02. LTE Dimensiong v.1

Radio Network Design for Rollout • • •

Link Budget Coverage Planning Capacity

Confidential © Nokia Siemens Networks

Project Request Dear GDC Hub Indonesia, Currently we are on bidding LTE project and Please kindly your support on dimensioning of LTE with requirement as attached. Looking forward your confirmation, Parameter

Thanks Jhon.


Project Creation


Radio Module Configuration


Area Definition


Link Budget Module General Parameters •

Operating Band: – 3GPP TS 36.104 specifies 19 operating bands for FDD – Dimensioning tool generalises these to 730, 750, 800, 850, 900, 1500, 1700, 1800, 1900, 2100 and 2600 MHz – Defined by customer

•

RF Unit: – Flexi RF modules FDD, 20W, 30W and 40W Flexi RRH, 0.1W Femto (in RL40) – Default SW license is for 20W (FDD), using any other pow er has additional SW license cost – Power is referred to the power at 1 singl e antenna connector – Usually defined by Customer

•

UE Power Class: – Defined by 3GPP Class 3: 23 dB m +/- 2 dBm.

•

Channel Bandwidth: – 3GPP TS 36.104 specifies values of 1.4, 3, 5, 10, 15 and 20 MHz – Defined by customer.


Note: RL10 supports 5, 10 and 20 MHz; RL20/RL30 additionally support 15MHz

Feature Activation


Services Define



Link Budget Module Transmitting End Tx Power per Antenna [dBm]

• DL: eNodeB

Antenna Gain [dBi] • DL: eNodeB

– Antenna gain changes with the antenna type and

– Automatically updated by the tool when

frequency band

– Common value: 18 dBi directional antenna

selecting the flexi RF module in General Parameters

– Typical value: 43dBm (20W) – Automatically updated by the tool when selecting the UE Power Class in General Parameters

– Typical value: 23dBm (UE Class 3) •

UL: UE

-

0 dBi for UE antenna CPEs: Variable gains

- Outdoor: 14 dBi - Indoor: 2 dBi Confidential © Nokia Siemens Networks

Link Budget Module Transmitting End Feeder Loss [dB] • 0.4 dB if Feederless solution (jumper looses) • 2 dB feeder solution w/o TMA

•

2.4 dB if feeders with used (2 dB feeders + 0.4dB additional jumpers for TMA TMA). Automatically updated if TMA is enabled

TMA (MHA) Insertion Loss [dB] • 0.5 dB assumed if TMA in use, otherwise 0 dB. Edita from parameters worksheet

• only considered in calculations if TMA is enabled • No TMA used with feederless solution

Body Loss [dB] (only UL)

• •

Otherwise (card) : 0dB

User EIRP [dBm] EIRP: Tx Power per Antenna + Antenna Gain – Feeder Loss – TMA Insertion Loss (if TMA is present) + Total Tx Power Increase


Rooftop Model site pricing comparison – Feederless and feeder solutio performance differences Feeder solution Feeder solution

Feederless Solution

(with <=15m 7/8” cables) Downlink Uplink 32W e.g.142 dB

Downlink Uplink 39W e.g.142 dB

LTE DL Loss 0.4 dB

UL Loss 0.4 dB

0.4 dB

(with >=20m 7/8” cables + M Dow nlin 21.5W

LTE

LTE

0.4 dB

0.4 dB

43W arr er n e

DL Loss 1.3 dB .

141.6dB ~ . m ~2.94sqkm

DL RF power l ost in anten na line 43W – 0.4dB = 39W (or 2 x 19.5W)  -10% when using feederless

Feederless provides:

0.5-1 dB

0.4 dB

Higher capacity Higher coverage  Better overall RF performance  Less sites Confidential © Nokia Siemens Networks

MHA

DL Loss 3.0 dB

0.4 dB

0.4 dB

43W

140.7 dB 1.16 km, 2.6 km2 (2.45 sqkm)

Carrier in eNB DL RF power l ost in anten na line 43W – 1.3dB = 32W (or 2 x 16W)  -25% when 7/8” cable 7.5m UL site area degra dation vs. feederless  -12% when 7/8” cable 7.5 m  -17% when 7/8” cable 15 m

Presentation / Author / Date

0.5 dB

>1.2 dB

7/8” 2.1.GHz 0.5 dB =7.5m 1 dB = 15m

 

UL Loss 1.3 dB (1.8dB)

43W Carrier in eNB

DL RF power l ost in ante nna li 43W – 3dB = 21.5W (or 2 x 10.75W  -44% when 7/8” cable 20m UL site area can be slig htly high feederless, but depends on antenna lin

