Coverage Evaluation for 5G Reduced Capability New Radio (NR-RedCap)
Saeedeh Moloudi, Mohammad Mozaffari, Sandeep Narayanan Kadan Veedu, Kittipong Kittichokechai, Y.-P. Eric Wang, Johan Bergman, Andreas Höglund
DDate of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier xx.xxxx/ACCESS.2020.DOI
Coverage Evaluation for 5G ReducedCapability New Radio (NR-RedCap)
SAEEDEH MOLOUDI , MOHAMMAD MOZAFFARI , SANDEEP NARAYANAN KADANVEEDU , KITTIPONG KITTICHOKECHAI , Y.-P. ERIC WANG , JOHAN BERGMAN , andANDREAS HÖGLUND Ericsson Research, Sweden, (e-mails: {saeedeh.moloudi, sandeep.narayanan.kadan.veedu, kittipong.kittichokechai, andreas.hoglund}@ericsson.com) Ericsson Research, Silicon Valley, Santa Clara, CA, USA, (e-mails: {mohammad.mozaffari, eric.yp.wang}@ericsson.com) Ericsson Business Unit Networks, Sweden, (e-mail: [email protected])
Corresponding author: Saeedeh Moloudi (e-mail: [email protected]).
ABSTRACT
The fifth generation (5G) wireless technology is primarily designed to address a widerange of use cases categorized into the enhanced mobile broadband (eMBB), ultra-reliable and low latencycommunication (URLLC), and massive machine-type communication (mMTC). Nevertheless, there are afew other use cases which are in-between these main use cases such as industrial wireless sensor networks,video surveillance, or wearables. In order to efficiently serve such use cases, in Release 17, the 3rdgeneration partnership project (3GPP) introduced the reduced capability NR devices (NR-RedCap) withlower cost and complexity, smaller form factor and longer battery life compared to regular NR devices.However, one key potential consequence of device cost and complexity reduction is the coverage loss. Inthis paper, we provide a comprehensive evaluation of NR RedCap coverage for different physical channelsand initial access messages to identify the channels/messages that are potentially coverage limiting forRedCap UEs. We perform the coverage evaluations for RedCap UEs operating in three different scenarios,namely Rural, Urban and Indoor with carrier frequencies 700 MHz, 2.6 GHz and 28 GHz, respectively. Ourresults confirm that for all the considered scenarios, the amounts of required coverage recovery for RedCapchannels are either less than 1 dB or can be compensated by considering smaller data rate targets for RedCapuse cases.
INDEX TERMS
I. INTRODUCTION
The fifth generation (5G) wireless technology enables a widerange of services with different requirements in terms ofdata rates, latency, reliability, coverage, energy efficiency,and connection density. Specifically, the 5G new radio (NR)primarily supports enhanced mobile broadband (eMBB) andultra-reliable and low-latency communication (URLLC) usecases [1, 2]. The 5G NR caters to flexibility, scalability, andefficiency with its unique features and capabilities. It sup-ports a wide frequency range, large bandwidth (BW), flexiblenumerology, dynamic scheduling, and advanced beamform-ing features which make it suitable for enabling various usecases with stringent data rate and latency requirements [1].Meanwhile, massive machine-type communication (mMTC)is supported by the low-power wide-area network (LPWAN)solutions such as long term evolution for machine-type com- munications (LTE-M) and narrowband Internet-of-Things(NB-IoT) [3–5]. In addition, there are still several other usecases whose requirements are higher than LPWAN (i.e., LTE-M/NB-IoT) but lower than URLLC and eMBB [6]. In orderto efficiently support such use cases which are in-betweeneMBB, URLLC, and mMTC, the 3rd generation partnershipproject (3GPP) has studied reduced capability NR devices(NR-RedCap) in Release 17 [6]. The RedCap study item hasbeen completed in December 2020 and is continued as a workitem [7].The NR-RedCap user equipment (UE) is designed to havelower cost, lower complexity (e.g., reduced bandwidth andnumber of antennas), a longer battery life, and enable asmaller form factor than regular NR UEs. These devicessupport all frequency range 1 (FR1) and frequency range2 (FR2) bands for both frequency division duplex (FDD)
VOLUME XX, 2016 a r X i v : . [ c s . I T ] J a n nd time division duplex (TDD) operations. One drawbackof complexity reduction in terms of device bandwidth re-duction or number of Rx/Tx antenna reduction for RedCapUEs is coverage loss which varies for different physicalchannels. To compensate for the coverage loss, differentcoverage-recovery solutions can be considered depending onthe coverage-limiting channels and the level of the neededcoverage recovery.Our aim in this paper is to investigate the impact ofthe complexity reduction on the coverage performance ofRedCap UEs, identify the corresponding coverage-limitingchannels, and evaluate the amount of coverage recoveryneeded for those channels. For that, we have consideredthe Rel-15 NR UEs as a reference UE, and compared thecoverage performance of RedCap UEs to the reference UEperformance for all NR physical channels used for the initialaccess, random access, as well as control and data channelsfor downlink (DL) and uplink (UL) transmissions.To evaluate the coverage