Positioning in 5G networks
Satyam Dwivedi, Ritesh Shreevastav, Florent Munier, Johannes Nygren, Iana Siomina, Yazid Lyazidi, Deep Shrestha, Gustav Lindmark, Per Ernström, Erik Stare, Sara M. Razavi, Siva Muruganathan, Gino Masini, ?ke Busin, Fredrik Gunnarsson
11 Positioning in 5G networks
Satyam Dwivedi, Ritesh Shreevastav, Florent Munier, Johannes Nygren, Iana Siomina, Yazid Lyazidi, DeepShrestha, Gustav Lindmark, Per Ernstr¨om, Erik Stare, Sara M. Razavi, Siva Muruganathan, Gino Masini, ˚AkeBusin, Fredrik Gunnarsson
Abstract —In this paper we describe the recent 3GPP Release16 specification for positioning in 5G networks. It specifiespositioning signals, measurements, procedures, and architectureto meet requirements from a plethora of regulatory, commer-cial and industrial use cases. 5G thereby significantly extendspositioning capabilities compared to what was possible withLTE. The indicative positioning performance is evaluated inagreed representative 3GPP simulation scenarios, showing a 90percentile accuracy of a few meters down to a few decimetersdepending on scenarios and assumptions.
I. I
NTRODUCTION
5G and digitalization are often closely related, and withposition information being central in digitalization, the intro-duction of 5G positioning is a vital step. In Release 16, LTEpositioning feature is extended to accommodate enablers of 5Gsuch as wideband signals, higher frequencies, multiple anten-nas, low latency and flexible architecture. This article presentsmethods, architecture, procedures, signals and measurementsfor 5G positioning recently specified in Release 16.5G is designed to address requirements and needs of indus-trial verticals. The 3rd Generation Partnership Project (3GPP)Service and System Aspects (SA) specifications and technicalreports on positioning requirements across industry verticalssuggests accuracy target from tens of meters for emergencycalls, to a few decimeters within indoor factory and onedecimeter for vehicle-to-everything (V2X) use cases [1], [2].The required user equipment (UE) positioning accuracy forLTE was motivated by regulatory positioning requirementsset by FCC [3], [4]. For 5G, the requirements are motivatedby commercial use cases. A few key enablers for precisepositioning in 5G include mmWave frequency bands, whichenables wideband signals, beamforming and precise angleestimation with multiple antennas [5], [6]. However, mostfeatures are also available at low and mid frequency ranges.In this paper, we evaluate positioning performance in differentfrequency ranges. The Global Navigation Satellite System(GNSS) is an example of a successful positioning technologybut restricted to outdoor scenarios. The 5G will bring highaccuracy positioning to indoor scenarios while also providingbetter positioning accuracy outdoors than possible with LTEor GNSS alone [7].It is expected that many features useful for positioning canbe extracted from the specified elements than only the intendedfeatures during the standardization. Such as, fingerprinting,radio network optimization, soft information extraction etcetera [8], [9]. It is also possible to use signals and measure-ments defined for mobility and radio resource management forpositioning. Such enablers can be suitable for positioning inthe contexts of radio network management and analytics. The UE
NG-RANng-eNB
TPTP gNB
TRPTRPNR-Uu NG-C
LTE - U u NG - C NLs
AMFLMF
NRPPaRRCLPP gNB gNB-CUgNB-CUgNB-DU gNB-DU
TRPTP RP TRPTP RP F - C F - C NG-C Xn-CNR-Uu NR-Uu
Fig. 1: UE positioning architecture and logical protocolsamong different entities applicable to NG-RAN. Protocolsbetween entities are shown using colored dashed lines.full positioning potential of the offerings from the standardis possible with a UE and the network cooperating with eachother. Elements of specifications can still be useful when suchcooperations are limited.Release 16 specifies positioning signals and measurementsfor the 5G New Radio (NR). In subsequent sections we explainthe new 5G positioning improvements in architecture, signalsand measurements. Moreover we also demonstrate possiblepositioning accuracy with these new capabilities.II. P
OSITIONING ARCHITECTURE AND SIGNALINGPROTOCOLS