Link Budget Module Receiving End Noise Figure [dB] • NF depends on the receiver equipment design and represents the additive noise generated by various HW components

DL: UE – Default value: 7dB (pessimistic) UL: eNodeB – Automatically updated by the tool. – Default values can be changed in the corresponding table inside the parameter sheet (see previous slide)

– Default values:  2 dB for eNodeB (FDD HW with TMA)  2.2 dB for eNodeB (FDD HW w/o TMA)  2.8 dB for eNodeB (TD-LTE HW with TMA)  3 dB for eNodeB (TD-LTE HW w/o TMA)


Additional Gains [dB]

– Possibility of considering additional gains or

losses. In case of additional losses the num entered must be negative

– Default value: 0dB

Link Budget Module System Overhead • Overheads are automatically calculated by the tool and indicate how many resources are left for data

• Total Number of PRBs per TTI: Depends on the available BW  1.4 MHz: 6 RBs  3 MHz: 15 RBs  5 MHz: 25 RBs  10 MHz: 50 RBs  15 MHz: 75 RBs  20 MHz: 100 RBs

•

NOTE: The eNodeB scheduler works with TTI (Transmission Time Intervals). Therefore, within the dimensioning tool context, the term RB is referred to 1ms (TTI) rather than 0.5ms periods as per the standard. RB within this context sho

be understood as a ‘scheduling resource block’ of 1ms interval in time domain and 12 subcarriers in frequency domai


Link Budget Module System Overhead • Cyclic Prefix (CP): – Two options:  Normal: 7 symbols/slot; 7x12: 84 RE per RB

6 symbols/ slot; 6x12: 72 RE per  Extended: RB

– Default: Normal – Extended CP Not common dimensioning case. Currently not supported. Use in cells with long e ay sprea

• Number of OFDM symbols per subframe: – Depends on the type of CP selected  Normal: 7 symbols per slot x 2 slots per

subframe :14 symbols  Extended: 6 symbols per slot x 2 slots per

subframe: 12 symbols

• Number of PDCCH Symbols per Subframe – PDCCH carries Downlink Control Information (DCI) – Signalled by the PCFICH under the indication of the eNodeB – Based on number of active connections (increase in active connections = increase in PDCCH signalling)

– Automatically updated by the tool when selecting the Bandwid – Possible values: 1 to 4 PDCCH symbols

 Dimensioning recommendation: 3 PDCCH symbols per fra


Link Budget Module System Overhead • Number of PRBs for PUCCH – PUCCH carries the Uplink Control Information (UCI) i.e. scheduling requests, HARQ ACK/NACKS, CQI and MIMO information (Rank Indication and Precoding Matrix Indication) – PUCCH PRBs are always allocated at the edges of the channel bandwidth to avoid fragmenting PRBs allocated to PUSCH – Automaticall u dated b the tool when selectin the Bandwidth – Recommendation (used by tool)  1 PUCCH PRB in 1.4 MHz bandwidth  2 PUCCH PRBs in 3 and 5 MHz bandwidth  4 PUCCH PRBs in 10 MHz bandwidth  6 PUCCH PRBs in 15 MHz bandwidth 

8 PUCCH PRBs in 20 MHz bandwidth

The scope of RACH Density and Number of PRBs for PUCCH in the tool is to calculate UL overheads Confidential © Nokia Siemens Networks

•

RACH Density for 10ms (frame) – RACH resources occupy 6PRB in frequency domain (1. and can occupy 1, 2 or 3 subframes (ms) in time doma – Density indicates how many RACH resources are used 10ms frame and it is part of the different preamble configurations – Recommended: 1 (1 RACH resource per frame)

Link Budget Module System Overhead Downlink Reference Signal - If 1 Tx antenna: 4 Reference Signals per RB - If 2 Tx antenna, there are 8 Reference Signals per Resource Block - If 4 Tx antenna, there are 12 Reference Signals per Resource Block Example below: Normal CP (84 RE) and 2Tx antenna, overhead = 8 / 84= 9.52 %

Primar S nchronization Si nal PSS - Occupies 144 Resource Elements per frame (20 timeslots) I.e. (62 subcarriers +10 DTx) x 2 times/frame Example below: Normal CP and 2Tx antenna, overhead = 144 / (84 × 20 × 50) = 0.17 %

