GSM Cell Planning and Optimization Study Case : Sragen Area: By Sumantri Pramudiyanto (+6281703544310) Jakarta, April 7 th , 2009 GSM Cell Planning and Optimization Study Case : Sragen Area Materi berikut merupakan open content, bersifat free utk didistribusikan Content: Content Cell Planning Process Idle Mode Operation BSS Parameter RF Optimization flow chart Study Cases of RF Optimization Cell Planning Process: Cell Planning Process Cell planning can be described briefly as all the activities involved in determining which sites will be used for the radio equipment, which equipment will be used and how the equipment will be configured Traffic and coverage Analysis: Traffic and coverage Analysis Collecting required data Making discussion with the client to know their demand. Analyzing traffic and coverage Required data : Geography data Demography data Data of around network Available frequency Number of customer which wish to cover Customer demand : GOS 2% 95 % Indoor coverage (C/I > 12 db) > 95% Nominal Cell Plan - Dimensioning (1): Nominal Cell Plan - Dimensioning (1) Geography and Demography data taken from sragen.go.id Table 1 Geographic + Demographic data Table 2 Demographic data per years Wide of area which will be covered Target subscriber After the data available, we need to divided the region into smaller cluster, then classify the subdistrict into cluster depend on traffic, contour area, and etc. From the demographic data (Table 2) , we able to calculate % growth of population. 573333 = 437556 x (1+r) 6 r = 0.046 Assume that, in 2010 the operator want to cover 10 % subscriber in the region, so the number of subscribers to be covered : Pt = 10 % x 57333 x (1+0.046) 3 Pt = 65629 If traffic allocation per subscriber equal to 60mE then total traffic in Sragen area = 65629 x 60 mE = 3937.74 E Nominal Cell Plan - Dimensioning (2): Nominal Cell Plan - Dimensioning (2) From the geographic data we can determine Erlang distribution by density in each cluster. The next step, we can calculate the number of required sites depend on traffic. Using erlang B table we can count number of sites for cluster sragen tengah (GOS 2%, 1585.85 Erlang ) 1586 TS ~ 227 TRX Sragen tengah locate in center of town and has high traffic we use configuration 5/5/5 so the number of sites required in this cluster : 227/15 TRX = 15 Sites With the same way we can calculate number of sites for the others cluster: Sragen Timur = 9 sites ( Config 4/4/4) Sragen Barat = 12 sites ( Config 4/4/4) Sragen Utara = 9 sites ( Config 4/4/4) Nominal Cell Plan – Link Budget: Nominal Cell Plan – Link Budget Link Budget Calculation is required to achieve system balance between uplink and downlink signal. Output from system balance is a cell size in every sites. We can calculate coverage area per sites in suburban and rural cluster by equation L = k x R 2 L site suburban = 1.95 x (1.865) 2 = 6.78 km 2 L site rural = 1.95 x (2.67) 2 = 13.90 km 2 The next step, calculating number of sites related to coverage. Choose the biggest one Nominal Cell Plan - Result: Nominal Cell Plan - Result Sragen Utara Sragen Timur Sragen Tengah Sragen Barat CW Test / Model Tuning: CW Test / Model Tuning In order to find out appropriate propagation model, RF Engineer should perform CW Test. Drivetesting should be performed encircle the route and represent all of azimuth. Input all of drivetesting result to the planning tools then conducting calibration to get appropriate propagation model. Survey: Survey Survey is required to ensure whether the nominal cell position can be realized or not ? In general, The RF Team should give alternative nominal or informed SAR ( ± 300m) to survey team, in case the nominal cannot be realized. The survey team should take the panoramic photo around the nominal (0-360 degree), and report to the RF team if appear obstacle around the nominal. Panoramic photo used to determine coverage target/azimuth Survey team also need to survey : road to nominal, electricity, space for equipment. Detail Planning: Detail Planning All of nominal coordinates must be fixed in detail planning phase. Scope of works in Detail Planning : Frequency Planning Parameter Planning (BSIC, MAList, MAIO, HSN, etc) Adjacency planning The main key in conducting frequency planning is avoiding co-channel and adjacent interference. For TCH Frequency planning, we able to apply SFH or Baseband hopping in order to reduce the interference. Detail Planning – Frequencies Allocation: Detail Planning – Frequencies Allocation SFH Pattern 1x1 will be applied in this implementation Number of Malist frequencies can be calculated by using equation : Maximum configuration for the sites 5/5/5, so that number of required frequencies : N freqs/site = (12 – 3 ) x 2 + 3.2 = 24 Frequencies FL = N TRX # Hoppers X 100% 3 sectors with 24 hopping frequencies : TCH = 3 TRXs FL = (3/24)*100% = 12,5% TCH = 4 TRXs FL = (4/24)*100% = 16,67% Detail Planning – Frequencies Allocation: Detail Planning – Frequencies Allocation MAIO = 0 8 16 MAIO Step = 2 