The 5G positioning architecture is derived from the 4Gpositioning architecture, with added modifications that areconsequent of the new logical nodes that are introduced tothe 5G Core Network (5GC). Figure 1 shows the Release16 positioning architecture for Next Generation Radio AccessNetwork (NG-RAN) that is applicable for positioning a UEwith NR gNB Transmission Reception Points (TRPs) or LTEng-eNB access with Transmission Points(TPs). Figure 1 alsoshows the Rel-16 NG-RAN split positioning architecture. Inthe gNB functional split, the gNB-Central Unit (CU) andgNB Distributed Units (DU) communicate via F1 interface.As shown in the figure, the gNB-CU terminates the connectionwith the 5GC and can be connected to one or multiple gNB-DU, which hosts the TP/RP/TRP. Figure 1 shows signalingprotocols among different entities in the 5G positioning archi-tecture. The gNB(gNB-CU)/ng-eNB exchanges the necessarypositioning information and measurements with the LocationManagement Function (LMF) which is in the 5GC, via the NRPositioning Protocol Annex (NRPPa) protocol [10]. In 4G, thepositioning support between a UE and the location server ishandled by the LTE Positioning Protocol (LPP). This protocolhas been extended to also support 5G positioning between a a r X i v : . [ c s . N I] F e b UE and LMF [11]. Whereas, the UE receives necessary radioconfiguration information from NG-RAN node over RadioResource Control (RRC) via NR-Uu or LTE-Uu interface.Reusing the LPP protocol also for 5G enables extensions ofboth 4G and 5G in the common protocol. Both the NRPPaand the LPP protocols are transported over the control planeof the NG interface (NG-C) via Access Mobility Function(AMF) [11].5G provides not only enablers for precise positioning assuch, but also introduces some new positioning methods.Positioning based on multi-cell round trip time (multi-RTT)measurements, multiple antenna beam measurements to en-able downlink angle of departure (DL-AoD) and uplink an-gle of arrival (UL-AoA) estimates has been introduced asnew concepts. While the multi-RTT positioning method isrobust against network time synchronization errors, anglebased methods are more relevant with usage of mmWave andmultiple antennas in 5G NR. For multi-RTT LMF initiatesthe procedure whereby multiple TRPs and a UE performthe gNB Rx-Tx and UE Rx-Tx measurements respectively.For multi-RTT, gNBs and UEs transmit downlink positioningreference signal (DL-PRS) and uplink sounding referencesignal (UL-SRS) respectively. gNB configures UL-SRS to theUE using RRC protocol. Whereas, LMF provides the DL-PRS configuration using LPP to the UE. The UE reports themeasurement results using LPP to the LMF and gNB reportsthe measurements using NRPPa to the LMF for UE locationestimation.For DL-AoD, UE provides the DL-PRS beam ReceivedSignal Received Power (RSRP) measurements to LMF overLPP. The gNB provides the beam azimuth and elevationangular information to LMF over NRPPa. In the UL AoApositioning method, the UE position is estimated based on ULSRS AoA measurements taken at different TRPs. TRPs reportAoA measurements to LMF over NRPPa. Using angle infor-mations, either AoD or AoA, along with other configurationand deployment informations such as TRP co-ordinates andbeam configuration details LMF estimates the UE location.The Release 16 5G positioning specifications also includeNR broadcast of positioning assistance data such as for GlobalNavigation Satellite Systems- Real Time Kinematics (GNSS-RTK). It enables assistance data to be either broadcasted asintroduced for LTE in Relase 15 or made available on demand.With the new on-demand System Information (SI) procedure,a UE may request positioning System Information Blocks(posSIBs) by means of an on-demand SI request (randomaccess procedure message 1 or 3) in RRC Idle/Inactivate statesand using on-demand connected mode procedure while inRRC Connected mode.Release 16 also enhances the scope of GNSS RTK AD withsupport for spatial atmospheric delay models, leveraged bythe models defined for Quasi-Zenith Satellite System (QZSS).These models enables the devices to compensate for theatmospheric delays of the satellite signals.III. P
OSITIONING SPECIFIC SIGNALS
Multiple reference signals are used for communication re-lated procedures. For the purpose of positioning, NR supports O n e P RB t f TRP1 TRP2 TRP3
Comb-6 DL-PRS Comb-4 UL-PRS
Fig. 2: Specific configurations of NR positioning referencesignals. The DL-PRS is arranged in a comb-6 pattern withthree TRPs. The UL-PRS from a UE has comb-4 pattern.two new reference signals, the DL PRS and the UL SRS forpositioning. Figure 2 shows example within slot configurationsof DL PRS and UL SRS for positioning. The CSI-RS and SSBsignals used for radio resource management (RRM) can alsobe used as part of the enhanced cell ID (E-CID) positioningmethod.