Secondary Synchronization Signal (SSS) – Identical calculation to PSS


Link Budget Module System Overhead Downlink PDCCH, PCFICH and PHICH - The combination of PDCCH, PCFICH and PHICH is able to occupy the first 1, 2 or 3 time domain symbols per TTI - The number of RE occupied per 1 ms TTI is given by (12 × y – x), where: • y depends upon the number of occupied time domain symbols per TTI (1, 2 or 3) • x depends upon the number of RE already occupied by the Reference Signal x = 2 for 1 transmit antenna x = 4 for 2 transmit antenna x = 4 for 4 transmit antenna when y = 1 x = 8 for 4 transmit antenna when y = 2 or 3 Example in screen shot illustrates the case for normal CP, 2 Tx and the first 3 time domain symbols occupied: overhead = (12


×

3 - 4) / (12

×

7 × 2) = 19.05%

Link Budget Module System Overhead Uplink Reference Signal • The ‘Demodulation Reference Signal is sent within the 4th time domain RE of each RB occupied by the PUSCH • Occupies all RBs not used by the PUCCH. For a 1.4 MHz Channel Bandwidth, the PUCCH occupies 1 RB per

The RE per RB (5 × number 12)/(6 × of 84) = 11.9 % is 84 when using the normal CP. This means the overhead generated by the Ref. Sig • For the normal cyclic prefix: (( (( (( (( ((


( ( ( ( (

% % % % %

Link Budget Module System Overhead Uplink PRACH • PRACH uses 6 Resource Blocks in the frequency domain. • The location of those resource blocks is dynamic. Two parameters from RRC layer define it: Configuration for Timing, 1 of 4 PRACH – PRACH preambles can be send Index in any: radio frame selecting or only inbetween even numbered ones durations and defining if PRACH

– PRACH Frequency offset: Defines the location in frequency domain

• PRACH overhead is calculated as: 6RBs * RACH Density / (#RB per TTI)* 10 TTIs per frame – RACH density: how often are RACH resources reserved per 10 ms frame i.e. for RACH density: 1 (RACH re reserved once per frame)

ChanneB lW 1.4 M Hz M3Hz M5Hz 1M 0 Hz 1M 5 Hz 2M 0 Hz


(6 (6 (6 (6 (6 (6

Overhead × 1) / (6 × 10) = 10 % × 1) / (15 × 10) = 4 % × 1) / (25 × 10) = 2.40 % 1) / (50 × 10) = 1.20 % 1) / (75 × 10) = 0.8 % × 1) / (100 × 10) = 0.6 %

×

×

Link Budget Module System Overhead Uplink

PUCCH • Ratio between the number of RBs used for PUCCH and the total number of RBs in frequency dom per TTI % % % % % %

Additional Overhead (%) • Tool allows to consider additional overheads not included in the overhead section


Link Budget Module Capacity • Modulation and Coding Scheme – 3GPP TS 36.211 specifies modulation schemes of QPSK, 16QAM and 64QAM for the Physical DL and Shared Channel

automa tically selec ts theUser bestThroughput possi ble MCS for DL and UL (automatic link adaptation) maximizin – Tool MAPL for a certain Cell Edge

• Service Type – Two possible options:  Data  AMR for different codecs (VoIP)

– Default: Data – Typical dimensioning cases w ill be for data. However, customer may require specific dimensioning fo

VoIP: RNT DIM 9.0 offers the possibility to do the dimensioning for VoIP as cell-edge service. Default i case is: AMR12.2


Modulation and Coding Scheme (MCS) 3GPP TS 36.213 specifies tables to: • link the MCS Index to a Modulation Order (modulation type) and TBS Index • link the TBS Index to a Transport Block Size (TBS) for a specific number of Physical Resource Blocks (PRB) Only a subset of the complete table (3 GPP TS 36.213 specifies 110 columns)

High MCS corresponds to high throughput

Modulation Order 2 QPSK 4 16QAM 6 64QAM ≡ ≡


≡

Link Budget Module Capacity • Cell Edge User Throughput (kbps) – Target throughput requirement to be achieved at the cell edge; minimum single UE throughpu requirement. Determines the service that can be provided at the cell border.

– It can limit the MCS to be used if the required cell edge user throughput is higher than the Max MCS Throughput

– Normally customer requirement – Tool automatically updates the MCS each time a different cell edge user throughput value is entered.


