CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS

COURSE CONTENTS
1: INTODUCTION
1.1 WHAT ARE RADIO NETWORK PERFORMANCE MEASUREMENTS
1.2 WHY MAKE RADIONETWORK PERFORMANCE MEASUREMENTS
1.3 HOW ARE RADIO NETWORKPERFORMANCE MEASUREMENTS MADE
2: KEY PERFORMANCE INDICATORS
2.1 HOW ARE DROPPED CALLS MEASURED?
2.2 HOW IS CALL COMPLETION MEASURED?
2.3 HOW IS TRAFFIC LEVEL MEASURED
2.4 HOW IS CONGESTION MEASURED
2.5 HOW IS HANDOVER PERFORMANCE MEASURED
3: WHERE ARE RADIO PERFORMANCE MEASUREMENTS AVAILABLE
4: COUNTERS AND MEASUREMENTS FOR VARIOUS FUNCTIONS
4.1 COUNTERS AND MEASUREMENTS FOR RANDOM ACCESS CHANNEL
4.2COUNTERS AND MEASUREMENTS FOR SDCCH CHANNEL
4.3 COUNTERS AND MEASUREMENTS FOR TRAFFIC CHANNELS
4.4 COUNTERS AND MEASUREMENTS FOR DROPPED CONNECTION CAUSES
4.5 COUNTERS AND MEASUREMENTS FOR HANDOVERS
4.6 COUNTERS AND MEASUREMENTS FOR LOCATING
4.7 COUNTERS AND MEASUREMENTS FOR ASSIGNMENT TO ANAOTHER CELL
4.8 COUNTERS AND MEASUREMENTS FOR INTRA CELL HANDOVER
4.9 COUNTERS AND MEASUREMENTS FOR IDLE CHANNEL MEASUREMENTS
4.10 COUNTERS AND MEASUREMENTS FOR LOAD SHARING
4.11 COUNTERS AND MEASUREMENST FOR OVERLAY/UNDRLAY SUB CELLS
5 USING RADIO NETWORK PERFORMANCE MEASURMENS
5.1 MANAGEMENT OVERVIEW REPORTS
5.2 NETWORK PLANNING
5.3 NETWORK OPTIMISATION
5.4 NETWORK ALARMS


1.1 WHAT ARE RADIO PERFORMANCE MEASUREMENTS ?
Radio performance measurements are measurements of events of interest
that occur within the GSM Base Station Subsystem of the GSM-XL network
These measurements are derived from counters that are automatically
collected by the Base Station Controllers (BSCs) in the BSS.
There are many events that are counted, that can be used to derive
measurements of such things as traffic level in cells, congestion, hand over
success rate, call completion rate and dropped call rates
This course will describe how these events are counted, then how these
counters are turned into useful information, then what this information
can be used for
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
1.2 WHY MAKE RADIO PERFORMANCE MEASUREMENTS ?
MANAGEMENT OVERVIEW AND REPORTING
NETWORK PLANNING
NETWORK OPTIMISATION
DETECTING FAULTY NETWORK ELEMENTS
Radio performance measurements can be used to produce management
overview reports of overall network performance, to ensure that service
level targets are being met
Traffic level measurements can be used for correct and future
dimensioning of network resources, such as radio transceivers
Radio performance measurements are used in the optimization process, as
a guideline to which cells need attention from the radio optimization group.
Radio performance measurements can be used to detect faults in the
system by checking for sudden changes in some measurements, or when
other measurements exceed a certain threshold.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
1.3 HOW ARE RADIO PERFORMANCE
MEASUREMENTS MADE?
Software processes inside each BSC step counters for predefined events
usually signaling events between the BSC and BTS or MSC, or internal
signal transfers within the BSC

Counters are transferred every 15 minutes to the IOG-11 in the BSC
The counters are formatted into reports by the IOG-11, and
transferred every hour to the Operational Support System (OSS)
The OSS uses information from the IOG-11 reports, and predefined
user formulas to produce useful measurements. Measurements are
archived in the OSS, and some reports post processed and posted on the
NOC intranet web page
The NOC auto reports system retrieves data from the OSS daily, and
post processes data into other reports ready for users to use
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
PERFORMANCE MEASUREMENT COLLECTION PROCESS
BSC
BSC
IOG-11

IOG-11

OSS
BTS
BTS
1. Signaling events in the BSS step various counters in the BSCs
2: Counters are transferred every 15 minutes to the IOG-11
3 Counters are processed as reports in the IOG-11
4 Reports are transferred to the OSS every hour
AUTO-reports
5 Reports are post
processed in the OSS
6: Reports are transferred
daily to NOC auto report
processing system
7: NOC auto report system produces
management and detailed reports for
various users and management
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.0 KEY PERFORMANCE INDICATORS
CALL COMPLETION RATE
When a subscriber enters a number on their MS, and presses the send key,
it is useful to know how often this process is successful. A call is defined as
completed when a connection through to the B party is successful, this means
that ringing tone is received from the other end. (And then maybe the person
who has been dialed answers)


PERCEIVED DROPPED CALL RATE
Mobile calls will drop out unexpectedly on occasions due to insufficient
radio signal strength or poor radio conditions preventing a proper
connection between the MS and BTS. Perceived dropped call rate is the
number of occurrences of this, as function of the total number of calls
established on the network

CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.0 KEY PERFORMANCE INDICATORS
DROPPED CONNECTION RATE
Traffic channels (TCH) or signaling channels (SDCCH) on the air interface can
be seized at call set up or due to incoming hand over. The number of call drops
as a function of the number of successful channel seizures is a useful indicator
of a cells performance.

ERLANG DROP CALL RATE
The number of drops as a function of the total TCH or SDCCH traffic
carried by the cell is another useful indicator of BSS performance.
TRAFFIC LEVEL
The amount of traffic that a cell carries is useful to know for network
dimensioning and forward planning, as well as optimization and fault
detection. Optimization efforts should be directed towards cells with high
traffic load. Cells with sudden changes in traffic level may have
developed a fault.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.0 KEY PERFORMANCE INDICATORS
CONGESTION
As traffic levels increase in the network, attempts to seize a channel will fail
because all resources are already occupied by other users. This is congestion.
Congestion can be expressed as congestion rate or congestion probability.
Congestion rate is the number of failed seizures as a function of total seizure
attempts. Congestion probability, or channel utilization, is calculated by
measuring carried traffic as a function of traffic capacity and grade of service


HANDOVER SUCCESS RATE
For a mobile phone network to be mobile, a method of handing over an
MS from one cell to another as the user moves through the network is
required. There is always a chance that hand over of an MS from the old
cell to the new cell will fail. The number of successful hand overs as
function of total hand over attempts is a useful performance indicator
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.1a HOW ARE DROPPED CALLS MEASURED ?
Under normal circumstances, a call is terminated by the mobile user pressing
the end key, resulting in a clear request message being sent to the MSC.
Alternatively if the called party hangs up, a clear request from the MSC will be
sent to the MS
This results in all transmission paths being cleared, and the release of radio
channels by the BTS and MS, as well as the termination of call charging
In some circumstances, radio communication between the MS and BTS is
interrupted, due to unfavorable radio conditions. If this results in a complete
signaling failure between the MS and BTS, then no Clear Request message can be
sent or received by the MS. A mechanism for the BTS and the MS to
automatically clear the call, release radio resources, and inform the MSC that the
call has been disconnected, exists.
A counter exists in the MS that counts the successful reception of down link
SACCH frames. The counter starts at a value of RLINKT (=20 in GSM-XL), and
is decremented by 1 every time a SACCH frame cannot be decoded, and
incremented by 4 up to a maximum value of RLINKT every time a SACCH
frame is successfully decoded. (A leaky bucket type counter).
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.1b HOW ARE DROPPED CALLS MEASURED ?
Whenever this counter reaches zero in the MS, an abnormal channel release
message is sent on the up link, the channel released, and the MS returns to idle
mode, after sending some form or message such as 3 beeps to the user
A similar leaky bucket counter RLINKUP exists in the BTS, counting the
successful reception of up link SACCH frames being received from the MS. If
this counter reaches zero, an abnormal channel release message is sent to the MS,
and the radio channel released. If the message is not received by the MS, the MS
will stop receiving SACCH frames from the BTS, and its own RLINKT counter
will expire.
Either event results in an abnormal channel release message being sent to the
BSC, informing the BSC that the MS has been lost.
The BSC is informed by the BTS of the timing advance being used by the MS.
If the timing advance reaches a value MAXTA, (=63 for GSM-XL), then the BSC
will clear the call, to prevent the MS interfering with an MS using the next
timeslot.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.1c HOW ARE DROPPED CALLS MEASURED ?
