MR Averaging and Voting Part1

HANDOVER WORKSHOP 1 Part 1 - MR Averaging & Voting
Overview
Measurement Reporting BSS Measurement Averaging Voting Mechanism Decision Processes Application of Different Averaging Algorithms
July 2000
CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering
MS / BSS Handover & Power Control
UL & DL Measurements Taken Averages Produced A Number of Averages Compared Against Threshold
Threshold Exceeded
Power Output Adjusted / Handover Triggered

July 2000
Information Sent to MS
Contents

July 2000
of SYS INFO 5
3 versions: 5, 5bis, 5ter BA bit - toggles every neighbour list update Contain list of all neighbour frequencies If <125 frequencies (eg PGSM) then use bitmap For GSM1800 can have up to 374 frequencies SI 5 - only one message needed SI 5bis extension message needed SI 5ter - dual band list needed
4
Information Sent to MS
Contents

of SYS INFO 6
Cell identity & Location Area: MCC, MNC, LAC Cell Options
PWRC DTX
(may, shall, shall not) Radio Link timeout in 4 frame (1920ms) blocks

NCC permitted, bitmap of 8 allowed NCC (upper 3 bits of BSIC) MS is only allowed to report neighbours with these NCCs
5
July 2000
MS Measurement Report
Ncell 1
Ncell 6
28 2 0 0 17 0 17 0 34 23 66
.......
28 26 66
DTX Used BA Used Power Level Timing Advance Rxlev & Rxqual Full Rxlev & Rxqual Sub RXLEV FREQ Index
BSIC
July 2000
MS Measurement Cycle
IDLE IDLE
SACCH
25
24
23 22
21
20 19
18
17
16 15
14
13
12 11
10
6 5
1 SACCH Multi-frame = 4 x 26 -Multi-frame
120 ms Multi-frame
Frame sent every 4.615ms 104 Multi-frame = 480ms

96 frames for Speech/ Data 4 frames for SACCH - Measurement report 4 idle frames to decode BSIC of neighbours
No DTX = 100 frames out of 104 (4 idles) DTX = 12 frames (4 SACCH + 8 SID)
7
July 2000
MS Measurement Cycle
Change from DL to UL frequency
1 2
4 5
7 0
2 3
4 5
0 1
2 3
4 5
7 0
DOWNLINK
1ms
2 ms
TDMA FRAME
Measure 1 neighbour frequency
6 7
1 2
4 5
7 0
1 2
5 6
7 0
1 2
4 5
UPLINK
SAMPLE PERIOD
July 2000
Decoding the BSIC
July 2000
Neighbour Reporting Process
MS to synchronise to and demodulate surrounding BCCH carriers & identify the BSIC MS shall use at least 4 spare frames per SACCH period to decode the BSIC. MS shall demodulate the SCH on the BCCH carrier of each surrounding cell and decode the BSIC as a minimum at least once every 10 seconds. If the MS detects one or more BCCH carriers, among the six strongest without BSIC, then the MS shall as a matter of priority attempt to decode their BSICs (devote all search frames for a period up to 5 seconds. If the BSIC on a surrounding cell BCCH carrier is not decoded for more than three successive attempts, it will be considered lost. When a BCCH carrier is found to be no longer one of the 6 strongest, timing and BSIC information shall be retained for at least 10 seconds. The reported parameters (RXLEV) shall be the average of the received signal strength measurement samples in dBm taken within the SACCH reporting period. The MS shall transmit the MR to the BSS in the next SACCH message block.
10
July 2000
Contents of Measurement Report-BTS

Measurement report number (0 to 255) RxLev UL (full and sub) DTX used RxQual UL (full and sub) BTS Power level (2dB steps below max power of HW) MS Power (reported by MS in layer 1 header) Timing Advance (reported by MS in layer 1 header) Measurement Report from MS Relative Timing Advance (full and sub)
July 2000
11
What Is Not in Measurement Reports