Link Budget Module Capacity: VoIP Dimensioning

•Default scenario for VoIP dimensioning is represented in scenario 2 of the tool •Service Type: AMR + codec •Cell Edge User Throughput: automatically updated based on codec according to values in the VoIP worksheet of

tool •VoIP Layer 2 Segmentation Order (UL) RL30: – Divides packets into segments on L2. Each segment is transmitted in a smaller Transport Block than the srcin one. MCS can be more robust and VoIP coverage increases – Ca acit decreases and cell ed e user throu h ut is automaticall ad usted because the additional RLC/ overhead – Not to be used together with TTI bundling (RL40)


Link Budget Module Capacity •

Residual BLER/Number of Transmissions: –

– – –

Defines the number of HARQ transmissions and a residual BLER after the last transmission Recommended value (data): 10% at 1st transmission because of the nature of link adaptation Recommended value (VoIP): 1% after the 4th transmission Tool also considers the possibility of BLER 1% and 2% at 2nd, 3rd and 4th transmissions but its use is recommended in particular cases not strictly related to an RFQ dimensioning (e.g. comparison betw LTE and GSM/UMTS link budgets on lower frequency bands or to show the potential of HARQ gain)


Link Budget Module Capacity: Number of PRBs per User TBS set • Number of user data bits transmitted to single user during one TTI (1 ms) • Transport Block occupies two resource blocks in time domain MCS = 5_QPSK  TBS_index = 5

384 / (100% - 10%) = 427 kbps …search for TBS in ITBS5 >= Air Interface #RB_used = 5  TBS = 424 bits 424 bits / TTI = 424bits / 1 ms = 424 kbps >= 427 kbps Conclusion: # RB used= 5


Link Budget Module Capacity Channel Usage per TTI Resource utilization by the user: how many PRBs are • allocated for PDSCH/PUSCH Ratio between Number of R B per User an d Total • number of RB available in the frequency domain Transport Block Size for PDSCH/PUSCH Defined by cell edge throughput and BLER • requirements • Determines the Number of RBs per User

Modulation Efficiency Transmitted bits per modulated symbol •

CR =

Effective Coding Rate

•

Coding rate applied on PDSCH/PUSCH w ith respect to the allocated resource blocks, TBS and overheads

TBS

# RB ⋅ # RE ⋅ (1 − overhead

)⋅M

order

TBS: transport block size [bits] Overhead: system overhead Modulation order: QPSK=2, 16QAM=4, 64QAM=6 #RE per RB: 168 normal CP, 144 extended CP


Link Budget Module Channel Channel Model Link level simulation results available for: • Enhanced Pedestrian A 5Hz (EPA 05) propagation channel: 5Hz Doppler shift (low speed mobiles) • Enhanced Typical Urban (ETU70) propagation channel: 70Hz Doppler shift valid for higher speed mobiles (>30km/h)

Doppler Freq = Carrier Freq * UE Speed / Speed Of Light E.g. If 2000MHz frequency band then 5Hz Doppler shift corresponds roughly to 3km/h

Antenna Configuration •

DL: 2Tx -2Rx refers to single stream 2x2 MIMO (transmit diversity only) because at cell edge is not likely to have Spatial

• •

Multiplexing (SM) When calculating capacities MIMO Spatial Multiplexing is considered UL: 2Rx is the default option in Flexi eNB


Link Budget Module Channel

Note: in 3GPP terminology, CL TxDiv is regarded as a variant of spatial multiplexing (single layer)

Tx/Rx Algorithm at eNB

• Allows to select the type of transmit diversity to be considered in calculations: Open Loop ( OL TxDiv) o Closed Loop (CL TxDiv)

• Both algorithms send one code word through the 2Tx using a pre-coding matrix wh en generating the inf that goes through each antenna Tx.

• In CL pre-coding matrix is based on feedback provided by UE (optimal for the radio conditions) • OL lacks the UE feedback therefore re-codin matrix is alwa s the same • Benefits: Improved cell edge performance (respect OL) i.e. better DL MAPL and better capacity • Recommendation: Select OL TxDiv (SFCB) if dimensioning is to be aligned with RL10, otherwise (RL2 RL30) select CL TxDiv (with PMI) as it provides better cell capacity results



FDPS (Frequency Domain Packet Scheduler) Type DL : Channel aware/ Channel unaware

• NSNs DL scheduler is channel aware (i.e. Proportional Fair in time and frequency domain) • Round Robin is the reference case in the tool for the FDPS channel aware gains UL : Channel unaware/ Interference aware (from RL30)

• Interference aware scheduler (IAS) improves the UL coverage based on IM value such as: – IM<=1 then IAS gain=0 – IM>1 then IAS gain=1, reflected in field FDPS gain field • Channel aware in UL is currently planned for RL40

FDPS Gain • Round Robin is the reference case in the tool for the FDPS channel aware gain

• Depends on the required capacity per user • FDPS Gain table specified for a 10MHz bandwidth. A scaling factor is applied for other bandwidths Confidential © Nokia Siemens Networks


DL Power Boosting and PDSCH Power Penalties • RL30 feature affecting the PCFICH, PHICH and cell specific Reference signal