Either radio link failure condition, or excessive timing advance, will result in
the BSC sending a clear request to the MSC. (Remember that the Clear request
normally comes from the MS or the MSC)
This abnormal condition, whereby the BSC sends the clear request, is counted
in the BSC. Counter TNDROP is stepped in the cell where the MS was connected,
and the connection was on a TCH. Counter CNDROP is stepped in the cell where
the MS was connected and the connection was on an SDCCH.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC MSC
RLINKT
RLINKUP
CLEAR REQUEST
MAXTA
DOWNLINK
UPLINK
HOW ARE DROPPED CALLS COUNTED ?
RLINKT SUPERVISES THE SUCCESSFUL DECODING OF DOWNLINK SACCH FRAMES IN THE MS,
RLINKUP SUPERVISES THE SUCCESSFUL DECODING OF UPLINK SACCH FRAMES IN THE BTS
ABNORMAL
CHANNEL
RELEASE
IF RLINKT REACHES ZERO, THE MS RELEASES THE RADIO
CHANNEL AND RETURNS TO IDLE MODE. RLINKUP WILL
THEN DECREMENT TO ZER0, AS NO MORE UPLINK SACCH
FRAMES ARE BEINGRECEIVED AND DECODED AT THE BTS
WHEN RLINKUP REACHES ZERO, BTS SENDS ABNORMAL
CHANNEL RELEASE MESSAGE TO THE MS AND BSC, AND
THEN RELEASES THE RADIO CHANNEL
CALL SUPERVISION FUNCTION IN THE BSC WILL SEND CLEAR
REQUEST TO THE MSC TO CLEAR THE CALL WHEN ABNORMAL
CHANNEL RELEASE RECEIVED OR TIMING ADVANCE > MAXTA.
DROPPED CALL COUNTER FOR THE CELL IS INCREMENTED.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.2a HOW IS CALL COMPLETION MEASURED ?
Call completion means that after pressing the send key, a connection through to
the called party, that results in ring tone being received, is successfully
established.
For this to happen, a move from the Random access channel, to the signaling
channel (normally SDCCH, but may be TCH for immediate assignment on TCH
traffic cases) and then on to a TCH in response to an assignment request from the
MSC is required.
Call completion applies to mobile originated as well as mobile terminated calls,
but failed call completion for MT calls are rarely noticed by the subscriber.
Reasons that calls may fail to complete include (but are not limited to )
Failure of the channel request message to get through on the RACCH. The
MS will retry a maximum of MAXRET (=7 for GSM-XL) times before giving
up and declaring a signaling failure condition, resulting in cell re selection.
Failure to set up on the allocated SDCCH after reception of Immediate
Assignment message from the BSC
Dropped call during signaling transfer phase on the SDCCH
Failure to assign a TCH after reception of assignment request from MSC
Signaling or other failure in the transit or terminating network
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC MSC
CHANNEL REQUEST
CHANNEL REQUIRED
IMMEDIATE ASSIGNMENT
IMMEDIATE ASSIGNMENT
SABM/UA ESTABLISH INDICATION
SCCP REQUEST/CONFIRM
CANNEL REQUEST FROM
UNKNOWN MS RECEIVED
CM + MM SIGNALLING PERFORMED ON THE SDCCH BETWEEN THE MS AND MSC
ASSIGNMENT REQUEST ASSIGNMENT COMMAND
THE BSC ALOCATES AN
SDCCH CHANNEL
SABM/UA
ASSIGNMENT COMPLETE ASSIGNMENT COMPLETE ASSIGNMENT COMPLETE
ASSIGNMENT COMMAND
SPEECH/DATA FRAMES EXCHANGED ON THE TCH BETWEEN THE MS AND THE MSC
THE MSC REQUESTS A TCH
FOR SPEECH/DATA TRANSFER
2.2B SIMPLIFIED CALL SET UP DIAGRAM
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.2C HOW IS CALL COMPLETION MEASURED ?
It is not possible to precisely measure call completion rate, as perceived by
the subscriber, for a number of reasons. Instead SDCCH and TCH
completion ratio are measured separately.
In theory, call completion rate could be calculated from the total number of
failed random accesses + total number of SDCCH completion failures + total
number of SDCCH drops + total number of TCH completion failures as a
function of total calls set up. There are two problems that prevent this.
1: Phase 1 establishment causes are ambiguous, so it is not possible to
determine at the RR signaling level if the channel request and resulting
SDCCH seizure is due to call set up, SMS or location updating.
2: The channel request message is only 8 bits long, so false random access
burst detection, which leads to a failed SDCCH allocation, can be quite
common, resulting in low SDCCH completion rate, even though this effect is
having no impact on subscriber service.
GSM-XL just use the TCH completion ratio as a measure of call
completion ratio, which yields a somewhat optimistic result.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.3A HOW IS TRAFFIC LEVEL MEASURED
The GSM BSS shares limited traffic resources amongst many users, relying
on the fact that all users do not make calls at the same time, and only make
calls of limited duration, typically 1 to 2 minutes. Hence a small number of
traffic channels can be shared amongst many users.
If 1 traffic channel is occupied for 1 hour, this is equal to 1 Erlang of traffic
Hence, the absolute maximum traffic level that a cell can carry in Erlang is
equal to the number of traffic channels
Traffic is measured in CME-20 BSS by scanning every cell, once every 10
seconds, and counting how many TCH (or SDCCH) are occupied. The result
is added to the previous result, to form an accumulation. At the end of the
measurement period, this accumulation is divided by the number of scanning
operations performed, to yield the average traffic level over the measurement
period.
Traffic measurements are forwarded to the IOG-11 every 15 minutes.


CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.3 TRAFFIC LEVEL MEASUREMENTS EXAMPLE
BTS
6 TRAFFIC
CHANNEL CELL
CHANNEL STATUS
= IDLE
= SEIZED
T=0
T+10
T+20
T+30
T+70
T+60
T+50
T+40
T+80
1 2
3 7
4 11
6 21
5 17
7 25
9 30
2 5
TIME NUMBER
OF SCANS
8 27
TRAFFIC LEVEL
ACCUMULATOR
AVERAGE TRAFFIC = 30/9 = 3.3 ERLANG
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.4a HOW IS CONGESTION MEASURED ?
Congestion occurs whenever an attempt to seize a channel fails because all
of the channels in a cell are already occupied. Seizures may be due to call set
up attempts, assignments from other cells or incoming hand over attempts.
Congestion may be measured as congestion ratio, by counting the number
of failed attempts as a function of total attempts, or as congestion probability,
by measuring the actual carried traffic and calculating the congestion
probability as a function of the number of available channels and the
required grade of service.
As the amount of traffic carried by a cell increases, the probability that
some seizures will fail due to congestion also increases. This probability is
usually expressed by the Erlang B formula, which yields the traffic carrying
capacity in Erlang of a group of channels, as a function of grade of service
GOS, which is the percentage of seizures that will fail. See appendix A
For example for a 2% GOS, (2 in 100 attempts will fail), 6 channels has a
capacity of 2.2 Erlang, 14 Channels has a capacity of 8.2 Erlang, 22 Channels
has capacity of 14 Erlang and so on. As the GOS is relaxed, more seizures are
allowed to fail and the traffic capacity increases.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.4b HOW IS CONGESTION MEASURED ?
Congestion probability is calculated by doing the reverse, measuring the
carried traffic , knowing the number of channels available, and working out
the GOS from the Erlang B formula, which will be the congestion probability
Channel utilization is a similar concept, and is the actual traffic divided by
the traffic capacity for the required GOS. For example a 6 channel cell, for
2% GOS that carries 3.3 Erlang has a channel utilization of 3.3/2.2 = 150%
Fixed wire line networks typically operate with a GOS of 1%. The CME-20
BSS network can operate with larger GOS values for a number of reasons.
Call set up failures due to congestion can be saved by function
Assignment to Another Cell for TCH, or Immediate Assignment
on TCH for SDCCH congestion cases.
Incoming hand over attempts that fail due to congestion, will not
normally be noticed by the user. (The BSC does not command a hand
over to a cell which has no channels available). When a hand over
allocation attempt fails due to congestion, timers in the BSC TALLOC (=
1 second for GSM-XL) for normal hand hovers, or timer TURGEN (=1
second for GSM-XL) for urgency condition hand over will start. When
the timer expires, a new allocation attempt will be made and eventually
hand over to the correct cell will occur.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.2c CONGESTION MEASUREMENT EXAMPLE
BTS
FAILED SEIZURES
DUE TO CONGESTION
NUMBER OF SEISURE ATTEMPTS= 1000
NUMBER OF FAILED SEISURE ATTEMPTS= 200
CONGESTION PROBABILITY FROM ERLANG B TABLE
FOR 3.3 ERLANG AND 6 CHANNELS = 0.07 = 7%
NUMBER AVAILABLE CHANNELS= 6
CARRIED TRAFFIC = 3.3 ERLANG
DESIRED GOS = 2%
CONGESTION RATIO = 200/1000 = 20%
CHANNEL UTILISATION = 3.3/2.2 = 150%
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.5a HOW ARE HAND OVERS MEASURED ?
When ever an SDCCH or TCH is allocated to an MS, a locating individual
is created in the BSC regional processor that controls the cell where the
connection is made.