Detection of neighbours/ interferers co-channel with serving cell Detection of co-channel neighbours/ interferers with similar RX Levels Cannot decode BSIC unless C/I > 7dB Detection of non-GSM neighbours MS only reports signals with de-codeable BSIC Cannot identify unknown cells, only know the BCCH/ BSIC
July 2000
12
BSS Processing & Threshold Comparisons
Reported by MS

DL Received Signal Strength (TCH/SDCCH) DL Quality (TCH/SDCCH) DL Surrounding Cell Rxlev (Neighbour BCCH) UL Received Signal Strength (TCH/SDCCH) UL Quality (TCH/SDCCH) MS-BTS Distance Interference Levels in Idle Timeslots
Performed by RSS

Processed value every SACCH multiframe (480ms)

13
July 2000
BSS Processing & Threshold Comparisons
RXQUAL Range
BER
0= less than 0.2% 1= 0.2 0.4% 2= 0.4 0.8% 3= 0.8 1.6% 4 = 1.6 3.2% 5 = 3.2 6.4% 6 = 6.4 12.8% 7 = greater than 12.8%
ASSUMED BER
0.14% 0.28% 0.57% 1.13% 2.26% 4.53% 9.05% 18.1%
14
July 2000
BSS Measurement Averaging
HREQAVE
The number of reported averages required to produce one rolling average as one instance of HREQT
HREQT
The number of averages retained in the Handover software task, needs to be at least equal to the n parameter for that particular threshold
INTAVE
The number of interference averages, reported by the DRCU, that are used to produce one rolling averages every 480 mS
HREQT & HREQAVE have a maximum value of 32, although

HREQT * HREQAVE for the same averaging process must equal 32 or less (HREQT * HREQAVE 32)
15
July 2000
BSS Measurement Averaging
HREQAVE and HREQT must be defined for:

Rxqual UL/DL: Handover (ho) & Power Control (pc) Rxlev UL/DL: Handover (ho) & Power Control (pc) Surrounding Cell DL: Handover (s_cell) Timing Advance: MS to BSS timing advance (rel_tim_adv)
INTAVE and HREQT must be defined for:
Idle Interference UL: Idle Channel Processing
Command: chg_act_alg_data <alg_name> <cell_id> <bin_num> <alg_num> <alg_data>,<index>

16
July 2000
Averaging (HREQAVE & HREQT)

MEASUREMENT REPORTS DLRxLev -80 -82 -81 -81 480ms -81 480ms -79 -82 -83 -83 -79 ....
....
hreqave = 4
-80 -81 -81 DLRxLev Re-averages
hreqt = 3
hreqt = 3
VOTING: p out of n re-averages below DL RxLev threshold needed to trigger HO!

July 2000
17
N and P Voting Mechanism
Voting occurs for each decision process every SACCH multiframe (480ms) N
Defines the number of BSS produced averages required for a decision to be made
Defines the minimum number of BSS produced averages, over threshold value, to cause a positive trigger
N = HREQT
18
July 2000
Voting (N & P)
Example of Voting for DL RxLev Handover
p out of n re-averages below threshold needed to trigger HO!

DLRxLev 63 -47dBm
n
L_RxLev_DL_H threshold for DL RxLev HO
n = HREQT p <= n
p
0 -110dBm
July 2000
19
HREQAVE/HREQT and N/P Relationship

HO Trigger N=3 P=2
-95 -96
DLRxLev MEASUREMENT REPORTS -85 -86 -87 -88 -88 -88 -89 -91 -91 -91 -91
480ms 480ms 480ms -92 -94 -94
HO Trigger N=3 P=3

-98 .... -99
-97 -97 -98
-86.5
-89
-91.25 -89 -91 -93
-91
-93
-97
HREQAVE = 4 HREQT = 3 Threshold L_RXLEV_DL_H = 20 (-90 dBm)
-87
-89
-92
-86
-89
-91
AVERAGES always truncated & rounded down. No floating point operations in HDPC
20
July 2000
Effect of Correct N & P Settings
Before
Modified N & P settings will help the effectiveness of power control
Mobile using less power
After
July 2000
21
Measurement Reporting & Processing
When RSS does not receive expected MR from MS (DL measurements), action is controlled by:
missing_rpt flag=1: report missing, no DL voting, no PBGT calculated. For UL, averaging and decision making will progress in the normal way missing_rpt flag=0: report repeated, voting continues, PBGT calculated
Flag should be set to 0, otherwise only UL intra-cell HO can be performed if UL fails!

