ilsa2 workshop v1
TRANSCRIPT
ILSA 2 workshop
1. Plan New sites (1800 MHz)( )2. Retune existing sites (900 MHz)
Copyright 2008 AIRCOM International
START Summaryy
Do analysis existing before Start AFP
AFPDefine reasonable size network (filters)Setup cost matrix based on local knowledge of the area or from previous interference problems. (separation list, cost matrix)Create Best server & Interference Array with y
Low resolution to see overview (for finding problem area)Higher resolution with reasonable size of network and problem to do AFPproblem to do AFP
Run – Check – Run … until get acceptable resultsAFP Can take hours to weeksILSA will not stop until Cost = 0 (only obtain for new networkILSA will not stop until Cost = 0 (only obtain for new network with low capacity and wide bandwidth)
Do Post Anal sis after Finish AFP
Copyright 2008 AIRCOM International
Do Post Analysis after Finish AFP
Process ISet Up Frequency BandSetup Prerequisite
Import traffic (if used)Import traffic (if used)Use XML Editor put in column “TRAF”
Import NBR from live or create from NBR WizardC t FiltCreate Filter
PlanRead only
Create Best ServerCreate Interference Table#HO count if used
Check Existing Frequency PlanBest serverWorst connectionWorst connectionWorst interfererTotal interference
Copyright 2008 AIRCOM International
Process II
Run ILSA 211. Initialize2. ILSA Menu
Set Plan ListSet Plan ListSet Cost MatrixSet OptionSet OptionRunStop
Continue
▫ Check Frequency Assignment ▫ Apply to Database▫ Check Interference Array & Statisticy▫ OK?▫ NOK?
END
Copyright 2008 AIRCOM International
Frequency Assignment Strategiesq y g gAvailable Spectrum for Indosat GSM1800
Ch: 549 – 573 (5 MHz)Ch: 549 573 (5 MHz)Mainly used for Microcells and Indoor sites
Ch: 712 – 786 (15 MHz)Mainly used for the Macrocell layerMainly used for the Macrocell layer
dedicated band for BCCH channels is used in most regions
BCCH1800 755 785BCCH1800 :755-785TCH1800 : 712 - 753
5 BCCHMicrocell
1 guard ch between BCCH &TCH
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573
4 BCCHIndoor
7 + 9 TCH Chs:Microcell
42 Chs :Macro Layer
BCCH &TCH712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
Indoor
Copyright 2008 AIRCOM International
1 guard ch with other operator
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786
31 BCCH carriers, Macro Layer
Frequency Band from strategyq y gy
Copyright 2008 AIRCOM International
Frequency Bandq y
Copyright 2008 AIRCOM International
Array Setting for 1800y g
Copyright 2008 AIRCOM International
ILSA WORKSHOP
CHECK EXISTING NETWORKBEFORE AFP NEW SITE 1800
ILSA WORKSHOP
BEFORE AFP NEW SITE 1800
Copyright 2008 AIRCOM International
Create Best Server Microsoft Office Excel Worksheet
Copyright 2008 AIRCOM International
Worst Connection
Copyright 2008 AIRCOM International
Average Connectiong
Copyright 2008 AIRCOM International
Worst Interferer
Copyright 2008 AIRCOM International
Total Interference
Copyright 2008 AIRCOM International
Create NBR for new sites
Copyright 2008 AIRCOM International
Create New Site 2G GSM1800
Copyright 2008 AIRCOM International
Create New sites GSM 1800 –No FreqNo Freq
Copyright 2008 AIRCOM International
Create New sites GSM 1800 –when BCCH Allocatedwhen BCCH Allocated