A. Downlink Positioning Reference Signal, DL-PRS
A NR DL-PRS can be configured at two levels, withina slot and at multi slot level. Within a slot, the startingresource element in time and frequency from a TRP can beconfigured. Across multiple slots, gaps between PRS slots,their periodicity and density within a period can be configured.Here we explain some of the salient features of the DL-PRSspecified in Release 16 [12].1)
Maximum Bandwidth:
The PRS footprint on the timefrequency grid is configurable with a starting physical resourceblock (PRB) and a PRS bandwidth. The PRS may start at anyPRB in the system bandwidth and can be configured with abandwidth ranging from to PRBs in steps of PRBs.This amounts to a maximum bandwidth of about
MHz for kHz subcarrier spacing and to about MHz for kHzsubcarrier spacing. The flexible bandwidth configuration al-lows the network to configure the PRS while keeping out ofband emissions to an acceptable level. The large bandwidthallows a very significant improvement in time-of-arrival (TOA)accuracy compared to LTE.2)
Resources and resource sets:
The PRS can be transmittedin beams. A PRS beam is referred to as a PRS resource whilethe full set of PRS beams transmitted from a TRP on the samefrequency is referred to as a PRS resource set as illustratedin Fig. 3. The different beams can be time-multiplexed acrosssymbols or slots. To assist UE RX beamforming, the DL PRScan be configured to be quasi-co-located (QCL) Type D witha DL reference signal from a serving or neighboring cell,signaling that the same RX beam used by the UE to receivesaid reference signal can be used to received the configuredPRS. The beam structure of the PRS improves coverageespecially for mm-wave deployments and also allows for AoDestimation, e.g. the UE may measure DL PRS Received SignalTime Difference (RSTD) per beam and report the measuredRSTD including DL PRS Resource id (beam id) to the LMF. UE NG-RANng-eNB
TPTP gNB
TRPTRPNR-Uu NG-C
LTE - U u NG - C NLs
AMFLMF
NRPPaRRCLPP gNB gNB-CUgNB-CUgNB-DU gNB-DU
TRPTP RP TRPTP RP F - C F - C NG-C Xn-CNR-Uu NR-Uu
Fig. 1: UE positioning architecture and logical protocolsamong different entities applicable to NG-RAN. Protocolsbetween entities are shown using coloured dashed lines. A R e s ou r ce s e t o f D L - P R S DL-PRS resources ✓ U L - S R S D L - P R S R T T ✓ = Azimuth angle of departure (AOD) ✓ = Zenith angle of departure (ZOD) ⇢, ( ✓ , ) are distance and angles of arrival in polar coordinates ⇢ TRPTRP
Fig. 3: An illustration showing a few positioning elements withNR. Beams as resources and set of beams as resource sets areshown. The newly supported positioning methods, multi-RTTand angle-based positioning methods are illustrated.3)
Repetition and Periodicity:
In order to improve position-ing accuracy, more measurements can be collected. Measure-ments are collected per resource. Hence, repeated transmis-sion of PRS resources helps to collect more measurements.Measurements can be collected per resource. The repetitionof resources can be done in two ways, repeat before sweepand sweep before repeat as shown in the Fig.4. The amountand type of repetition can be configured with parametersfor configuring the gaps between resources ( T P RS g ap ) and thenumber of resource repetition ( T P RS r ep ) within a period ofresource set ( T P RS p er ). The DL PRS resources can be repeated upto times within a resource set period, either in consecutiveslots or with a configurable gap between repetitions. Theresource set period in FR1 ranges from to milli-seconds [12]. For example, dense set of measurements can becollected by settings T P RS g ap = 1 , T P RS r ep = 32 and T P RS p er = 4 .4) Interference suppression:
The DL PRS is designed toallow the UE to perform accurate TOA measurements inpresence of interfering DL PRSs from nearby TRPs. Eachsymbol of the DL PRS has a comb-structure in frequency,i.e. the PRS utilizes every N th subcarrier. The comb value N can be configured to be , , or . The length of the PRSwithin one slot is a multiple of N symbols and the positionof the first symbol within a slot is flexible as long as the slotconsists of at-least N PRS symbols. It allows accumulationof contiguous sub-carriers across a slot which improves cor-relation properties for TOA estimation. The resource elementpattern can be shifted in frequency with a frequency offset of to N − subcarriers thus allowing N orthogonal DL PRSsutilizing the same symbols. All configurable patterns coverevery subcarriers in the configured bandwidth over the patternduration which give maximum measurement range for theTOA measurement in scenarios with large delay spreads. TheDL-PRS is QPSK modulated by a standardized 31-bit Goldcode sequence initialized based on a DL PRS sequence IDtaking values from to . As an example, the Fig.2 shows comb-6 DL-PRS and the pattern repeats after symbols.Figure 2 shows multiplexing of three base stations to avoidinterference among them. The NR PRS comb-12 configurationallows for twice as many orthogonal signals as the comb-6LTE PRS which is useful to mitigate interference. Further, thelength of the NR PRS can be flexibly configured down to 2symbols which can be useful e.g. in indoor scenarios wherecoverage is not an issue.Besides a comb structure allowing multiplexing of multipleTRPs in a slot, muting of signals can also be used as a wayto mitigate interference. As shown in Fig.4, muting can beused either at the repetition level, where each repetition canbe individually muted within a periodic occasion, or at theoccasion level, where the whole periodic DL PRS occasion(including all repetitions) can be muted.5) Hierarchical structure:
The DL-PRS configuration isprovided in a hierarchy as shown in Fig.5. There can be atmost frequency layers and each frequency layer has at most TRPs. Each TRP per frequency layer can have DL-PRSResource sets thus resulting in a total of resource sets perTRP and each resource set can have up to resources. Eachresource corresponds to a beam. Having different resourcesets per frequency layer per TRP allows gNB to configure oneset of wide beams and another set of narrow beams for eachfrequency layer. B. Uplink signal, UL-SRS for positioning
The SRS for positioning is a reference signal based onthe SRS for communication. Although the signals have a lotin common, SRS for positioning and for communication areconfigured separately and with different properties specific totheir usage. The UL-SRS shown in the Fig.2, is a comb-4signal.1)
Resources and resource sets:
The SRS for positioningis configured in a resource, which in turn can be part ofa resource set. a resource correspond to an SRS beam, andresource sets correspond to a collection of SRS resource (i.e.beams) aimed at a given TRP.2)
Coverage features
The SRS resource is defined as acollection of symbols transmitted on the time-frequency NRgrid. Like the DL PRS, the SRS resources for positioningare transmitted on a single antenna port, and can be placedto begin on any symbol in the NR uplink slot. In the timedomain, the SRS resources for positioning can span 1,2,4,8,12consecutive OFDM symbols which provide enough coverageto reach all TRPs involved in the positioning procedures.Contrary to the SRS for communication, repetition is not sup-ported in an SRS for positioning resource. Similar to SRS forcommunication, the SRS for positioning is using Zadoff-Chusequences as a base signal, to ensure low-PAPR transmissionfrom the UE. The particular sequence used to generate an SRSsymbol depends on configuration parameters and sequencehopping is supported as for the SRS for communication.3)
Interference-free UE multiplexing
Since 3GPP Release15, NR has used a comb structure for the SRS, so that onlya fraction of the OFDM subcarriers are occupied by a givenSRS resource. For the SRS for communication, the comb size R e s ou r ce R e s ou r ce R e s ou r ce Periodicity of resource set T P RS p er Repeat before sweep T P RS g ap = 1 , T P RS r ep = 2 Repetition
Periodicity of resource set T P RS p er Sweep before repeat T P RS g ap = 4 , T P RS r ep = 2 Periodicity of resource set
Muting within a resource set
Muting
A set
Muting of resource sets
Fig. 4: Repetition and muting of PRS resources and resource sets. PRS resources and resource sets can be repeated forimproving accuracy by enabling collection of more measurements. However, they can be muted too for reducing interference.is either 2 or 4, meaning that the SRS occupies 1 subcarrier outof 2 or 4, respectively. In Release 16, several enhancementswere added in the specification of the SRS for positioning. Thecomb size K T C for the SRS for positioning is 2,4,8 and newcomb patterns was specified in order for all the subcarriers tobe sounded in one resource. Different resource element offsetspattern can be configured as a function of the comb size K T C and the number of symbol in the resource N SRSsymb [12]. For UEmultiplexing, SRS for positioning can be configured with aninitial comb offset and a specific cyclic shift. The comb offsetcan take any value in. [0 , K T C − ].IV. M EASUREMENTS FOR P OSITIONING
