Exceed Connection Limit Optimization in 3G

Node B resources – spread factor limitation

• DL and UL
• Counter: pmNoFailedRabEstAttemptExceedConnLimit – MO UtranCell.
• Non-Guaranteed and Guaranteed Service Classes
• DL Non-Guaranteed, HO/non-HO Service Class
• Limited by sf8Adm (8), sf16Adm (16) and sf32Adm (32)
• DL Guaranteed, HO/non-HO Service Class
• Limited by sf16gAdm (16)
• UL Non-Guaranteed, HO/non-HO Service Class
• Limited by sf4AdmUl (1000), sf8AdmUl (50) and sf16AdmUl (50)
• UL Guaranteed, HO, non-HO Service Class
• Limited by sf8gAdmUl (8)
• Counter: pmNoFailedRabEstAttemptExceedConnLimit, MO Class UtranCell

pmNoFailedRabEstAttemptExceedConnLimit

• Exceed Connection Limit, is mainly due to the exceed in the usage"number of available connections" of a specific SF,check if the SF settings in UL and DL are set correctly i.e: SF16AdmUl=50,SF32Adm=32 etc
• If they are set to their maximum limits the only solution will be going for second carrier expansion
• Sometimes it's better to give a small value for sf4AdmUl and sf8Adm. It will help save SF resources to allow more UE connections with lower speed.
• Exceed Connection Limit is pegged when the number of a certain connection "64,128,384" in UL or DL exceeds its admission threshold"sf8Adm, sf16Adm and sf32Adm,etc

Optimization

• In most cases increasing SF16admUl value will reduce the admission blocking. This is because in 3G networks usually the most PS combination can be R99/HS where there is high traffic in UL in R99. Accordingly increasing sf16admUl usually is the 1st admission threshold to increase.
• To solve this first you need to identify which RAB is pegging the exceed connection limit and you need to check if the SFAdm settings are set correctly for this RAB or they might be increased to overcome the blockings, if the SFAdm are set to their maximum values the only solution to solve this is to perform second carrier expansion.
• It is recommended to configured the sfXXAdm either uplink or downlink to the default value. For instance, sf4AdmUl shall be configured to 1000 for the software version of P7 onward, sf8AdmUl to 8 and so on. The intention is not to limit the connection by the spreading factor admission setting.
• In past experience, we did changed before the sf4AdmUl from 1 to 1000 and as well as the sf32AdmUl from 32 to 50 does really reduced the exceeded connection limit significantly in two different cases.
• tune the rateselectionpsinteractive in order to reduce the blocking; Identify the cells and then set 8 for dlprefrate and ulprefrate where channel dch is used knowing thatthe feature Flexible Initial Rate Selection, PS Interactive is enabled. Change from ulPrefRate/dlPrefRate of 64 to 16 we noticed about a 10% decrease in Channel Element consumption across the network. There was also a big reduction in channel switching from 64 to 16Kbits in the uplink.

Restrict Access to a particular RAB

• UeRc disabling under RNCfeatureRABcombinationxxx or for a particular sfxx >>sfxxadm=0
• by setting these SF Adm to 0 will block user of accessing these RAB according their SF: sf8Adm (PS XX/384)
• sf16Adm (PS XX/128),
• sf16gAdm (Streaming XX/128),
• sf32Adm (PS XX/64),
• sf4AdmUl (PS 384/XX),
• sf8AdmUl (PS 128/XX),
• sf8gAdmUl (Streaming 128/XX),
• sf16AdmUl (PS 64/XX).
• Unfortunately we don't have specific parameter for RAB PS 16, but as mentioned above we can disabled them from RNC Feature State.

WRAN CAPACITY MANAGEMENT

Admission Control – DCRNC002 settings

•DL Ch Code – dlCodeAdm (80).

•DL Power Adm – pwrAdm (75).

•DL ASE Adm – aseDlAdm (500).

•UL ASE Adm – aseUlAdm (500).

•DL SF Adm – sf8Adm (8), sf16Adm (16), sf32Adm (32) and sf16gAdm (16).

•UL SF Adm – sf4AdmUl (1000), sf8AdmUl (50), sf16AdmUl (50) and sf8gAdmUl (8).

•CPM Adm – compModeAdm (15)

•HS Adm – hsdpaUsersAdm (20 or 25) and maxNumHsdpaUsers (16, 26 or 32).

•EUL Adm – eulServingCellUsersAdm (20, 25 or 32) and eulNonServingCellUsersAdm (100). Included 2ms and 10ms TTI.

•DL RBS HW – dlHwAdm(99 or 100).

•UL RBS HW – ulHwAdm (99).