RF Network Design
Network Planning
Introduction • The high level life cycle of the RF network planning process can be summarised as follows :•
To help the operator to identify their RF design requirement • Optional • Discuss and agree RF design parameters, assumptions and objectives with the customer • Coverage requirement • Traffic requirement • Various level of design (ROM to detail RF design)
Comparative Analysis
• Site Realisation
Issuing of search ring Cand. assessment Site survey, design, approval Drive test (optional)
• • •
RF Design requirement
• RF Design Implementation
RF Design
• • •
Frequency plan Neighbour list RF OMC data Optimisation
Comparative Analysis • This is an optional step • This is intended to :o Help an existing operator in building/expanding their network o Help a new operator in identifying their RF network requirement, e.g. where their network should be built • For the comparative analysis, we would need to :o Identify all network that are competitors to the customer o Design drive routes that take in the high density traffic areas of interest o Include areas where the customer has no or poor service and the competitors have service
Comparative Analysis • The result of the analysis should include :• For an existing operator o All problems encountered in the customer’s network o All areas where the customer has no service and a competitor does o Recommendations for solving any coverage and quality problems • For a new operator o Strengths and weaknesses in the competitors network o Problem encountered in the competitors network
RF Network Design Inputs • The RF design inputs can be divided into :o Coverage requirements Target coverage areas Service types for the target coverage areas. These should be marked geographically Coverage area probability Penetration Loss of buildings and in-cars o Capacity requirements Erlang per subscriber during the busy hour Quality of service for the air interface, in terms GoS Network capacity
RF Network Design Inputs • Available spectrum and frequency usage restriction, if any • List of available, existing and/or friendly sites that should be included in the RF design • Limitation of the quantity of sites and radios, if any • Quality of Network (C/I values) • Related network features (FH, DTX, etc.)
Coverage Design Inputs by BSNL • Coverage Thresholds o
o
o o
Indoor Coverage : Signal Level measured at street better than –65 dBm. Indoor coverage to be provided in commercial complexes, hotels,technology parks etc. In Car Coverage: Signal Level measured at street better than –75 dBm. In Car coverage to be provided in residential areas, highways, tourist spots etc. Outdoor Coverage : Signal level measured at street better than –85 dBm. All remaining areas to be covered with Outdoor coverage. These are general guidelines for planning , specific areas not provided.
Capacity Design Inputs by BSNL
• • • • •
Frequency spectrum available 6.2 MHz (31 channels). Average traffic per sub for RF design : 50 mErlang. Synthesizer frequency hopping can be used. GOS: 2% Existing network Database o Total No. of sites with configuration o Site details eg location(Lat-Long), Antenna height ,azimuth, etc.
RF Network Design • There are 2 parts to the RF network design to meet the :o Capacity requirement o Coverage requirement • For the RF Coverage Design
CW Drive Testing
Propagation Model
Digitised Databases
RF Coverage Design
Customer Requirements
Link Budget
CW Drive Testing • CW drive test can be used for the following purposes :o Propagation model tuning o Assessment of the suitability of candidate sites, from both coverage and interference aspect • CW drive test process can be broken down to :•
Test Preparation
• • •
Propagation Test Data Processing
• • •
Equipment required BTS antenna selection Channel selection Transmitter setup Receiver setup Measurement averaging Report generation
• • •
Power setting Drive route planning Test site selection
• •
Drive test Transmitter dismantle
CW Drive Testing - Test Preparation • The test equipment required for the CW drive testing :o Receiver with fast scanner Example : HP7475A, EXP2000 (LCC) etc. The receiver scanner rate should conform to the Lee Criteria of 36 to 50 sample per 40 wavelength o
CW Transmitter Example : Gator Transmitter (BVS), LMW Series Transmitter (CHASE), TX-1500 (LCC) etc.
o
Base Station test antenna DB806Y (Decibel-GSM900), 7640 (Jaybeam-GSM1800) etc.
