communications, June 11th, 2018

Single e path (LOS) with 6 beam ams

23

Position error bound [m] Orientation error bound [deg]

Summer School on 5G V2X Communications, June 11th, 2018

• Beam centers marked with black circles.

Four paths s (including LOS) with 6 beam ams

24

Scatterer locations Position error bound [m] Orientation error bound [deg]

Summer School on 5G V2X Communications, June 11th, 2018

0.05 0.1 0.15 0.2 0.25 0.3 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Indep epend endent path assu sumption

25 Summer School on 5G V2X Communications, June 11th, 2018

Uplinks s vs downlink: : PEB (delay ay + AOD (DL) or AOA (UL) L))

26 Summer School on 5G V2X Communications, June 11th, 2018

Uplinks s vs downlink: : OEB (AOA and AOD)

27 Summer School on 5G V2X Communications, June 11th, 2018

Impac act of more e recei eive e antenn ennas: as: PEB

28

Uplink sensitive to orientation UL positioning takes advantage of AOA with number of receive antennas

Summer School on 5G V2X Communications, June 11th, 2018

DL is better UL is better Rx gain increases, but not the spatial diversity for the AOD.

𝑂𝑈 = 144

Impac act of more e transm smit antenn ennas as (fixed ed number of beam ams) s)

29

More TX antennas: more beamforming gain, more AOD accuracy Many TX antennas: too narrow beams, miss the BS More TX antennas: but PEB depends on AOA, only the additional transmit gain is exploited. In the UL, it does not matter if you don’t have many antennas at the User Device.

Summer School on 5G V2X Communications, June 11th, 2018

𝑂𝑆 = 144

Impac act of number er of BS beam ams

30

Poor coverage Saturation: more beams, each with lower power More TX antennas needs more beams for full coverage

Summer School on 5G V2X Communications, June 11th, 2018

Downlink transmission is assumed.

Single-base ase station local alizat ation: : clock offset set

31

• Most existing studies ignore the tx-rx synchronization
• The existence features in the environment permits the clock

ck synch chronization using on one-wa way tra ransmissions.

80 1 60 2 40

x

3 20

z

4

y

• 20
• 10

10 20 30

• 20

40 5 50 60 70 80 6

©

x UE x BS

↵ UE

x VA ,1 x VA ,2 x VA ,3 x VA ,4

↵ ehicle’

−

↵

−

✓ φ − ⌧ ⌧ ✓ φ ⌧ ✓ φ − ⇠ ⌃ ⌃

refine

−

↵ fill ficient

fil modification.

Summer School on 5G V2X Communications, June 11th, 2018

Single-base ase station local alitzat ation: : clock offset set

32

• Bias estimation accuracy

2 4 6 8 10 12 14 10-2 10-1 100 101 102

• When LOS is present, 1

NLOS is needed.

• When LOS is absent, 3

NLOS are needed (or 2 if map information is available).

• TOA estimation: 0.1 m st.dev.
• Angle estimation: 0.01 rad

st.dev.

Single-base ase station local alizat ation: : clock offset set

33

• Second alternative: use two-way transmissions

Summer School on 5G V2X Communications, June 11th, 2018

Single-base se station mapping: : clock offse set

34 Summer School on 5G V2X Communications, June 11th, 2018

Centralized Localization Protocol (CLP) Distributed Localization Protocol (DLP)

Single-base se station mapping: : clock offse set

35 Summer School on 5G V2X Communications, June 11th, 2018

• Z. Abu-Shaban, H. Wymeersch,
• T. Abhayapala, G. Seco-

Granados, “Single-Anchor Two- Way Localization Bounds for 5G mmWave Systems: Two Protocols”, arXiv:1805.02319.

CDF of PEB with UE orientation angles of 0, and NUE = NBS =144 antennas, NB = 25 beams.

Single-base ase station mapping: : clock offset set

36

• UE and BS have a

12x12 array.