22
July 2000
Measurement Reporting & Processing

Repeated MEASUREMENT REPORTS DLRxLev -80 -81 -81 480ms -81 480ms -79 -79 -83 -83 -79 ....
....
-80 -80 -81
Report missing FLAG=1
DLRxLev Re-averages
No DL voting No PBGT calculated

hreqave = 4
July 2000
hreqt = 3
Report missing FLAG=0

Voting continues PBGT calculated
23
BSS Averaging Mechanism

<index> 0 1 2 <alg data> 0 31 0 31 0 255 ALL ALL Only Rxqual_pc/ho Note: Only required if quality weighting decision_alg_num=1 Only Rxqual_ho Note: This can only be set to 1 if decision_alg_num = 1 Rxqual_ho Interfere_ho Rxqual_ho Interfere_ho 24 Averaging Mechanisms
Parameter Hreqave Hreqt
qual_weight 3 0/1
qual_pwr_flag
qual_margin_flag adj_qual_margin 4 6 0/1 -63 to +63 dBs
July 2000
BSS Averaging Mechanism

Description Weighting factor when using algorithm number 1. The weighted algorithm will only apply when the reported averages are full values i.e. not sub (DTX). It causes the most recent reported averages to be repeated a number of times, as per the weighting factor, before an average is calculated by the averaging mechanism. If enabled, will allow a handover only when BTS or mobile is operating at full power. Qual_power_flag cannot be set to 1 if decision_alg_num equals 0. If disabled, the call will be handed over to the best neighbor available, which may have a weaker signal level than the serving cell. Additional receive level that the best neighbor may be below the serving cell. used for rxqual_dl_ho rxqual_dl_pc rxqual_ul_ho rxqual_ul_pc
Parameter qual_weight
qual_power_flag
rxqual_dl_ho rxqual_ul_ho rxqual_dl_ho rxqual_ul_ho
qual_margin_flag
adj_qual_margin
rxqual_dl_ho rxqual_ul_ho 25
July 2000
BSS Decision Process

Averaging Mechanism rel_tim_adv surround_cell Decision process decision_1_tim_adv_av_alg decision_1_p_bgt_rxlev_av_h (neighboring cell of Criteria 2) decision_1_ncell_rxlev_av_calc (Criteria 1) decision_1_p_bgt_rxlev_av_h (serving cell of Criteria2) decision_1_dl_rxlev_av_h (DL rxlev Handover) decision_1_dl_rxlev_av_ih (DL interference Handover) decision_1_ul_rxlev_av_h (UL rxlev Handover) decision_1_ul_rxlev_av_ih (UL interference Handover) decision_1_dl_rxlev_av_p decision_1_ul_rxlev_av_p decision_1_dl_rxqual_av_h decision_1_ul_rxqual_av_h decision_1_dl_rxqual_av_p decision_1_ul_rxqual_av_p
rxlev_dl_ho rxlev_ul_ho rxlev_dl_pc rxlev_ul_pc rxqual_dl_ho rxqual_ul_ho rxqual_dl_pc rxqual_ul_pc
July 2000
26
Pointing Averaging Mechanisms to Decision Processes (HO)
Activity
Link
Enable
Averaging
Decision
Threshold
Voting
Notes
Quality Handover
Up Down
ul_rxqual_ho_allowed dl_rxqual_ho_allowed
rxqual_ul_ho rxqual_dl_ho
decision_1_ul_rxqual_av_h decision_1_dl_rxqual_av_h
l_rxqual_ul_h l_rxqual_dl_h
P6, N6 P6, N6
Used by interference handover as well
Level Handover
Up Down
ul_rxlev_ho_allowed dl_rxlev_ho_allowed
rxlev_ul_ho rxlev_dl_ho
decision_1_ul_rxlev_av_h decision_1_dl_rxlev_av_h
l_rxlev_ul_h l_rxlev_dl_h
P5, N5 P5, N5
--
July 2000
27
Pointing Averaging Mechanisms to Decision Processes (PC)
Activity
Link
Enable
Averaging
Decision
Threshold
Voting
Power Control due to Level
Up Down
ms_power_control_allowed bts_power_control_allowed
rxlev_ul_pc rxlev_dl_pc
decision_1_ul_rxlev_av_p decision_1_dl_rxlev_av_p
l_rxlev_ul_p u_rxlev_ul_p l_rxlev_dl_p u_rxlev_dl_p
P1, N1 (increase) P2, N2 (decrease ) P3, N3 (increase) P4, N4 (decrease )
Power Control due to Quality
Up Down
ms_power_control_allowed bts_power_control_allowed
rxqual_ul_pc rxqual_dl_pc
decision_1_ul_rxqual_av_p decision_1_dl_rxqual_av_p
l_rxqual_ul_p u_rxqual_ul_p l_rxqual_dl_p u_rxqual_dl_p
July 2000
28
Averaging & Voting Mechanism