o
Accessories Including flexible coaxial cable/jumper, Power meter, extended power cord, GPS, compass, altimeter
CW Drive Testing - Test Preparation • Base Station Antenna Selection o The selection depends on the purpose of the test o For propagation model tuning, an omni-directional antenna is preferred o For candidate site testing or verification, the choice of antenna depends on the type of BTS site that the test is trying to simulate. For Omni BTS : Omni antennas with similar vertical beamwidth For sectorised BTS Utilising the same type of antenna is preferred Omni antenna can also be used, together with the special feature in the post processing software like CMA (LCC) where different antenna pattern can be masked on over the measurement data from an omni antenna
CW Drive Testing - Test Preparation • Test Site Selection • For propagation model tuning, the test sites should be selected so that :o They are distributed within the clutter under study o The height of the test site should be representative or typical for the specific clutter o Preferably not in hilly areas • For candidate site testing/verification, the actual candidate site configuration (height, location) should be used. • For proposed greenfield sites, a “cherry-picker” will be used.
CW Drive Testing - Test Preparation • Frequency Channel Selection o The necessary number of channels need to be identified from the channels available With input from the customer o The channels used should be free from occupation From the guard bands Other free channels according to the up-to-date frequency plan o
The channels selected will need to be verified by conducting a pre-test drive It should always precede the actual CW drive test to verify the exact free frequency to be used It should cover the same route of the actual propagation test A field strength plot is generated on the collected data to confirm the channel suitability
CW Drive Testing - Test Preparation • Transmit Power Setting
• For propagation model tuning, the maximum transmit power is used • For candidate site testing, the transmit power of the test transmitter is determined using the actual BTS link budget to simulate the coverage • On sites with existing antenna system, it is recommended that the transmit power to be reduced to avoid interference or intermodulation to other networks. • The amount of reduction is subject to the possibility if separating the test antenna from the existing antennas
CW Drive Testing - Test Preparation • Drive Route Determination o The drive route of the data collection is planned prior to the drive test using a detail road map Eliminate duplicate route to reduce the testing time o For propagation model tuning, each clutter is tested individually and the drive route for each test site is planned to map the clutter under-study for the respective sites. o It is important to collect a statistically significant amount of data, typically a minimum of 300 to 400 data points are required for each clutter category o The data should be evenly distributed with respect to distance from the transmitter o In practice, the actual drive route will be modified according to the latest development which was not shown on the map. The actual drive route taken should be marked on a map for record purposes.
CW Drive Testing - Propagation Test •
Transmitter Equipment Setup o Test antenna location Free from any nearby obstacle, to ensure free propagation in both horizontal and vertical dimension For sites with existing antennas, precaution should be taken to avoid possible interference and/or inter-modulation o
Transmitter installation
o
A complete set of 360º photographs of the test location (at the test height) and the antenna setup should be taken for record
CW Drive Testing - Propagation Test • Scanning Receiver Setup - HP 7475A Receiver Example
HP 7475A Receiver
CW Drive Testing - Propagation Test • Scanning Receiver Setup o The scanning rate of the receiver should always be set to allow at least 36 sample per 40 wavelength to average out the Rayleigh Fading effect. For example: scanning rate = 100 sample/s test frequency = 1800 MHz therefore, to achieve 36 sample/40 wavelength, the max. speed is =
o
It is recommended that : Beside scanning the test channel, the neighbouring cells is also monitored. This information can be used to check the coverage overlap and potential interference Check the field strength reading close to the test antenna before starting the test, it should approach the scanning receiver saturation
CW Drive Testing - Propagation Test • Drive Test o Initiate a file to record the measurement with an agreed naming convention o Maintain the drive test vehicle speed according to the pre-set scanning rate o Follow the pre-plan drive route as closely as possible o Insert marker wherever necessary during the test to indicate special locations such as perceived hot spot, potential interferer etc. o Monitor the GPS signal and field strength level throughout the test, any extraordinary reading should be inspected before resuming the test. • Dismantling Equipment o It is recommended to re-confirm the transmit power (as the pre-set value) before dismantling the transmitter setup
Measurement Data Processing • Data Averaging o This can be done during the drive testing or during the data processing stage, depending on the scanner receiver and the associated post-processing software o The bin size of the distance averaging depends on the size of the human made structure in the test environment • Report Generation o For propagation model tuning, the measurement data is exported into the planning tool (e.g. Asset) o Plots can also be generated using the processing tool or using MapInfo o During the export of the measurement data, it is important to take care of the coordinate system used, a conversion is necessary if different coordinate systems are used.