Copyright 2008 AIRCOM International
Select returnable area (1)( )Use Polygon selection
Copyright 2008 AIRCOM International
Select returnable area (2)( )
Copyright 2008 AIRCOM International
Crate Filter from Selection Expertp
Copyright 2008 AIRCOM International
Filters (new sites + read only)( y)
Copyright 2008 AIRCOM International
Run Prediction for 2 new sites
Copyright 2008 AIRCOM International
Create New Best Server Coverage+ 2 new sites+ 2 new sites
Copyright 2008 AIRCOM International
New Service Area
Copyright 2008 AIRCOM International
Create Interference table (1)( )
Copyright 2008 AIRCOM International
Create Interference table (2) : Select View( )
Copyright 2008 AIRCOM International
Create Interference table (3) : Select Filter( )
Copyright 2008 AIRCOM International
Create Interference table (4)Select parametersSelect parameters
Copyright 2008 AIRCOM International
Create Interference table (5)Select parametersSelect parameters
Copyright 2008 AIRCOM International
Save to File
Copyright 2008 AIRCOM International
Adding 6 frequencies without retuning surrounding sites
1ST AFP TRIALADD 2 SITES NO RETUNE
Adding 6 frequencies without retuning surrounding sites
ADD 2 SITES NO RETUNE
Copyright 2008 AIRCOM International
RUN ILSA2: Initialisation
Database from Filter or XML
One of theseO e o t ese1. Interference Array in Memory2. .ait file (prediction)3. .msmt file (measurement)
#HO occured
Copyright 2008 AIRCOM International
Network: FilterWhen click import filter …
Copyright 2008 AIRCOM International
Load Interference Table
One of theseO e o t ese1. Interference Array in Memory2. .ait file (prediction)3. .msmt file (measurement)
Copyright 2008 AIRCOM International
ILSA MENU: Edit plan listp
Copyright 2008 AIRCOM International
First Try to Allocation only for New Sitey y
Copyright 2008 AIRCOM International
Cost Matrix Setting : Carrier Costsg
If we define frequency group already and allow all frequencies to beand allow all frequencies to be allocated, this should be blank.
Copyright 2008 AIRCOM International
Cost Matrix Setting : Carrier Layer Costsg y
We picked up one layer BCCH only!
Copyright 2008 AIRCOM International
Cost Matrix Setting : Equipment Costsg q p
In case we have hierarchy or micro, macro to define cost of ,allocation
Copyright 2008 AIRCOM International
Cost Matrix Setting : Neighbour Costsg gAvoid Adj-channel in immediate NBRCo is forbidden
Avoid co-channel in 2nd Order NBRAdj is OK
Copyright 2008 AIRCOM International
Cost Matrix Setting : Exception Costsg p
If you don’t have exception in cell layer defined, this will be blanklayer defined, this will be blank
Copyright 2008 AIRCOM International
Cost Matrix Setting : Filter Priorityg y
The priority of the filter expressed as a p y pvalue. Filters with a higher value have precedence over filters with a lower value.
Copyright 2008 AIRCOM International
Cost Matrix Setting : Handover Count & Intermod CostsIntermod Costs
Copyright 2008 AIRCOM International
Cost Matrix Setting : Separation Costsg p
Copyright 2008 AIRCOM International
Optionsp
Copyright 2008 AIRCOM International
Run … Start APF
ILSA will stop only if Cost = 0 after allocated all frequencies required.