Locating is the software process in the BSC central processor (CP) and
regional processors (RP) that controls the hand over process, based on
measurement reports being received from each MS. When ever a locating
individual decides that a hand over is required for the MS that is being
supervised, an allocation request is sent to the CP. If resources are available
in the target cell, the CP will execute a hand over to the target cell.
When a hand over is required, the BSC will activate a channel on the new
cell, and send a HANDOVER COMMAND to the MS on the old cell,
instructing the MS to change to the channel and timeslot of the new cell. The
MS then moves, and tries to establish a connection on the new cell. If L2
messaging is set up OK, and a SACCH established on the new cell, the MS
sends a HANDOVER COMPLETE message to the BSC,and the BSC will then
release the channel on the old cell.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.5b HOW ARE HAND OVERS MEASURED ?
If the MS cannot establish a connection on the new cell, after a period of
time (about 250mSec for GSM-XL), the MS will return to the channel and
timeslot of the old cell and attempt to re establish a connection there. If this is
successful, a HANDOVER FAILURE message is sent to the BSC, and the call
continues on the old cell. The BSC will then release the channel on the new
cell.
If the MS cannot re establish the connection on the old cell, the call is lost,
and the dropped call counter of the old cell stepped.
The BSC counts the signaling events HANDOVER COMMAND,
HANDOVER COMPLETE and HANDOVER FAILURE, as well as
counting allocation attempts from each RP, even when these fail due to
congestion, as well as the reasons for the allocation, including normal path
loss hand over, normal signal strength hand over and urgency hand overs
due to bad up link or down link quality or excessive timing advance.
Handover counters are made per neighbor relation (n-cell)
There are separate counters for internal n-cells (neighbors between cells on
the same BSC) and external n-cells (neighbors between cells on different
BSCs)
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.5c NORMAL HAND OVER EXAMPLE
CELL-A
CELL-B
BSC
RP RP
CP
2:LOCATING INDIVIDUAL
IN RP PROCESSES REPORTS
AND REQUESTS HANDOVER
FROM CP WHEN REQUIRED
3: CP CHECKS CAPACITY
IN CELL B, ACTIVATES
CHANNEL IN CELL B, AND
HANDOVER REASON
COUNTER IS STEPPED
1: MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
5: MS ESTABLISHES SACCH ON
CELL B, AND SENDS HANDOVER
COMPLETE MESSAGE TO BSC
6: CP DE-ACTIVATES CHANNEL
IN CELL A, AND HANDOVER
SUCCESS COUNTER IS STEPPED
7: NEW LOCATING
INDIVIDUAL IS CREATED
IN CELL Bs RP
4: HANDOVER COMMAND
MESSAGE IS SENT BY CP
TELLING MS TO SWITCH
TO CELL B, HANDOVER
ATTEMPT COUNTER IS
STEPPED
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
2.5d HAND OVER FAILURE EXAMPLE
CELL-A
CELL-B
BSC
RP RP
CP
2:LOCATING INDIVIDUAL
IN RP PROCESSES REPORTS
AND REQUESTS HANDOVER
FROM CP WHEN REQUIRED
3: CP CHECKS CAPACITY
IN CELL B, ACTIVATES
CHANNEL IN CELL B, AND
HANDOVER REASON
COUNTER IS STEPPED
1: MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
5: MS ATTEMPTS TO ESTABLISH
CONNECTION WITH CELL B, BUT FAILS
7: CP DE-ACTIVATES CHANNEL
IN CELL B, AND HANDOVER
REVERSION COUNTER IS
STEPPED
8: LOCATING CONTINUES
ON CELL A RP
4: HANDOVER COMMAND
MESSAGE IS SENT BY CP
6:MS RE-ESTABLISHES CONNECTION
ON CELL A AND SENDS HANDOVER
FAI LURE MESSAGE TO THE BSC
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
3.0 WHERE ARE PERFORMANCE MEASUREMENTS AVAILABLE
Performance measurements are available on the NOC Web page, under
detailed statistics : //http/noc4web
NOC AUTO REPORTS also provides detailed statistics
BSC System Reports: Provides various tables of information grouped
by region, or BSC, as well as network summaries
BSC Extension System reports: Provides all information on a per cell
and per neighbor relation for all cells in the network
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CNROCNT
RAACCFA
RAEMCAL
RACALRE
RAANPAG
RAOSREQ
RAOTHER
'Channel required' received from BTS in response to Channel req on RACCH
Stepped at message 'Channel required' received from BTS cannot be decoded
Successfully decoded Random Access with cause code Emergency call
Successfully decoded Random Access with cause code Call estblishment
Successfully decoded Random Access with cause code Answer to Page
Successfully decoded Random Access for other service request
Successfully decoded Random Access that does not fit above cases
4.1a COUNTERS FOR RANDOM ACCESS MEASUREMENTS
When an MS wants to access the radio network for the first time, for mobile originated or mobile
terminated call set up, SMS transmission or reception, or location updating, a channel request
message is sent over the Random Access Control Channel (RACCH
Channel request messages are only 8 bits long, and contain a 3 bit cause code, which gives some
information as to what reason is for requesting the radio channel
When the BTS successfully decodes a channel request message from an MS, a channel required
message is sent to the BSC. Reception of these channel required messages steps various counters in
the BSC as detailed below
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC MSC
CHANNEL REQUEST
CHANNEL REQUIRED
CHANNEL ACTIVATION
CHANNEL ACT ACK
IMMEDIATE ASSIGNMENT
IMMEDIATE ASSIGNMENT
SABM
UA ESTABLISH INDICATION
SCCP REQUEST
SCCP CONFIRM
CM SERVICE REQUEST
CCALLS
CNROCNT
RAACCFA
RAEMCAL
RACALRE
RAANPAG
RAOSREQ
RAOTHER
CMSESTB
4.1b RANDOM ACCESS COUNTERS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.1c FORMULAS FOR RANDOM ACCESS CHANEL MEASUREMENTS
RANDOM ACCESS
CHANNEL
MEASUREMENTS FORMULA DESCRIPTION
rach_att CNROCNT Number of random accesses
racch_to_sdcch_rati
o (CCALLS / CNROCNT) * 100%
Percentage of random access that lead to an SDCCH
seisure
rach_failed_ratio (RAACCFA/CNROCNT) * 100% Percentage of failed random accesses
rach_emergency_rat
io (RAEMCAL/CNROCNT) * 100% Percentage of random accesses for reason emergency call
rach_callreestablish
_ratio (RACALRE/CNROCNT) * 100%
Percentage of random accesses for reason call
establishment
rach_anspaging_rati
o (RAANPAG/CNROCNT) * 100% Percentage of random accesses for reason answer to page
rach_otherservice_r
atio (RAOSREQ/CNROCNT) * 100%
Percentage of random accesses for reason other service
request
rach_othercases_rat
io (RAOTHER/CNROCNT) * 100% Percentage of random accesses for reason other cases
THESE MEASUREMENTS ARE MADE PER CELL
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.2 a COUNTERS FOR SDCCH CHANNELS IN EACH CELL - OBJECT TYPE CELLSSDCCH
Counter CCALLS is stepped every time an Access Grant message is sent to
allocate an SDCCH
Counter CMSESTAB is stepped every time an SDCCH seizure leads to the
successful set up of a signaling connection to the MSC
Counter CCONGS counts failed allocations due to congestion, CTCONGS
is stepped every second that all SDCCH are seized in the cell
Counter CNDROP counts dropped SDCCH connections
Counters CTRALACC and CNSCAN count SDCCH traffic level
Counters CAVAACC and CAVASCAN count SDCCH availability
CCALLS Stepped when ever a Channel request message is received from BTS and can be processed
CMSESTAB Stepped at message SCCP CONFIRM received from the MSC
CCONGS Stepped every time an attempt to seize an SDCCH fails due to no SDCCH channels available
CTCONGS Stepped once a second when all SDCCH are seized and or blocked
CNDROP The number of dropped SDCCH connections
CTRALACC Every 10 seconds, the no. of SDCCH seized in the cell is added to the previous result
CNSCAN Number of times CTRALACC is stepped in the measurement period
CNUCHCNT The number of SDCCH channels defined in the cell
CAVAACC Every 10 seconds, the no. of SDCCH available (not blocked) is added to the previous result
CAVASCAN Number of times CAVACC is stepped in the measurement period
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC MSC
CHANNEL REQUEST
CHANNEL REQUIRED
CHANNEL ACTIVATION
CHANNEL ACT ACK
IMMEDIATE ASSIGNMENT
IMMEDIATE ASSIGNMENT
SABM/UA ESTABLISH INDICATION
SCCP REQUEST
SCCP CONFIRM
CCALLS
CMSESTAB
4.2b SDCCH COUNTERS
CLEAR REQUEST (BY BSC)
CNDROP
CTRALACC, CNSCAN
COUNTS TRAFFIC
CAVACC, CAVASCAN
COUNTS AVAILABILITY
CCONGS COUNTS
CONGESTION EVENTS
CTCONGS COUNTS
CONGESTION TIME
8 X SDCCH CHANS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
SDCCH CHANNEL
MEASUREMENTS FORMULA DESCRIPTION
sdcch_att CCALLS Number of SDCCH seisures
sdcch_suc CMSESTB Number of successful channel establishment on SDCCH
sdcch_suc_ratio (CMSESTB/CCALLS) * 100% Ratio of SDCCH seisures to successful SDCCH channel establishment
sdcch_drop CNDROP Number of dropped SDCCH connections
sdcch_drop_ratio (CNDROP/CMSESTB) * 100% Dropped SDCCH in realtion to SDCCH channels established
sdcch_erldrop_ratio
(CNDROP * CNSCAN * 60) / (CTRALACC *
PLENGTH) Dropped SDCCH in relation to traffic carried
sdcch_traffic CTRALACC/CNSCAN Carried SDCCH traffic in Erlang
sdcch_avail (CAVACC/(CAVASCAN * CNUCHNT)) * 100% SDCCH availability
sdcch_cong_ratio (CCONGS/CCALLS) * 100% SDCCH congestion rate
sdcch_congtime CTCONGS Total time in seconds that all SDCCH were seised
sdcch_congtim_ratio (CTCONGS) / (PLENGTH * 60) *100% % of measurement interval that all SDCCH were seised
sdcch_meanholdtim (CTRALACC/CNSCAN)*60*60/CMESTB Mean holding time of SDCCH
4.2c SDCCH MEASUREMENT FORMULAS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.3a COUNTERS FOR TRAFFIC CHANNEL MEASUREMENTS
Counter TCALLS is stepped by internal software signals, every time a TCH is required, including
attempts that fail due to congestion.