Compared to LTE, more positioning measurements havebeen standardized for NR. Furthermore, some of the mea-surement types (e.g., Rx-Tx time difference) have been stan-dardized not only for the serving cell but also for neighborcells or TRPs. The following downlink measurements basedon PRS can be configured in the UE by LMF: RSTD, UE Rx-Tx time difference, and PRS-based Reference Signal ReceivedPower (PRS-RSRP) measurements, unlike in LTE where onlyRSTD were possible for PRS. In addition, the UE may alsobe requested to report some RRM measurements to support E-CID positioning method: Synchronization Signal based RSRP(SS-RSRP), Synchronization Signal based Reference SignalReceived Quality (SS-RSRQ), Channel-State Information Ref-erence Signal based RSRP (CSI-RSRP), and Channel-StateInformation Reference Signal based RSRQ (CSI-RSRQ). Thefollowing UL measurements can be configured by LMF andreported by gNB: UL Relative Time of Arrival (UL-RTOA),gNB Rx-Tx time difference, Sounding Reference Signal RSRP(SRS-RSRP), Azimuth and Zenith of Angle of Arrival (A-AoA and Z-AoA, respectively). The 3GPP standard supportsthe above measurements in all supported bands and both lowerand higher frequency ranges (FR1 and FR2, respectively), overthe corresponding bandwidths within the operating frequencybands, which are up to
MHz in FR1 and
MHz in FR2.The range of reportable absolute values for power-basedmeasurements is [ − − dBm, with dB resolution, sim-ilar to that in LTE. The range of reportable absolute values fortiming-based measurements is [ − in T c units,with a flexible resolution step of kT c ,where T c correspondsto . ns and k is an integer in the interval [2; 5] for FR1and [0; 5] for FR2. The value of k can be configured by LMF or adjusted by the UE, e.g., to account for the NR subcarrierspacing and the UE reporting capability.The above measurements can be performed for serving andneighbor TRPs and can be used for a variety of RAT-dependentand hybrid positioning methods, standardized or not. Someexamples of the standardized methods are [13], • DL Time Difference of Arrival (DL-TDOA) - based onRSTD and optional complimentary PRS-RSRP; • DL Angle of Departure (DL-AoD) - based on PRS-RSRP; • UL-TDOA - based on UL-RTOA and optional compli-mentary SRS-RSRP; • UL-AoA – based on A-AoA and Z-AoA and optionalcomplimentary SRS-RSRP measurements; • Multi-RTT – based on gNB Rx-Tx and UE Rx-Tx timedifference measurements and optional complimentaryPRS-RSRP, SRS-RSRP, A-AoA, and Z-AoA measure-ments.Accurate and timely measurements are necessary to ensurereliable positioning, therefore there are standardized require-ments on the maximum allowed time during which the mea-surements are to be performed (a.k.a. measurement period)and the maximum allowed error for the reported measurements(a.k.a. measurement accuracy). The measurement period spansthe time necessary to obtain no more than four measurementsamples, while the achieved accuracy should not be worse thanthe corresponding measurement accuracy requirement, whichis expected to be more stringent than that in LTE.V. S
IMULATIONS AND D ISCUSSIONS
Based on above discussions on various aspects of NRpositioning as specified in Release 16, we demonstrate theperformance of positioning in this section. We have evaluated5G positioning performance in three different scenarios using3GPP channel models Urban Macro (UMa), Urban Micro(UMi) and Indoor Open Office (IOO) [14]. The scenarios forevaluation is taken from agreed evaluation scenarios duringthe Release 16 positioning standardization. Figure 6 showssimulation results for UMa and UMi scenarios with differentdimensions. The UMa and UMi scenarios have m × m and m × m areas respectively. The inter-sitedistances(ISD) are , and meters in UMa, UMiand IOO scenarios respectively. The UMa and UMi scenariosconsists of hexagonal cells, with each cell consisting of sectors. Hence the number of TRPs in UMa