Propagation Model • Standard Macrocell Model for Asset o Lp (dB) = K1 + K2 log(d) + K3 Hm + K4 log(Hm) + K5 log(Heff) o + K6 log(Heff) log(d) + K7 Diffraction + Clutter factor o where Lp, Diffraction, Clutter factor are in dB o d, Hm, Heff are in m o It is based on the Okumura-Hata empirical model, with a number of additional features to enhance its flexibility o Known to be valid for frequencies from 150MHz to 2GHz o Applies in condition : Base station height : 30 - 200 m Mobile height : 1 - 10 m Distance : 1 - 20 km o An optional second intercept and slope (K1, K2) for the creation of a twopiece model with the slope changing at the specified breakpoint distance.
Morphology Class
Link Budget Link Budget Element of a GSM Network BTS Antenna Gain
LNA (optional)
Feeder Loss ACE Loss
Diversity Gain
BTS Transmit Power
BTS Receiver Sensitivity
Max. Path Loss
Fade Margin
Penetration Loss MS Antenna Gain, Body and Cable Loss Mobile Transmit Power
Mobile Receiver Sensitivity
Link Budget • BTS Transmit Power o Maximum transmit power o GSM900 and 1800 networks use radios with 46dBm maximum transmit power • ACE Loss o Includes all diplexers, combiners and connectors. o Depends on the ACE configuration o The ACE configuration depends on the number of TRXs and combiners used
Link Budget • Mobile Transmit Power o GSM900 : Typical mobile class 4 (2W) o GSM1800 : Typical mobile class 1 (1W)
• Mobile Receiver Sensitivity o The sensitivity of GSM900 and GSM1800 mobile = -102 dBm
Link Budget • Diversity Gain o Two common techniques used : Space Polarisation o Reduce the effect of multipath fading on the uplink o Common value of 3 to 4.5 dB being used • BTS Receiver Sensitivity o Depends on the type of propagation environment model used, most commonly used TU50 model o BTS : Receiver Sensitivity for GSM900 = -107 dBm
Link Budget • Feeder Loss o Depends on the feeder type and feeder length o The selection of the feeder type would depends on the feeder length, I.e. to try to limit to feeder loss to 3 -4dB. • BTS Antenna Gain o Antenna gain has a direct relationship to the cell size o The selection of the antenna type depends on : The morphology classes of the targeted area and coverage requirements Zoning and Local authority regulations/limitations o Common antenna types used : 65º, 90º, omni-directional antennas with different gains
Link Budget • Slow Fading Margin o To reserve extra signal power to overcome potential slow fading. o Depends on the requirement of coverage probability and the standard deviation of the fading o A design can take into consideration : both outdoor and in-building coverage, which utilises a combined standard deviation for indoor and outdoor (Default value = 9dB) Only outdoor coverage (Default value = 7dB) Pathloss slope used, 45dB/dec (Dense Urban), 42dB/dec (Urban), 38dB/dec (Suburban) and 33dB/dec (Rural)
Link Budget • Penetration Loss o Penetration loss depends on the building structure and material o Penetration loss is included for in-building link budget o Typical value used for Asia-Pacific environment (if country specific information is not available) : Dense Urban : 20 dB Urban : 18 dB Suburban : 15 dB Rural : 9 dB • Body Loss o Typical value of 3dB body loss is used • MS Antenna Gain o A typical mobile antenna gain of 2.2 dBi is used
Link Budget • Link Budget Example (GSM900)
Antenna • Antenna Selection o Gain o Beamwidths in horizontal and vertical radiated planes o VSWR o Frequency range o Nominal impedance o Radiated pattern (beamshape) in horizontal and vertical planes o Downtilt available (electrical, mechanical) o Polarisation o Connector types (DIN, N) o Height, weight, windload and physical dimensions
Antenna • The antenna selection process o Identify system specifications such as polarisation, impedance and bandwidth o Select the azimuth or horizontal plane pattern to obtain the needed coverage o Select the elevation or vertical plane pattern to be as narrow as possible, consistent with practical limitations of size, weight and cost o Check other parameters such as cost, power rating, size, weight, mounting capabilities, wind loading, connector types, aesthetics and reliability to ensure that they meet system requirements