TMESTB is stepped every time a TCH is successfully seized, by counting when a signaling channel
is successfully established on the SACCH
TASSALL counts assignment requests (of TCH) received from the MSC
TCASSAL counts assignment complete messages at assignment to same cell, or hand over complete,
cause = directed retry, at assignment to another cell
CELLTCH TCH counters
TCALLS Stepped every time an attemp is made to seize a TCH, including at congestion
TMSESTB Stepped when a successf ul RR session established on SACCH, ie TCH seized successf ully
TASSALL Stepped at message Assignment Request received from the MSC
TCASSALL Stepped at message Assignment Complete sent to MSC, or Handover Complete f or ASWCL
TNDROP Stepped when Clear request message sent to MSC by the BSC
TCONGS Stepped every time a TCH seisure f ails due to all TCH already being seized or blocked
TTONGS Steps every 1 second that all TCH in cell are seized
TTRALACC Every 10 seconds, the number of seized TCH is added to the previous result
TNSCAN The number of times that TTRALACC is stepped
TNUCHCNT No of TCH defined in cell (may be incorrect if SY hopping used)
TAVAACC Every 10 seconds, the number of available (not blocked) TCH is added to the previous result
TAVASCAN The number of times TAVASCAN is stepped
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC MSC
ASSIGNMENT COMMAND
SABM/UA
ASSIGNMENT COMPLETE ASSIGNMENT COMPLETE
ASSIGNMENT REQUEST
TASSAL
TCASSAL
4.3b TCH COUNTERS FOR NORMAL CALL SET UP
CLEAR REQUEST (BY BSC)
TNDROP
TTRALACC, TNSCAN
COUNTS TRAFFIC
TAVACC, TAVASCAN
COUNTS AVAILABILITY
TCONGS COUNTS
CONGESTION EVENTS
TTONGS COUNTS
CONGESTION TIME
N x TCH CHANS
ASSIGNMENT COMMAND
ASSIGNMENT COMPLETE
MS MOVES FROM SDCCH TO
TCH AND ESTABLISHES RR
SIGNALING ON THE SACCH
TMSESTB
SIGNALLING CONNECTION ON SDCCH ALREADY ESTABLISHED BETWEEN MS AND MSC
TCALLS IS STEPED
EVERY TIME A
TCH IS WANTED
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.3c TCH CHANNEL MEASUREMENT FORMULAS
CELLTCH FORMULA DESCRIPTION
tch_att TCALLS The total number of TCH seizure attempts
tch_suc TMSESTB The total number of times a TCH is successfully seized
asg_att TASSAL The total number of "Assignment Command" messages from MSC
asg_suc TCASSAL
The total number of "Assignment Complete" messages for normal call
set up or "handover Complete= directed retry" for ASWCL
tch_suc_ratio (TMSESTB/TCALLS) * 100% Ratio of successful TCH seisures to all TCH seisure attempts
asg_suc_ratio (TCASSAL/TASSAL) *100% Ratio of successful TCH assignments to total assignment attempts
tch_traffic TTRALACC/TNSCAN The traffic carried in the cell in Erlang
tch_drop TNDROP The number of dropped calls
tch_drop_ratio (TNDROP/TMSESTB) * 100% Dropped TCH as a percentage of successful TCH seisures
tch_erldrop_ratio
(TNDROP * TNSCAN * 60) / (TRRALACC *
PLENGTH) Dropped TCH as a percentage of total traffic
tch_drop_asg_ratio (TNDROP/TCASSAL) * 100% Dropped TCH as a percentage of call set up (assignments)
tch_avail (TAVACC/(TAVASCAN * TNUCHNT)) * 100% The availability of TCH during the measurement period
tch_meanholdtime (TTRALACC/TNSCAN)*60*60/TMSESTB The mean holding time of TCH in seconds
tch_cong_prob Erlang fuction (tch_traffic+tch_avail)
Congestion calculated as a function of offered traffic vs available
channels
tch_cong_ratio (TCONGS/TCALLS) * 100%
The number of failed seisures due to congestion as a percentage of
seisure attempts
tch_congtime TTONGS The time in seconds when all TCH are siezed
tch_congtime_ratio (TCONGS) / (PLENGTH * 60) *100% Percentage of time that all TCH are seized
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.4a COUNTERS FOR DROPPED CONNECTION REASONS
When ever a connection is established on an SDCCH or TCH, measurement reports are sent
on the SACCH to the BSC and a locating individual is started, so hand over can be invoked
when required.
If a connection is dropped, the locating individual can report on the measured radio
environment at the time, so some reason behind the dropped connection can be deduced.
If the filtered down link signal strength of the MS was less than MSRXMIN when the call
dropped, or the estimated up link signal strength was less than BSRXMIN when the call
dropped, then low signal strength reason is declared
If the timing advance reached the value of MAXTA, excessive timing advance reason is
declared
If the filtered up link quality exceeded QLIMUL or the filtered down link quality exceeded
QLIMDL at the time of the dropped connection, then bad quality reason is declared.