and UMiscenarios are . Whereas, in IOO it is . These scenarios PFL TRP ID TRP ID TRP ID Set ID R ID R ID R ID Set ID Set ID Set ID PFL TRP ID TRP ID Set ID R ID R ID R ID Set ID TRP ID PFL TRP ID TRP ID Set ID Set ID R ID R ID R ID TRP ID PFL TRP ID TRP ID TRP ID Set ID Set ID R ID R ID R ID Fig. 5: NR DL-PRS configuration hierarchy. The configuration hierarchy allows network to provide assistance data in astructured format. It also enables the UE to locate the resources unambiguously to perform measurements.are described in [14], [15]. The evaluations are done for UMaand UMi in FR1 and the IOO scenario is evaluated both inFR1 and FR2. The evaluations are done both in presenceand absence of interferences. Positioning methods in UMaand UMi scenarios is DL-TDOA based positioning. Whereas,in IOO we have also shown performances using multi-RTTand uplink angle of arrival (UL-AOA) based methods. Thenetwork is assumed ideally time synchronized for TDOAbased positioning evaluations. The simulations are done formaximum bandwidth corresponding to
PRBs both inuplink and downlink.
A. TRPs and the downlink simulation parameters
The downlink transmitted powers are , and dBmrespectively in the three scenarios. The carrier frequencychosen in FR1 is GHz and in FR2 it is GHz. Thesubcarrier spacings for FR1 and FR2 is chosen to be kHzand kHz respectively. The DL-PRS used in simulationhas comb-12 configuration spanning across symbols ofa slot. The receiver noise figure for receiving uplink signalsis assumed dB. The TRP heights are − m uniformlydistributed in UMa, m in UMi and m in IOO. The TRPantennas are explained in [15]. B. The UE and the uplink simulation parameters
The UL-PRS for multi-RTT and UL-AoA methods havecomb-2 configuration spanning across symbols of a slot.The receiver noise figure is assumed dB. The UE speeds areassumed to be km/h in UMa and km/h in UMi and IOO.The UE antenna is assumed dual-polarized isotropical.In Fig.6, different curves correspond to different cases forevaluation. The performances are shown in bracketsinside the legends of the figure. Many cases are evaluated inthe IOO scenario. Following observations can be drawn fromthe evaluations in IOO scenario1) Larger bandwidth in FR2 results in better performance.2) Multi-RTT has higher accuracy than TDOA based meth-ods with signals being transmitted in DL and UL. It alsorelaxes requirements on network time synchronization.3) Performance inside the convex-hull region of the IOOdeployment is better than outside convex hull region.This shows importance of deployment in positioning. Maximizing the convex hull region while deploying basestations would improve the positioning performance.4) Due to comb-12, twelve orthogonal DL-PRS avoids in-terference in IOO scenario with twelve TRPs as shownin Fig.6. Hence no performance difference is observedbetween interference-free and interference cases in IOO.In UMa and UMi scenarios, DL-PRS from multiple TRPsdo interfere in interference case as number of orthogonalDL-PRS is twelve and number of TRPs is twenty-one.5) × angle based positioning corresponds to the case with antenna elements to estimate angle of arrival in UL.Angle based positioning accuracy improves with numberof antenna elements.These evaluations are under the deployment and parameterassumptions discussed above and shown in Fig.6. Better orworse accuracy can be obtained by different set of theseselections. VI. C ONCLUSIONS
EFERENCES[1] 3GPP, “Service requirements for the 5G system,” 3rd GenerationPartnership Project (3GPP), Technical Specifications (TS) 22.261, 12,version 18.1.0.[2] ——, “Study on Communication for Automation in Vertical domains(CAV) ,” 3rd Generation Partnership Project (3GPP), Technical Report(TR) 22.804, 06, version 16.3.0.[3] X. Lin, J. Bergman, F. Gunnarsson, O. Liberg, S. M. Razavi, H. S.Razaghi, H. Ryd´en, and Y. Sui, “Positioning for the internet of things:A 3GPP perspective,”
IEEE Communications Magazine , vol. 55, no. 12,pp. 179–185, 2017.[4] S. M. Razavi, F. Gunnarsson, H. Ryd´en, ˚A. Busin, X. Lin, X. Zhang,S. Dwivedi, I. Siomina, and R. Shreevastav, “Positioning in cellularnetworks: Past, present, future,” in , 2018, pp. 1–6.[5] S. Parkvall, E. Dahlman, A. Furuskar, and M. Frenne, “NR: The new 5Gradio access technology,”
IEEE Communications Standards Magazine ,vol. 1, no. 4, pp. 24–30, 2017.