• 25 beams in both

links providing SNR>17dB is 90% of locations.

• f=38 GHz
• Z. Abu-Shaban, H.

Wymeersch, T. Abhayapala,

• G. Seco-Granados, “Single-

Anchor Two-Way Localization Bounds for 5G mmWave Systems: Two Protocols”, arXiv:1805.02319.

Summer School on 5G V2X Communications, June 11th, 2018

Outline

• 5G positioning: 5 selling points
• 5G positioning: performance bounds
• 5G positioning: algorithms
• Conclusions
• References

37 Summer School on 5G V2X Communications, June 11th, 2018

Algorithms

• Observation model
• We want to estimate TOA, AOA, AOD and L
• We know that each path provides independent information
• Use sparsity to extract channel parameters
• Idea:
• Same for transmitter side
• So

38

Nr x Dr Dr x 1 Each column in Ur is the response to a possible angle 1-sparse if (i) true angle is in dictionary and (ii) all columns are orthogonal If dictionary is well chosen: 1-sparse matrix with row = AOA, column = AOD L-sparse

Summer School on 5G V2X Communications, June 11th, 2018

Choosi sing a good dictionar ary

• For ULA
• When angles take on value
• DFT matrix is a reasonable dictionary

39

Exactly on a DFT angle

Summer School on 5G V2X Communications, June 11th, 2018

• In practice: angles not exactly
• n the DFT grid:

approximately sparse

• Example: 64 antennas

Channel el estimat ation algorithm

• OFDM signal on subcarrier k:
• Vectorize
• Introduce
• We can recover AOA / AOD by solving
• Can be solved with OMP
• Then recover gains (closed form) and delays (line search per path)
• Refine by adapting dictionary or post-processing (e.g., SAGE)

40

Ideally L sparse with entries For DFT dictionary: size Nr x Nt Try to find sparse vectors with common support

Summer School on 5G V2X Communications, June 11th, 2018

From angles es & delays ays to positions

• We have estimates of
• Use transformation to
• Apply Extended Invariance Principle (EXIP) and solve the

following non-linear least square problem

41

BS user BS user

Summer School on 5G V2X Communications, June 11th, 2018

minimize෥

𝜽 ෝ

𝜽− 𝜽 ෥ 𝜽

𝐗 2

Initialize by

1. When LOS path exists: use (AOD,delay) of path with shortest delay to recover position, then (AOD,AOA) to recover orientation. Then virtual anchors (or scatterer locations) are easily recovered path by path. 2. When LOS path does not exist: try all possible

• rientations: each (AOD,AOA,delay) gives rise to a
• path. Intersection of two paths is position. Evaluate

cost for each guess. 3 paths are needed in total. 3. When not know if LOS exists: try (1) and (2) and evaluate cost.

Results: s: 4 path chann annel el (including LOS)

42

−30 −25 −20 −15 −10 −5 5 10 10

−6

10

−4

10

−2

10 10

2

−30 −25 −20 −15 −10 −5 5 10 10

−4

10

−2

10 10

2

SNR (in dB) SNR (in dB) REB PEB RMSE( ˆ α) [rad] RMSE(ˆ p) [m] RMSE( ˆ α) RMSE(ˆ p)

τ τ θ θ θ τ τ θ ∈ τ τ ∈ τ τ π

− π τ − τ

θ θ θ θ θ θ fi π λ θ − θ θ θ θ θ θ efi θ θ θ θ θ θ θ θ θ θ θ θ

OEB PEB

• A. Shahmansoori, G.E. Garcia, G. Destino, G. Seco-Granados, H. Wymeersch

“Position and Orientation Estimation through Millimeter Wave MIMO in 5G Systems”, IEEE Transactions on Wireless Communications, vol. 17, no. 3, pp. 1822-1835, Mar 2018.