performance, we have followedtwo main steps: 1) performed link-level simulations (LLSs)to obtain the required SINR, considering performance targetssuch as block error rate (BLER) for the different physi-cal channels; 2) used the LLS results and performed link-budget evaluation for both reference UE and RedCap UE.Finally, considering maximum isotropic loss (MIL) as acoverage-evaluation metric, we have identified the referenceUE channel with the lowest MIL as the bottleneck channel,(i.e. the channel that is limiting Rel-15 coverage), and thecorresponding MIL as a coverage threshold. Any RedCapchannel with MIL smaller than the threshold is consideredas coverage limiting channel and needs coverage recovery.Our results show that for RedCap UEs operating in FR1bands, PUSCH and Msg3 need approximately 3 dB and 0.8dB coverage recovery, respectively. In FR2, the impact ofcomplexity reduction is more considerable for DL channels.Based on our results, PDSCH and Msg4 require 3.4 dBand 0.5 dB coverage recovery, respectively. It should benoted that the required coverage recovery for data channelscan be compensated by reducing the data rate targets. Ourresults demonstrate key tradeoffs and guidelines needed fordesigning the NR-RedCap.The rest of the paper is organized as follows. First, weprovide a list of main abbreviations (see Table 1) usedthroughout the paper. In Section II, we provide an overviewof NR-RedCap UEs and their key features. In Section III,a detailed description and results of LLSs are presented.Subsequently, Section IV covers our link budget evaluations.Finally, the concluding remarks are provided in Section V. II. REDUCED CAPABILITY NEW RADIO DEVICES(NR-REDCAP)
The use cases envisioned for RedCap include industrial wire-less sensor network (IWSN), video surveillance cameras,and wearables (e.g., smart watches, rings, eHealth relateddevices, medical monitoring devices, etc.). The specific re-quirements of these use cases are summarized in Table 2. TABLE 1:
List of abbreviations.
Abbreviation DefinitionBW BandwidthBWP Bandwidth part in NRCRC Cyclic redundancy checkCORESET Control resource setDCI Downlink control informationDL DownlinkDMRS Demodulation reference signalFDD Frequency division duplexFDRA Frequency domain resource allocationLLS Link-level simulationLPWAN Low-power wide-area networkLTE-M Long term evolution for machine-type communicationsMCS Modulation and coding schemeMCL Maximum coupling lossMIL Maximum isotropic lossMPL Maximum path lossMsg2 Message 2 for random access response over PDSCHMsg3 Message 3 for scheduled UL transmission over PUSCHMsg4 Message 4 for contention resolution PDCCH or PDSCHNB-IoT Narrowband Internet-of-ThingsNR New RadioOFDM Orthogonal frequency-division multiplexingPDCCH Physical downlink control channelPDSCH Physical downlink shared channelPRACH Physical random-access channelPRB Physical resource blockPUCCH Physical uplink control channelPUSCH Physical uplink shared channelRx ReceiverSCS Subcarrier spacingSSB Synchronization signal blockTBS Transport block sizeTDD Time division duplexTDRA Time domain resource allocationTx TransmitterUE User equipmentUL Uplink
As can be seen from Table 2, the requirements on data rate,latency and reliability are diverse for RedCap use cases.Furthermore, these requirements differ significantly from therequirements for LPWAN use cases, currently addressed byLTE-M and NB-IoT. Thus, NR-RedCap is not intended forLPWAN use cases and is mainly intended “mid-range" IoTmarket segment.In addition to the use case specific requirements in Table 2,the following generic requirements are common to all Red-Cap use cases [6]: • Lower device cost and complexity as compared to high-end eMBB and URLLC devices of Release-15/Release-16. • Smaller device size or compact form factor, and • Support deployment in all FR1/FR2 bands for FDD andTDD.In order to meet the above generic requirements, and morespecifically the one on device complexity and device size, thefollowing features have been considered in the RedCap studyitem [6]:a) Reduced number of UE receiver (Rx) and/or transmitter(Tx) branches, VOLUME XX, XXXX
ABLE 2:
Use case specific requirements for RedCap.
IWSN(non-safety) IWSN(safety) Videosurveillance WearablesData rate(reference bit rate) UL: < 2 Mbps UL: < 2 Mbps UL : 2-4 Mbps UL : 2-5Mbps, DL :5-50 MbpsLatency < 100 ms 5-10 ms < 500 ms -Battery life Few years - - 1-2 weeksReliability 99.99% 99.99% 99%-99.9% -Note 1: High-end video e.g. for farming would require 7.5-25 Mbps.Note 2: Peak data rate of the wearables is up to 50 Mbps for uplink.Note 3: Peak data rate of the wearables is up to 150 Mbps for downlink. TABLE 3:
Estimated relative UE cost reduction for reduced num-ber of UE Rx branches.
Reduced numberof Rx branches FR1 FDD(2Rx to 1Rx) FR1 TDD(4Rx to 2Rx) FR1 TDD(4Rx to 1Rx) FR2 (2Rx to1Rx)Case 1 26% 31% 46% 31%Case 2 37% 40% 60% 40%Case 1: Total cost reduction (without DL MIMO layer reduction).Case 2: Total cost reduction (with DL MIMO layer reduction, i.e. numberof MIMO layer equal to number of Rx branches).
TABLE 4:
Estimated relative UE cost reduction for reduced maxi-mum UE bandwidth.