123 45 67 m/ mUMa/UMi m/ mUMa/UMiy[m] x[m]ISD = m/ m m my[m] x[m]ISD = mConvex hull region mIOO scenario Fig. 6: Simulation results for 3GPP scenarios showing per-formance curves of various cases. The performances areshown in brackets inside the legend box. [6] I. Rahman, T. Chapman, M. A. Kazmi, and F. Ghasemzadeh, “Enablingthe potential of 5G: Solutions to the technical challenges of the diverse5G bands,”
IEEE Wireless Communications , vol. 27, no. 2, pp. 6–11,2020.[7] J. A. del Peral-Rosado, F. Gunnarsson, S. Dwivedi, S. M. Razavi, O. Re-naudin, J. A. L´opez-Salcedo, and G. Seco-Granados, “Exploitation of3D city maps for hybrid 5G RTT and GNSS positioning simulations,”in
ICASSP 2020 - 2020 IEEE International Conference on Acoustics,Speech and Signal Processing (ICASSP) , 2020, pp. 9205–9209.[8] R. Di Taranto, S. Muppirisetty, R. Raulefs, D. Slock, T. Svensson, andH. Wymeersch, “Location-aware communications for 5G networks: Howlocation information can improve scalability, latency, and robustness of5G,”
IEEE Signal Processing Magazine , vol. 31, no. 6, pp. 102–112,2014.[9] A. Conti, S. Mazuelas, S. Bartoletti, W. C. Lindsey, and M. Z. Win,“Soft information for localization-of-things,”
Proceedings of the IEEE ,vol. 107, no. 11, pp. 2240–2264, 2019.[10] 3GPP, “NG-RAN; NR Positioning Protocol A (NRPPa),” 3rd GenerationPartnership Project (3GPP), Technical specification (TS) 38.455, 10,version 16.1.0.[11] ——, “NG Radio Access Network (NG-RAN); Stage 2 functionalspecification of User Equipment (UE) positioning in NG-RAN,” 3rdGeneration Partnership Project (3GPP), Technical Specification (TS)38.305, 10, version 16.2.0.[12] ——, “NR; Physical channels and modulation,” 3rd Generation Partner-ship Project (3GPP), Technical Specification (TS) 38.211, 10, version16.3.0.[13] ——, “NR; Physical layer measurements,” 3rd Generation PartnershipProject (3GPP), Technical Specification (TS) 38.215, 10, version 16.3.0.[14] ——, “Study on channel model for frequencies from 0.5 to 100 GHz,”3rd Generation Partnership Project (3GPP), Technical Report (TR)38.901, 10, version 16.1.0.[15] ——, “Study on NR positioning support,” 3rd Generation PartnershipProject (3GPP), Technical Report (TR) 38.855, 10, version 16.0.0.
Satyam Dwivedi ([email protected]) isa senior researcher at Ericsson research and teamleader for positioning research and standardizationteam. He also works on propagation. He held posi-tion as researcher at the Royal Institute of Technol-ogy (KTH) in Stockholm, Sweden. He holds Mas-ter’s and PhD from the Indian Institute of Science,Bangalore, India.
Ritesh Shreevastav ([email protected]) is a seniorresearcher and 3gpp RAN2 delegate at Ericssonresearch. He received MSc. in Telecommunicationsfrom Queens University Belfast, UK (2006) andResearch Masters in Computer Science from TrinityCollege Dublin, Ireland (2008). His interests arePositioning, Machine Type Communications andCoverage enhancements.