Dropped calls that meet none of the above criteria are bundled in to other reasons
There are separate counters for dropped SDCCH connections and dropped TCH connections
The locating process knows the power capability of the MS, so there are separate dropped
connection due to low signal strength counters for MS power classes 1, 2, 3, 4 and 5
All MS in GSM-XL are class 4
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELLTCHDR Dropped TCH connection reasons
TDISQA Stepped at dropped connection on TCH when Bad Quality urgency condition exists
TDISSS1 Stepped when a class 1 MS drops on TCH and meaured signal strength < MSRXMIN
TDISSS2 Stepped when a class 2 MS drops on TCH and meaured signal strength < MSRXMIN
TDISSS3 Stepped when a class 3 MS drops on TCH and meaured signal strength < MSRXMIN
TDISSS4 Stepped when a class 4 MS drops on TCH and meaured signal strength < MSRXMIN
TDISSS5 Stepped when a class 5 MS drops on TCH and meaured signal strength < MSRXMIN
TDISTA Stepped at dropped connection on TCH when excessive TA urgency condition exists
CELLCCHDR Dropped SDCCH connection reasons
CDISQA Stepped at dropped connection on SDCCH when Bad Quality urgency condition exists
CDISSS1 Stepped when a class 1 MS drops on SDCCH and meaured signal strength < MSRXMIN
CDISSS2 Stepped when a class 2 MS drops on SDCCH and meaured signal strength < MSRXMIN
CDISSS3 Stepped when a class 3 MS drops on SDCCH and meaured signal strength < MSRXMIN
CDISSS4 Stepped when a class 4 MS drops on SDCCH and meaured signal strength < MSRXMIN
CDISSS5 Stepped when a class 5 MS drops on SDCCH and meaured signal strength < MSRXMIN
CDISTA Stepped at dropped connection on SDCCH when excessive TA urgency condition exists
4.4b COUNTERS FOR DROPPED CONNECTION REASONS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
MSC
CLEAR REQUEST
1: WHEN AN SDCCH OR TCH CONNECTION IS ESTABLISHED, A SACCH IS ESTABLISHED, AND A
LOCATING INDIVIDUAL ON THE CELLS RP BEGINS PROCESSING MEASUREMENT REPORTS FROM MS
ABNORMAL
CHANNEL
RELEASE
BSC
RP
2: IF THE RADIO CONNECTION IS DROPPED, THE BTS AND BSC INITIATE
ABNORMAL CALL DISCONECTION PROCEDURE
3: LOCATING IS STOPPED AT ABNORMAL CHANNEL RELEASE, AND
DROPPED CALL REASONS ARE REPORTED IF APPROPRIATE
CNDROP, TNDROP
CDI SSS,TDI SSS
CDI SQA, TDSI SQA
CDI STA, TDI STA
4.4c COUNTERS FOR DROPPED CONNECTION REASONS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.4d MEASUREMENT FORMULAS FOR DROPPED CONNECTION REASONS
sdcch_drop_ss_ratio
((CDISSS1+CDISS2+CDISS3+CDISS4+CDISS5)/CN
DROP)) * 100% % of SDCCH drops at bad signal strength
sdcch_drop_qa_ratio (CDISQA/CNDROP) * 100% % of SDCCH drops at bad quality urgency
sdcch_drop_ta_ratio (CDISTA/CNDROP) * 100% % of SDCCH drops at excessive timing advance
sdcch_drop_ot_ratio
((CNDROP-CDISSS1-CDISSS2-CDISSS3-CDISSS4-
CDISSS5-CDISQA-CDISTA)/CNDROP) * 100% % of SDCCH drops due to unknown reasons
tch_drop_ss_ratio
((TDISSS1+TDISS2+TDISS3+TDISS4+TDISS5)/TN
DROP)) * 100%
Number of drops at bad signal strength (TDISSS) as a percentage of
number of drops
tch_drop_qa_ratio (TDISQA/TNDROP) * 100%
Number of drops at bad quality (TDISQA) as a percentage of number of
drops
tch_drop_ta_ratio (TDISTA/TNDROP) * 100%
The number of drops at excessive timing advance as a percentage of
number of drops
tch_drop_ot_ratio
((TNDROP-TDISSS1-TDISSS2-TDISSS3-TDISSS4-
TDISSS5-TDISQA-TDISTA)/TNDROP) * 100%
The difference between drops due to (signal strength + quality + timing
advance) and total drops, as a percentage of total drops
Dropped connection reasons are measured separately for SDCCH
and TCH connections
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.5a COUNTERS FOR HANDOVER EVENTS
Handover events are counted for each neighbor relationship defined in the BSS
There are separate object types, using the same counter names for internal and external
neighbor relationships. Internal N-cells are n-cells between cells on the same BSC, external n-
cells are n-cells between cells on different BSCs
NCELLREL Intra BSC handover counters, between cells on the same BSC (INT)
HOVERCNT Stepped at message Handover Command sent to MS on old cell
HOVERSUC Stepped at message Handover Complete received from MS on new cell
HORTTOCH Stepped at message Handover Failure received from MS on old cell
NECELLREL Inter BSC handover counters between cells on different BSCs (EXT)
HOVERCNT Stepped at message Handover Command sent to MS on old cell
HOVERSUC Stepped at message Handover Complete received from MS on new cell
HORTTOCH Stepped at message Handover Failure received from MS on old cell
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELL-A
CELL-B
BSC
RP RP
CP
MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
MS ESTABLISHES SACCH ON
CELL B
HANDOVER COMMAND
4.5b COUNTERS FOR HANDOVER EVENTS, NORMAL HANDOVER
HANDOVER COMPLETE
HOVERCNT
HOVERSUC
LOCATING INDIVIDUAL
IN RP PROCESSES REPORTS
AND REQUESTS HANDOVER
FROM CP WHEN REQUIRED,
CP SENDS HANDOVER
COMMAND TO THE MS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELL-A
CELL-B
BSC
RP RP
CP
MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
MS ATTEMPTS TO ESATLISH
CONNECTION ON CELL B BUT
FAILS, AND THEN RETURNS
TO THE OLD CHANNEL
HANDOVER COMMAND
4.5c COUNTERS FOR HANDOVER EVENTS, HANDOVER FAILURE
HOVERCNT
HORTOCH
LOCATING INDIVIDUAL
IN RP PROCESSES REPORTS
AND REQUESTS HANDOVER
FROM CP WHEN REQUIRED,
CP SENDS HANDOVER
COMMAND TO THE MS
HANDOVER FAI LURE
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.5d MEASUREMENT FORMULAS FOR HANDOVER EVENTS
PER N-CELL MEASUREMENTS
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER N-CELL)
ttype Ext or Int External for inter BSC and int for intra BSC handovers
HOVERCNT HOVERCNT
The number of handover attempts per n-cell (Handover command
messages sent)
PER_SUCC (HOVERSUC / HOVERCNT) *100% Percentage of successful handovers in relation to handover attempts
PER_REV (HORTOCH / HOVERCNT) * 100%
Percentage of reversions to original channels in relation to handover
attempts
PER_LOST
(HOVERCNT - HORTOCH - HOVERSUC) *100
(HOVERCNT)
Difference between successful handovers and handover reversions in
relation to handover attempts
The number of hand over attempts is counted per n-cell
The percentage of successful hand over attempts, ie hand over commands that lead to hand
over completions is measured
The percentage of hand over reversions is measured ie hand over commands that fail and the
MS reverts back to the old cell
The percentage of hand overs lost is measured, by measuring the difference between hand
over commands and hand over completions + hand over reversions
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.5e MEASUREMENT FORMULAS FOR HANDOVER EVENTS
PER CELL MEASUREMENTS
It is useful to sum the hand over performance of all neighbor to a cell together, so hand over
performance on a per cell basis can be seen.
Total incoming and outgoing hand overs are measured, as well as total incoming hand over
reversions and lost, per cell
SCELL is serving (old) cell in the neighbor relation, EXTCELL or INTCELL is the target
(new) cell in the neighbor relation for external and internal neighbors
Cells with high percentage of incoming hand over reversions need attention !
SUMINCHOATT
SUM HOVERCNT For all INTCELL and EXTCELL =
CELL
The sum of all incoming handover attempts from neighbour internal and
external cells (Handover command messages)
PERINC_SUC
SUM PER_SUCC For all INTCELL and EXTCELL =
CELL
The sum of all successful incoming handovers from neighbouring
internal and external cells.(Handover Complete) Ie SUM of HOSSUC for
INTCELL and EXTCELL = this cell as a percentage of Sum incoming
PERINC_REV
SUM PER_REV For all INTCELL and EXTCELL =
CELL
The sum of all handover reversions from neighbouring internal and
external cells (Handover failure message)
PERINC_LOST
SUM PER_LOST For all INTCELL and EXTCELL =
CELL
(Sum incoming handover attempt)-(Sum successful incoming handover) -
(Sum incoming handover failure) as a percentage of Sum incoming
handover attempt.
