Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors

Speech Quality

Viewing 15 posts - 1 through 15 (of 27 total)
  • Author
  • #51221


    It is often said that HR decreases the speech quality. So, do the same principal of speech quality 0-7 hold good for HR also. I mean do we have a
    different scale for HR w.r.t. speech quality.



    RXQUAL(0-7) is not the measure of “speech quality” instead it is the measure of Received Signal Quality and it is calculated from BER(Bit Error Rate). Speech quality is measured differently by different systems e.g. TEMS SQI, MOS, SQI in stats and it is calculated from FER(frame erasure rate) and several other factors. RXQUAL doesn’t depend on the speech codec used. But SQI definitely depends on the codec. You can say HR gives worse speech quality than FR and EFR at same RXQUAL. So, there is no point of having different RXQUAL scale for different speech codecs.



    SQI depends on FER
    RXQUAL depends on BER.

    Is there any equation between SQI and RXQUAL?
    If No, then should a poor RXQUAL be a cause of worry.
    If yes, then what can be done to improve RXQUAL to finally improve SQI.Or what can be the causes of poor BER(thus poor RXQUAL).

    Also, What is the scale or range for SQI from worst to good?



    There is a correlation between FER and BER, which is empirical. It is noted that :

    with No Hopping, an acceptable voice quality (FER) occurs when RXQUAL is below 4 (BER).

    With Hopping, acceptable VQ occurs when RXQUAL below 5.

    The SQI is between 0 and 30, or something like this. It is well explained in the TEMS user guide.

    30 is very good, and 0 is really poor 🙂


    Is there any equation between SQI and RXQUAL?
    – Rxqual affects SQI but the opposite is wrong

    If No, then should a poor RXQUAL be a cause of worry.
    -yes if the coding doesnt not coevr this poorness.

    If yes, then what can be done to improve RXQUAL to finally improve SQI.Or what can be the causes of poor BER(thus poor RXQUAL)


    hi every body…
    who can help me for this problem?
    -whats the reason of handover revision?


    i’ve never heard about handover revision in GSM, nor UMTS.


    handover revision, I think he means the case in which the mobile fails to handover to another cell and thus stays in the current serving one.


    hi pix..
    we have five type of handover in our system GSM:
    1.handover success.
    2.handover failure.
    3.handover attempt.
    4.handover ping-pong.
    5.handover revision.
    the great problem is handover revision.i think this occure if we have long distance between two cells.


    I think he means to say “handover reversion” i.e. reversion to old channel after failed HO attempt. The question can be re-phrased as: What are the reasons of poor HSR? 🙂


    hhaaa.. i’m so blur.

    HO Reversion to Old Channel, aka “ROC”

    R.O.C. is not a bad thing really : it just means that your HO was attempted towards the target cell, but for some reason it was failed (poor radio in target cell or hardware problem, probably).
    Then the MS manages to roll back to the previous serving cell and get back to its old TCH channel.

    Beware : congestion in target cell will not lead to ROC : the HO is not yet attempted. The failure occurs before the attempt.

    As a conclusion, I would say that your HO are triggered too early, and towards bad target cells (poor radio condition).


    Although Reversion to old channel doesn’t lead to very bad subscriber perception like a call drop but still it is a handover failure. Handover is mostly triggered when there is urgency condition in source cell or target cell is better than source cell. And if the MS fails to complete the handover, this is not a good sign for the network.


    SQI Vs. RxQUAL.

    SQI: estimates the speech quality in cellualr networks as pereceived by human listener , the SQI includes :
    3- HO data events
    4- the distribution of BER over time
    5- the speech codes been used

    RxQual: is obtianed by transfering the BER into scale value ( 0-7 ).( only take the average BER

    drawback :
    1- doesn’t include the distribution of BER over time
    2- NO mention Of FER
    3- frames lost during HO
    4- the choise of speech codec


    thanks amkam
    SQI is definitely a better estimation of the voice quality than the RxQual, but it clearly doesn’t replace a subjective test (yet).


    hi guys, pls help me…
    -i found good RxQual (less than 1)but worse SQI=speech quality index (less than 21), how can that be?

    -is that connected with speech codec?

    -what is the differences and advantages of AMR, EFR, Half rate and Full rate for GSM Network

Viewing 15 posts - 1 through 15 (of 27 total)
  • The forum ‘Telecom Design’ is closed to new topics and replies.