Reduced UEbandwidth FR1 FDD(100 MHz to20 MHz) FR1 TDD(100 MHz to20 MHz) FR2 (200MHz to 100MHz) FR2 (200MHz to 50MHz)Total cost reduction 32% 33% 16% 24% b) UE bandwidth reduction,c) Half-duplex FDD,d) Relaxed UE processing time,e) Relaxed UE processing capability.The complexity reduction features which are expected tohave the largest impact on coverage performance are (a) re-duced number of UE Rx/Tx branches and (b) UE bandwidthreduction. Therefore, in what follows, we describe thesefeatures in more detail. More details on features (c), (d) and(e) are provided in TR 38.875 [8].The reduction of minimum number of Rx and/or Txbranches relative to that of a reference Rel-15 NR UE willlower the cost and complexity of the RedCap UEs. Thereference NR UE supports 2Rx/1Tx branches in FR1 FDDbands, 4Rx/1Tx branches in FR1 TDD bands, and 2Rx/1Txbranches in FR2 bands [8]. For RedCap UEs, the config-uration for Rx and Tx branches that were considered are1Rx/1Tx and 2Rx/1Tx, in both FR1 and FR2. Furthermore,carrier aggregation is not considered. The cost reduction,relative to that of the reference NR UE and in terms ofmodem bill of materials, from reducing the minimum numberof Rx branches is summarized in Table 3 [8]. In FR1, thereduction of number of Rx branches is also beneficial interms of reducing the device size. In FR2, however, thereduction of number of Rx branches may not provide muchbenefit in terms of reducing the device size as the antennaseparation is in the order of the wavelength.In addition to the reduction in number of Rx branches,UE bandwidth reduction is another important feature thatcan considerably bring down the cost and complexity of theRedCap UE. For the estimation of relative cost/complexitysaving due to UE bandwidth reduction, a Release 15 NR UE is used as a reference. The maximum bandwidth capabilityof the reference UE is assumed to be 100 MHz in FR1 and200 MHz in FR2, for both uplink and downlink. For RedCapUEs, the bandwidth reduction options considered during thestudy item [6] are 20 MHz in FR1 and 50 or 100 MHz inFR2. The cost reduction, relative to that of the reference NRUE, is summarized in Table 4 [8].As shown in Table 3 and Table 4 , the reduction of numberof Rx branches and UE bandwidth will lead to cost savingbenefits for the RedCap UEs. The drawback, however, isthat the performance and consequently the coverage of theUEs can be negatively impacted. In the following sections,we evaluate the coverage impacts that entail from the useof these complexity reduction features. In addition to thecomplexity reduction features, the coverage analysis in FR1also takes into consideration reduced antenna efficiency dueto size limitations for devices such as wearables. The antennaefficiency loss is limited to 3 dB, and is considered for bothuplink and downlink channels in the link budget evaluations.
III. LINK LEVEL SIMULATIONS
In order to evaluate the impact of the UE complexity re-duction on coverage of RedCap physical channels, as thefirst step we have performed link-level simulations (LLS)to obtain the required SINR for the physical channels underperformance target for the both reference UEs and RedCapUEs. Then, the outcomes of the LLSs are used to performthe link budget evaluation to find coverage limiting channels.As it is expected that the coverage of a physical channelis affected by complexity reduction differently in differentfrequency bands, we have performed the LLSs for threedifferent scenarios:1) FR1, Rural with the carrier frequency of 0.7 GHz,2) FR1, Urban with the carrier frequency of 2.6 GHz,3) FR2, Indoor with the carrier frequency of 28 GHz.Any of the UL and DL initial access messages or physicalchannels can be potentially coverage limiting for RedCapUEs, therefore, we have considered LLSs for the followingmessages and channels [1]: • Synchronized signal block (SSB), including primarySS (PSS), secondary SS (SSS) and physical broadcastchannel (PBCH), is periodically transmitted on DL toinitial cell search (in this paper mainly consider PBCH),and carries the information that UE needs in order toconnect to the network, • Physical random-access channel (PRACH), is used byUE for transmission of preamble over UL, • Message 2 (Msg2) or random-access response, is trans-mitted on DL for indicating reception of the preambleand sending time alignment information, • Message 3 (Msg3) is used by UE to transmit informationsuch as a device identity that is needed for the nextmessage over PUSCH, • Message 4 (Msg4) transfers the UE to the connectedstate,
VOLUME XX, XXXX ABLE 5:
Link-level simulations assumptions for reference UE.
Carrier frequencies Rural: 700 MHz (FDD)Urban: 2.6 GHz (TDD)Indoor: 28 GHz (TDD)BWP BW Rural: 20 MHzUrban: 100 MHzIndoor: 100 MHzSCS Rural: 15 kHzUrban: 30 kHzIndoor: 120 kHzFrame structure for TDD Urban: DDDDDDDSUU (S: 6D:4G:4U)Indoor: DDDSU (S: 10D:2G:2U)Number of gNB TX chains Urban: 4Rural, Indoor: 2Number of gNB RX chains Urban: 4Rural, Indoor: 2Number of UE TX chains Rural, Urban, Indoor: 1Number of UE RX chains Urban: 4Rural, Indoor: 2Channel model Rural, Urban,: TDL-C, NLOSIndoor: TDL-A, NLOSUE antenna correlation Rural, Urban, Indoor: LowDelay spread Rural, Urban: 300 nsIndoor: 30 nsUE velocity Rural, Urban, Indoor: 3 km/h
TABLE 6:
Link-level simulations assumptions for RedCap.