Copyright 2008 AIRCOM International
p y q q
Result in : Carrier Assignmentsper Layerper Layer
Copyright 2008 AIRCOM International
Apply to Databasepp y
Copyright 2008 AIRCOM International
See new Frequency from Mapviewq y p
Copyright 2008 AIRCOM International
Visual Check Site 1_S1_
Copyright 2008 AIRCOM International
Visual Check Site 1_S2_
Copyright 2008 AIRCOM International
Visual Check Site 1_S3_
Copyright 2008 AIRCOM International
Visual Check Site 2_S1_
Copyright 2008 AIRCOM International
Visual Check Site 2_S2_
Copyright 2008 AIRCOM International
Visual Check Site 2_S3_
Copyright 2008 AIRCOM International
AFP for 2 new sites but no retune
CHECK AFTER 1ST TRIALINTERFERENCE ARRAY
AFP for 2 new sites but no retune
INTERFERENCE ARRAY CHECK
Copyright 2008 AIRCOM International
Worst Interferer Before
Copyright 2008 AIRCOM International
Worst Interferer After Add 2 sites
Copyright 2008 AIRCOM International
Comparison : Before vs Afterp
added 6 frequencies for 2 sites without retune any surrounding sites
No degradation (interference) • Improve probably because adding
Copyright 2008 AIRCOM International
when add new sitesp p y g
new 2 sites
Statistically comparison in overviewStatistically comparison in overview
BeforeCategory name Category value Covered area(%)
After add 2 sitesCategory name Category value Covered area(%)
very good 12.00 <= x dB 96.78%
good 9.00 <= x < 12.00 dB 98.76%
very good 12.00 <= x dB 96.58%
good 9.00 <= x < 12.00 dB 98.50%
d dmedium 6.00 <= x < 9.00 dB 99.67%
bad 3.00 <= x < 6.00 dB 99.95%
very bad 0.00 <= x < 3.00 dB 100.00%
medium 6.00 <= x < 9.00 dB 99.28%
bad 3.00 <= x < 6.00 dB 99.49%
very bad 0.00 <= x < 3.00 dB 99.54%y y
• Improve probably because adding new 2 sitesImprove probably because adding new 2 sites• add 6 frequencies for 2 sites without retune any surrounding sites
Copyright 2008 AIRCOM International
Adding 6 frequencies and retune surrounding
2ND AFP TRIALADD 2 SITES WITH RETUNE
Adding 6 frequencies and retune surrounding
ADD 2 SITES WITH RETUNE
Copyright 2008 AIRCOM International
Plan ListAllow surrounding to be returnableAllow surrounding to be returnable
Copyright 2008 AIRCOM International
Progress… 1. Rung 1. Run2. Stop3. Check Assignment4. Apply databasepp y5. Create Interference Array6. Visual & Statistical check7. OK? -> Stop8. NOK? -> Continue
Copyright 2008 AIRCOM International
Stop & Checkp
Copyright 2008 AIRCOM International
Time Consuming & Loop Check(Process II)(Process II)
Run ILSA 211. Initialize2. ILSA Menu
Set Plan ListSet Plan ListSet Cost MatrixSet OptionSet OptionRunStop
Continue
▫ Check Frequency Assignment ▫ Apply to Database▫ Check Interference Array & Statisticy▫ OK?▫ NOK?
END
Copyright 2008 AIRCOM International
Retune 900 MHz
3RD AFP TRIALRetune 900 MHz
Copyright 2008 AIRCOM International
Analyze existing plan
BCCH 900 + TCH 900INTERFERENCE ANALYSIS
Analyze existing plan
INTERFERENCE ANALYSIS
Copyright 2008 AIRCOM International
Frequency Assignment Strategiesq y g g
Available Spectrum for Indosat GSM900Ch: 1 – 50 (10 MHz)
Shared between the Macrocell, Microcells and Indoor sitesFor the broadcast channel (BCCH)( )
required cluster size BCCH channel > required cluster size TCH channelsdedicated band for BCCH channels is used in most regionsdedicated band for BCCH channels is used in most regions
Total Indosat GSM 900 bandwidth (10 MHz = 50 carriers)Common band:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
20 BCCH carriers28 TCH carriers 1 guard ch between BCCH &TCH
1 guard ch with other operator
Copyright 2008 AIRCOM International
Define Frequency Bandq y
Copyright 2008 AIRCOM International
Define Frequency Bandq y
Copyright 2008 AIRCOM International
Array Setting 900 (BCCH+TCH)y g ( )
Copyright 2008 AIRCOM International
FH, FD, DTX, ,
Copyright 2008 AIRCOM International
Create Best Server 900
Copyright 2008 AIRCOM International
Create Worst Connection
Copyright 2008 AIRCOM International
Create Average Connectiong
Copyright 2008 AIRCOM International
Create Worst Interferer See differences !!See differences !!