Results: 3 path channel without LOS

43

−30 −25 −20 −15 −10 −5 5 10 10

−2

10

−1

10 10

1

−30 −25 −20 −15 −10 −5 5 10 10

−1

10 10

1

10

2

10

3

SNR (in dB) SNR (in dB)

REB PEB

RMSE(ˆ α) [rad] RMSE(ˆ p) [m]

RMSE( ˆ α), ∆ α [rad] = 0.01 RMSE( ˆ α), ∆ α [rad] = 0.05 RMSE(ˆ p), ∆ α [rad] = 0.01 RMSE(ˆ p), ∆ α [rad] = 0.05 − ∆ α

fic Ψ τ ˜ ℜ ˜ ∗ θ θ τ τ θ θ ℜ −˜ ∗ θ θ τ τ θ θ Ψ θ ˜ ℜ ˜ ∗ θ θ τ τ θ θ ℜ ˜ ∗ θ θ τ τ θ θ Ψ θ ˜ − ℜ ˜ ∗ θ θ τ τ θ θ ℜ ˜ ∗ θ θ τ τ θ θ

OEB PEB

• A. Shahmansoori, G.E. Garcia, G. Destino, G. Seco-Granados, H. Wymeersch

“Position and Orientation Estimation through Millimeter Wave MIMO in 5G Systems”, IEEE Transactions on Wireless Communications, vol. 17, no. 3, pp. 1822-1835, Mar 2018. Summer School on 5G V2X Communications, June 11th, 2018

From angles es & delays ays to positions: s: Bayesi esian an approac ach

44 Summer School on 5G V2X Communications, June 11th, 2018

• Perform belief propagation
• Needs schedule
• Allows the introduction of prior

information

• Facilitates tracking and

hybridization with other sensors

ficiently find ’ − − reflecting p(x UE, ↵UE, B, x VA ,1, . . . , x VA ,L − 1|Z) = p(x UE)p(↵UE)p(B)

L − 1

Y

l= 1

p(xVA ,l ) ⇥p(z0|xUE, ↵UE, B)

L − 1

Y

l= 1

p(zl |xUE, ↵UE, x VA ,l , B). ⌃ µ µ µ

φ

µ µ

⌧

µ

⌧

µ µ

⌧

⇥

⌧

⇠ ⇠

⌧

µ

⌧

µ

⌧

⌧

µ

⌧

µ

⌧

⇥µ

⌧

⌧

⇠ ficient –

modification, ⌃

• Factor graph. The objective is to obtain the posterior distribution of

the position parameters given the channel parameters.

From angles es & delays ays to positions: s: Associat ation

45 Summer School on 5G V2X Communications, June 11th, 2018

• We choose to separate positioning and simple DA, no clutter
• M VAs, L paths
• For each path l: existing VA m or potential new VA
• Create L x (M + L+1) matrix S, small constant a
• Find optimal association via Munkres algorithm
• Only use reliable associations for positioning

p(z1|LOS) p(z1|VA1) a a p(z2|LOS) p(z2|VA1) a a Path 1 Path 2 VA1 New VA New VA LOS

Single-base ase station posi sitioning

46 Summer School on 5G V2X Communications, June 11th, 2018

• R. M. Buehrer, H. Wymeersch and R.
• M. Vaghefi, "Collaborative Sensor

Network Localization: Algorithms and Practical Issues," in Proceedings of the IEEE, vol. 106, no. 6, pp. 1089- 1114, June 2018. 10-1 100 101 102 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Multipath helps! Maps help! LOS is good, but not needed Beamforming helps in the beams

Single-base ase station mapping

47

• R. M. Buehrer, H. Wymeersch and R.
• M. Vaghefi, "Collaborative Sensor

Network Localization: Algorithms and Practical Issues," in Proceedings of the IEEE, vol. 106, no. 6, pp. 1089- 1114, June 2018. 100 101 102 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

LOS is good, but not needed Location helps a little Loss if scatterers are not in beams

Complet ete e flowchar art

48 Summer School on 5G V2X Communications, June 11th, 2018

Data communication PRS / SRS design Channel estimation Data association Positioning and mapping Mobility Position information Position information Channel information Downlink/uplink communication

• PRS design: beamformers and signal’s

design.