• It is possible to boost the pow er of REs carrying the above control channels respect the REs carrying PDSCH

• Benefits: better detection of PCFICH, higher reliability of ACK/NACK and better channel estimation fro the RS ( i.e. may improve handover)

• Recommendation: Off , however if it needs to be ‘On’ the effects in LiBu are small i.e. small reduction DL MAPL that normally is not the limiting factor

• Below penalties are applied on PDSCH if DL power boosting is ‘on’


Link Budget Module Channel Number of Users per TTI (Loaded cell)

• Ratio between total number of RB available in the frequency domain and Number of RB per User

• Maximum number of users (100% load =100% resource utilization) which can be scheduled in the frequency domain in a single TTI are:

• 1.4 MHz: 1 • 3 MHz: 3 • 5 MHz: 7

• 10MHz:10 • 15 MHz: 15 • 20 MHz: 20

HARQ Gain

•Only applicable when using retransmissions •Gain is the SINR delta between the required SINR for BLER 10% after 1st transmission and the requir SINR to achieve the required BLER Confidential © Nokia Siemens Networks

Link Budget Module Channel Required SINR @ BLER10% • Value comes from system level simulations (SINR tables in the Parameters Sheet)

• Values is for 10% BLER after 1st Transmission • In order to get the required SINR, the following inputs must be determined:  Number of resource blocks  Antenna scheme  Channel model


Link Budget Module Channel Coding Rate Offset [dB]

• It compensates for SINR differences between the particular link budget case and the simulated one (link level). • Defined as: – SINR for the effective coding rate – min required SINR

Required SINR at Cell Edge [dB]

• Required signal level at the receiver compared to Maximum SINR at Cell Edge [dB] noise and interference in order to achieve the desired cell edge throughput requirement

• Final SINR at the cell edge taking into account

possible gains (e.g. FDPS gain and HARQ gain) and the coding rate offset


• Obtained from SL simulations (MoRSE S L simul

for 3GPP Macro Case 1 (ISD=500m) represents t 10th percentile of the SINR CDF • Input in the Interference Margin Formula

Link Budget Module Channel Cell Load (%)

• Cell load represents the resource utilization in terms of RBs • It refers to neighbour cells: no information about own cell load is considered in LiBu as intra-cell interference is not taken into account

• Affects the Interference Margin (IM) – High neighbour cell load increases the IM that in terms reduces the MAPL

interference level

• Recommended value: 50% (subject to change in future LTE releases) • Customer may provide this value • UL and DL cell load can have different values


Link Budget Module Channel Interference Margin (IM) • Relation between signals received with and without interference • DL: IM is defined by analytical methods (formula below) • UL: value is taken from simulations due to non-deterministic user’s distribution • Tool offers additional possibility of entering user defined values for DL and UL • The DL Interference Margin is defined as -10 LOG(1 – Load) where load is defined by: Req.SINR

Load

= 10

at Cell Edge 10

× Cell Load

× 10

−

Max. SINR

at Cell Edge 10

• From the formula above it shall be noted that Interference Margin is a function of required SINR, Ce Load and Maximum SINR at cell edge


Link Budget Module Channel Receiver Sensitivity [dBm] • Gives and indication of receiver’s ability for detection of low level signals Single RB bandwidth S

Rx

= − 174 dBm

/ Hz + 10 ⋅ log( 15 kHz

⋅ 12 ⋅ # RB ) + NF

+ SINR

Receiver bandwidth Noise ower Maximum Allowable Path Loss [dB] • Maximum allowable attenuation of the radio w ave traversing the air interface • Excludes clutter data (e.g. penetration looses, propagation data) – Tx EIRP – Rx Sensitivity + Rx Ant. Gain + Additional Gains - Interference Margin - Body Losses


Link Budget Module Propagation Model: Macro Case General Information

• Tool considers three deployment classes each one refers to a certain BTS Antenna Height [m], Average Penetration Loss [dB], Combined Standard Deviation [dB] and Cell Edge Probability [%]

• User can select one of these deployment class or enter the values manually MS Antenna Height [m]

• Default: 1.5 m

BTS Antenna Height [m]

• Default: 30 m


Link Budget Module Propagation Model: Macro Case

Average Penetration Loss (dB) •Depends on clutter type and frequency band •Recommendation: If not provided by values use the default ones according to the deployment scenario selected •Note: Default values are calculated for the reference of 1500 ≤ f ≤ 2600MHz. If using lower frequency bands these values are automatically corrected by a delta as per the graph below. This will have a results!

big impact in the site count

Delta values: • - 4dB for f < 700MHz • - 2dB for 700MHz <= f < 1500MH • +1dB for 2600MHz < f <= 3600MHz

