Penjelasan yang tidak pernah dipublish

Based on the new convention of Cell Congestion KPI Method, a cell could be considered as a congested cell if the cell fulfilled the congestion factor. There are two factors that contribute the congested cell verdict, i.e.:

 

1)      The status of cell congestion based on daily cell utilization

 

Average_Cell_Traffic

Average_Cell_Traffic =

 

Remarks:

-          For Daily Report or daily cell utilization, the granularity period = 24 hours

-          For Weekly Report or weekly cell utilization, the granularity period = 24 * 7 hours

-          For Monthly Report or monthly cell utilization, the granularity period =

24 * 7 * number of days in appropriate month

 

 

Cell Utilization

For Full Rate:

 

Or

 

For Half Rate:

 

 

Or

 

E.g. Table Erlang with GOS 2% of FR TCH

TRX

TCH

Erlang

1

6

2.28

2

12

6.61

3

19

12.33

4

26

18.38

5

33

24.63

6

40

31.00

7

48

38.39

8

55

44.94

9

63

52.48

10

71

60.08

11

79

67.73

12

87

75.42

 

 

Example based on TCH Full Rate:

Cell A has two TRXs and implemented in Full Rate. The total traffic in one day is 11.19 Erl

So the cell utilization for cell A on that day is:

 = 7.054% (Assumption the measurement period = 24 hours)

 

Note:

The granularity period is not always 24 hours, it depends on the measurement period

 

Based on the new convention of Cell Decongestion KPI Method, if the cell utilization in one day is more than 60% (for Java Area) or 70% (for outside Java), it could be considered that the utilization of that cell is Congested Utilization. Due to the calculation, it could be concluded, the utilization of Cell A on that day is Normal Utilization.

 

 

2)      The status of cell congestion based on peak hourly utilization

 

 

Hourly cell utilization could be calculated from the hourly traffic volume.

For example (based on TCH Full Rate):

Traffic Cell A (same cell with the above sample) on 12 PM – 01 AM = 0.82 Erl

So the hourly utilization for cell A on that time is:

 = 12.42%

And so forth for the next every hour (every hour has utilization; Erlang per hour could be seen on Metrica data sheet hourly_traffic)

 

If in one day, there are 2 (or more) peaks hourly utilization (> 100%); it could be categorized as Peak Congestion. But if in one day there is no peak or only one peak hourly utilization (>100%), it could be categorized as Normal Congestion.

 

 

The Congested Cell Final Conclusion:

A cell could be categorized as a CONGESTED CELL, if in two weeks period, there are 6 days (non consecutive) occurred when the cell stated as congested utilization based on daily cell utilization AND stated as Peak Congestion based on hourly utilization.

And for the calculation of congested cells percentage in one region is:

 

 

The calculation for TCH Half Rate is same with the calculation for TCH Full Rate.

 

 

 

 

 

TCH Available

Total number of TCH available will be used in calculation of cell utilization, but first, it converted to Erlang Table GOS 2%

 

The TCH available general formula could be calculated as follows:

1. For TCH Full Rate:

 

2. For TCH Half Rate:

For Half Rate, TCH Available depends on the Vendor Half Rate Specification.

2.1 For Nokia:

For Dual Rate:

The dual rate could be activated per Time Slot (TS), 1 TS dual rate = 2 TCH.

 

Example Nokia Dual Rate, TCH Available Calculation:

Cell B activated as dual rate and has Configuration 2 TRXs. So the TCH Available is:

Full Rate:

TCH Available = (2 * 8 TS) – (1 TS for BCCH + 1 TS for SDCCH + 2 TS for Dedicated PDCH)

                        = 12 TCH

 

Half Rate:

TCH Available = (2 * 8 TS) – (1 TS for BCCH + 1 TS for SDCCH + 4 TS for Dedicated PDCH)

                        = 10 TS Total

= 8 TS FR + 2 TS DR (the time slot sharing could be seen in the table below)

                        = 8 TCH FR + 4 TCH DR

                        = 12 TCH

 

 

Because the logical channel configuration could be different, thus the total number of TCH for Full Rate could be same with the total number of TCH for Half Rate.