– Trade-off between accuracy and coverage – Trade-off between com and loc.

• Extend the position and mapping to more

complex environment: clutter,

Outline

• 5G positioning: 5 selling points
• 5G positioning: performance bounds
• 5G positioning: algorithms
• Conclusions
• References

49 Summer School on 5G V2X Communications, June 11th, 2018

Conclusions

• Radio signals have always provides location information
• 5G has important advantages for localization
• Improved resolvability in time, angle for

– Single anchor localization, tracking – Single anchor mapping – SLAM – Location-aided communication – Radar

50

5G

Summer School on 5G V2X Communications, June 11th, 2018

Challenges

• Good geometric mmWave channel models for

positioning, including blockage

• Database of location-based channel measurements
• Design of precoding and combining for positioning,

mapping

• Pilot design for positioning, mapping
• Fast algorithms for positioning, tracking, mapping
• Online synchronization for positioning
• Multi-user positioning, resource allocation for MU

positioning

• Using position information for communication

(Location-aided communications)

• Calibration of references (location, time)
• Practical demonstration

51

5G

Summer School on 5G V2X Communications, June 11th, 2018

Outline

• 5G positioning: 5 selling points
• 5G positioning: performance bounds
• 5G positioning: algorithms
• Conclusions
• References

52 Summer School on 5G V2X Communications, June 11th, 2018

References

Basics of radio-based positioning

• F. Gustafsson and F. Gunnarsson, "Mobile positioning using wireless networks: possibilities and fundamental

limitations based on available wireless network measurements," in IEEE Signal Processing Magazine, vol. 22,

• no. 4, pp. 41-53, July 2005.
• H. Liu, H. Darabi, P. Banerjee and J. Liu, "Survey of Wireless Indoor Positioning Techniques and Systems," in

IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), vol. 37, no. 6, pp. 1067-1080, Nov. 2007.

• J. A. del Peral-Rosado, J. A. López-Salcedo, F. Zanier and M. Crisci, "Achievable localization accuracy of the

positioning reference signal of 3GPP LTE," 2012 International Conference on Localization and GNSS, Starnberg, 2012, pp. 1-6.

• Sahinoglu, Zafer, Sinan Gezici, and Ismail Guvenc. "Ultra-wideband positioning systems." Cambridge, New

York (2008). 5G properties for positioning

• H. Wymeersch, G. Seco-Granados, G. Destino, D. Dardari, and F. Tufvesson, “5G mm-Wave Positioning for

Vehicular Networks”, IEEE Wireless Communication Magazine, vol. 24, no. 6, pp. 80-86, Dec 2017.

• Ping Zhang, Jian Lu, Yan Wang, Qiao Wang, Cooperative localization in 5G networks: A survey, ICT Express,

Volume 3, Issue 1, March 2017, Pages 27-32, ISSN 2405-9595.

• A. Dammann, R. Raulefs, S. Zhang. "On prospects of positioning in 5G." Communication Workshop (ICCW),

2015 IEEE International Conference on. IEEE, 2015.

• A. Shahmansoori, G. Seco-Granados, H. Wymeersch. "Survey on 5G Positioning." Multi-Technology
• Positioning. Springer International Publishing, 2017. 165-196.
• Han, Y., Shen, Y., Zhang, X. P., Win, M. Z., & Meng, H. (2016). Performance limits and geometric properties of

array localization. IEEE Transactions on Information Theory, 62(2), 1054-1075. 53 Summer School on 5G V2X Communications, June 11th, 2018

References

Single anchor positioning

• Li, Ji, Jean Conan, and Samuel Pierre. "Mobile terminal location for MIMO communication systems." IEEE

Transactions on Antennas and Propagation 55.8 (2007): 2417-2420. Cm Cm-wave 5G G positioning

• V. Savic, E. G. Larsson. "Fingerprinting-based positioning in distributed massive MIMO systems." Vehicular

Technology Conference (VTC Fall), 2015 IEEE 82nd. IEEE, 2015.