BWP BW Rural: 20 MHzUrban: 20 MHzIndoor: 100 MHzNumber of UE RX chains Rural, Urban, Indoor: 2 and 1 • Physical downlink control channel (PDCCH), is mainlyused for transmission of control information such asscheduling decisions, • Physical downlink shared channel (PDSCH), is mainlyused as the main transmission of DL unicast data, • Physical uplink control channel (PUCCH) is used byUE to send information such as acknowledgments andchannel-state reports, • Physical downlink shared channel (PUSCH), is the up-link counterpart of PDSCH.Our general simulation assumptions for the reference UEare listed in Table 5. To investigate the impact of the com-plexity reduction on the BLER performance of RedCap UEs,we have also performed the LLSs for RedCap UEs consid-ering the parameters shown in Table 5, except that the UEbandwidth and the number of Rx branches are reduced asreported in Table 6.Our channel-specific assumptions, the required perfor-mance targets such as the BLER performance are discussedseparately for each channel in the following sections. More-over, the SINR requirements for meeting BLER targets fordifferent channels and signals in FR1 and FR2 are summa-rized in Tables 16-18.
A. SSB
Based on the assumptions reported in Table 7, we haveperformed the LLSs for both reference UE-SSB and RedCap- TABLE 7:
Channel-specific parameters for SSB.
Channel AssumptionsSSB (Residual) frequency offset (UE): 0.1 ppmSS burst set periodicity: 20 msPrecoder: Precoder cyclingNumber of transmissions (shots): 4BLER target: 1% -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2
SNR [dB] -2 -1 B L E R t r an s m i ss i on s FIGURE 1: BLER performance of SSB, 700 MHz.SSB.Our simulation results are shown in Figures 1-3 for Rural,Urban, and Indoor scenarios, respectively. For Rural scenarioat carrier frequency of 700 MHz, the performance (at 1%BLER) degrades by 4.4 dB considering the complexity re-ductions for RedCap UEs.Based on the results shown in Figure 2, the performancelosses for PBCH (after 4 transmissions, at 1% BLER) in-curred from reducing the number of receiver branches for aRedCap UE with respect to the reference NR UE are 3.0 dBand 6.9 dB for a 2 Rx and 1 Rx RedCap UE, respectively,for Urban scenario. For the Indoor scenario in the FR2 band,as it is shown in Figure 3, reducing the Rx branches to 1 theBLER performance degrades by 3.7 dB at 1% BLER.
B. PRACH
Table 8 represents our assumptions for LLS of PRACH. Themiss detection rate for the PRACH of Rural, Urban, andIndoor scenarios are shown in Figure 4. In the uplink, thenumber of Tx branches is the same at the reference NR UEand the RedCap UE. Furthermore, as shown in Table 8, thePRACH BW for each PRACH occasion in the frequencydomain is less than that of the RedCap UE BW in all theconsidered scenarios. Therefore, the link performance ofRedCap-PRACH is identical to that of the reference UE-PRACH. VOLUME XX, XXXX
14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4
SNR [dB] -2 -1 B L E R t r an s m i ss i on s FIGURE 2: BLER performance of SSB, 2.6 GHz. -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4
SNR [dB] -2 -1 B L E R t r an s m i ss i on s FIGURE 3: BLER performance of SSB, 28 GHz.TABLE 8:
Channel-specific parameters for PRACH.
Channel AssumptionsPRACH PRACH format:- Rural: Format 0 (1.25 KHz SCS), BW = 1.04875 MHz- Urban: Format B4 (15 KHz SCS), BW = 2.085 MHz- Indoor: Format B4 (120 KHz SCS), BW = 16.68 MHzNumber of transmissions: 1Rx combining: non-coherent combining of branchesPropagation delay (RTT):- Rural: uniformly distributed [0, 23] µs (ISD 6 km)- Urban: uniformly distributed [0, 2.7] µs (ISD 700 m)- Indoor: uniformly distributed [0, 0.077] µs (ISD 20 m)Frequency error: 0.1 ppm at UE, none at gNBPerformance target: 10% and 1% missed detectionat 0.1% false alarm probability
C. MSG2
Our simulation assumptions for performing Msg2 LLSs, areshown in Table 9. For Msg2 we have considered the payloadsize of 9 bytes and MCS index of 0 from Table 5.1.3.1-1 in -25 -20 -15 -10 -5
SNR [dB] -3 -2 -1 M i ss ed de t e c t i on r a t e FIGURE 4: Missed detection rate of PRACH.TABLE 9:
Channel-specific parameters for Msg2.
