RTWP Issue

What is RTWP?

If you work with UMTS,'ve probably heard someone talk about RTWP. Its definition can be found in a dictionary of acronyms, such as [Only registered and activated users can see links] Received Total Wideband Power.

Represents a measure of UMTS technology: the total level of noise within the UMTS frequency band of any cell.

RTWP is related to uplink interference, and its monitoring helps control the call drops - mainly CS. It also has importance in the capacity management, as it provides information for the Congestion Control regarding Uplink Interference.

In UMTS, the uplink interference may vary due to several factors, such as the number of users in the cell, the Service, Connection Types and Conditions of Radio, etc.

As our goal is to always be as simple as possible, we will not delve in terms of formulas or concepts involved. We will then know the typical values, and know what must be done in case of problems

Typical Values

Ok, we know that RTWP can help us in checking the uplink interference, then we need to know its typical values

In a network is not loaded, normal, acceptable RTWP Average value is generally around -104.5 and -105.5 dBm.

Values around -95 dBm indicate that the cell has some uplink interferers.

If the value is around -85 dBm, the situation is ugly, with strong uplink interferers

Usually we have High, Low and Medium measures of RTWP. However, the maximum and minimum values are recommended only as auxiliary or reference, since they may have been caused by a peak of access, or even been forced to have a momentary value due to some algorithm i.e..

Thus, the value that helps us, and has the most accurate information is the same Mean RTWP!

For cases in which cell has two carriers, the difference between them RTWP should not exceed 6 dB.

Based on these typical values, most vendors have an alarm: RTWP "Very High. "

What to do in case of problems?

We have seen that RTWP can cause performance degradation, mainly CS Call Drops. Note: Actually, it's not RTWP that causes performance degradation. What happens is that when its value is 'bad', it's actually indicating the

presence of interference - the latter being responsible for degradation.

But what can we do when we find bad values?

If RTWP is not at acceptable levels, some actions should be taken.

The first thing to do is check if there is a configuration issue with the RNC or NodeB. This is the most common case, especially in cases of new activations.

Once verified the parameter settings, the next step is the physical examination, especially jumpers and cables, often partially reversed. It also should be checked if there is faulty transmitters, or any other problem that could generate intermodulation between the NodeB and the antenna.

If the parameter settings and hardware are ok, the chance is very high that we have external interference, such as a Interferer Repeater.

In cases where there may be external interference, we must begin to act after such a prioritization based on how much this is affecting the cell KPI's across the network, if it carry high traffic, major subscribers, etc.

Note: There are many forms of interference in the uplink, both internal and external. Only a few are listed above. The deepening of all possibilities is beyond the goal of being simple to teach the concepts, but this is a suggestion for whoever wants to deepen the study, identification and elimination of interference.

In practice

to find - and eliminate - problems of interference is one of the biggest challenges in our area. For being such a complex problem, we recommend that be collected enough data for each investigation. Insufficient data collected can lead to erroneous conclusions, further worsening the problem .

The uplink interference may appear only in specific periods. Thus, it is recommended that data be collected from at least one week (7 days) for every 24 hours. Usually this amount of data is sufficient. In the figure below, we see different days and times - colorful - a fictional example where the interference occurred.

Data should be collected for the suspicious cell, but also for its adjacent cells, allowing it to make a triangulation increasing the chances of locating the source of interference.

Another way to locate the source of interference is to do a test in field. An antenna guy must gradually change the azimuth of the antenna, while another professional do RTWP measurements. That is, through the information directing the antenna and the respective values of RTWP, you can draw conclusions very good.

It is obvious that changing the online system may not be a good practice, and tests can be made with a Yagi antenna and a Spectrum Analyzer.

Vendors offer several ways to measure RTWP, using the OSS, performance counters and logs.

Possibility of External Interference:

In one cell site with high RTWP connect the spectrum analyzer to the cell site antena. Check the noise level. Turn off your 4G base station and check again your noise level.

If your 4G site are turn off you noise level should be low if not you have an external interference.

If you have external interference use the spectrum analyzer to meter the noise in 360°, take note where you found the high noise value. Do the same in at least 3 cell sites with high RTWP.

With this information and using triangulation you can discover where is the source of interference.

Internal or External interference?

In case of RTWP occurence in the UMTS network checking the system against intermodulation should be done in the first step. The test is very simple - during monitoring RTWP in the observed cell set the maximum tx power in that cell. When RTWP rises as tx power rises intermodulation is suspected. It can be either internal or external but it's simple to differentiate it - external interference reacts to the tilt change... From my experience the most cases of internal intermodulation is due to improper connector fixing - corroded surface acts as a mixer. I've also found several cases of external intermodulation - 100% of them were GSM or/and DCS repeaters with too small attenuation set.

Searching external interference in UMTS network with a spectrum analyzer is tricky to some extent - it isn't rare the interference is broadband and it's difficult to see it unless the noise floor on the SA display is around -130 [dBm] or lower so preamplifier should be used along with pass band filter to avoid intermodulation in the spectrum analyzer alone and false readout in the result.

Update: In case of RTWP affected cells neighbours list should be also checked. UE connected to other cells can cause elevated RTWP in case of missing neighbours between these (affected and serving UE) cells, especially with lower band (eg 900 MHz) UMTS systems.

Possible of PARAMETER SETTING cause:

Don't know if it is still relevant

We are facing similar problem. Although average RTWP is within limits -> -105..-103, the maximum is much higher up to -80. The peaks are very evident in RTWP monitor - they get 1s peaks of high level (up to -80 and in one particular indoor cell up to -65).

Spectrum analyser on the indoor cell did show any issues - thermal noise level, even we were seeing -65dBm on RTWP monitor. Then it was considered, that Path budget is too good and UE cannot downregulate power because of limitation (by 3gpp it is -51dbm). Therefore 20db attenuation has been added, however no change in rtwp monitor - not even decrease in peaks by 20dB.

Then we have limited particular types of traffic - and found the reason --> RRC setups, i.e. RACH procedure. Huawei parameters allow ramping up to max what 3Gpp permits - 20 x 2dB, backoff is the smallest of all vendors i've seen -> 20 (e// has 27 and nsn 25). Therefore we tuned RACH procedure - limited ramping, increased back off. That caused Rach failures altogether for some mobiles.

Note, that we have on many cells, that does not seem to be installation issue, and spectrum analyser does not show anything. Disabling RACH had made RTWP monitor chart clean. I not regard it as an issue, however this RTWP monitor value is transferred to SIB7 what causes just generation of interference. Moreover this likely is causing another issue - high R99 UL RLC retransmission rate and even some RRC setup complete failures.

We are waiting for RND reply, however i would suspect the issue is in RRU software (MRFU is doing just the same). Software version we are running is R13. 

The problem have been identified. Our setup is to have 3x of RRC release complete to be sent by UE. However RNC was setup to release RL even after reception of first RRC RELEASE CMP message. Therefore UE will have RL lost and would be ramping UL power causing uncontrolled UL interference. 

After RNC was setup to delay RL, RTWP issue was solved.

How to check RNC parameter for RRC_CONN_REL_CMP? Thanks I think he means the PROCESSSWITCH3_UM_RRCRELCMP_RLDEL_DELAY_SWITCH in URRCCTRLSWITCH?

I'm surprised it made such a big difference, from the information I saw, this problem only affected certain mobiles, I think the iPhone was singled out as one.