•


+2dB for f > 3600MHz

Link Budget Module Propagation Model: Macro Case Combined Standard Deviation (dB) • Combined slow fading standard deviation

σ indoor

=

2

σ outdoor

2

+ σ building

Location Probability (%) • Probability for a user to be located in the cell area or at the cell edge

• Also Shadow Fading Margin or Slow Fading Margin • Difference between the signal level necessary to cover the cell with a certain probability of coverage and the average signal level at the cell edge • Calculated using the standard deviation and location probability requirement


Link Budget Module Propagation Model: Macro Case

Gain Against Shadowing (multi cell coverage) • Since the UE can be standing at the edge of two or more cells the slow fading margin can be smaller because on one of the cells needs to be offering sufficient signal strength at any point in time

• Automatically calculated by the tool. Computation based on modified Jake’s formula Maximum Allowable Path Loss [dB] (clutter considered) • Propagation data is included in the calculation • – – • Base for cell range calculations


Link Budget Module Propagation Model: Macro Case Propagation Model • Tool offers the possibility of two propagation models: – Cost 231 Model (one and two slopes) – User Defined • Recommended input: Cost 231/two slopes for all clutter types unless the customer provides the propagation model data


Link Budget Module Propagation Model: Macro Case • Clutter correction Factors

• Modified Cost231-Hata L = A + B log

h  h   f   d    − 13 . 82 log  BS  − a  MS  + s log   + L clutter  MHz   m   m   km 

F r e q u en c y 1 5 0 - 1 5 0 0M H z 1 50 0- 2 00 0M Hz

A

− L clutter =   −   −

B

6 9 .5 5

2 6 .1 6

4 6 .3

3 3 .9

• •

• UE Height Correction Factors  3.2[lg(11. 75h MS )] 2 − 4.97 DU, U a ( h MS ) =  − − − [ 1 . 1 lg( f ) 0 . 7 ] h [ 1 . 56 lg( f ) 0 . 8 ] SU , R MS 

•

Slopes

•



 h BS   44 . 9 − 6 . 55 log  m  , s =  1   47 . 88 + 13 . 9 log  f  − 13 . 82 log  h BS   × ,     MHz   m   log50



3 0 

d ≥ 1km d < 1km

2   2 ⋅  lg  f   + 5 . 4   28    (4 . 78 [lg (f )]2 − 18 . 33

(4 . 78 [lg (f )]

2

   log (f ) + 40 .

− 18 . 33 log

(f ) + 35 .

>= < Two slope is an extension of one slope model – If cell range >1km results are the same for two slope models (same formula used) – If cell range <1 km then two slope model pr results Recommended value: 2 slopes for all clutter types

Link Budget Module Site Count: Macro Case Cell Range • Calculated based on the modified Cost231-Hata formula for each clutter type: L = A + B log

  

f MHz

   − 13 . 82 log   

Site Layout Options: • Omni • 6 sectors • 3-sector antenna BW> 90o • 3-sector antenna BW<= 90o (default)

h BS m

 − 

 

a

h MS m

 + 

s log

  

d km

 + 

L clutter

Additional information provided by the tool: • Cell area (km2) • Site area (km2)

Deployment Area (sq Km) • Rough site count estimation, considering ONLY the coverage conditions, not the capacity constraints. For a (coverage and capacity) dimensioning refer to the figures in the site count sheet



Capacity Module


Capacity Calculations • DL/UL Cell throughput automatically calculated by tool • Algorithm calculates the Average Cell throughput (capacity)for a single cell • Capacity is based on:

Level simulations (spectral efficiency) for 800MHz, 2100MHz and 2600MHz to cover for differe • System bands

• LiBu inputs: operating band, channel bandwidth, channel model, cell load, scheduler and inter-site distance (ISD).

• • Cell load represents the load of the neighbour cells as per Link budget cell load. Values lower than 100% will provide better throughputs as the interference w ill be lower

• Load in victim cell is considered by default 100% ( i.e. 100% PRB utilization) as it provides better throughputs. It is not recommended to change this value

• The Deployment class should be aligned with the one in the L iBu ( i.e. basic/mature or high end)


Capacity Calculations Methodology • Four representative site grids (defined by the inter site distance) have been simulated in dynamic system level environment (MoRSE) for 800, 2100 and 2600MHz bands

• UL & DL spectral efficiency figures have been gathered for all available channel bandwidth configurations (1.4 … MHz) and for three scenarios according to the penetration losses:

