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

FER and SQI

Viewing 12 posts - 16 through 27 (of 27 total)
  • Author
    Posts
  • #51279
    Bilal
    Guest

    But SQI statistics don’t depend on number of handovers so there should be no need to adjust the stats for handovers.

    #51280
    Pix
    Guest

    Akefa,

    Ok I understand your point of view. But I personally still consider a HO as a thing that should be avoided if possible.
    In the previous post, I did a typo mistake : read “30” instead of “0”.

    Bilal,

    Actually, yes, SQI is bad when there is a HO (according to akefa. myself, i really can’t remember)

    so the question is whether or not remove bad SQI due to HO or not… In my opinion, no : the HO is a degradation in the voice quality, but for akefa, yes : the SQI shall not reflect poorly a handover.

    #51281
    Bilal
    Guest

    Pix,

    yes, HO definitely affects speech quality but does it affect FER?

    #51282
    Pix
    Guest

    Bilal,

    It’s what I’ve been saying for the past 3 days : There is no speech frame during a handover ! How do you want to compute a frame erasure rate, if there is no frame to look at ? ๐Ÿ™‚

    A handover = a blank of 40ms to 80ms in the speech.

    #51283
    Bilal
    Guest

    Exactly this is the point i wanted to make. HO doesn’t affect FER and SQI stats are totally dependant on FER so?
    HO doesn’t impact SQI stats. It is SQI value in TEMS which is impacted by handovers not the SQI stats in the BSS.

    I asked the question just to make my point clear ๐Ÿ™‚

    #51284
    Pix
    Guest

    Ok, it’s clear for me too, now ๐Ÿ™‚

    The thing is that for Akefa and I, SQI is a metric that belongs to TEMS investigation, and reflects other elements than just FER.

    In which BSS have you seen SQI ? Ericsson ? What is the difference between “BSS SQI” and FER ? Is there a linear correlation between the two ?

    #51285
    Bilal
    Guest

    Yes, SQI counters are there in Ericsson. I don’t about the exact algorithm but Ericsson documentation says it does consider some radio parameters in addition to FER. The outcome of the algorithms is an SQI that is measured in the MNRU scale dBQ which is an absolute scale for measuring speech quality(ITU-T Recommendation P.810 Modulated Noise Reference Unit ). The output values are from -40 to +50 and the higher value the better speech quality.

    SQI values are then divided into the three categories “Good”, “Acceptable” and “Unsatisfactory”. The thresholds that classifies the SQI values into these three categories are 13.5 dBQ and 22.5 dBQ, i.e. everything below 13.5 dBQ is “Unsatisfactory”, everything between 13.5 and 22.5 dBQ is “Acceptable” and everything above 22.5 dBQ is “Good”.

    #51286
    Pix
    Guest

    Thanks for these interesting details ๐Ÿ™‚ It would be interesting to know the algorithm behind the SQI

    #51287
    Djordje
    Guest

    Hello guys,

    Does anybody know exact formula for SQI-MOS computation for WCDMA?

    #51288
    khalil
    Guest

    Hi,
    We have an Ericsson GSM network with 2 media gateway and BSC’s connected to both.I wonder if it’s possible to know by command on the BSC, to which media gateway a ralt2 device (A interface) is connected.

    Thanks

    #51289
    Dharmendra
    Guest

    Hi PIX,
    Can you pleas tell me what is the relationship between Electrical tilt and VSWR because as i heard ,VSWR reduces as we increases Electrical tilt

    #51290
    pix
    Guest

    hello,

    i’ve never heard such thing. If there is an effect, then it has to be very small and without any practical benefit.

Viewing 12 posts - 16 through 27 (of 27 total)
  • The forum ‘Telecom Design’ is closed to new topics and replies.