Selasa, 21 Juni 2011

Call drop reason analyzing


1 Equipments problem
The method for equipments problem can refer to the section in call failure analyzing. TRX, CE, vocoders and transmission links are the key points which should be cared for.


2 Call drop caused by overstepping the coverage
Typical case:
MS Rx power is about -100dBm or less;
MS Tx power tends to maximum 23dBm;
The strongest pilot Ec/Io < -15dB or less;
MS Tx_Adj maintains in normal range: 0~-10dB; MS goes into system searching mode after call drop, and can not find system or system signal is very weak and is very easy to be lost.


Analyzing:
In the edge of coverage area, for the forward and reverse links are all very bad, the call drop is one normal phenomenon
When pilot Ec/Io decreases to some extent, the forward link quality will be bad greatly and can not be demodulated well, F-FER will increase very fast. If MS oversteps coverage area too long (more than 5s), MS fading timer expires in 5s, MS will initialize again, that is call drop. MS will go into system searching mode after initialization, because MS has overstepped coverage area, it can not find service system or system signal is very weak and is very easy to be lost.


Note:
The above time is 5 seconds, in fact MS oversteps the coverage area in shorter time (less than 5s) also can cause call drop. In this case usually the reason is BTS call drop mechanism is faster than MS fading timer. When MS oversteps the coverage area, reverse link is also weak and R-FER is also high, then BTS call drop mechanism will be triggered. BTS will release forward link in short time (<5s), when reverse link is too bad. Even if MS returns to coverage area at this time, the call drop will happen also. Because the forward link does not exist, so the call drop must happen, though the pilot has been resumed.

Optimization method:
The ultimate way for this kind of problem is add new BTS or repeater in blind coverage area or poor coverage area.
If adding new BTS is not possible, other methods also can be used to improve coverage, like increase antenna height, select large gain antenna, and adjust antenna azimuth and down tilt. But these methods can not solve problems ultimately, and it should be very careful when change these parameters.

Call drop caused by access and handoff collision


Typical case:
MS origination call may be failed in poor coverage area in DT test, or call drop will happen very soon after successful origination. MS Rx level and pilot Ec/Io are all low, and MS Tx power is very high. As MS moves, MS Rx power will become larger and larger, but pilot Ec/Io is too bad to satisfy the demodulation requirement;
MS initializes again after call drop and stays in one new strong pilot; If MS originates successfully in good coverage area and passes the same DT rote, call drop will not happen;


Analyzing:

If system can not support access handoff (access handoff also need MS to support), then during MS access procedure, MS can only start handoff after finish access. The above case is belonging to call drop caused by access and handoff collision. If one MS starts origination in the edge of coverage area, because this is near to another cell, it is possible to meet handoff. But our system does not support access handoff, access and handoff will have collision and access has priority in system. Because access need some time, so MS Rx power will become higher and higher, but pilot Ec/Io will become lower and lower. At this time, the objective handoff pilot is becoming strong interference, when current pilot Ec/Io is lower enough; the forward link will become very bad and can not be demodulated successfully. The access will be failed. After access failure, MS will stay in new pilot (It is the objective handoff pilot). If MS originates successfully in good coverage area and passes the same DT rote, call drop will not happen, because at this time MS can carry the normal handoff procedure.


Note:

Access handoff can be supported after 5.4 version;
For terminals, 1X handset can support, but 95 handset can not.


Optimization method:
If system and terminals can support access handoff, this kind of problem will not happen;
Adjust network structure and soft handoff area. Extend soft handoff area in the area where the above problems are heavy, then MS will firstly handoff to another cell before access and MS will have enough time to finish access.

Call drop caused by forward and reverse imbalance


Typical case:
MS received power and strongest pilot Ec/Io maintain in good state, such as Rx Power>-100dBm, Ec/Io>-15dB;
As MS moves, MS reverse Tx power is increasing continuously to 23dBm, until call drop happen;
As MS moves, MS Tx_Adj is increasing continuously and it is one positive value, until call drop happen;
After call drops, MS will initialize and stay in previous pilot. There is signal in MS, but origination call is difficult, even if origination is successful, call drop is very possible.