Channel AssumptionsMsg2 FDRA: 12 PRBs (considering TBS scaling factor is 0.25)TDRA: 12 OFDM symbolsWaveform: CP-OFDMDMRS: Type I, 3 DMRS symbol, no multiplexing with dataPayload/MCS index: 9 bytes/MCS0Number of transmissions: No HARQRx combining: MRCPrecoder: Precoder cycling, PRB bundle size of 2BLER target: 10% [9]. We have also considered TBS scaling of 0.25, so that asmaller TBS can be assigned to a given MCS and a givennumber of PRBs, by considering a TBS scaling factor incomputing N info as [9]: N info = SN RE Q m v, (1)where S , N RE , R , Q m , and v are the scaling factor, thenumber of available resource elements, code rate, modulationorder, and the number of transmission layers, respectively.Figures 5-7 show the BLER performance of Msg2 atcarrier frequencies of 700 MHz, 2.6 GHz and 28 GHz,respectively. As it is shown in Figure 5, at BLER performanceof 10%, by reducing the number of UE Rx branches to 1,Msg2 performance is degraded by 6.5 dB for Rural case.Based on our results shown in Figure 6, at carrier fre-quency of 2.6 GHz and BLER performance of 10%, Msg2performance is respectively degraded by 3.1 dB and 3.4 dBfor reducing the number of UE Rx branches from 4 to 2 andfrom 2 to 1. As it is shown in Figure 7, at carrier frequenciesof 28 GHz and BLER performance of 10%, reducing thenumber of UE Rx branches to 1, Msg2 performance isdegraded by 3.8 dB. D. MSG3
Table 10 shows the assumptions for LLS of Msg3. TheBLER performance of Msg3 is shown in Figure 8. Similar
VOLUME XX, XXXX
12 -10 -8 -6 -4 -2 0
SNR [dB] -2 -1 B L E R FIGURE 5: BLER performance of Msg2, 700 MHz. -12 -10 -8 -6 -4 -2 0
SNR [dB] -2 -1 B L E R FIGURE 6: BLER performance of Msg2, 2.6 GHz. -12 -10 -8 -6 -4 -2SNR [dB]10 -2 -1 B L E R FIGURE 7: BLER performance of Msg2, 28 GHz.to PRACH, the Msg3 performance of the RedCap UE is thesame as that of the reference NR UE. This is because the TABLE 10:
Channel-specific parameters for Msg3.
Channel AssumptionsMsg3 FDRA: 2 PRBsTDRA: 14 OFDM symbolsWaveform: DFT-s-OFDMDMRS: Type I, 3 DMRS symbol, no multiplexing with dataPayload/MCS index: 56 bits/MCS0Number of transmissions: No HARQRx combining: MRCNo frequency hoppingBLER target: 10% -12 -10 -8 -6 -4 -2 0 2 4 6
SNR [dB] -2 -1 B L E R ( i n i t i a l t r an s m i ss i on ) FIGURE 8: BLER performance of Msg3.TABLE 11:
Channel-specific parameters for Msg4.
Channel AssumptionsMsg4 FDRA (reference UE):- Rural: 36 PRBs- Urban: 36 PRBs- Indoor: 37 PRBsTDRA: 12 OFDM symbolsWaveform: CP-OFDMDMRS: Type I, 3 DMRS symbol, no multiplexing with dataPayload/MCS index: 130 bytes /MCS0Number of transmissions: No HARQRx combining: MRCPrecoder: Precoder cycling; PRB bundle size of 2BLER target: 10%
BW of Msg3 is assumed to be smaller than the RedCap UEBW, and the number of the Tx branches of the RedCap UE isidentical to that of the reference UE.
E. MSG4
Our simulation assumptions for LLS of Msg4 are shown inTable 11. Figures 9-11 show the BLER performance of Msg4at carrier frequencies of 700 MHz, 2.6 GHz and 28 GHz,respectively. Based on our simulation results in Figure 9,by reducing the BW and the number of UE Rx branchesto 1, Msg4 performance is degraded by 4.1 dB at carrierfrequencies 700 MHz. As it is shown in Figure 10, at carrier VOLUME XX, XXXX
10 -8 -6 -4 -2 0 2
SNR [dB] -2 -1 B L E R FIGURE 9: BLER performance of Msg4, 700 MHz. -10 -8 -6 -4 -2 0 2 4
SNR [dB] -2 -1 B L E R FIGURE 10: BLER performance of Msg4, 2.6 GHz.TABLE 12:
Channel-specific parameters for PDCCH.
Channel AssumptionsPDCCH DCI payload size: 40 bits+CRCAggregation level (AL): 16CORESET: 2 symbols and 48 PRBsPrecoding: Precoder cycling at CCE level (REG bundle = 6)BLER target: 1% frequency of 2.6 GHz and BLER performance of 10%, Msg4performance is respectively degraded by 3.5 dB and 4 dB forreducing the number of Rx branches from 4 to 2 and from 2 to1. For Indoor scenario, by reducing the BW and the numberof UE Rx branches to 1, at BLER performance of 10%, Msg4performance is degraded by 4 dB.
F. PDCCH
We have performed the LLS for PDCCH channel based onthe assumptions reported in Table 12, and our simulationresults are shown in Figures 12-14, respectively for carrierfrequencies 700 MHz, 2.6 GHz, and 28 GHz. -10 -8 -6 -4 -2 0 2 4
SNR [dB] -2 -1 B L E R FIGURE 11: BLER performance of Msg4, 28 GHz. -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2
SNR [dB] -2 -1 B L E R FIGURE 12: BLER performance of PDCCH, 700 MHz.For Rural scenario at carrier frequency of 700 MHz, theperformance (at %1 BLER) degrades by 3.5 dB consideringthe complexity reductions for RedCap UEs.The performance losses for PDCCH (at 1% BLER) in-curred from reducing the number of receiver branches for aRedCap UE with respect to the reference NR UE are 3.2 dBand 6.2 dB for a 2 Rx and 1 Rx RedCap UE, respectivelyat carrier frequency of 2,6 GHz. In FR2 band at carrierfrequency of 28 GHz, the performance loss is 3.9 dB byreducing the number of Rx branches to 1 for RedCap UE.