Antenna • System Specification o Impedance and frequency bandwidth is normally associated with the communication system used o The polarisation would depends on if polarisation diversity is used • Horizontal Plane Pattern o Three categories for the horizontal plane pattern : Omnidirectional Sectored (directional) Narrow beam (highly directional) • Elevation Plane Pattern o Choosing the antenna with the smallest elevation plane beamwidth will give maximum gain. However, beamwidth and size are inversely related o Electrical down tilt o Null filling
Nominal RF Design Link Budget
Propagation model
Coverage requirements
Site radius
Nominal RF Design (coverage)
Traffic requirements
Maximum path loss
•
Typical site configuration • Transmit Power
•
•
•
•
Antenna configuration (type, height, azimuth) Site type (sector, omni)
Traffic requirements
Standard hexagon site layout Friendly, candidate sites Initial site survey inputs Coverage
•
•
site
Recalculate the site radius using the number of sites from the traffic requirement Repeat the nominal RF design
count
Traffic site count
Traffic > Cov.
Cov. > Traffic
Nominal site count
Nominal RF Design • Calculation of cell radius o A typical cell radius is calculated for each clutter environment o This cell radius is used as a guide for the site distance in the respective clutter environment o The actual site distance could varies due to local terrain • Inputs for the cell radius calculation :o Maximum pathloss (from the link budget) o Typical site configuration (for each clutter environment) o Propagation model
Nominal RF Design • There are different level of nominal RF design :o Only using the cell radius/site distance calculated and placing ideal hexagon cell layout o Using the combination of the calculated cell radius and the existing/friendly sites from the customer The site distance also depends on the required capacity • In most mobile network, the traffic density is highest within the CBD area and major routes/intersections • The cell radius would need to be reduce in this area to meet the traffic requirements • BASED ON THE SITE DISTANCE & THE COVERAGE REQUIREMENTS CELL COUNT BASED ON COVERAGE IS CALCULATED.
Nominal RF Design • Cell count based on traffic is derived based on capacity inputs: Capacity requirements GOS Spectrum availability Freq. Hopping techniques • If the total sites for the traffic requirement is more than the sites required for coverage, the nominal RF design is repeated using the number of sites from the traffic requirement o Recalculating the cell radius for the high traffic density areas o The calculation steps are : Calculate the area to be covered per site Calculate the maximum cell radius Calculate the site distance
Site Realisation • After completion of Nominal design based on cell count ( coverage & capacity requirements) , search rings for each cell site issued. • Nominal design is done , with the existing network in place(existing BTS). Existing site location remain unchanged , azimuth , tilts as per the new design requirements. • Based on the search ring form physical site survey is undertaken.
Site Realisation Search Ring Form • • • • • • • •
Site ID Site Name Latitude/Longitude Project name Issue Number and date Ground height Clutter environment Preliminary configuration o o o o
Number of sector Azimuth Antenna type Antenna height
• Location Map & SR radius • Search ring objective • Approvals
Site Realisation Release of Search Ring
Suitable Candidates?
Y
Candidates Approved?
N
N
Problem identifying candidate
Next candidate
N
Y
All parties agreed at Caravan
Arranged Caravan
N Caravan next candidate
Exhausted candidates
N Y Discuss alternative with customer
Driveby, RF suggest possible alternative
N
Issue design change
N
Cell split required
Y
Candidate approved?