Activity Quality Handover Link Up Down Enable ul_rxqual_ho_allowed dl_rxqual_ho_allowed Averaging rxqual_ul_ho rxqual_dl_ho Decision decision_1_ul_rxqual_av_h decision_1_dl_rxqual_av_h Threshold l_rxqual_ul_h l_rxqual_dl_h Voting P6, N6 P6, N6 Notes Used by interference handover as well -Relevant quality handover must trigger first Uplink only No voting as lastest hreqt used --
Level Handover Interference Handover
Up Down Up Down
ul_rxlev_ho_allowed dl_rxlev_ho_allowed interfer_ho_allowed interfer_ho_allowed
rxlev_ul_ho rxlev_dl_ho rxlev_ul_ho rxlev_dl_ho
decision_1_ul_rxlev_av_h decision_1_dl_rxlev_av_h decision_1_ul_rxlev_av_ih decision_1_dl_rxlev_av_ih
l_rxlev_ul_h l_rxlev_dl_h u_rxlev_ul_ih u_rxlev_dl_ih
P5, N5 P5, N5 P7, N7 P7, N7
Distance Handover Power Budget Handover
--
ms_distance_allowed
rel_tim_adv
decision_1_tim_adv_av_alg
ms_max_range handover_margin
P8, N8 --
Source pwr_handover_allowed Neighbor
rxlev_dl_ho decision_1_pbgt_rxlev_av_h surround_cell/p er neighbor decision_1_ul_rxlev_av_p
Up Power Control due to Level Down Up Down ---
ms_power_control_allowed rxlev_ul_pc bts_power_control_allowed rxlev_dl_pc ms_power_control_allowed rxqual_ul_pc bts_power_control_allowed rxqual_dl_pc --surround_cell rxlev_dl_ho surround_cell
Power Control due to Quality Criteria 1 Criteria 2
l_rxlev_ul_p u_rxlev_ul_p l_rxlev_dl_p decision_1_dl_rxlev_av_p u_rxlev_dl_p l_rxqual_ul_p decision_1_ul_rxqual_av_p u_rxqual_ul_p l_rxqual_dl_p decision_1_dl_rxqual_av_p u_rxqual_dl_p decision_1_ncell_rxlev_av_calc -decision_1_pbgt_rxlev_av_h --
P1, N1 (increase) P2, N2 (decrease) P3, N3 (increase) P4, N4 (decrease) ---
--
---
Note:PBGT HO - per neighbor averaging as defined in the add-neighbor is only active if the chg_element use_neighbor_pbgt_hreqave is set to 1
July 2000
29
Application of Different Averaging Algorithms
Criteria 2 (serving cell):
chg_ele decision_1_p_bgt_rxlev_av_h chg_act_alg_data rxlev_dl_ho
DL Rxlev handover:
chg_ele decision_1_dl_rxlev_av_h chg_act_alg_data rxlev_dl_ho
DL Interference handover:

chg_ele decision_1_dl_rxlev_av_ih chg_act_alg_data rxlev_dl_ho

30
July 2000
Criteria 2: decision_1_pbgt_rxlev_av_h = 1
DL RxLev HO: decision_1_dl_rxlev_av_h = 1
DL Interf HO: decision_1_dl_rxlev_av_ih = 1