G. PDSCH
Table 13 show our assumptions for LLS of PDSCH. It isworth to mention that our assumptions on data rate target isbased on agreements from [6] and for the given number ofPRBs, we have selected the smallest MCS index, from table5.1.3.1-1 in [9], that satisfies our data rate constraints.The BLER performances for PDSCH at carrier frequenciesof 700 MHz, 2.6 GHz and 28 GHz are show in Figures 15-17, respectively. As it is shown in these figures, at the carrier
VOLUME XX, XXXX
12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2
SNR [dB] -2 -1 B L E R FIGURE 13: BLER performance of PDCCH, 2.6 GHz. -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
SNR [dB] -2 -1 B L E R FIGURE 14: BLER performance of PDCCH, 28 GHz.frequency of 700 MHz and 10% BLER performance, by re-ducing the number of UE Rx branches to 1, the performanceof PDSCH is degraded by 3.8 dB.As it is shown in Figure 16, at the carrier frequency of 2.6GHz and BLER performance of 10%, PDSCH performanceis respectively degraded by 3 dB and 3.2 dB for reducing thenumber of Rx branches from 4 to 2 and from 2 to 1. As itis shown in Figure 17, For a RedCap UE with 1 Rx branchand operating at the carrier frequency of 28 GHz the PDSCHperformance is 4 dB worse than that of the reference UE at10% BLER.
H. PUCCH
Table 14 shows the channel-specific parameters and perfor-mance targets for PUCCH. The LLS results for PUCCH atcarrier frequencies of 700 MHz, 2.6 GHz and 28 GHz areshown in Figures 18-23. The results show that there is nosignificant performance impact due to complexity reductionin terms of reduced BW as the PUCCH frequency resource TABLE 13:
Channel-specific parameters for PDSCH.
Channel AssumptionsPDSCH FDRA (reference UE):- Urban: 200 PRBs- Indoor: 60 PRBs- Rural: 40 PRBsFDRA (RedCap UE):- Urban: 51 PRBs- Indoor: 30 PRBs- Rural: 40 PRBsTDRA: 12 OFDM symbolsWaveform: CP-OFDMDMRS: Type I, 2 DMRS symbol, no multiplexing with dataTarget data rate/TBS/MCS (reference UE):- Urban: 10 Mbps/TBS =5640/MCS0- Indoor: 25 Mbps/TBS = 3624/MCS3- Rural: 1 Mbps/TBS = 1128/MCS0Target data rate/TBS/MCS (RedCap UE):- Urban: 10 Mbps/TBS = 1480/MCS0- Indoor: 25 Mbps/TBS =3240/MCS6- Rural: 1 Mbps/TBS = 1128/MCS0Number of transmissions: No HARQRx combining: MRCPrecoder: Precoder cycling; PRB bundle size of 2BLER target: 10% -10 -8 -6 -4 -2 0 2
SNR [dB] -2 -1 B L E R FIGURE 15: BLER performance of PDSCH, 700 MHz.TABLE 14:
Channel-specific parameters for PUCCH.
Channel AssumptionsPUCCH FDRA: 1 PRBTDRA: 14 OFDM symbolsPayload and format:- 2 bits (A/N) for format 1- 4/11/22 bits (A/N+SR/UCI) for format 3Frequency hopping: At UL BWP edgeDMRS:- Format 1: every even symbol according to the specification- Format 3: Additional DMRS configured (4 symbols)Performance target:- Format 1: 1% D2A and Aerr, 0.1% N2A- Format 3: BLER 1% VOLUME XX, XXXX
SNR [dB] -2 -1 B L E R FIGURE 16: BLER performance of PDSCH, 2.6 GHz. -6 -4 -2 0 2 4 6
SNR [dB] -2 -1 B L E R FIGURE 17: BLER performance of PDSCH, 28 GHz.TABLE 15:
Channel-specific parameters for PUSCH.
Channel AssumptionsPUSCH FDRA:- Urban: 30 PRBs- Indoor: 66 PRBs- Rural: 4 PRBsTDRA: 14 OFDM symbolsWaveform: DFT-s-OFDMDMRS: Type I, 2 DMRS symbol, no multiplexing with dataTarget data rate/TBS/MCS: using MCS Table 6.1.4.1-2 (TS38.214 [])- Urban: 1 Mbps/TBS =552/MCS3- Indoor: 5 Mbps/TBS = 736/MCS1- Rural: 100 kbps/TBS = 128/MCS6Rx combining: MRCNo frequency hoppingBLER target: 10% spans only 1 PRB. Since a single UE transmit antenna isassumed in the simulation for both RedCap and NR referenceUE, there is no performance impact related to the reductionof the number of UE antennas. -16 -14 -12 -10 -8 -6 -4 -2 0
SNR [dB] -2 -1 B L E R PF3, 4 bitsPF3, 11 bitsPF3, 22 bits
FIGURE 18: BLER performance of PUCCH format 3, 700MHz. -14 -12 -10 -8 -6 -4 -2 0 2
SNR [dB] -4 -3 -2 -1 A C K , N A , D T X e rr o r Ack error (Aerr)Nack-to-ACK (N2A)DTX-to-ACK (D2A)
FIGURE 19: BLER performance of PUCCH format 1, 700MHz. -14 -12 -10 -8 -6 -4
SNR [dB] -2 -1 B L E R PF3, 4bits, 20 MHz BWPF3, 11bits, 20 MHz BWPF3, 22bits, 20 MHz BWPF3, 4bits, 100 MHz BWPF3, 11bits, 100 MHz BWPF3, 22bits, 100 MHz BW
FIGURE 20: BLER performance of PUCCH format 3, 2.6GHz.