Exhausted candidates
Y
Y
Additional sites required
Y
N
Y Produce Final RF Design
Site Realisation • Candidate Assessment Report-Site Survey Forms o Site survey Forms for all suitable candidates for the search ring o For each candidates : Location (latitude/longitude) Location map showing the relative location of the candidates and also the search ring Candidate information (height, owner etc) Photographs (360º set, rooftop, access, building) Possible antenna orientations Possible base station equipment location Information for any existing antennas Planning reports/comments (restrictions, possibilities of approval etc.)
Site Realisation-Site Survey Form • Final RF Configuration Form o Base Station configuration Azimuth Antenna height Antenna type Down tilt Antenna location Feeder type and length BTS type Transmit power Transceiver configuration
Traffic Engineering
Spectrum Available
Reuse factor
Traffic Requirement Maximum number of TRX per cell
Channel loading
No of TCH available
Traffic offered
Subscriber supported
Traffic Engineering • Traffic Requirement • The Erlang per subscriber • Grade of Service (GoS) o GoS is expressed as the percentage of call attempts that are blocked during peak traffic o Most cellular systems are designed to a blocking rate of 1% to 5% during busy hour
Traffic Engineering • Frequency Reuse o In designing a frequency reuse plan, it is necessary to develop a regular pattern on which to assign frequencies o The hexagon is chosen because it most closely approximated the coverage produced by an omni or sector site o Common reuse factor : 4/12, 7/21
Traffic Engineering • Channel Loading o As the number of TRX increases, the control channels required increases accordingly o The following channel loading is used for conventional GSM network o For services such as cell broadcast, additional control channels might be required
Traffic Engineering • After determining the number of TCH available and the traffic requirements, the traffic offered is calculated using the Erlang B table o For example, for a 2% GoS and 3 TRX configuration, the traffic offered is 14 Erlang o If the traffic per subscriber is 50mE/subscriber, then the total subscribers supported per sector = 280 • For a uniform traffic distribution network, the number of sites required for the traffic requirement is :-
Traffic Engineering • Erlang B Table
Traffic Engineering • If a traffic map is provided, the traffic engineering is done together with the coverage design • After the individual sites are located, the estimated number of subscribers in each sector is calculated by :o Calculating the physical area covered by each sector o Multiply it by the average subscriber density per unit area in that region o The overlap areas between the sectors should be included in each sector because either sector is theoretically capable of serving the area • The number of channels required is then determined by :o Calculating the total Erlangs by multiplying the area covered by the average load generated per subscriber during busy hour o Determine the required number of TCH and then the required number of TRXs o If the number of TRXs required exceeded the number of TRXs supported by the available spectrum, additional sites will be required
SWAP PLAN • Why do we need a swap plan? To reduce mix of different vendor BTS within a large city/ area o Reduce Inter MSC HO. o Better maintenance efficiency
Swap Strategy o No. of existing BTS sites with configuration known o No. of new sites with configuration known.