July 2000
31
Criteria 2: decision_1_pbgt_rxlev_av_h = 0
DL RxLev HO: decision_1_dl_rxlev_av_h = 1
DL Interf HO: decision_1_dl_rxlev_av_ih = 2

July 2000
32
Criteria 1 for every handover:

chg_ele decision_1_ncell_rxlev_av_calc chg_act_alg_data surround_cell
Criteria 2 for every handover:

chg_ele decision_1_p_bgt_rxlev_av_h chg_act_alg_data surround_cell
July 2000
33
Criteria 1: decision_1_ncell_rxlev_av_calc = 0
Criteria 2: decision_1_p_bgt_rxlev_av_h = 0
July 2000
34
Criteria 1: decision_1_ncell_rxlev_av_calc = 0
Criteria 2: decision_1_p_bgt_rxlev_av_h = 1
July 2000
35

MR Averaging and Voting Part1

Uploaded by

Copyright:

Available Formats

MR Averaging and Voting Part1

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MR Averaging and Voting Part1

Uploaded by

Copyright:

Available Formats

HANDOVER WORKSHOP 1 Part 1 - MR Averaging & Voting

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

MS / BSS Handover & Power Control

Power Output Adjusted / Handover Triggered

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

1 SACCH Multi-frame = 4 x 26 -Multi-frame

Frame sent every 4.615ms 104 Multi-frame = 480ms

Measure 1 neighbour frequency

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

Decoding the BSIC

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

Neighbour Reporting Process

Contents of Measurement Report-BTS

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

What Is Not in Measurement Reports

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

BSS Processing & Threshold Comparisons

Processed value every SACCH multiframe (480ms)

BSS Processing & Threshold Comparisons

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

BSS Measurement Averaging

HREQT & HREQAVE have a maximum value of 32, although

BSS Measurement Averaging

HREQAVE and HREQT must be defined for:

INTAVE and HREQT must be defined for:

Idle Interference UL: Idle Channel Processing

Command: chg_act_alg_data <alg_name> <cell_id> <bin_num> <alg_num> <alg_data>,<index>

Averaging (HREQAVE & HREQT)

VOTING: p out of n re-averages below DL RxLev threshold needed to trigger HO!

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

N and P Voting Mechanism

p out of n re-averages below threshold needed to trigger HO!

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

HREQAVE/HREQT and N/P Relationship

480ms 480ms 480ms -92 -94 -94

HO Trigger N=3 P=3

-97 -97 -98

-91.25 -89 -91 -93

HREQAVE = 4 HREQT = 3 Threshold L_RXLEV_DL_H = 20 (-90 dBm)

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

Effect of Correct N & P Settings

Modified N & P settings will help the effectiveness of power control

Mobile using less power

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

Measurement Reporting & Processing

Flag should be set to 0, otherwise only UL intra-cell HO can be performed if UL fails!

Measurement Reporting & Processing

-80 -80 -81

Report missing FLAG=1

No DL voting No PBGT calculated

Report missing FLAG=0

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

BSS Averaging Mechanism

Command: chg_act_alg_data <alg_name> <cell_id> <bin_num> <alg_num> <alg_data>,<index>

Parameter Hreqave Hreqt

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering

BSS Averaging Mechanism

Command: chg_act_alg_data <alg_name> <cell_id> <bin_num> <alg_num> <alg_data>,<index>

rxqual_dl_ho rxqual_ul_ho rxqual_dl_ho rxqual_ul_ho

CUSTOMER SUPPORT SERVICES DIVISION Network Operations Engineering