• M. Koivisto et al., "Joint Device Positioning and Clock Synchronization in 5G Ultra-Dense Networks," in IEEE

Transactions on Wireless Communications, vol. 16, no. 5, pp. 2866-2881, May 2017.

• N. Garcia, H. Wymeersch, E. G. Larsson, A. M. Haimovich and M. Coulon, "Direct Localization for Massive

MIMO," in IEEE Transactions on Signal Processing, vol. 65, no. 10, pp. 2475-2487, May15, 15 2017. Mm Mm-wave 5G G positioning in delay domain

• K. Witrisal, S. Hinteregger, J. Kulmer, E. Leitinger and P. Meissner, "High-accuracy positioning for indoor

applications: RFID, UWB, 5G, and beyond," 2016 IEEE International Conference on RFID (RFID), Orlando, FL, 2016, pp. 1-7.

• E. Leitinger, et al. "Evaluation of position-related information in multipath components for indoor positioning."

IEEE Journal on Selected Areas in communications 33.11 (2015): 2313-2328.

• K. Witrisal, et al. "High-Accuracy Localization for Assisted Living: 5G systems will turn multipath channels from

foe to friend." IEEE Signal Processing Magazine 33.2 (2016): 59-70.

• C. Gentner, et al. "Multipath Assisted Positioning with Simultaneous Localization and Mapping." IEEE

Transactions on Wireless Communications 15.9 (2016): 6104-6117. 54

References

Mm Mm-wave positioning in delay and angle domain

• Z. Abu-Shaban, H. Wymeersch, T. Abhayapala, G. Seco-Granados, “Single-Anchor Two-Way Localization

Bounds for 5G mmWave Systems: Two Protocols”, arXiv:1805.02319.

• Z. Abu-Shaban, Xiangyun Zhou, T. Abhayapala, G. Seco-Granados, H. Wymeersch, "Error Bounds for Uplink

and Downlink 3D Localization in 5G mmWave Systems", IEEE Transactions on Wireless Communications, May 2018.

• A. Shahmansoori, G. E. García, G. Destino, G. Seco-Granados, H. Wymeersch, "Position and Orientation

Estimation through Millimeter Wave MIMO in 5G Systems", IEEE Transactions on Wireless Communications,

• vol. 17, no. 3, pp. 1822-1835, Mar 2018.
• A. Guerra,F. Guidi, D. Dardari. "Single anchor localization and orientation performance limits using massive

arrays: MIMO vs. beamforming." IEEE Transactions on Wireless Communications, May 2018.

• H. Wymeersch, N. Garcia, Hyowon Kim, G. Seco-Granados, Sunwoo Kim, Fuxi Wen, Markus Frohle, “5G

mmWave Downlink Vehicular Positioning”, submitted to IEEE Globecom, 2018.

• Z. Lin, T. Lv and P. T. Mathiopoulos, "3-D Indoor Positioning for Millimeter-Wave Massive MIMO Systems," IEEE

Transactions on Communications, 2018.

• D. Wang, M. Fattouche and X. Zhan, "Pursuance of mm-Level Accuracy: Ranging and Positioning in mmWave

Systems," IEEE Systems Journal, 2018.

• M. Ruble, I. Güvenç, “Wireless Localization for mmWave Networks in Urban Environments”, arXiv:1805.11208.
• Y. Wang, Y. Wu and Y. Shen, "Multipath Effect Mitigation by Joint Spatiotemporal Separation in Large-Scale

Array Localization," IEEE GLOBECOM 2017. 55 Summer School on 5G V2X Communications, June 11th, 2018