For Example BSNL UP(W) Circle
UP(W) Circle Network Diagram Haryana Saharanpur
Uttaranchal
Muzaffarnagar Bijnor
NCR
Meerut
Nepal Ghaziabad
Moradabad
Delhi Noida Bulandshahr
Rampur Bareilly Budaun
Haryana
Pilbhit
Mathura Aligarh Etah
Rajasthan
Agra
UP(E) Mainpuri
Etawah
Nokia BTS
Ericcsson BTS
All DHQ on Nokia
UP(W) Circle Network Distribution • Major Cities /SSA’s to be deployed on Nokia BTS o o o o o o o o o
DHQ of all SSA’s Meerut Agra Mathura Noida Ghaziabad Muzaffarnagar Aligarh Bulandshahar
• SSA’s except DHQ’s deployed on Ericsson BTS o o o o o o o o o o
Bijnor Bareilly Moradabad Etah Etawah Rampur Pilbhit Badaun Mainpuri Saharanpur
HW & Rly Plan for UPW NH-58 Haryana Saharanpur
Uttaranchal
Muzaffarnagar Bijnor Meerut Ghaziabad Delhi NH-02
Moradabad
Noida
Nepal
Rampur Bulandshahar Pilbhit Badaun Bareilly
69 Ericsson HW Site 56 Nokia HW Site
Haryana
National HW
Aligarh Etah Mathura
Railways
State Highway
Agra Rajasthan
Mainpuri
UP(E)
District Border
NH-91 Etawah NH-03
NH-24
Sl NO
SSA
PH-IV PLANNED SWAP NOKIA NOKIA WITH ERICSSON A
B
SWAP EXISTING SWAP SUMMARY ERICSSON ERICSSON
TOTAL NOKIA
TOTAL ERICSSON
Highways Nokia
GRAND TOTAL
E
F
G
H
(A+D-B)
(C-D+B)
WITH NOKIA C
D
(E+F+G)
1
Agra
74
2
43
37
109
8
8
125
2
Aligarh
40
4
27
19
55
12
1
68
3
Badaun
16
10
11
3
9
18
1
28
4
Bareilly
45
11
27
17
51
21
2
74
5
Bijnor
39
32
16
3
10
45
0
55
6
Bulandshahar
27
3
17
12
36
8
1
45
7
Etah
17
12
10
3
8
19
3
30
8
Etawah
29
21
16
4
12
33
0
45
9
Ghaziabad
27
1
15
9
35
7
0
42
10
Mainpuri
22
17
12
2
7
27
0
34
11
Mathura
34
1
22
17
50
6
7
63
12
Meerut
68
5
30
26
89
9
11
109
13
Moradabad
73
35
33
16
54
52
9
115
14
Muzaffarnagar
48
10
17
13
51
14
3
68
15
Noida
12
0
8
6
18
2
0
20
16
Pilbhit
11
6
6
2
7
10
5
22
17
Rampur
20
13
11
3
10
21
0
31
18
Saharanpur
31
18
16
9
22
25
5
52
Total
633
201
337
201
633
337
56
1026
UP(W) Circle 24volt BTS Distribution • Before Swap 24volt’s (40) BTS status o Agra – 9 o Aligarh – 2 o Bareilly – 5 o Mathura – 2 o Meerut – 3 o Moradabad – 6 o Saharanpur – 4 o Bijnor – 2 o Bulandshahar – 2 o Etah – 1 o Etawah –3 • After Swap 24volt’s (40) BTS status Pilibhit oo Agra – 1– 1 o o o o o o
Moradabad – 16 Saharanpur – 1 Bijnor – 17 Etah – 1 Etawah – 3 Bulandshahr – 1
•
Out of 40 sites 31 have been swapped to o Bijnor – 16 o Moradabad – 15
•
Out of 40 sites 9 left as it is (No Swap) o Agra - 1 o Moradabad – 1 o Saharanpur – 1 o Bijnor – 1 o Bulandshahr – 1 o Etah – 1 o Etawah – 3
Advanced Network Planning Steps
Parameter Planning
• Parameter planning means creating a default set of BSS parameters. • The most important parameters to plan for: o frequencies o BSIC o LAC o handover control parameters o adjacent cell definitions.