SFH Allocation untill configuration 5/5/5 Detail Planning – HSN Planning: Detail Planning – HSN Planning Drivetest for QOD Program.ppt / 24.04.2008 / VS HSN used to the parameter that differentiates the hopping algorithm between two cells having the same MAList. We can choose best pairs HSN to reduce collision frequencies between server and adjacent. Detail Planning – Coverage Result: Detail Planning – Coverage Result NCC = 4,5 NCC = 2,3 NCC = 0,1 NCC = 5,6 Detail Planning – Interference Prediction: Detail Planning – Interference Prediction Installation and System Tuning: Installation and System Tuning After Installation done, the installation team need to conduct commisioning (VSWR Measurement, check hardware installation ) The Drive test team also need to verify whether the BTS serving target correctly or not Then, Acceptance Test Procedure is conducted to check how well the KPI meets the demand. Pre Launch Optimization performed to achieve the KPI Target for new site or TRX expansion. IDLE MODE OPERATION Normal Cell Selection: IDLE MODE OPERATION Normal Cell Selection Search all the RF channels , take samples during 3-5 s and calculate averages. And put them in ascending order with respect to signal level. Then tune to the strongest RF channel. Search for the frequency correction burst in that carrier in order to varify if it is a BCCH carrier Camp on the cell Try to synchronize to the carrier and read the BCCH data. Is it a BCCH carrier? Is it a correct PLMN ? Is the cell barred? Is C1>0 Tune to the next highest RF channel which is not tried before No No No No Yes Yes Yes Yes C1 = (A - Max(B,0)) A = Received Level Average - p1 B = p2 - Maximum RF Output Power of the Mobile Station p1 = rxLevelAccessMin Min. received level at the MS required for access to the system p2 = msTxPowerMaxCCH Max. Tx power level an MS may use when accessing the system Cell Reselection : Cell Reselection C1 + cellReselectOffset - temporaryOffset*H(penaltyTime-T) T < = penaltyTime C2 = C1 + cellReselectOffset …………………………………………. T > penaltyTime 1 when T < = penaltyTime H(x) = 0 when T > penaltyTime Cell Reselection Histerysis: Cell Reselection Histerysis BSS Parameter: BSS Parameter BTS Parameter RxLevAMI (0-63) Minimum signal strength for access the BTS in idle mode. CRESOFF (Cell Reselection Offset) (0-25) used for C2 Calculation, normally used in dual band network (GSM<>DCS) RACHBT (RACH Busy Threshold) (0-127) defines a threshold for the signal level on the RACH HRACTT1 (0-100) Half Rate Activation Threshold T3212 (0-255) Parameter LUP Periodically MAXRETR (1,2,4,7) Maximum Retransmission on RACH SDCCHCONGTH (0-100) SDCCH Congestion threshold RDLNKTO (0-15) Timer for Radiolink timeout Power Control Parameter LOWTLEVD/U (0-63) the lower threshold of the received signal level on the downlink/uplink for power increase UPTLEVU/D (0-63) defines the upper threshold of the received signal level on the uplink/downlink for power reduction LOWTQUAD/U ( 0-7) the lower threshold of the received signal quality on the downlink for power increase UPTQUAU/D (0-7) defines the upper threshold of the received signal quality on the uplink for power reduction PWRINCSS (DB 2,4,6) defines the step size used when increasing the MS transmit power PWREDSS (DB 2,4) defines the step size used when reducing the MS transmit power BSS Parameter (2) : BSS Parameter (2) Handover Parameter HOLTHLVDL/UL (0-63) defines the receive signal level threshold on the downlink /uplink for inter-cell level handover decision. HOLTHQUDL/UL (0-7) defines the receive signal quality threshold on the downlink/uplink for inter-cell quality handover decision Adjacent Parameter RXLEVMIN the minimum received signal level the adjacent cell must provide to be regarded as a suitable target cell for handover HOM Handover margin for better cell LEVHOM parameter defines the handover margin for handovers due to uplink level or downlink level QUALLEVHOM this parameter defines the handover margin for handovers due to uplink quality or downlink quality RF Optimization Flow Chart: RF Optimization Flow Chart Start Identify KPI Formula Identify the problems SDSR Problems ? Check TRX Quality Check Alarm Check Interference Check all others cause (Radio link Failure, T200, Transcoder) from statistics HOSR Problems ? Check SDCCH Blocking Check TCH Blocking Check Alarm Check co-channel and co-BSIC Check Interference Check Neighbor Relation Check Handover Failure Per Cause Check Handover Parameter DCR Problems Check TRX Quality Check Alarm Check co-channel and co-BSIC Check TA Check Interference problems Check Malist, MAIO and HSN Check measurement from statistics Take Action list Y Y Conclusion : Conclusion Every New Network need good plan for avoid problems that will be arise. Commonly there are six step in conducting planning. The most critical problems in performing cell planning process is interference. the New Sites onair need to be optimized to achieve the KPI Pre Launch Optimization is done for new sites on air or expansion sites.