Analyzing:
In the above case, MS Rx power and pilot Ec/Io are all good, and this means the forward link is good. But in MS origination procedure MS is increasing transmitting power until maximum, and this means the reverse link is bad. That is forward and reverse imbalance. Tx_Adj>0 also means forward is better than reverse.
Because forward is better than reverse, so in the edge of coverage BTS can not receive the signal from MS correctly and BTS call drop mechanism is triggered, and then call drop will happen.


Note:
Forward and reverse link imbalance includes two kinds: forward is better than reverse; reverse is better than forward.
The above case is the former. It popular in practice and effect is large. Because subscriber can not tolerate that call failure and call drop happen when there is signal in MS, so this case should be cared for very much.


Optimization method:
Find the rootstock of imbalance:
For the case that forward is better than reverse:
Judge if the cell power configuration is too high;
Judge if the pilot gain is too high;

Judge if there is reverse interference;
For the case that reverses is better than forward:
Judge if the cell power configuration is too low;
Judge if the pilot gain is too low;
Judge if there is forward interference;

Call drop caused by forward interference


Typical case:
As MS moves, its Rx power is increasing;
As MS moves, the strongest pilot Ec/Io is decreasing and is not enough for demodulation requirement, F-FER increases very fast until call drop;
As MS moves, MS Tx power is normal and will not tend to maximum;
After call drops, MS will initialize and stay in previous pilot, but origination call is difficult and call drop is very easy to happen. MS may stay in one new pilot and the signal stabilization is decided by interference source.


Analyzing:
MS Rx power is increasing, but the pilot Ec/Io is decreasing, this means there is forward interference. When MS moves to interference continuously, the pilot Ec/Io will decrease continuously, until F-FER increases very fast and MS fading timer expires, MS will initialize again, that is call drop.


Note:
For the forward interference source, there are two kinds: external interference; internal interference.
External inherence: the radio signal from other system drops into MS receive bandwidth, MS will re-initialize after call drop and stay in previous pilot. Call drop is very easy to happen again, even if origination call is successful. Internal interference: usually it is handoff which causes call drop. When MS moves to a new cell, for some reasons (wrong neighbor list setting; too narrow search window) handoff is failed, and the objective cell pilot will be a strong pilot interference. This kind of interference belongs to internal interference. The difference from external interference is: MS will re-initialize after call drop and stay in one new pilot (the objective handoff cell pilot) very well.


Optimization method:
External interference: Inform customer and they connect with government to clear; Internal interference: Check background configuration and adjust the related parameters.

Jumat, 17 Juni 2011

Call drop caused by reverse interference

Typical case:
As MS moves, its Rx power and pilot Ec/Io keeps in normal level;
As MS moves, MS Tx power increases and tends to the maximum 23dBm, until call drop;
As MS moves, MS Tx_Adj increases and it will be positive value, until call drop;
MS re-initializes after call drop and will stay in previous pilot. There is signal in MS, but origination call is difficult or call drop is very easy to happen after successful origination.


Analyzing:
In the above case, MS Rx power and pilot Ec/Io are all good, and this means the forward link is good. But in MS origination procedure MS is increasing transmitting power until maximum, and this means the reverse link is bad. That is there is reverse interference. Tx_Adj>0 also means forward is better than reverse. As MS moves continuously, MS R-FER will increase higher and higher. BTS can not receive the signal from MS correctly and BTS call drop mechanism is triggered, and then call drop will happen.


Note:
Call drop caused by reverse interference is very similar to forward and reverse imbalance, in fact reverse interference is just one of reasons that lead to forward and reverse imbalance. For the reverse interference source, there are two kinds: external interference; internal interference. External interference: other system radio signal may drop into BTS received
bandwidth; Internal interference: there may be illegal MS which try to access continuously. Because MS Tx power is limited, the number of illegal MS must be very large.

Optimization method:
External interference: Inform customer and they connect with government to clear; Internal interference: Usually this case seldom happens, but there is possibility, for example like repeaters with MS module, which repeaters use MS module for
background observation. But sometimes, if MS is not assigned number, its origination will be illegal, and then this will become strong interference. The solution is to close them or assign numbers to them.