I. PUSCH
Our assumptions for performing PUSCH LLSs are shownin Table 15. Figure 24 and Figure 25 show the BLER per-
VOLUME XX, XXXX
14 -13 -12 -11 -10 -9 -8 -7 -6
SNR [dB] -3 -2 -1 A C K , N A , D T X e rr o r Aerr, BW 20 MHzAerr, BW 100 MHzN2A, BW 20 MHzN2A, BW 100 MHzD2A, BW 20 MHzD2A, BW 100 MHz
FIGURE 21: BLER performance of PUCCH format 1, 2.6GHz. -14 -12 -10 -8 -6 -4 -2 0
SNR [dB] -2 -1 B L E R PF3, 4bits,
FIGURE 22: BLER performance of PUCCH format 3, 28GHz. -7 -6 -5 -4 -3 -2
SNR [dB] -3 -2 -1 A C K , N A , D T X e rr o r Aerr, BW 100 MHzAerr, BW 200 MHzN2A, BW 100 MHzN2A, BW 200 MHzD2A, BW 100 MHzD2A, BW 200 MHz
FIGURE 23: BLER performance of PUCCH format 1, 28GHz.formance and data rate of the PUSCH for different carrierfrequencies. Similar to other uplink physical channels con-sidered in this paper, the number of Tx branches is the same -14 -12 -10 -8 -6 -4 -2 0 2
SNR [dB] -2 -1 B L E R ( i n i t i a l t r an s m i ss i on ) FIGURE 24: BLER performance of PUSCH. -18 -16 -14 -12 -10 -8 -6 -4 -2
SNR [dB] -2 -1 D a t a r a t e [ M bp s ] FIGURE 25: Data rate for PUSCH.at the reference UE and the RedCap UE. Furthermore, asshown in Table 15, the PUSCH transmission BW is assumedto be less than that of the RedCap UE BW in Urban, Indoorand Rural scenarios. Therefore, the link performance will beidentical for the RedCap UE and the reference UE.
IV. LINK BUDGET EVALUATION
Link budget evaluation is used to investigate coverage bytracking the transmitted power, the gains and the losses alongthe transmission path and power is sufficient so that thesystem can operate acceptably. Coverage can be expressedby different metrics such as maximum coupling loss (MCL),maximum path loss (MPL) and maximum isotropic loss(MIL) [10]. Among these metrics, MIL and MPL includethe antenna gains. However, compared to MPL, MIL ismore straightforward to compute as it does not considerparameters such as shadow fading and penetration margins.Therefore, in this paper, we have used MIL as the key cov-erage evaluation metric. Considering the simulation resultsand the corresponding performance targets for the differentphysical channels, the required SINRs to fulfill these targetsare reported in Tables 16-18, respectively, for Rural, Urban, VOLUME XX, XXXX
ABLE 16:
Required SINR (dB), 700 MHz.
BW= 20 MHz, 2Rx BW= 20 MHz, 1RxSSB -7.3 -2.9PRACH -9.2 -9.2Msg2 -9.6 -5.8Msg3 -1.5 -1.5Msg4 -5.9 -2.2PDCCH -6.6 -3.1PDSCH -5.6 -2PUCCH (2 bits) -2.9 -2.9PUCCH (11 bits) -3.4 -3.4PUCCH (22 bits) -0.9 -0.9PUSCH -2.4 -2.4
TABLE 17:
Required SINR (dB), 2.6 GHz.
BW= 100 MHz, 4Rx BW= 20 MHz, 2Rx BW= 20 MHz, 1RxSSB -11 -8 -4.1PRACH -17.5 -17.5 -17.5Msg2 -9.6 -6.5 -3.1Msg3 -6 -6 -6Msg4 -6.6 -3.1 0.9PDCCH -9.2 -6 -3PDSCH -5.7 -2.7 0.5PUCCH (2 bits) -6.6 -6.6 -6.6PUCCH (11 bits) -7.3 -7.3 -7.3PUCCH (22 bits) -5.3 -5.3 -5.3PUSCH -10.5 -10.5 -10.5
TABLE 18:
Required SINR (dB), 28 GHz.
BW= 100 MHz, 2Rx BW= 100 MHz, 1RxSSB -8.2 -4.5PRACH -12.2 -12.2Msg2 -9.4 -6Msg3 -1.8 -1.8Msg4 -5.4 -1.4PDCCH -6 -2.1PDSCH -2.3 1.7PUCCH (2 bits) -3.02 -3.02PUCCH (11 bits) -3.37 -3.37PUCCH (22 bits) -0.91 -0.91PUSCH -9.4 -9.4
TABLE 19:
Link budget assumptions.