Copyright 2008 AIRCOM International
Total Interferer
Copyright 2008 AIRCOM International
Consider only BCCH900 layer, Find Worst Interferer
BCCH 900 ANALYSISConsider only BCCH900 layer, Find Worst Interferer
Copyright 2008 AIRCOM International
Array Settingy g
Copyright 2008 AIRCOM International
Array Settingy g
Copyright 2008 AIRCOM International
Best Server
Copyright 2008 AIRCOM International
Finding Worst Interfererg
Copyright 2008 AIRCOM International
Finding Worst Interfererg
Pointing to area and see worst interferer• Retune Manually• AFP
Copyright 2008 AIRCOM International
Use AFP ILSA to retune BCCH 900 in area affected by worst interferer
4TH AFP TRIALRETUNE BCCH 900
Use AFP ILSA to retune BCCH 900 in area affected by worst interferer
RETUNE BCCH 900
Copyright 2008 AIRCOM International
Trial Area Red – 1st TargetBlue – 2nd Target gBlack – Read only
Copyright 2008 AIRCOM International
SteppCreate Filter Retune_Inner
1st Target to be retunedC Fil R OCreate Filter Retune_Outter
2nd Tier to be RetunableCreate Filter Read_only
Only to be considered as interferer or victim but will not be retuned.
Create Best ServerCreate Interference TableRun ILSAInitializeInitializePlan ListSet up Cost MatrixRunRun ….
Check assignment & Result Apply to DatabaseCh k i t & R lt
Copyright 2008 AIRCOM International
Check assignment & Result
Filters Red – 1st TargetBlue – 2nd Target gBlack – Read only
Copyright 2008 AIRCOM International
Create Interference Table
Copyright 2008 AIRCOM International
Run ILSA
Copyright 2008 AIRCOM International
Select Filters to load into ILSA
Copyright 2008 AIRCOM International
Select Cells Layersy
Copyright 2008 AIRCOM International
Select Interference table from file
Copyright 2008 AIRCOM International
Plan List
Copyright 2008 AIRCOM International
Filter Priorityy
The priority of the filter expressed as a value. Filters with a higher value have precedence over filters with a lower valueprecedence over filters with a lower value.
Copyright 2008 AIRCOM International
Separation Costp
Copyright 2008 AIRCOM International
Run (1)( )
Copyright 2008 AIRCOM International
Run (2)( )
Copyright 2008 AIRCOM International
Results (Inner)Results (Inner)After 10 min After 20 min
Copyright 2008 AIRCOM International
Results OutterResults OutterAfter 10 min After 20 min
Copyright 2008 AIRCOM International
Continue… tuningg
Copyright 2008 AIRCOM International
30 min after Run ILSA•Stop•Apply to database•Create new worst interferer again from new frequencies•Create new worst interferer again from new frequencies•Visual check
Interference area is decreased
Copyright 2008 AIRCOM International
END Summaryy
Do analysis existing before Start AFP
AFPDefine reasonable size network (filters)Setup cost matrix based on local knowledge of the area or from previous interference problems. (separation list, cost matrix)Create Best server & Interference Array with y
Low resolution to see overview (for finding problem area)Higher resolution with reasonable size of network and problem to do AFPproblem to do AFP
Run – Check – Run … until get acceptable resultsCan take hours to weeksILSA will not stop until Cost = 0 (only obtain for new networkILSA will not stop until Cost = 0 (only obtain for new network with low capacity and wide bandwidth)
Do Post Anal sis after Finish AFP
Copyright 2008 AIRCOM International
Do Post Analysis after Finish AFP
Thank you
Learning by doingg y g
Copyright 2008 AIRCOM International
Problem definition, Arrays
SUPPORT SLIDESProblem definition, Arrays
Copyright 2008 AIRCOM International
PROBLEM DEFINITION It is sensible to consider the size of the problem that we are attempting to solve. Consider a small example
network. Its characteristics are as follows: S = Number of Sites = 10 C = Number of Cells = 30 = 3 cells / site A = Number of Carrier Allocations Required = 60 = 2 carriers / cellF = Number of Available Frequencies = 20
Therefore Clearly checking every possible frequency plan for a reasonable sizeTherefore, Clearly checking every possible frequency plan for a reasonable size network would take a prohibitive period of time, even if the period of time taken to check a plan were only of the order of 1milli-second. This being the case, algorithms need to be developed to focus processing power on high probability
Copyright 2008 AIRCOM International
algorithms need to be developed to focus processing power on high probability solutions. These algorithms are generally known as search algorithms.