Call drop caused by F-TCH power limitation


Typical case:
As MS moves, MS Rx power and pilot Ec/Io is decreasing, but they are still enough to maintain the link, like: Rx power>-100dBm, Ec/Io >-15dB; As MS moves, its reverse Tx power is increasing, but it is still enough to maintain the
link; As MS moves, Tx_Adj can maintain the normal state; When MS is far away from BTS, F-FER will increase very fast, until call drop; MS will re-initialize after call drop and stay in the previous pilot.


Analyzing:
Traffic channel gain can be set in background, if the setup value is too low, system can not assign enough power resource to traffic channel. In the above case, according to DT test it can be observed that Ec/Io and Rx power are all higher than threshold (Ec/Io>-15dB, Rx>-100dBm); But when MS is far away from BTS, F-FER increases very fast until call drop. The reason may be forward traffic channel gain is too low and the power is not enough. After call drop, MS will re initialize and stay in previous pilot. The confirm method is use the instrument which can measure code domain power (like
Viper) to test.

Optimization method:
Check background parameters setup and adjust forward traffic channel maximum gain.

Call drop caused by handoff failure
Handoff failure is one of main reasons for call drop, when MS moves from one cell to another cell, handoff will happen; if handoff is failed, signal of the current cell will become worse and worse, F-FER and R-FER will be very high, then call drop will happen when MS fading timer expires.

CDMA2000 Optimization Procedure

CDMA2000 Optimization Procedure.
Wireless network Optimization can be devided in to three layers:
  • Device Layer
  • Network Layer
  • Resource Utilization Layer

We can conduct some methods to examine problems on those layers, i.e.,

  • Drive tests and Analysis
  • Signaling Tracking
  • OMC analysis
  • Synthesis

Let’s we go in to deep to each layer above,

  1. Device Layer
    • Antenna and Feeder Cable Fault
    • Transmission Fault
    • GPS Fault
    • Wireless Configuration
    • Office Direction Problem
    • Termination Problem
  2. Network Layer
    • Solving Dropped Call Problem with Drive Test and Analysis Method
    • Solving Problem with OMC Analysis Method
    • Improving Coverage with Synthess Method
  3. Resource Utilization Layer
    • Network Block Optimization

CDMA Performance Indicators
RF Performance indicators captured by drive-test activity. It show the CDMA RF environment to guide Optimization and Troubleshooting in air interface. Some parameters indicate uplink conditions, some downlink, and some both. These parameters collected at the subscriber side, so it’s easy to capture using commercial handset equipment without BSC’s assist.
Basic knowledge about CDMA spread spectrum signal characteristics such as: channel definitions, power control system, call processing flow, signal behaviour in noise and interference, and RF units ( transmitter and receiver) are needed, to analyse the parameters below.

  • FER (Frame Error Rate) -> an excellent call quality summay statistic, it’s the end result of the whole transmission link
    • reverse channel -> realized on the Base Station
    • forward channel -> realized at handset
    • if FER is good, any other problems aren’t having much effect
    • if FER is bad, we have to check other indicators to analyze the network problem, because FER is just the end-result of the problem
  • Mobile Receiver Power (Rx)-> Received Power at the handset (dBm). It should be noted that Received Power is Important, but it’s exact value isn’t critical.
    • High Rx value (-35 dBm or higher) could cause overload condition in Amplifier sensitivity, intermod and code distortion on received CDMA signals.
    • Low Rx value (-105 dBm or weaker would leave too much noise in the signal after de-spreading, resulting symbol errors, bit errors, bad FER and other problems.
  • Ec/Io -> We can’t just use the handset’s power level to guide handoffs because it represents the total power measurement from all sectors reaching the handset. To measure the the signal level of each sector individually, we have to use each sector’s pilot (Walsh 0) as a test signal to guide handoffs.
    • Ec/Io is a parameter, represents pilot cleannesses.
    • foretells the readability of the associated traffic channels to guide soft handoffs decision
    • derived from: ratio of good to bad energy seen by search correlator at the desired PN offset -> Ec/Io(forward) = Pilot Energy / (Paging + Synch + Traffic) Energy. Can be degraded by strong RF from other Cells, Sectors (imperfect PN orthogonality could cause -20 dB degradation) and also by noise.
    • in Light condition (without traffic) Ec/Io = 3 dB, in Heavy loaded Ec/Io = -7 dB
    • in a clean Situation which theres a sector dominant, Ec/Io just as good as it was when transmitted. But in “Pilot Pollution” condition, mobile hears a ’soup’ made up of all the overlapped sectors signal. So Io is the sum of all the signal received by MS, and Eo is the energy of desired sector’s Pilot signal. The Large Io will overrides the weak Ec -> Ec/Io is too Low!
  • Handset Transmitter power (TxPO)
    • TxPO is the actual RF power output of the handset transmitter(max= 23 dBm), including combined of open loop power control and closed loop power control (TxGA)
    • this is the simple formula : TxPO = -(Rx dBm) - C + TxGA ( C= +73 for 800 MHz systems, and C=+76 for 1900 MHz systems) to reach balance link.