SUMOUTHOATT SUM HOVERCNT For all SCELL = CELL
Sum of all outgoing handover attempts to other neighbouring internal
and external cells. Ie sum of HOATT for this cell = SCELL for both
internal and external n-cells
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.6a COUNTERS FOR LOCATING FUNCTION - HANDOVER REASONS
Whenever a locating individual sends a hand over request to the CP for execution, the reason
for the hand over request is counted
These counters are stepped even if the request does not lead to a hand over, due to congestion
in the target cell or other reason
There are separate counter object types for internal and external neighbors
Counter HODUPFT counts ping pong events. When a normal hand over occurs, if another
normal hand over back to the old cell occurs within 10 seconds, HODPUFT is stepped. This
should not be confused with a hand over failure (reversion), where the MS could not establish a
connection on the new cell and reverts to the old cell
NCELLRXT Handover causes for intra BSC handovers, cells on same BSC (INT)
HOTOLCL Stepped when handover to better cell, L-cell criterion met
HOTOKCL Stepped when handover to better cell, K-cell criterion met
HOUPLQA Stepped when handover to due to bad uplink quality. BTS(RxQual) > QLIMUL
HODWNQA Stepped when handover due to bad downlink quality. MS(RXQual) > QLIMDL
HOEXCTA Stepped when handover due to excessive timing advance. TA > TALIM
HODUPFT Stepped when a normal handover back to old cell from new cell occurs within 10 sec
NECELLRXT Handover causes for inter BSC handovers, cells on different BSC (EXT)
HOTOLCL Stepped when handover to better cell, L-cell criterion met
HOTOKCL Stepped when handover to better cell, K-cell criterion met
HOUPLQA Stepped when handover to due to bad uplink quality. BTS(RxQual) > QLIMUL
HODWNQA Stepped when handover due to bad downlink quality. MS(RXQual) > QLIMDL
HOEXCTA Stepped when handover due to excessive timing advance. TA > TALIM
HODUPFT Stepped when a normal handover back to old cell from new cell occurs within 10 sec
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELL-A
BSC
RP
CP
MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
HANDOVER COMMAND
4.6b COUNTERS FOR LOCATING - HAND OVER REASONS
HOVERCNT
LOCATING INDIVIDUAL
IN RP PROCESSES
REPORTS
AND REQUESTS
HANDOVER
FROM CP WHEN
REQUIRED,
HOTOLCL
HOTOKCL
HOUPLQA
HODWNQA
HOEXTA
RP SENDS HANDOVER
REQUEST TO CP, AND
HAND OVER REASON
COUNTERS ARE STEPPED
TCALLS
TCALLS IS ALSO STEPPED
EVEN IF NO HANDOVER
COMMAND IS SENT BY CP
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELL-A
CELL-B
BSC
RP RP
CP
1: A SUCCESSFUL HANDOVERE IS EXECTUTED FROM CELL A TO CELL B
HANDOVER COMMAND
4.6c COUNTERS FOR LOCATING - HAND OVER PING-PONG
HANDOVER COMPLETE
HODUPFT
2: A TIMER IS STARTED IN THE CP. IF A HAND OVER FROM CELL B TO
CELL A OCCURS WITHIN 10 SECONDS, PING-PONG COUNTER IS STEPPED
HANDOVER COMMAND
HANDOVER COMPLETE
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.6d MEASUREMENT FORMULAS FOR LOCATING : N-CELL MEASUREMENTS
The measurements are made per n-cell
PER_DUPFT is percentage of hand overs that result in a hand back to the old cell within 10
seconds (ping-pong effect_
PER_DECISION is the total number of hand over requests sent to the CP as a function of
actual hand over commands sent. Should be ~100% until cell experiences congestion
Other measurements are hand over reasons. LCL is path loss, KCL is normal signal strength,
UPLQA is up link quality, DWNQA is down link quality, EXCTA is excessive timing advance
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER N-CELL)
PER_DUPFT (HODUPFT) / (HOVERSUC) * 100%
Ratio of handback to normal cell within 10 seconds to succesful
handovers
PER_DECISION
HOVERCNT * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA) Ratio of actual handover attempts to total of all handover decisions
PER_LCL
HOTOLCL * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA)
Ratio of handover decisions due to L cell path loss criterion to sum of all
handover decisions
PER_KCL
HOTOKCL * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA)
Ratio of handover decisions due to K cell signal strength criterion to sum
of all handover decisions
PER_UPLQA
HOUPLQA * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA)
Ratio of handover decisions due to uplink bad quality urgency criterion
to sum of all handover decisions
PER_DWNQA
HODWNQA * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA)
Ratio of handover decisions due to downlink bad quality urgency
criterion to sum of all handover decisions
PER_EXCTA
HOEXCTA * 100%
(HOTOLCL+HOTOKCL+HOUPLQA+HODWNQA+HOEXCTA)
Ratio of handover decisions due to excessive timing advance urgency
criterion to sum of all handover decisions
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.6e MEASUREMENT FORMULAS FOR LOCATING : CELL MEASUREMENTS
As with hand over events, it is useful to sum hand over reasons together on a per cell basis
Outgoing hand over reasons on a per cell basis can be useful indicators of a cells performance
Cells with large percentage of outgoing urgency hand overs need attention!
PEROUT_DUPFT SUM PER_DUPFT For all SCELL = CELL
Sum of all outgoing successful handovers that handback to the serving
cell within 10 seconds as a function of all successful outgoing handovers
PEROUT_DECISION SUM PER_DECISION For all SCELL = CELL
Ration of the sum of all outgoing handover attempts to sum of all
outgoing handover decisions
PEROUT_KCL SUM PER_kcl For all SCELL = CELL
Sum of all outgoing handover attempts due to signal strength K cell
criterion as a percentage of Sum outgoing handover.
PEROUT_LCL SUM PER_LCL For all SCELL = CELL
Sum of all outgoing handover attempts due to path loss L cell criterion
as a percentage of Sum outgoing handover
PEROUT_DWNQA SUM PER_DWNQA For all SCELL = CELL
Sum of all outgoing handover attempts due to Down link Quality urgency
condition, as a percentage of Sum outgoing
PEROUT_UPLQA SUM PER_UPLQA For all SCELL = CELL
Sum of all outgoing handover attempts due to Up link Quality urgency
condition, as a percentage of Sum outgoing
PEROUT_EXCTA SUM PER_EXCTA For all SCELL = CELL
Sum of all outgoing handover attempts due to excessive Timing
Advance urgency condition, as a percentage of Sum outgoing
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER CELL)
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.7a COUNTERS FOR ASSIGNMENT TO ANOTHER CELL
Assignment to another cell occurs when a call is set up on the SDCCH of one cell, and
allocated a TCH on a different cell. The TCH allocation is executed with a HAND OVER
process, as opposed to the ASSIGNMENT process, when the TCH is on the same cell as the
SDCCH.
Assignment to another cell can be to a better cell, when locating finds a better cell to connect
to during the signaling phase of the call, when the connection is on the SDCCH.
Assignment to another cell can be to a worse cell, when there is TCH congestion on the serving
cell, which would otherwise result in call set up failure.
Assignment to other cells are collected per neighbor relation, and different object types exist
for internal and external neighbors, as with hand over counters
HOASBCL Stepped when assignment to a better cell during call setup is attempted
HOASWCL Stepped when assignment to a worse cell during call setup is attempted
HOSUCBCL Stepped when assignment to a better cell during call setup is successful
HOSUCWCL Stepped when assignment to a better cell during call setup is successful
HOASBCL Stepped when assignment to a better cell during call setup is attempted
HOASWCL Stepped when assignment to a worse cell during call setup is attempted
HOSUCBCL Stepped when assignment to a better cell during call setup is successful
HOSUCWCL Stepped when assignment to a better cell during call setup is successful
NECELLRXT Handover causes for inter BSC handovers, cells on different BSC (EXT)
NCELLRXT Handover causes for intra BSC handovers, cells on same BSC (INT)
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BSC MSC
HANDOVER COMMAND
SABM/UA
HANDOVER COMPLETE
HANDOVER COMPLETE
(CAUSE= DIRECTED RETRY)
ASSIGNMENT REQUEST
TASSAL
TCASSAL
4.7b COUNTERS FOR ASSIGNMENT TO ANOTHER CELL
HANDOVER COMMAND
HANDOVER COMPLETE
MS MOVES FROM SDCCH ON
CELL A TO THE TCH ON
CELL B
SIGNALLING CONNECTION ON SDCCH ALREADY ESTABLISHED BETWEEN MS AND MSC ON CELL A
CELL-B
CELL-A
HOASBCL,HOASWCL
HOSUCBCL,HOSUCWCL
BSC LOCATING ASSIGNS
CONNECTION TO CELL B
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.7c MEASUREMENT FORMULAS ASSIGNMENT TO ANOTHER CELL
Assignment attempts to better and worse cell are counted
The success rate of assignment to better or worse cell is measured
Measurements are made per n-cell
The per n-cell measurements are summed for outgoing events for each cell as well
HOASBCL HOASBCL Number of assignment handovers to better cell attempts
PER_SUCBCL (HOSUCBCL / HOASBCL) *100% Success rate of assignment handovers to better cell
HOASWCL HOASWCL Number of assignment handovers to worse cell attempts
PER_SUCWCL (HOSUCWCL / HOASWCL ) * 100% Success rate of assignment handovers to worse cell
SUMOUT_ASBCL
SUM HOASBCL for all INTCELL & EXTCELL =
CELL Number of assignment attempts to a better cell during call set up
PEROUT_SUCBCL
SUM PER_SUCBCL for all INTCELL and EXTCELL
= CELL
Number of successful assignment attempts to a better cell as a
percentage of number of assignments to better cell
SUMOUT_ASWCL
SUM HOASWCL for all INTCELL & EXTCELL =
CELL Number of assignment attempts to a worse cell during call set up
PEROUT_SUCWCL
SUM PER_SUCWCL for all INTCELL and EXTCELL
= CELL
Number of successful assignment attempts to a worse cell as a
percentage of number of assignments to worse cell.
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER CELL)
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER N-CELL)
ttype Ext or Int External for inter BSC and int for intra BSC handovers
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.8a COUNTERS FOR INTRA CELL HAND OVER
Intra cell hand over is an auxiliary radio network function, that moves a connection from one
timeslot to a different timeslot within a cell, when conditions of high signal strength and bad
quality exists.
An intra cell hand over may be triggered by bad quality on the up link, down link or both
An intra cell hand over is executed by locating sending a request to the CP, and the CP sending
an ASSIGNMENT COMMAND to the MS, telling the MS to move to a different timeslot on the
same cell
If the assignment is unsuccessful, the MS will revert to the old timeslot and send
ASSIGNMENT FAILURE back to the BSC
HOINUQA Stepped when an intra cell handover is attempted due to bad upink quality
HOINDQA Stepped when an intra cell handover is attempted due to bad downlink quality
HOINBQA Stepped when intra cell handover due to both uplink and downlink quality occurs
HOINSUC Stepped when intra cell handover is successful
HOINBOCH Stepped when intracell handover fails and MS returns to old channel
CELLEVENT Subcell change and intra cell handover counters
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
CELL-A
BSC
RP
CP
MS SENDS MEASUREMENT
REPORTS TO THE BSC ON
THE SACCH
ASSI GNMENT COMMAND
4.8b COUNTERS FOR INTRA CELL HANDOVER
LOCATING PROCESS
DETECTS BAD QUALITY
AND HIGH SIGNAL
STRENGTH, INTRA CELL
HAND OVER CRITERION
IS MET.