Best Server Array y
This array displays the signal strength of the best serving cell at each pixel on the Map View. This decision is based on parameters specified in the Site Database window and in the Array Settings dialog boxin the Array Settings dialog box
Copyright 2008 AIRCOM International
Interference Arrayy
When creating one of the Interference arrays, ASSET requires a Best Server array in memory. If this is not the case, a Best Server array will be automatically created. However if you later create subsequent InterferenceHowever, if you later create subsequent Interference arrays after making changes to the network, ASSET does not automatically create a fresh Best Server array.not automatically create a fresh Best Server array.
Copyright 2008 AIRCOM International
Per Carrier Interference Array y
For all the interference calculations, ASSET generates an i t di t i t l ll d ' i i t f ' Fintermediate internal array called a 'per carrier interference array'. For each pixel in the array, the serving sub-cell is determined, and for each carrier of the serving sub-cell the worst carrier to interference (C/I) (lowest numerical value) and the total C/I is calculated, taking into consideration all co- and adjacent carriers from all interfering sub-cells. The total C/I is determined by summing the interfering signals in watts and then later converting back to dB. The result is an array such that for each pixel, a list is obtained of serving carriers plus thesuch that for each pixel, a list is obtained of serving carriers plus the worst and total C/I for each carrier. You cannot currently visualise this intermediate array, which no longer exists when all the other selected arrays have been createdlonger exists when all the other selected arrays have been created.
Copyright 2008 AIRCOM International
Worst Connection Array y
For each pixel, the serving sub-cell is determined, and for each hopping carrier group the average carrier to interference (C/I) is calculated from the corresponding pixel in the 'per carrier interference array' by convertingpixel in the per carrier interference array , by converting total C/I to BER and calculating the mean. The mean Bit Error Rate is converted back to dB and the hoppingError Rate is converted back to dB and the hopping carrier group with the lowest resultant C/I is presented, that is, it corresponds to the worst of the mean connection C/I values. This interference array type was designed for networks using frequency hopping although it also works for nonusing frequency hopping, although it also works for non-hopping networks. In a non-hopping network, the carrier group can be considered to contain just a single carrier in
Copyright 2008 AIRCOM International
g oup ca be co s de ed to co ta just a s g e ca ethe above description.
Average Connection Array g y
For each pixel, the serving sub-cell is determined, and for each hopping carrier group the average carrier to interference (C/I) is calculated from the corresponding pixel in the 'per carrier interference array' by convertingpixel in the per carrier interference array by converting total C/I to BER and calculating the mean. The mean BER is converted back to dB and the average value for allBER is converted back to dB and the average value for all hopping carrier groups is presented.
This interference array type was designed for networks using frequency hopping, although it also works for non-hopping networks In a non hopping network the carrierhopping networks. In a non-hopping network, the carrier group can be considered to contain just a single carrier in the above description.
Copyright 2008 AIRCOM International
t e abo e desc pt o
Worst Interferer Array y
For each pixel, the carrier with the worst carrier to interference (C/I) is determined from the corresponding total C/I value in the 'per carrier interference array'. The result is the worst C/I and the sub cell from which theresult is the worst C/I and the sub-cell from which the interference originates.
Copyright 2008 AIRCOM International
Total Interference Array y
For each pixel, the total carrier to interference (C/I) is calculated by summing the total C/I per carrier. This array is applicable to both fully-loaded frequency hopping and non hopping networks The calculated C/I is NOT merelynon-hopping networks. The calculated C/I is NOT merely as experienced by any individual subscriber, but rather it represents the total of the interference experienced byrepresents the total of the interference experienced by ALL subscribers at each pixel.
Copyright 2008 AIRCOM International
Check List
FiltersSelection expertStaticDynamic Filters byDynamic Filters by
Area ParameterParameterDevices
ArrayBest server, Interference
NBRCreate NBRCreate NBR
Copyright 2008 AIRCOM International