– Outdoor only: 0 dB penetration loss (all UEs located outside) – Outdoor-to-Indoor Basic & Mature: penetration loss 20dB for ISD=500m,1732m; 10dB for ISD=3000m and 5 ISD=9000m. (UEs located in buildings)

– Outdoor-to-Indoor High End: penetration loss 20dB for ISD=500m,1732m,3000m,9000m. (UEs located in buildings)

• When the channel bandwidth and the ISD are known from the link budget scenario, the spectral efficiency is interpolated/extrapolated based on a look-up table obtained from the simulator


Capacity Calculations Inputs Most capacity inputs are imported from LiBu sheet based on the scenario. Fu rther tuning of parameters is possible:

Extended UL MCS Range (RL30 onwards)

• Allows for signalling of MCS21…24 as 16QAM instead of 64QAM w hich improves the data rate of Category 1…4 terminals (otherwise limited to MCS20-16QAM since they don’t support


Capacity Calculations Inputs Victim Cell Always Fully Loaded (100% PRB utilization)

•Recommendation: Yes (100% PRB utilization) as it provides the best throughput values

6-sector Deployment •In a 6 sector deployment the average cell capacity per cell is lower due to the interference increase of having 6 sectors

•Note however that the overall site capacity is higher than the 3 sector deployment •Default: n/a depends if link budget is done for 6 sector sites or not


Capacity Calculations Outputs

• DL and UL spectral efficiency for the particular Link Budget scenario (inter-site distance and bandwidth) i calculated by interpolation with the simulated results.

Purple bars obtained from simulations. Yellow bars have been interpolated based on simulation results. Confidential © Nokia Siemens Networks

Downlink Results

Effect of cell load on capacity calculations • Cell load impacts the resource utilization AND the inter-cell interference level • Simulated spectral efficiency (SE) figures consider 100% load in all cells: – Best case from the resource utilization point of view (all resources -RBs- are utilized) – Worse case from the interference point of view

• Tool considers by default 100% cell load w hen calculating cell capacity • If cell load is other than 100% (i.e. normal case if cell load is taken from Link Budget) the final spectral efficiency throughput figures in the tool reflect the improvement (by using an scaling factor) in SE/capacity due to the of interference as load will be less than simulated 100%

Inter-cell interference significantly impacts the average cell throughput in tight grids (interference limited scenarios) In typical noise limited scenarios (ISD >3km), the effect is neglectable


Cell Capacity Calculation Example

Example for ISD=500m, 10MHz, 2x2MIMO, 50% load

1. Cell capacity is estimated based on link budget scen (ISD and channel bandwidth). MIMO gain is applied case of 2 TX antennas at eNB.

2. Spectral efficiency figures havethem beenaccording simulated to forthe load case. It is needed to scale resource utilization and inter-cell interference level.

Step1: SE = interpolate_SE(ISD, channel_bandwidth) Step2: C = SE x channel_bandwidth Step3: C = C x (1 + MIM O_gain(ISD)) Step4: C = C x load x scaling_factor(load)

Step1: interpolate_SE(500m, 10MHz) = 1.19bps/Hz Step2: C = 1.19bps/Hz x 10MHz = 11.9Mbps Step3: C = 11.9Mbps x (1+ 20%) = 14.28Mbps Step4: C = 14.28Mbps x 50% x 1.37 = 9.8Mbps Confidential © Nokia Siemens Networks

Traffic Modelling


Traffic Model Scope: To calculate the total amount of offered traffic data in the BH (Total Offered Traffic)

Tool offers three ways to introduce traffic data based on customer inputs: 1. Import traffic from the ‘Traffic Model’ Sheet – Customer provides their own traffic model and traffic figures are entered in the Traffic Model (TM) Sheet 2. Directly enter traffic w ith the User Defined Option – User calculates traffic fi ures outside the tool and enters the total avera e traffic fi ures er user in the Count Sheet – Can be used if the customer provides total data figures 3. Use NSN Traffic Model (TM) – Customer doesn’t provide any traffic data. Possible to use NSN Default values in Site Count Sheet

NSN traffic model: https://sharenet-ims.inside.nokiasiemensnetworks.com/Overview/413311463 Confidential © Nokia Siemens Networks

Traffic Model From Traffic Model Sheet • Represent the total traffic (sum of traffic from all

services) for DL and UL per user in the BH • Used as an input in the Site Count Sheet when opti ‘Import from TM Sheet’ option is selected

