UMTS Reselection Parameters

General parameters with typical network values of the three types of reselection in UMTS is presented below:

3G to 3G

• qrxlevmin = -109 dBm     (min acceptable threshold of cell RSCP)
• qqualmin = -18 dB           (min acceptable threshold of cell quality EcNo)

• sintersearch =  10 dB   (f1,f2 case)
• sintrasearch =  14 dB    (same frequency)

3G to 2G

• ssearchrat = 2dB            (indicates when to trigger measurements on GSM cells, increasing this value quickens while decreasing it delays the process)
• treselect = 2s
• qoffset1sn = 7dB for gsmrelation; 0 dB for utranrelation                                                       (Source and target signal offset when reselection strategy is RSCP)
• qoffset2sn = 0 dB                 (Source and target signal offset when reselection strategy is EcNo)

2G to 3G

• FDDqRSCPmin = -102 dBm
• FDDqmin = -12 dB

Equation

The equations on which the decision is made using these parameters are:

For EcNo based reselection:

Squal <= sintrasearch     OR        sintersearch           OR            ssearchrat

where Squal = qmeasured  -    qqualmin

Thus the equation becomes:

qmeas - qqualmin <= sintrasearch

qmeas  <= qqualmin + sintrasearch

Similarly, EcNo values can be replaced with RSCP for the RSCP based reselection.

UMTS - Types of congestion

UMTS cells can have below congestion types:

1. CE (Channel Elements) Congestion:  This happens when resources on the hardware are fully utilized. One solution is to add extra cards/boards in the Node B to increase capacity. For example, ZTE has BPK1 and BPKd cards that comprises of 384 and 768 CE respectively. CE congestion is always monitored by observing UL CE (since UL CE consumption is always greater than that in the DL)
2. Code Congestion:  High number of users could result in code congestion. Parameters such as minHSDPA users can be decreased to make room for cell DCH users. Carrier expansion can be done to accommodate larger  number of users.
3. TCP Congestion:  This represents that Total Carrier Power limit has been reached. Addition of Power Amplifier, new nodes/sites, decreasing CPICH power, or increasing cell transmission power are some of the remedial steps for rectification. 
4. Iub Congestion:  Congestion results due to fully utilized transmission media which can be increased to resolve the issue. 

UMTS Drive Test Basic Analysis

A basic overview of UMTS drive test analysis comprises of two key KPIs which are RSCP and EcIo corresponding to coverage and quality. The common cases and their remedies are presented below:

1. Bad RSCP and bad EcIo (Coverage): This means weak coverage of the serving cell which should be improved to better the EcIo
2. Good RSCP and bad EcIo (Pilot Pollution/High Cell Load): If the issue is due to multiple servers (pilot pollution) having RSCP within 3dB, dominance of one cell should be formed. For high cell load problem, carrier expansion or traffic balancing are some ways of improving the EcIo.
3. Bad RSCP and good EcIo: Low cell load or no handover in the area represents this condition. 

An example of pilot pollution issue is shown below:

GSM TCH Channel Allocation Priorities - Zero Traffic Analysis

Zero traffic means the hardware is not carrying traffic even if there is blocking on the cell. Below are the steps to check for this issue:

1- Filter out TRXs having zero attempts/traffic
2- Check those cells TCH availability - if there is low TCH, then resolve low TCH first
3- Examine TCH blocking on these cells - if there is zero blocking, then the traffic is not high enough
4- Inspect the below TCH channel allocation priorities to finalize the zero traffic TRXs and recommend for health check

Note: TRXs can have zero or low assignments with normal ranges of availability if there is no blocking.

UMTS Paging Success Rate

Paging success rate is a KPI that can be checked if there are complaints involving user not responding messages or missed call notifications. If paging success rate is degraded, it is mainly due to poor coverage. In such case, below actions may be taken:

• increase minimum required received level (QRxLevmin) (eg:-111 to -105) and minimum required quality level (qqualmin) (eg: -18 to -15) of worst cells
• increase PICH power by a few dB (eg: -7 to -5) and increase downlink DCH AC threshold (eg: 85 to 95)
• reduce detected preamble threshold (dB) (eg: -21 to -24)
• increase PRACH initiation Tx power constant value (dB) (eg: -21 to -18)
• Reduce Location update timer both at the core and RNC (say from 3 hour to 1 hour)

Paging flow chart is given below for understanding.

Note:
Paging type 1 = Idle mode
Paging Type 2 = Dedicated mode

UMTS CE Consumption - Uplink and Downlink

Uplink CE consumption is more than that of downlink. The reason is that CEs are not utilized  by HSDPA being a shared channel. HSUPA and R99, on the other hand, consume dedicated CS resources for their respective sessions.

UMTS - Analysis for Second Carrier Expansion

Second Carrier expansion can be proposed if there is low user throughput due to high number of users. One way of finding Node B/Cells that requires such an expansion is presented below. Extract the cell hourly level performance statistics of last 10 days and check for below criteria:

• RoT (Rise Over Thermal) >= 6dB and Cell DCH users >= 16
• HS User Throughout is less than 500 kbps and Average HSDPA users are larger than 16
• Non HSDPA TCP Utilization avg >=55% or max>=90%  and Cell DCH users >=16 

If these conditions are fulfilled by cells in at least 10 hours, the cell should be recommended for second carrier expansion. 

Note that second carrier expansion will not increase the peak user throughput, but the average user throughput is enhanced and so is the user QoS. The peak throughput is 21 Mbps for UMTS, however with Dual Carrier technology, the theoretical throughput limit is increased to as high as 42 Mbps.

UMTS Inter Vendor Hard Handover Issue Analysis

Hard handover could be an issue if two or more vendors are involved. Following steps can be taken to identify and kill the problem:

1. Neighbor configuration verification (LAC/RAC)
1. Definitions in both source and target RNC
2. Core network (MSC) definitions
2. Parameters check related to Hard Handver
1. Switch (Some vendors have Hard Handover Switch)
2. Threshold settings (could be too tight to initiate Handover)

A solved case scenario:

Ericsson to ZTE hard handovers were not taking place (there were no attempts at all). 

• Neighbors were correctly defined at both RNCs
• Core definitions were verified as well
• RNC trace analysis showed relocation failure message from MSC to RNC with eRANAP cause as "misc_unspecified_failure"

Source RNC Trace log

Core/MSC trace showed that relocation_request was being forwarded by the MSC to target RNC when relocation_required message is sent by source RNC. However, when target RNC did not respond with relocation_request_ack, the MSC sent the relocation_preparation_failure with "unspecified failure". This showed that target RNC (Ericsson) had some functional issue. On further investigation, it was found that Ericsson RNC had a switch parameter for hard handover which was disabled. Enabling it rectified the problem and hard handovers began to take place.

GSM - Low Paging Sucess Rate Issue

Paging discards can be a source low PSR or users complaining about getting "user not responding" messages as they call party B in GSM. The steps that can be taken to reduce or eliminate discards are as follows:

1- LAC Splitting
2- Cell paging capacity
      - can be increased of worst offenders
      - Shifting of SD/PD timeslots to non-BCCH TRXs
3- Audit of Periodic Location Update BSC Timer Value (t3212)
      - Can be set to 6 hours
      - This timer value at the MSC should be greater than the BSC value
4- Activation of Precise Paging Feature
      - Paging is done only on cells where there is high probability of finding subscribers according to the location record saved in the BSC