Minggu, 05 Juni 2011

What is RTWP?

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.

Conclusion

In this brief tutorial, we learn what is RTWP, and that the ideal typical value is about -104.5 dBm and -105.5 dBm.

As the RTWP is directly related to Uplink Interference - and we know that interference is the main cause of performance degradation - have concluded that improving RTWP, ie making is as close as possible to -105 dBm, improving the Call Drop Rate!

IMPORTANT : Seizing the opportunity, see what was stated at the start of this tutorial - dictionary - by describing RTWP. Remember that this site has been the subject of a very interesting tutorial in the Tips Section. If you have not visited this section of the portal yet , I strongly recommend, because it has many issues that help in our growth in telecom and IT area.

What is 0 dBm?

Watt (W) and miliWatt (mW)

First of all, to understand what it means for example 0 dBm, we at least have to know the basic unit of power, the Watt. By definition, 1 Watt means 1 Ampere (A) current in 1 Volt (V) voltage, or in mathematical terms P = VA. It is interesting to note that the amount of power radiated by an antenna is very small in terms of Watts, but it is enough to reach several miles.

And as the signs are very small, is more common to refer to them in terms of prefix, such as military or milliwatts (mW), which means 1 / 1000 (one thousandth) of Watts.

Mathematics


Besides the signals were rather small, it - as well as other quantities of physics such as electricity, heat or sound - propagate nonlinearly. It would be more or less like compound interest on a loan.

Or brought into our world of engineers, imagine a cable for transmitting 100 watts, with a loss of 10% per meter. If the spread was linear, the final 10 meters would have no more power!

Only it's not how it happens. In the first meters, have 10% less power, which is 90 watts. And this is the value that 'enter' on the cable until the next meter. Thus, the second meter, we would have 10% less of that power or 81 watts (= 90 - 9). Repeating this calculations, you see that in fact the power never reaches zero, as it would if calculations were linear. (At the end of the cable have actually 34.86 Watt)

To solve problems o deal with this - and make our lives easier - we need to know the logarithms. We saw this in school, but there are people who do not like to hear. The good thing is that we need not know much about them, just understand what they are.

Just understand that if we transformed the magnitudes in logarithms, the calculations become addition and subtraction rather than multiplication and division.

Of course, in order to do the calculations by adding and subtracting, we must make the necessary conversions. But with the help of a calculator or Excel, is not that complicated.

Decibels (dB)

By definition, we have:

Sure, we say that working with logarithms (or decibels) is much easier - and the common good. But by the formulas above, still can not understand. So the best way to understand why we use dB (decibels), is seeing how they help us through a practical example.

Consider a standard wireless link, where we have a transmitter (1) and a receiver (5), Antennas (3), Cables, Jumpers and Connectors (2) and Free Space (4).

Using real values, and without using the help of dB, let's do the math and see, from the transmitted power, how much power we have at receiver. So with dummy values, but close to reality, we have:

  • Transmitter Power = 40 Watts
  • Cables and connectors loss = - 0.5 (Half Power)
  • Antenna Gain = 20 + times in the Power
  • Free Space Loss = - 0 000 000 000 000 000 1 Power

Note: This amount of loss in free space is quite big. And it is obtained based on distance and other factors. For now, just accept that it is a practical value of loss of RF for a given distance of our link.