• Tool reflects the most common inputs that define each service • By default all applications except flat rate are off. It is up to the user w hat applications to select Flat Rate • Frequently used when no particular service is specified but just a generic application • Typical for operator policies offering a peak rate in the subscriber contract but assuming that not all subscribers will use the available resources simultaneously • Link: Defined for UL and DL separately • Subscription Rate: Peak data rate expected by an active user during the BH • Overbooking Factor (OF): Throughput reduction. Fraction of total throughput. OF can also be interpreted proportion of users active and that need to be scheduled (see note for this slide)


Traffic Model From Traffic Model Sheet VoIP • Link: Both (default) when VoIP is part of the TM because VoIP is a two-way application • Call attempts: call attempts during BH

• Call Duration [s]: sum of the durations of all calls during

BH • Data Rate [kbps]: data rate depends on the codec • Service Activity: Activity factor. Normally less than 50%

Services listed as VoIP, Streaming, Www, etc. are examples of traffic model formats: a set of input parameters for data volume computation Since the customers provide this information in very different , one should choose the most appropriate format tomanner calculate data volume


1024kbps/128kbps flat rate subscription with overbooking of 25 (all subscribers use1/25 subscription rate or 1/25th of subscribers are usi the purchased subscription flat rate)

Traffic Demand


Traffic Demand = BHCA x Call Duration x Bearer Rate/ 8

Traffic Calculation


Equipment Calculation


Transport Configuration


Last Mile Capacity


Site Count Sheet Overview Calculates the total number of sites required to serve certain area while fulfilling the coverage and capacity requirements Inputs: • Population and geographical data • Subscriber distribution • Site area (from link budget) • simulations • Average data volume per subscriber for DL and UL

Outputs: • Site count for Capacity (UL and DL) and Coverage is calculated for each clutter type and for each phase • Others: – Amount of required FSMx – Throughput per eNode B


Site Count Sheet Phase: • Population and penetration rate are normally customer inputs. Data is usually provided yearly or quarterly.

• Both inputs allow for the calculation of LTE subscribers. NOTE: If LTE subscribers are provided directly then, enter the subscribers in the population field and use 100% for penetration rate

Area Size: • Customer provided input per clutter type Geographical Subscriber Distribution: • % of subscribers for each clutter type • Customer provided input per clutter type

•

Together with the Area size allows for the calculation of the number of subscribers per clutter type (dense urban, urban,…)


Number of Subscribers: • Subscriber distribution per clutter type • Calculated as: Total number of LTE subscribers * Geographical Subscriber Distribution

Site Count Sheet • Site Capacity is calculated based on the number of cells per site and the cell throughputs

• Cell throughputs are the figures calculated in the Cell Capacity Sheet

• Number of capacity sites: =

• Number of coverage sites: S

= R


e

e

Site Count Sheet Sites (Final Figure) • Total number of sites is defined by the maximum between the amount of sites needed for coverage -in blue capacity (UL and DL) -in orange -:

– Also calculated for all defined clutter types and phases


Site Count Sheet Baseband dimensioning • This module allows to estimate HOW MANY sites are required taking into account the HW (System Module) Limitations • Current version contains updated figures for RL10, RL15TD,RL20 and RL30.

System Module: Options: – FSME: high capacity system module – – FSMF+FBBB: only for TDD RL25TD • FSME is the only one supported by RL10/RL15TD. From RL30 is possible to use FSMD

Product Release: the number of supported active users • As necessary to specify the RL xx release . per FSME module (see next slide) changes with the releases it

• Recommended: n/a (it needs to be in line with the dimensioning/features used)


Site Count Sheet Baseband dimensioning Active Subscribers • Flexi SM processing power has a strict limitation for the number of active UEs w hich can be handled • UE in E-UTRAN RRC_Connected and w ith DRB (Data Radio Bearer) established but with or w ithout data be transmitted in the buffer

– Term refers to terminals actively using applications as w ell as those which do not need to be conside for scheduling

Share of active Subscribers • Percentage of active subscribers which should be handled by the eNB

• Share of Active S ubscriber values have been calculated for each of NSN Traffic Models defined in the tool:

– Voice Dominant: 11% – Data Dominant: 40% – Voice & Data Mix: 30% • If a default traffic model is not used user should assume 30% Share of Active S ubscribers for dimensioning Confidential © Nokia Siemens Networks

Site Count Sheet Baseband dimensioning #Sites (Baseband) • Number of Sites required based on the number of active users: Subscribers x ShareOfActiveSubscribers #Sites =

#MaxActiveSubscribers x NoOfCellsPerSite

# Sites Final Site Count: • Maximum figure (limiting) between # sites needed for capac ty, coverage an ase an



DL/UL Throughputs per eNB (Mbps/site) • Based on the total offered traffic and the final # of sites

Thank You


02. LTE Dimensiong v.1

Recommend Documents