 

 

Full Rate Configuration 2 TRX

 

 

 

 

 

 

 

 

TS1

TS2

TS3

TS4

TS5

TS6

TS7

TS8

 

TRX1

BCCH

SDC

PDC dedicated

PDC dedicated

TCH/F

TCH/F

TCH/F

TCH/F

= 4 TCH

TRX2

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

= 8 TCH

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Half Rate Configuration 2 TRX

 

 

 

 

 

 

 

 

TS1

TS2

TS3

TS4

TS5

TS6

TS7

TS8

 

TRX1

BCCH

SDC

PDC dedicated

PDC dedicated

PDC dedicated

PDC dedicated

DR

DR

= 4 TCH

TRX2

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

TCH/F

= 8 TCH

 

 

 

2.2 For Alcatel

 

d. For Siemens

For TCH Available total could be seen in Metrica (Data Sheet Daily Analysis)

TCH Available total = TCH_FR + TCH_HR

 

d. For Ericsson & Huawei

The Data already provided by Vendor

PSR (Paging Success Rate)

Paging digunakan untuk mengetahui keberadaan MS di suatu location area (LA), paging biasanya di-trigger ketika ada panggilan atau sms yang akan masuk ke sisi penerima. Agar network dapat memanggil atau mengirim pesan ke sisi penerima (called party), maka network melakukan paging terlebih dahulu untuk mengetahui keberadaan MS called party.

Saat melakukan paging, BSC (akan tetapi trigger paging dilakukan dari MSC) akan membroadcast B# yang dituju ke seluruh location area dari data terakhir MS B# berada, yang diperoleh dari HLR. Jadi paging dapat diasumsikan seperti Guru memanggil seorang murid bernama Budi di ruang kelas.

Guru = BSC

Kegiatan memanggil = paging

Murid bernama budi = MS B# (called party)

Ruang kelas = location area

Jika pada paging pertama gagal maka akan dilakukan proses paging retry, jumlah banyaknya retry dapat diset. Jika sampai pada last paging retry tidak ada respons (paging response dari MS B#), maka dapat diasumsikan bahwa paging tersebut tidak berhasil. Nilai keberhasilan paging diperoleh dari jumlah paging response. Jadi nilai paging success diperoleh dari jumlah paging response.

Berikut ini beberapa hal yang mempengaruhi nilai keberhasilan paging:

1.  Coverage footprint makin banyak daerah yang tidak tercakup sinyal (blank spot), maka PSR makin buruk. Knapa? Karena MS yang dituju tidak dapat memberikan respons diakibatkan MS tsb tidak menerima paging dari network.

2.     LAI (Location Area Identity)/ LAC (Location Area Code) definition

a. Pendefinisian LAC yang tidak jelas (double LAC atau justru LAC yang terdelete akibat suatu aktivitas) menyebabkan gagal paging.

b. Location area yang terlalu luas (paging tidak efektif krn bila paging tidak sampai, akan terjadi pengulangan yang berarti paging attempt akan makin besar) atau terlalu kecil (akibatnya pelanggan terlalu cepat berpindah LAC) dapat menyebabkan penurunan PSR.

3.    Timer setting for paging timer untuk paging yang terlalu pendek bisa menyebabkan paging gagal, karena sebelum paging respons diterima, paging sudah dianggap timeout. Oleh karena itu perlu dicari nilai optimum untuk paging timer.

4.  Rendahnya nilai LUSR juga dapat mempengaruhi rendahnya nilai PSR. Hal ini disebabkan gagalnya location update menyebabkan tidak ada perubahan database HLR dari location area MS B#, padahal MS B# telah berubah location area, sehingga BSC salah melakukan paging pada location area MS B# yang lama

FAQ:

  1. Apakah high traffic atau occupancy mempengaruhi nilai PSR? Sebenarnya pada saat high traffic/occupancy, MSC CP Load sangat tinggi (overload), begitu pula dengan resource baik signaling maupun trafik semua sudah habis digunakan. Sehingga kemampuan MSC untuk melakukan PSR pun semakin rendah