RA4120-60A

LTE RPESS
LTE FDD Coverage Planning - Cell Range

© Nokia 2014 - RA41207EN60GLA0

Index
 LTE/EPC Overview
 LTE Air Interface
 Air Interface Overheads
 RRM overview
 LTE Link Budget
 Radio Planning – Coverage Planning Cell Range
 Radio Planning – Capacity
 LTE Performance Simulations
 Nokia LTE Solution
 Initial Parameters Planning

5 © Nokia 2014 - RA41207EN60GLA0

 Module Objectives
•After completing this module, the participant will be able to:
• Describe the main propagation models
• Calculate the cell range
• Compare the cell range of LTE with other technologies

6 © Nokia 2014 - RA41207EN60GLA0

 Coverage Dimensioning

• Propagation Models/ Cell Range Calculation

• Coverage Calculation
• Example: Service Dependent Coverage
• Comparison: LTE vs. other technologies

7 © Nokia 2014 - RA41207EN60GLA0

Propagation Modelling
Propagation modeling belongs to the technology independent area. Good propagation modeling is crucial for
exact network planning and dimensioning. Nokia recommendations:
• Checks should be made to ensure that the digital terrain map (DTM) is up-to-date and is accurate in terms of
clutter, vectors, and ground height. The DTM resolution should be about 20 m for urban and suburban areas and
about 50 m for rural areas.
• The number of clutter categories should be approximately 10.
• If an operator specifies the propagation model, then the results should be compared with those generated by an
Nokia model to identify any significant differences.
• Building vectors should be used if the radio network plan includes micro cells with antenna which are below the
roof-top.
• RF measurements used to tune a propagation model should be based upon at least eight sites, with a further two
sites measured for subsequent model validation. Drive routes should be planned to remain within the 3 dB beam
width of the antenna and should avoid elevated sections of road, tunnels or cuttings. The drive route should
include a balance of line-of-site and non-line-of-site locations.
• Measurement data should be filtered, binned and analyzed prior to completing tuning. Each clutter type should
have at least 400 data samples after binning. The most appropriate type of propagation models should be
determined during the tuning procedure. After tuning, the models should be validated and periodically audited.

8 © Nokia 2014 - RA41207EN60GLA0

Propagation Models
• Empirical
An equation based on extensive empirical measurements is created. Those models
can be used only in the environments similar to the examined one. The small
changes in the environment characteristic can cause enormous errors in the
prediction of wave propagation.

• Semi-empirical Combination of empirical and

deterministic models (e.g. empirical COST
Hata can be combined with the theoretical
knife edge model).

• Deterministic
Wave propagation is described by means of rays travelling between transmitted and receiving
antenna and coming in to reflections, scattering, diffractions, etc . Those methods, generally based
on ray optical techniques, give a very accurate description of the wave propagation but require a
large computation time.
9 © Nokia 2014 - RA41207EN60GLA0
Propagation Models used in Nokia tools
• Okumura-Hata

Statistical
- The most commonly used statistical model
• Walfish-Ikegami
- Statistical model especially for urban environments 

to be tuned!
• Juul-Nyholm
- Same kind of a prediction tool as Hata, but with
different equation for predictions beyond radio
horizon (~20km)

Deterministic
• Ray-tracing
- Deterministic prediction tool for
microcellular environments

10 © Nokia 2014 - RA41207EN60GLA0

Applicability of the Okumura-Hata and Walfisch-Ikegami propagation models

Okumura-Hata Walfisch-Ikegami
Frequency Range 150 MHz to 1.0 GHz 800 MHz to 2.0 GHz
1.5 to 2.0 GHz

eNodeB Antenna Height 30 to 200 m 4 to 50 m

above roof-top above roof-top
Applicability of the Okumura-Hata and
UE Antenna Height 1 to 10 m 1 to 3 mpropagation models
Walfisch-Ikegami

Range 1 to 20 m 30 m to 6 km
Applicable to Macrocells Macrocells

11 © Nokia 2014 - RA41207EN60GLA0

 Propagation Models – Okumura-Hata & COST Hata model
•In order to fit the Okumura-Hata model into the operation frequencies of 3G, some additional measurements and
adjustments were done in the framework of European Cooperation in the Field of Scientific & Technical Research
(COST)
•The validity range for the extended model:
- Frequency f: 150 MHz – 2000 MHz
- Distance R: 1-20 km
- BS height hBS: 10-200m
- MS height hMS: 1-10m correction
area type
•The correction factor c present in the model depends on area type factor [dB]
dense urban areas -3
city center areas 0
 2  f 
2  log10    5.4 suburban areas 12,27
Correction    28  for suburban areas
4.78  log 2  f   18.33  log  f   44.94 rural areas 32,52
Factor  10 10 for rural areas

12 © Nokia 2014 - RA41207EN60GLA0

Propagation Models – Okumura-Hata & COST Hata model

L = A + B  log10 (f) - 13.82  log10 (h BS ) - a(h MS )  [44.9 - 6.55  log10 (h BS )]  log(R)

