Generic selectors
Exact matches only
Search in title
Search in content
Search in posts
Search in pages

Ping Pong Handovers and C/I

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #61488 Reply
    Saadi
    Guest

    During the drive test log files analysis using TEMS investigation (GSM network), it was observed in the GSM Line chart, that whenever a handover takes place, the C/I,SQI and RxQual at that instant also degraded. It is also a general observation, the more the number of handovers, the larger the bad quality samples. My question is, why does the C/I degrade at the instance the handover command is sent? Is it because of the BIT steal mode? or is it because of the break before make concept in GSM?

    #61489 Reply
    Pix
    Guest

    Saadi,

    I believe it is a TEMS issue (or a TEMS choice). There is no link degradation during HO : the C/I and RXQual should be stable.
    However the speech frames are lost during the HO execution. And -perhaps- during the HO, to reflect that loss, TEMS artificially degrades the measurements.

    In other words : Voice Quality is degraded during HO, but not the link quality. Even though TEMS is saying otherwise…
    cheers
    pix

    #61490 Reply
    Bijoy
    Guest

    Saadi
    You can enable handover power boost feature to overcome this.In this case,during HO both MS & BTS radiates at MAx. power & the C/I degradation considerably reduces.However,you cant reduce it fully because this is what actully happens during any HO.(No TEMS bug atleast)

    Br\\
    Bijoy

    #61491 Reply
    Pix
    Guest

    Hi Bijoy, you are back ๐Ÿ™‚

    Why would C/I and rxqual degreade during HO ? I didn’t get your explanation.

    When MS goes in the new cell, both BTS and MS are transmitting at max power (by default), because noone knows what power reduction to apply. That’s how it works in ALU, but it’s probably 3GPP compliant.

    And during the HO execution itself, there are no speech frames, therefore there are no valid frames on which to do the measurements.

    #61492 Reply
    Saadi
    Guest

    Thanks Pix and Bijoy for your replies.

    @Bijoy:Is the handover power boost feature Vendor specific or is it a GSM standard? Furthermore, the name suggests that the feature would increase the MS or BTS powers during handovers, don’t you think this will further increase the interference?

    #61493 Reply
    Bijoy
    Guest

    Saadi
    Yes,it is vendor specific.Regarding Interference,no such impact will be there.

    Thanks Pix
    By default,both transmit at max power,for sure.But in some vendors,we can play more:-)

    Regarding C/I degradation.the SACCH frames are filled with dummy burst during HO command,which causes a minor degradation in C/I & Rx_Qual.Since MS measures all the DL bursts,it is shown in TEMS.However,these measurements are excluded in BSC,so your BSC RX_Qual will not get impacted.

    Br\\
    Bijoy

    #61494 Reply
    Pix
    Guest

    Bijoy, thanks !

    #61495 Reply
    AliAsgher
    Guest

    * Dummy bursts in SACCH frame. ๐Ÿ™‚

    Dear Bijoy, are you able to “SEE” the RF waves as they travel by?

    Very nice information. Thanks

    #61496 Reply
    Pix
    Guest

    Bijoy,

    More questions from my confused mind : you are saying the MS is still measuring the old SACCH while performing the HO (between HO command and the HO Access) ?

    Or are those SACCH sent on the new cell ? I can’t see the point of sending those dummy SACCH…

    And I can’t see why the MS would measure “bad” rxqual on those dummy SACCH. It is not because they are “dummy” that they are “poor”.

    What do you think ?

    Cheers
    pix

    #61497 Reply
    Bijoy
    Guest

    Exactly Pix
    It is not mandatory that Rx_Qual will be bad.During HO command,the MS is measuring nothing,so it is pointless to say that Rx_qual is good or bad,as no SACCH is there in DL at the time of HO command.What happens is for the DL SACCH which is scheduled around the time BSC gives the HO command to MS,it is this SACCH Block which is filled with dummy SACCH,TEMS/MS just decodes it & gives the Rx_Qual value,whatever it is.

    Please let me know if any doubt still persists

    Br\\
    Bijoy

    #61498 Reply
    Pix
    Guest

    Hi Bijoy,

    Well, yeah, my doubts are very persistant. I can’t take your word on that, you know me ๐Ÿ™‚

    So … how do you know about dummy SACCH ? Can you see those dummy SACCH messages in TEMS ?

    In 3GPP, they don’t mention any dummy messages like this. On the opposite, it is stated that:

    44.006
    “Both layer 3 procedures, dedicated channel assignment and handover, are initiated and controlled by a network entity (BSS or MSC). This entity therefore will suspend the flow of layer 3 messages towards the MS after the ASSIGNMENT or HANDOVER COMMAND message has been issued. When the data link on the new channel has been established, the message flow may resume.”

    Messages are suspended … ok, but it doesn’t mean the BTS doesn’t continue sending bummy blocks. Yet there is no benefit to use a frequency for nothing ! So i am still doubting ๐Ÿ™

    pix

    #61499 Reply
    Bijoy
    Guest

    Thanks Pix for ur doubt.
    I didn’t said that these SACCH frames comes after the HO command.In fact,it comes alongwith the HO command or even before.Dummy bits are inserted for the Inband signallingin SACCH or Normal burst.(In case of HO command,it is FACCH).This will result in deterioration of speech quality, since either or
    both of the 57 bit information blocks are used for signalling instead of user data.

    Also,in case of non-sync neighbours,it takes at least 300 ms for the MS to tune on it.Here also,dummy SACCH comes in play.

    I think you have to go beyond 3Gpp to comprehend this.

    Br\\
    Bijoy

Viewing 12 posts - 1 through 12 (of 12 total)
Reply To: Ping Pong Handovers and C/I
Your information:




<a href="" title="" rel="" target=""> <blockquote cite=""> <code> <pre class=""> <em> <strong> <del datetime="" cite=""> <ins datetime="" cite=""> <ul> <ol start=""> <li> <img src="" border="" alt="" height="" width="">