HOI NUQA
HOI NDQA
HOI NBQA
RP SENDS INTRA CELL
HANDOVER REQUEST TO
CP, IHO COUNTERS STEP
CP SENDS ASSIGNMENT
COMMAND TO MS TO
MOVE MS TO ANOTHER
TIMESLOT
ASSI GNMENT COMPLETE
ASSI GNMENT FAI LURE
HOI NSUC
HOI NBOCH
MS MOVES TO NEW TIMESLOT AND
SENDS ASSIGNMENT COMPLETE.
IF MS CANNOT ESTABLISH CONNECTION
ON NEW TIMESLOT,
IT GOES BACK TO THE OLD TIMELSLOT
AND SENDS ASSIGNMENT FAILURE
TCALLS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.8c MEASUREMENT FORMULAS FOR INTRA CELL HAND OVER
The number of successful intra cell hand overs is counted
The percentage of successful intra cell hand overs is measured in relation to total intra cell
hand over attempts
The percentage of reversions to old channel are measured in relation to total iho attempts
The percentage of iho attempts due to up link, down link or both in relation to total iho
attempts is measured.
There is no counter for intra cell hand overs executed, (ASSIGNMENT COMMANDS sent),so
ho_in_lost ratio may be wrong if cell has congestion. This is because an intra cell hand over
request may be denied by the CP if all timeslots are seized.
ho_in_suc HOINSUC Number of successfulintra cell handover attempts performed in the cell
ho_in_suc_ratio
HOINSUC *100%
(HOINUQA+HOINDQA+HOINBQA)
Number of successful intra cell handovers as a percentage of intra cell
handover attempts
ho_in_rev_ratio
HOINBOCH *100%
(HOINUQA+HOINDQA+HOINBQA)
Number of reversions to old channel as a percentage of intra cell
handover attempts
ho_in_lost_ratio
(HOINUQA+HOINDQA+HOINBQA-HOINSUC-
HOINBOCH) (HOINUQA+HOINDQA+HOINBQA)
*100%
The difference between successful + reverted intra cell handover
attempts as a percentage of intracell handover attempts
ho_in_dwnqa_ratio
HOINDQA *100%
(HOINUQA+HOINDQA+HOINBQA)
Number of intra cell handover attempts due to Down link quality as a
percentage of number of intra cell handover attempts
ho_in_uplqa_ratio
HOINUQA *100%
(HOINUQA+HOINDQA+HOINBQA)
Number of intra cell handover attempts due to Up link quality as a
percentage of intra cell handover attempts
ho_in_badqa_ratio
HOINBQA *100%
(HOINUQA+HOINDQA+HOINBQA)
Number of intra cell handover attempts due to both uplink and downlink
quality as a percentage of intra cell handover attempts
HANDOVER
MEASUREMENTS FORMULA DESCRIPTION (MEASUREMENTS ARE PER CELL)
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.9a COUNTERS FOR IDLE CHANNEL MEASUREMENTS
Idle channel measurements are measurements made by the BTS of the up link channel. When
a timeslot at the BTS is idle (not seized), measurements of the received power level are made.
This power measured is back ground noise or interference in the up link
Measurements are averaged over a period of time, and each time slot is placed in an
interference band. When measured power changes from one interference band to another, this
change is reported by the BTS to the BSC. The BSC collects statistics from each BTS, of the
number of times each TS was in each interference band
Separate measurements are made for overlay and underlay sub cells
IDLECH Counters for Idle channel measurements
NOACC The total number of idle channel measurements collected
ITOSIB1 The number of idle channel measurements in interference band 1 in the overlaid subcell
ITOSIB2 The number of idle channel measurements in interference band 2 in the overlaid subcell
ITOSIB3 The number of idle channel measurements in interference band 3 in the overlaid subcell
ITOSIB4 The number of idle channel measurements in interference band 4 in the overlaid subcell
ITOSIB5 The number of idle channel measurements in interference band 5 in the overlaid subcell
ITUSIB1 The number of idle channel measurements in interference band 1 in the underlay subcell
ITUSIB2 The number of idle channel measurements in interference band 2 in the underlay subcell
ITUSIB3 The number of idle channel measurements in interference band 3 in the underlay subcell
ITUSIB4 The number of idle channel measurements in interference band 4 in the underlay subcell
ITUSIB5 The number of idle channel measurements in interference band 5 in the underlay subcell
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC
RP
CP
4.9b COUNTERS FOR IDLE CHANNEL MEASUREMENTS
I TOSI B1, I TUSI B1
I TOSI B2, I TUSI B2
I TOSI B3, I TUSI B3
I TOSI B4, I TUSI B4
I TOSI B5, I TUSI B5
EVERY FEW SECONDS, THE POWER
LEVEL MEASURED IN IDLE TIMESLOTS
IS MEASURED BY THE BTS. BASED ON
THE POWER MEASURED, EACH TS IS
PLACED IN AN INTERFERENCE BAND
THE CP ACCUMULATES
NUMBER OF ICM
MEASUREMENTS IN EACH
BAND FOR EVERY CELL
BAND 1 =-110dBm to -105dBm
BAND 2 =-104dBm to -99dBm
BAND 3 =-99dBm to -92dBm
BAND 4 =-91dBm to -84dBm
BAND 5 =>-83dBm
CHANGES FROM 1 BAND
TO ANOTHER ARE SENT
FROM THE BTS TO BSC
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.9c MEASUREMENT FORMULAS FOR IDLE CHANNEL MEASUREMENTS
The number of accumulations of idle channel measurements are counted
The number of measurements, and the percentage of total measurements in each band
are measured
Measurements for underlay sub cells only are currently made
idlech_acc_num NOACC Number of 10 second accumulations of idel channel measurements
idlech_itub1 ITUSIB1 Number or measurements in interference band 1
idlech_itub1_ratio
(ITUSIB1) / * 100%
(ITUSIB1+ITUSIB2+ITUSIB3+ITUSIB4+ITUSIB5) % of total measurements in interference band 1
idlech_itub2 ITUSIB2 Number or measurements in interference band 2
idlech_itub2_ratio
(ITUSIB2) / * 100%
(ITUSIB1+ITUSIB2+ITUSIB3+ITUSIB4+ITUSIB5) % of total measurements in interference band 2
idlech_itub3 ITUSIB3 Number or measurements in interference band 3
idlech_itub3_ratio
(ITUSIB3) / * 100%
(ITUSIB1+ITUSIB2+ITUSIB3+ITUSIB4+ITUSIB5) % of total measurements in interference band 3
idlech_itub4 ITUSIB4 Number or measurements in interference band 4
idlech_itub4_ratio
(ITUSIB4) / * 100%
(ITUSIB1+ITUSIB2+ITUSIB3+ITUSIB4+ITUSIB5) % of total measurements in interference band 4
idlech_itub5 ITUSIB5 Number or measurements in interference band 5
idlech_itub5_ratio
(ITUSIB5) / * 100%
(ITUSIB1+ITUSIB2+ITUSIB3+ITUSIB4+ITUSIB5) % of total measurements in interference band 5
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.10a COUNTERS FOR CELL LOAD SHARING
Cell load sharing is a method to reduce temporary congestion on a cell by dynamically
reducing the hand over hysterysis border of a cell when the traffic load in a cell reaches a
certain level. Mobiles located near the borders with surrounding cells will then be handed over
to these surrounding cells, thus shedding some of the traffic load on the congested cell.
Counters are collected per cell, from object type CELLEVENT
Hand overs due to load sharing and O+M intervention are counted.