Florent Munier (fl[email protected]) isa senior researcher at Ericsson research involvedin 3GPP NR RAN1 standardization. He received aMSc in electrical and communication engineeringfrom Napier University in Edinburgh, scotland in2001, and a PhD in communication systems fromChalmers University of technology, G¨oteborg, Swe-den in 2007. His current research interests are NRstandardization of positioning and broadcasting.
Johannes Nygren ([email protected])is a senior researcher at Ericsson research. He de-fended his PhD on Networked Control Systems atUppsala University in 2016. His research interestsinclude robust estimation, channel estimation forpositioning, and sensor fusion.
Iana Siomina ([email protected]) is cur-rently Senior Specialist in RRM Performance, Er-icsson, Sweden. She has been working at EricssonResearch since 2006, and is a 3GPP delegate ofRRM and positioning. She has Ph.D. in Intra In-formatics from Link¨oping University (2007), M.Sc.in Mathematics from KTH Royal Institute of Tech-nology (2002), and M.Sc. in Computer Science fromStockholm University (2004).
Yazid Lyazidi ([email protected]) re-ceived his Master degree from Centrale SupelecFrance in 2014 and his PhD degree from UPMCSorbonne Universities Paris in 2017. He is a stan-dardization researcher at Ericsson Sweden.
Deep Shrestha ([email protected]) is asenior researcher at Ericsson research, Link¨oping,Sweden since 2018. Deep is involved in positioningrelated topics for 5G standardisation and developingsensing and localization concepts for B5G RadioAccess Technology. Deep did PhD in Signal Theoryand Communications at Universitat Polit`ecnica deCatalunya (UPC).
Gustav Lindmark ([email protected])is a researcher at Ericsson research working with3GPP standardization for positioning. In 2020 hereceived PhD degree in Electrical Engineering fromUniversity of Link¨oping. He previously worked withsoftware development in Automotive industry and atEricsson.
Per Ernstr¨om ([email protected]) is prin-cipal researcher at Ericsson AB focusing on 5Gpositioning. He was manager for the Ericsson 3GPPRAN standardization program 2010-2016. He re-ceived M.S. from KTH royal Institute of technologyin 1989 and Ph.D. in theoretical particle physicsfrom Stockholm University 1994 and had a researchfellowship at the Nordic center for theoretical physicat the Niels Bohr Institute in Copenhagen 1994 to1996.
Erik Stare ([email protected]) holds a M.Sc.degree (1984) in telecommunication from KTH(Royal Institute of Technology), Sweden. Between1987-1992 (at Swedish Telecom) and 1992-2018 (atTeracom) he was deeply involved in developing sys-tems and standards for digital terrestrial broadcast-ing. Since 2018 he is with Ericsson research holdingMaster Researcher position, working on positioningand a team leader for 5G/NR multicast/broadcaststandardization while being a 3GPP RAN1 delegate.
Sara Modarres Razavi ([email protected])(PhD) isa senior researcher at Ericsson research involvedin 3GPP positioning standardization since Rel.13.She is currently the project manager of LTE andNR 3GPP standardization project. She holds MScin Hardware for Wireless Communication fromChalmers University of Technology (2008) andPhD in Infra-Informatics from Link¨oping University(2014) in Sweden.
Siva Muruganathan ([email protected]) is currentlya researcher and 3GPP RAN1 delegate withEricsson Canada, Ottawa, Ontario. He previouslyheld research/postdoctoral positions at BlackBerryLimited, CRC Canada, and the University ofAlberta, Canada. His recent standardization work isin the areas of MIMO, positioning and air-to-groundcommunications.
Gino Masini ([email protected]) is princi-pal researcher with Ericsson in Sweden. He receivedhis MSc in Electronics Engineering from Politecnicodi Milano in 1996, and his MBA from SDA BocconiSchool of Management in Milano in 2008. He joinedEricsson in 1999, working with microwave radiolinks and MMIC development. Since 2009 he hasworked with 4G and 5G RAN architecture and iscurrently in his second term as 3GPP RAN WG3Chairman. ˚Ake Busin ([email protected]) received hisM.Sc. degree electrical engineering from KTH RoyalInstitute of Technology in Stockholm, in 1984. Heis currently employed with Ericsson AB, Stockholm,where he is working as a Senior Specialist LocationBased Services.