The link with the absolute values in Watt would then be as below.

We can work this way, of course. But you must agree that it is not very friendly.

Now, if we use the proper conversion of power, gain and loss for dB, we can simply add and subtract.

It was so much easier, isn't it?

Now we just need to know the formulas to do the conversions.

Converting with Formulas in Excel

Considering that the amount of wattage is in cell B3, the formula for convrting W in dBm is:

= 10 * ( LOG10 ( 1000 * B3 ) )

And the formula to reverse - convert dBm to Watt, considering that our power in dBm is in cell B6 is:

= 10 * ( LOG10 ( 1000 * B6 ) )

And as a result, we have calculated values.

Note that in case we are using the 1000 value in the formulas, for wearing the Watt, but we want the result in dBm.

To calculate (convert) db to ratio, or ratio to db, the formulas do not include the value of 1000.

Calculations without using a calculator

Of course, we will use calculators in the projects and programs such as Excel. But we also know how to do calculations (conversions) without using a calculator. If anyone tells you that the power is + 46 dBm, you need to know what that means in terms of Watts.

For this, there are certain tricks that can be used to arrive at least an approximate match.

For this, a good way is to memorize the equivalent to multiplying factors in dB, as in the table below (at least those that are in bold).

With the corresponding values of dB and multiplier factor, we convert eg +46 dBm to mW.

Answer: First, we expressed 46 values that we already know by heart.

So 46 = 10 + 10 + 10 + 10 + 3 + 3

That is, we multiply the reference value (1 mW) for four times the factor of 10 and twice the factor of 2.

What gives us

1mW x 10 x 10 x 10 x 10 = 1 000 0 mW

1 000 0 mW x 2 x 2 = 40000 mW = 40 W

Ie, + 46 dBm is equal to 40 watts.

Conclusion

Well, I think now you have given to see that when we do the calculations in dB everything is easier. Moreover, the vast majority of Telecom equipment has specifications of its units in dB (Power, Gain, Loss, etc.).

In short, just use basic math to understand the values and reach the final figures.

When we say that such a signal is attenuated by 3 dB, means that the final power is half the initial power. Likewise, if a given power is increased by 3 dB means twice the power.

A good practice, irrespective of how you will work with the calculations is to store at least some values such as 0 dBm = 1 mWatt (our initial question), 30 dBm = 1 Watt, and in our example, 46 dBm = 40 Watt.

So you can quickly learn, for example, the equivalent for the calculations.

For example, 43 dBm = 46 dBm - 3 dB. That is, half the power of 46 dbm. Then, 43 dBm = 20 Watts!

Just finally, in our example, the received power was - 84 dBm, remember?

In this case, doesn't need memorizing. Just so you know which is equivalent to a very low power, but enough for a good example for GSM call.

Sabtu, 04 Juni 2011

Senior 2G RF Optimization Engineer (2G RNO)

GCI INDONESIA, PT

Responsibilities:

- RF Optimization

- RF technical support

- RNO/P Training

- DT Data post-processing &analysis, RNO analysis report;

- Monitoring OMC KPI / KPI Analysis Report

Requirements :.

- At least 2 years of RF optimization working experience with 2G RF experience;

- Good knowledge of GSM ;

- Have strong ability of RF troubleshooting.

- Have strong knowledge of Optimization procedure;

- Bachelor’s degree in telecommunication or related field of study;

- Be excellent in communication in English;

- Be familiar with tools/software such us Aglient / TEMS / Actix / Mapinfo / GPS / compass / gradienter and so on;

- Good in communication skill and be able to convince the customer;

- Have HUAWEI Project experience in RNP / RNO is preferred;

- Have working experience in Operators is preferred;

- Have project management experience is preferred;

- Willing to travel frequently

* Please give a remark in your subject of applying email: “Apply for 2G RNO Engineer *

** Please send yor lates photo and resume to hr.gciindo@gmail.com **

My Headlines