LSTIME Number of times that cell load sharing threshold was met in the cell
TIME Total time in seconds that load sharing threshold was met in the cell
HOATTLS Number of handover attempts from the cell due to load sharing
HOSUCLS Number of successful handover attempts from the cell due to load sharing
HOATTBL Number of handover attempts from the cell due to O+M intervention
HOSUCBL Number of successful handover attempts from the cell due to O+M intervention
CELLEVENT Subcell change and intra cell handover counters
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
BSC
RP
CP
4.10b COUNTERS FOR CELL LOAD SHARING
LSTI ME, TI ME
HOATTLS, HOSUCLS
HOATTBL, HOSUCBL
1:WHEN THE NUMBER OF IDLE CHANNELS IN THE CELL
FALLS BELOW A SET LIMIT, CELL LOAD SHARING IS
INITIATED
THE CP COUNTS THE
NUMBER OF HAND OVER
ATTEMPTS DUE TO LOAD
SHARING, AND THE
NUMBER OF SUCCESSFUL
HAND OVERS
2: KHYST AND LHYST ARE GRADUALLY REDUCED TO
OFFLOAD MS LOCATED NEAR CELL BORDERS TO
SURROUNDING CELLS
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.10c MEASUREMENT FORMULAS CELL LOAD SHARING
The percentage of time that cell load sharing threshold was met or exceeded is measured
The number of load sharing hand overs and the success rate is measured
The number of O+M hand overs and the success rate is measured
lsh_time_ratio (LSTIME/TIME) * 100% Percentage of the time that cell load sharing criterion was met
ho_lsh_att HOATTLS Number of handover attempts due to load sharing
ho_lsh_suc_ratio (HOSUCLS/HOATTLS) * 100% Success rate of load sharing handovers (outgoing)
ho_onm_att HOATTBL Number of handovers due to O + M intervention
ho_onm_suc_ratio (HOSUCBL/HOATTBL) * 100% Success rate of O + M handovers
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.11a COUNTERS FOR OVERLAY/UNDERLAY SUBCELLS
The CME-20 BSS allows for cells to be partitioned into overlay/underlay sub cell structures, to
increase capacity. The overlay sub cell is operated at a lower ERP to allow a tighter frequency
re-use pattern to be utilized. Mobiles that are close to the BTS are handed down to the low
power overlay sub cell, as mobiles move away from the cell, and the path loss increases, they are
handed back up to the standard underlay cell
Sub cell changes are executed by the BSC sending an ASSIGNMENT COMMAND to the MS
to move the MS from a timeslot on the underlay sub cell, to a timeslot on the overlay sub cell, or
vice versa. The operation is similar to an intra cell hand over. Counters are available to count
such sub cell change events.
There are also counters in object type CELLTCH and CELLSDCCH to count traffic,
availability and congestion in the overlay sub cell, where a sub cell exists. The normal SDCCH
and TCH counters by default count traffic in the underlay sub cell, or where no sub cell
structure is defined
Sub cell changes can occur due to timing advance or path loss
CELLEVENT Subcell change and intra cell handover counters
HOAATOL Stepped when an assignment from UL to OL is attempted
HOSUCOL Stepped when an assignment from UL to OL is successful
HOAATUL Stepped when an assignment from OL to UL is attempted
HOSUCUL Stepped when an assignment from OL to UL is successful
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.11b COUNTERS FOR OVERLAY/UNDERLAY SUBCELLS
SDCCH AND TCH TRAFFIC, AVAILABILITY AND CONGESTION
CELLSDCCH SDCCH counters
CCALLSSUB As for CCALLS but for overlaid subcell
CMSESTBSUB As for CMSESTB but for overlaid subcell
CCONGSSUB As for CCONGS but for overlaid subcell
CTCONSUB As for CTCONGS but for overlaid subcell
CTRALSUB As for CTRALACC but for overlaid subcell
CNUCHSUB As for CNUCHCNT but for overlaid subcell
CAVASUB As for CAVAACC but for overlaid subcell
CELLTCH TCH counters
TCALLSSUB As for TCALLS but in overlaid subcell
TMSESTBSUB As for TMESTB but for overlaid subcell
TCONGSSUB As for TCONGS but in overlaid subcell
TTCONSUB As for TTONGS but for overlaid subcell
TAVASUB As for TAVACC but for overlaid subcell
TTRALSUB As for TTRALACC but for overlaid subcell
TNUCHSUB As for TNUCHCNT but for overlaid subcell
Note: There is no dropped call counter for overlay sub cells
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
4.11c COUNTERS FOR OVERLAY/UNDERLAY SUB CELLS
HOAATUL
THE BSC ORDERS A SUB CELL CHANGE FROM
OVERLAY TO UNDERLAY WHEN PATH LOSS
OR TIMING ADVANCE CRITERIA MET
OL
UL
BSC
ASSIGMNENT COMMAND
ASSIGMNENT COMPLETE
HOSUCUL
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
BTS
4.11d COUNTERS FOR OVERLAY/UNDERLAY SUB CELLS
HOAATOL
THE BSC ORDERS A SUB CELL CHANGE FROM
UNDERLAY TO OVERLAY WHEN PATH LOSS
OR TIMING ADVANCE CRITERIA MET
OL
UL
BSC
HOSUCOL
ASSIGMNENT COMMAND ASSIGMNENT COMPLETE
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
4.11e MEASUREMENT FORMULAS FOR OVERLAY/UNDERLAY SUB CELLS
The number of attempts to move from overlay to underlay, and from underlay to
overlay is measured
The percentage successful of each is also measured
Traffic, availability and congestion measurements are not currently produced, but
formulas are the same as for normal cell measurements
ho_tol_att HOAATOL Number of handover attempts from underlaid to overlaid subcell
ho_tol_suc_ratio (HOSUCOL/HOAATOL) * 100% Success rate of UL - OL handovers
ho_tul_att HOAATUL Number of handover attempts from overlaid to underlaid subcell
ho_tul_suc_ratio (HOSUCUL/HOAATUL) * 100% Success rate of OL - UL handovers
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
5.1a MANAGEMENT OVERVIEW REORTS
Reports are available that summarize key performance indicators of the network. Cells
are grouped by BSC and by geographical region. KPI reported include:
Subscriber perceived dropped call rate
Call completion ratio *
Traffic level
Congestion
Handover Success Rate
* Call completion ratio is measured as assignment success ratio on TCH channels, which
ignores call set up failures due to incomplete or dropped SDCCH seizures, hence this
number is very optimistic.
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
Subscriber Perceived Quality at 11Nov1998
NE Name %CallEst %PerceiveDrop Call Attempts Call Establis Call Drop
------- -------- ------------- ------------- ------------- ----------
Network 99.33 2.51 1484437 1474528 36944
BSC1 99.34 2.25 352276 349956 7887
BSC2 99.39 2.11 363150 360926 7625
BSC3 99.33 2.64 223815 222315 5874
BSC4 99.21 3.02 241494 239586 7247
BSC5 99.35 2.48 226915 225430 5598
BSC6 99.39 3.56 76787 76315 2713
5.1b MANAGEMENT OVERVIEW REORTS - EXAMLE
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
5.2a NETWORK PLANNING
Average daily traffic and cell busy hour traffic is measured for each cell. This can be
used to calculate cell utilization. Once a cell reaches a certain utilization level, additional
transceivers can be added so that congestion in the future, as traffic level grows.
Once the point where no more transceivers can be added is reached, planning can be
made for addition of more macro-cells (cell splitting) or micro cells to provide additional
capacity where required
Cell busy hour is used, not system busy hour. Different cells will be at their busiest at
different times of the day. Some cells will be busy in the morning, others in the afternoon,
and so on
Dropped call rate and dropped call reasons can be used as a justification for more cell
sites. Regions that have high dropped call rate, with most of the drops being due to signal
strength, require more cell sites to increase coverage density
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
5.3a NETWORK OPTIMIZATION
Performance measurements can be used as an aid to network optimization work
Performance measurements complement optimization activities, they are not a
replacement for field optimization work
Optimization can be concentrated on areas with high traffic, as this will have the biggest
impact on subscriber quality. Eg a cell that only carries 20 calls a day, of which 10 drop,
will have a dropped call rate of 50%. But why waste time on that cell !
Efforts should be concentrated on cells with the following
High traffic and high Erlang drop call rate
Cells with high level on intra cell hand overs
Cells with low level of incoming hand over success rate
Cells with high level of outgoing hand overs due to urgency conditions
Cells with short mean holding time
Cells with high levels of dropped calls due to quality or unknown reasons
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
5.3b NETWORK OPTIMIZATION
MEASUREMENT IDEAL GOOD NEEDS WORK
SDCCH COMPLETION RATE 100% >85% <50%
ASSIGNMENT COMPLETION RATE *1 100% >97% <90%
TCH COMPLETION RATE 100% >97% <90%
SDCCH DROPPED RATE 0% <2% >10%
TCH DROPPED RATE 0% <2% >5%
SDCCH/TCH DROPS SS 100% >90% >80%
SDCCH/TCH DROPS QA 0% <10% >20%
SDCCH/TCH DROPS TA 0% <10% >20%
SDCCH /TCH DROPS OT 0% <10% >20%
CONGESTION RATIO 0% 5% 30%
CONGESTION PROBABILITY 0% <2% >5%
TCH MEAN HOLDING TIME NA >30 SEC <20 SEC
INCOMING H/O SUCCESS 100% >97% <90%
INCOMING H/O REVERSION 0% <3% >10%
OUTGOING H/O DUE K + L 100% >80% <60%
OUTGOING H/O UPLINK QA 0% <10% >20%
OUTGOING H/O DOWNLINK QA 0% <10% >20%
OUTGOING H/O TIMING ADVANCE 0% <10% >20%
INTRA CELL HANDOVER RATE 0% <10% >20%
ICM MEASUREMENTS IN BAND 1 100% >99% >95%
ICM MEASUREMENTS OTHER BAN 0% <1% <5%
ASSI'MNT TO OTHER CELL SUC RATE 100% >90% >80%
CME-20 R6 RADIO NETWORK PERFORMANCE MEASUREMENTS
5.4a NETWORK ALARMS