BSS Parameter • Relevant BSS parameter for NW planning o frequency allocation plan o transmit power o definition of neighbouring cells o definition of location areas o handover parameters o power control parameters o cell selection parameters
Handover Types • • • • •
Intracell same cell, other carrier or timeslot Intercell between cells (normal case) Inter-BSC between BSC areas Inter-MSC between MSC areas Inter- PLMN e.g. between AMPS and GSM systems
intracell intercell
inter-BSC
Handover Criteria 1. Interference, UL and DL 2. Bad C/I ratio 3. Uplink Quality 4. Downlink Quality 5. Uplink Level 6. Downlink Level 7. Distance 8. Rapid Field Drop
9. MS Speed 10. Better Cell, i.e. periodic check (Power Budget, Umbrella Handovers) 11. Good C/I ratio 12. PC: Lower quality/level thresholds (DL/UL) 13. PC; Upper quality/level thresholds (DL/UL)
Location Area Design 1/2 • Location updating affects all mobiles in network o LocUp in idle mode o LocUp after call completion • Location updating causes signallingmajor road and processing load within the network (international LocUpdate !) • Avoid oscillating LocUpdate • Trade-off between Paging load and Location Update signalling
Location area 2
Location area 1
Location Area Design 2/2 • Different MSC can not use the same LAC. • Location areas are important input for transmission planners o should be planned as early as possible. • Never define location area borders along major roads! • Dual band or microcellular networks require more attention on LAC planning o co-located DCS and GSM cells are defined to the same LAC o same MSC to avoid too much location updates which would cause very high SDCCH blockings
Network Optimisation
What is network optimisation?
Network Optimisation is:
• Improving network quality from a subscribers point of view. • Improving network quality from an operators point of view.
What is network quality?
Overall Network Quality • H/W Failure • Network
O P E R A T O R
C U S T O M E R
NETWORK
SERVICES MOBILE COST
Configuration • Network Traffic • Spectrum Efficiency • Coverage yes/no • Service Probability • Quality • Call Set Up Time • Call Success Rate • Call Completion Rate • Mail Box, Data, Fax, etc. • Customer Care • Faulty H/W or S/W • Mobile Quality • Misuse of Equipment H/W Costs Subscription/Airtime costs Additional Services Costs Network Equipment Costs Maintenance Costs Site Leasing Costs
Tools for Optimisation Cell Planning Tools • Prediction • Simulation
Network Measurement Tools • Propagation • Drive test
Network Management System • Network configuration • BSS parameter data • Network performance
Performance Feedback • Network is under permanent change o ==> detect problems and symptoms early! OM C
It´s far too late when customers complain!
field tests
customer complaint s
Optimize compared to what?
Key Performance Indicators, KPI • KPIs are figures used to evaluate Network performance. o post processing of NMS data or o drive test measurements data • Usually one short term target and one long term target. o check the network evolution and which targets are achieved • KPIs calculated with NMS data o network performance on the operator side. • KPIs from drive test o performance on the subscribers side • Usually turn key projects are evaluated according to some predefined KPIs figures like drop call rate
Network Performance Evaluation with NMS • The most reliable KPIs to evaluate the network performance with NMS are: o SDCCH and TCH congestion o Blocking percentage [%] o Drop call rate [%] o Handover failure and/or success rate o Call setup success rate o Average quality DL and UL • The targets are always defined by the customer but the following figures can be considered as satisfactory results: Item limit Target Lowest acceptable Dropped calls: <2 % 4 % Handover success >98 % 96 % Good Qual samples (0..5) >98 % 95 %
Drive Test Measurements • Evaluate network performance from the subscriber point of view • KPIs information: o DL quality, call success rate, handover success rate, DL signal level o not statistically as reliable as NMS information • Added value of drive test measurement : o find out the geographical position of problems like bad DL quality to look for a possible interference source in the area o compare the performance of different networks o display the signal level on the digital maps to individuate areas with lack of coverage eventually improve the propagation model o verify the neighbour list parameter plan
Optimisation Process
• There are not strict processes for optimization because the activity is driven by the network evolution.
Optimisation Process: Young Network Case • In a young network the primary target is normally the coverage. • In this phase usually there is a massive use of drive test measurement o check the signal and o the performance of the competitors
MMAC
GPS NMS X
•
Optimisation Process: Mature Network Case In a mature network the primary targets are quality indicators
drop call rate, average quality, handover failures. • Important use the information from NMS o a general view of the network performance. • Drive test measurements are still used o but not in a massive way o in areas where new sites are on air o where interference and similar problems are pointed out by NMS data analysis. o