+ CorrectionFactor  MorphoCorrection

 R  .............
69.99 150 MHz  f  1500 MHz 
A= 
46.30 1500 MHz  f  2000 MHz 

26.16 150 MHz  f 1500 MHz 

B= 
33.90 1500 MHz  f < 2000 MHz 

a(h MS ) = [1,1 log10 (f) - 0,7]  h MS - 1.56  log10 ( f )  0.8

13 © Nokia 2014 - RA41207EN60GLA0

Link Budget
Propagation Model: Modified Cost231-Hata
 f  h  h   d 
L  A  B log   13.82 log BS   a MS   s log   Lclutter
 MHz   m   m   km 
Frequency A B
150-1500 MHz 69.55 26.16
1500-2000MHz 46.3 33.9
3 DU
0 U

 2

  2  lg


f 
  5 .4  SU
Clutter correction Lclutter   
 28 
 
    
Term Lclutter 
 4.78 lg f   18.33 log f
2
 40.94  RURAL

 
 4.78 lg f   18.33 log f
2
 35.94  ROAD

UE Height Correction Factors a(hMS)

3.2[lg(11. 75hMS )]2  4.97 DU,U
a(hMS )  
[1.1lg( f )  0.7]hMS  [1.56 lg( f )  0.8] SU
14 © Nokia 2014 - RA41207EN60GLA0
Link Budget Propagation Model: Modified Cost231-Hata

Slopes: 1 or 2 slopes model

• 1 slope for d ≥ 1km & 2 slopes for d < 1km
• 2 slope is an extension of 1 slope model for d < 1km
– If cell range > 1 km results are the same for 1 slope & 2 slope
models ( same formula used)
– If cell range < 1 km then 2-slope model provides better results
• Recommended: 2 slopes model for all clutter types

  hBS 
 44 .9  6.55 log m , d  1km
  
s  
 

 f   hBS   1
  47 .88  13 .9 log MHz   13 .82 log m     log50 , d  1km
     

15 © Nokia 2014 - RA41207EN60GLA0

Propagation Models – Walfish-Ikegami
•Model for urban macrocellular propagation
- Antenna close to roof-top level
•Assumes regular city layout (“Manhattan grid”)
•Total path loss consists of two parts:

LOS NLOS
• line-of-sight loss • roof-to-street diffraction and scatter loss
• mobile environment losses
d

h
w
16 © Nokia 2014 - RA41207EN60GLA0
b
Propagation Models – COST Walfish-Ikegami model

•This semi empirical model is the special adaptation of Walfish-Bertoni model, prepared
especially for the typical antennas placement in 3G (below the roof top).
•The validity range:
- Frequency: 800 MHz- 2000 MHz
- BS height: 4 – 50 m (above roof-top)
- MS height: 1 – 3 m
- Distance: 0.02 – 5 km
•Path loss with LOS between MS & BS

LLOS  42.6  20 log10 ( f )  26 log10 ( R )

LOS: Line-off-sight  R  .............

17 © Nokia 2014 - RA41207EN60GLA0
Propagation Models – Walfish-Ikegami
•Line-of-sight path (LOS)
- Use free space propagation
- Applicable for microwave & satellite links
•“Non-line-of-sight” path (NLOS)
- Heavy diffraction, refraction situations
- Great uncertainties in modelling
- COST Walfish-Ikegami model includes model for NLOS prediction
- Use ray-tracing models
• Needs detailed building databases (vectorial information)

“Manhattan grid”
model

18 © Nokia 2014 - RA41207EN60GLA0

Propagation Models – COST Walfish-Ikegami model
•Path loss without LOS between MS & BS (continue)
L1  L11  k a  k d  lg(d )  k f  lg( f )  9 lg(b)
 18 lg(1  hBS  h), hBS  h
L11  
0, hBS  h
54, hBS  h


k a  54  0.8  ( hBS  h), d  0.5
hBS  h and




54  0.8  ( hBS  h)  d
0.5
, hBS  h
and d  0.5

18, hBS  h

kd   h  h
18  15  BS , hBS  h

 h
 4  0.7 f
  925  1, Medium sized cities and suburban centres
  
kf  
 4  0.7 f  1,

  925  Metropolitan centres

Mean building spacing: b [m]

Mean building height: h [m]

20 © Nokia 2014 - RA41207EN60GLA0

Propagation Models – Microcell

Ray tracing Ray

launching

Rx
Tx

Tx
Very accurate methods, but due to the complexity of the algorithms computer power consuming.
Digital maps with a high accuracy are required.

21 © Nokia 2014 - RA41207EN60GLA0

 Coverage Dimensioning
• Propagation Models/ Cell Range Calculation
• Coverage Calculation
• Example: Service Dependent Coverage
• Comparison: LTE vs. other technologies

22 © Nokia 2014 - RA41207EN60GLA0

 Coverage Area – Coverage Area in Dimensioning
•After cell radius has been determined, cell area can be calculated
•When calculating cell area, traditional hexagonal model is taken into account

R
R

Omni- or Bi-sector Tri-sector

6-sectored Site
A = 1,73 R2
2 A = 1,95 R23
2
A = 2,6 R1
Please note that for one hexagon the radius R is equal to the site length !!