Parameter name ValuegNB total transmit power for carrier bandwidth (dBM) Rural: 49Urban: 53Indoor: 23Number of gNB TXRUs Rural: 2Urban: 64Indoor: 2UE total transmit power (dBm) for carrier bandwidth (dBM) Rural: 23Urban: 23Indoor: 12 and Indoor scenarios.The SINR values shown in these tables are used to performlink budget evaluation based on the template [11]. Table 19shows the key assumptions that we have considered in ourlink budget evaluations. It should be noted that for RedCapUEs operating in FR1 band (Rural and Urban), due to devicesize limitations, we have considered additional 3 dB antennainefficiency compared to the reference NR UEs.In Figures 26-28 the coverage of different RedCap phys-ical channels in terms of MIL are compared to that of thecorresponding NR channels at carrier frequencies of 700MHz 2.6 GHz and 28 GHz, respectively. At each of the FIGURE 26: MIL for 700 MHz.FIGURE 27: MIL for 2.6 GHz.FIGURE 28: MIL for 28 GHz.scenarios, the NR physical channel with the lowest MIL isconsidered as coverage bottleneck channel, i.e. the corre-sponding value is the minimum acceptable MIL and the Rel-15 NR coverage limit is assumed to be given by this MIL. Wehave considered this MIL value as the minimum acceptableMIL also for RedCap channels and used it as a threshold toidentify the RedCap physical channels that need coveragerecovery. Any RedCap channel whose MIL is worse thanthat of the threshold MIL needs coverage recovery and theamount of required coverage compensation is the differenceof the RedCap-channel-MIL and the threshold MIL.As it can be seen in Figure 26 and Figure 27, for the ref-erence UE operating in Rural and Urban scenarios, PUSCHis the bottleneck channel and has the lowest MIL value(MIL = 142.8 dB for Rural and MIL = 143.9 dB forUrban). For RedCap UEs operating in Rural scenario, allthe physical channels and initial access messages, exceptPUSCH and Msg3, have MIL larger than the threshold value.Based on our results for Rural case, PUSCH and Msg3need 3 dB and 0.8 dB coverage compensation. For RedCapUEs operating in Urban scenarios, only PUSCH needs 3 dBcoverage compensation.For the reference UE in Indoor scenario, as it is shownin Figure 28, PUSCH is the bottleneck channel withMIL = 127.7 dB. For RedCap UE with 1 Rx branch, coveragecompensations of approximately 3.4 dB and 0.5 dB arerespectively, needed for PDSCH and Msg4.
VOLUME XX, XXXX . CONCLUSIONS In this paper, we have investigated the coverage performanceof the NR-RedCap UEs and identified the physical channelsthat limit the coverage of these devices. We first have pro-vided an overview of the NR-RedCap and discussed its usecases, requirements, and main features. Then, for differentdeployment scenarios and carrier frequencies (FR1 and FR2),we have evaluated the link performance of RedCap UEs andperformed link-budget evaluations for all physical channelsand messages for DL and UL transmissions.Our results have shown that for RedCap UEs operatingin FR1 band, the PUSCH can limit the coverage, and itneeds 3 dB coverage recovery. It is worth to highlight twoobservations; first, the 3 dB coverage loss resulting from theUE antenna efficiency loss due to device size limitations;second, by reducing the data rate target for RedCap UEsin UL, no coverage recovery is needed. For the Rural case,a small amount of coverage compensation (approximately0.8 dB) is needed for Msg3. For RedCap UEs operatingin FR2 band, the impact of complexity reduction is moreconsiderable for DL channels, and PDSCH and Msg4 arethe channels that may need coverage recovery. However, theamount of coverage-compensation needed for Msg4 is lessthan 0.5 dB and by considering smaller data rates no coveragerecovery is needed for PDSCH.
REFERENCES [1] E. Dahlman, S. Parkvall, and J. Skold, 5G NR: Thenext generation wireless access technology. AcademicPress, 2018.[2] S. Ahmadi, 5G NR: Architecture, Technology, Imple-mentation, and Operation of 3GPP New Radio Stan-dards. Academic Press, 2019.[3] GSMA, “Mobile IoT in the 5G future-NB-IoT and LTE-M in the context of 5G,” White paper, 2018.[4] O. Liberg, M. Sundberg, E. Wang, J. Bergman, andJ. Sachs, Cellular Internet of Things: Technologies,Standards, and Performance. Academic Press, 2017.[5] M. Mozaffari, Y.-P. E. Wang, O. Liberg, andJ. Bergman, “Flexible and efficient deployment of NB-IoT and LTE-MTC in coexistence with 5G new radio,”in Proc. IEEE INFOCOM, April 2019, pp. 391–396.[6] 3GPP Tdoc RP-201677, Ericsson, “Revised SID onstudy on support of reduced capability NR devices,”July 2020.[7] 3GPP Tdoc RP-202933, Ericsson and Nokia, “NewWID on support of reduced capability NR devices,”Dec. 2020.[8] 3GPP TR 38.875, “Study on support of reduced capa-bility NR devices (Release 17),” Dec. 2020.[9] 3GPP TS 38.214, “NR; physical layer procedures fordata (Release 16),” 2019.[10] 3GPP TR 38.830, “Study on NR coverage enhance-ments (Release 17),” Dec. 2020.[11] 3GPP Tdoc R1-2007481, Moderator (Ericsson, Apple,Qualcomm), “Feature lead (FL) summary 4 for redcap evaluation templates, RAN1 102-e,” Aug. 2020.12