23 © Nokia 2014 - RA41207EN60GLA0

 Capacity Dimensioning Process – Overview
Site Area Area Size Subscribers Traffic Model Site Capacity

Subscribers Subscriber
Density Data Volume in
BH

Total Offered Traffic

# Coverage # Capacity
Sites Sites
Max
BH = Busy Hour

# Sites

• Outputs:
• Site count for capacity and coverage
• The final number of sites is the bigger number from capacity and coverage point of view
• The calculation could be done for each clutter type and for different phases (for example years)
24 © Nokia 2014 - RA41207EN60GLA0
The Number of Sites due to Coverage
Site Area Area Size Subscribers Traffic Model Site Capacity

Subscriber Data
Subscribers
Volume in BH
Density

Total Offered Traffic

# Coverage # Capacity
Sites Sites
Max

# Sites

25 © Nokia 2014 - RA41207EN60GLA0

The Number of Sites due to Coverage
Area Size (Km²)
• this is the planned area
• typically defined for each clutter type
• the customer may provide this value
Site Area (Km²)
• this is the site area calculated from the link budget and using the propagation model
• depends on the number of cells per site (typical 3 cells per site)

Number of sites due to coverage:

# Sites due to Coverage = Roundup (Area Size / Site Area)

Example:
• Planned area is 100 Km²
• Site Area is 10 Km²
• The number of sites due to coverage is 100 Km²/ 10 Km² = 10

26 © Nokia 2014 - RA41207EN60GLA0

Coverage Area – Hexagons vs. Cells

• Three hexagons • Three cells

27 © Nokia 2014 - RA41207EN60GLA0

Coverage Area – DL simulation

28 © Nokia 2014 - RA41207EN60GLA0

Coverage Area – UL simulation

29 © Nokia 2014 - RA41207EN60GLA0

Tools
There are definitely not a large number of commercial radio network planning
tools which may be used for either the path loss based approach or the LTE
simulation based approach.

However, almost any planning tool might be used for LTE DL coverage estimation
as propagation phenomenon is more or less technology independent. Trying to
simulate UL or even DL/UL throughput presents a bigger challenge, but certain
workarounds are feasible.

Currently, there are three LTE RF planning tools commercially available:

• Atoll (www.forsk.com)
• Mentum Planet (www.mentum.com)
• Asset (www.aircominternational.com)

30 © Nokia 2014 - RA41207EN60GLA0

 Coverage Dimensioning

• Propagation Models/ Cell Range Calculation

• Coverage Calculation
• Example: Service Dependent Coverage
• Comparison: LTE vs. other technologies

31 © Nokia 2014 - RA41207EN60GLA0

Exercise
• Compare the UL & DL Pathloss and Cell Range for the following services:
• DL: VoIP 12,2 Kbps
• 1024 Kbps, 2048 Kbps and 4096 Kbps
• UL: VoIP 12,2 Kbps
• 64 Kbps, 384 Kbps and 1024 Kbps

• For the pathloss use the results and the assumptions from chapter 5: LTE Link Budget
• Building Penetration Loss, Shadowing Margin & Gain Against Shadowing are considered

• For Coverage assume:

• BTS Antenna Height: 30 m
• UE Antenna Height: 1,5 m
• Propagation Model: Modified Cost 231-Hata with 2 slopes

32 © Nokia 2014 - RA41207EN60GLA0

Solution – DL

4096 kbps data:

2048 kbps data: L = 134.02 dB
VoIP 1024 kbps data: L = 140.79 dB 1.43 km
12.2 kbps: L = 144.52 dB 2.24 km
L = 148.06 dB 2.85 km
3.60 km

33 © Nokia 2014 - RA41207EN60GLA0

 Solution – UL

VoIp 384 kbps data: 1024 kbps data:

12,2 kbps data: L = 132.46 dB L = 128.37 dB
64 kbps: L = 135.24 dB 0.99 km
1.30 km
L = 138.14 dB 1.55 km
1.88 km

34 © Nokia 2014 - RA41207EN60GLA0

 Coverage Dimensioning

• Propagation Models/ Cell Range Calculation

• Coverage Calculation
• Example: Service Dependent Coverage
• Comparison: LTE vs. other technologies

35 © Nokia 2014 - RA41207EN60GLA0

Comparison: LTE vs. other technologies --- Assumptions

36 © Nokia 2014 - RA41207EN60GLA0

Comparison: LTE vs. other technologies

37 © Nokia 2014 - RA41207EN60GLA0

Comparison: LTE vs. other technologies

- LTE deployment on existing grid should allow to achieve:

• 1024 kbps in DL & 64 kbps in UL at the cell-edge

- LTE has an additional potential for improvement:

• UL limitation can be reduced by features such as:
- TTI bundling
- enhanced frequency-selective scheduling
- interference rejection/cancellation
- enhanced/optimized power control

38 © Nokia 2014 - RA41207EN60GLA0

© Nokia 2014 - RA41207EN60GLA0