RF data rate

Discussion in 'Wireless & RF Design' started by EngIntoHW, Sep 25, 2010.

  1. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    In an article I read about WCDMA, it is said that:
    What are those engineering design considerations for such trade off?
     
  2. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    System usage vs required data rate.
    Longer codes = more reliable, but a lower data rate.
    Shorter codes = less reliable, but a higher data rate.

    You can't engineer for this, it's a market based decision, and can change depending on your customer base.
     
    EngIntoHW likes this.
  3. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Thanks Sceadwian.

    I'm curious about what is this market based decision based on?

    At first I thought that maybe transmitting a long code at high data rate consumes more energy than transmitting the same code at low data rate, but the energy consumption should be the same, shouldn't it?

    So, what could be possible reasons for that decision?
     
  4. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    You have to look at the total bandwidth available. Sure you can give it ALL to one user and it'll be massive, cell networks are dramatically increasing in density though, there are users in the same cell as you now that are fighting for the same bandwidth, in a major city it's severely different that in the boonies. For your given network the number of slots are varied, what happens if the number of users is greater than the number of slots available? Someone's gonna get shafted, and someone is going to lose a call, or get slow data rates. Above the networks critical bandwidth level the system will fold, AT&T has problems because of this and the Iphone. They sold a device capable of more bandwidth than the user density in many locations could support.

    This isn't exclusive to CDMA, all networks experience this same problem if they're usign the same lines. Look back to the days of the original non switched Ethernet networks. It's 10mbit. for the WHOLE network, if two devices on the network try to transmit/receive at 10mbit, NEITHER will be able to communicate effectively.
     
    Last edited: Sep 25, 2010
    EngIntoHW likes this.
  5. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi Sceadwian,
    Thank you for helping out understanding it.

    In the following picture, UE #2, #3 and #4 have longer codes than UE #1, therefore their data rate (and bandwidth) is smaller.
    However, as you can see, the 1MHz bandwidth of the 800MHz RF channel is still as busy as it'd be if all users used the whole 3.84MHz bandwidth.

    However, you say that it's better having 1MHz fully occupied than 3.84MHz fully occupied?

    [​IMG]

    What would be the reason for having a limitted amount of slots available for users in WCDMA system?
    A WCDMA system allows users to share the same bandwidth and transmit at the same time, so allegedly, there's no slots limit.
    So what causes the restriction after all?


    It's a great example, thanks :)
     
    Last edited: Sep 25, 2010
  6. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    Consider this, if the total output power of all four signals is exactly the same the signal with the larger bandwidth is going to have it's power distributed more evenly across it's spectrum while the lower bandwidth signal is going to have it's power in a smaller portion of the spectrum, if the total system power is identical then that means there is more power concentrated in the smaller bandwidth signal, so while the total power will be the same the peak value in the lower bandwidth signal will be higher, this means increased reception range.

    It's not always better it just has to be weighed in a given application.


    No the number of slots aren't unlimited not in practice theory isn't reality practical implementation requires paying attention to the real world, there is a finite amount of spectrum each channel must occupy some of that bandwidth if you have a total spectrum allowance of 10mhz and each channel is using 1mhz you can't have more than 10 channels in that band it just can't happen. WCDMA allows effective sharing of this bandwidth it can not possibly however increase it beyond what is actually there.
     
    Last edited: Sep 26, 2010
    EngIntoHW likes this.
  7. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Thanks sceadwian!
    I see what you're saying, but how does it relate to the size of the packet (length of the code)?
    Yes, I agree that there's a limitted amount of channels per total bandwidth.
    For example, if the total bandwidth is 15MHz, and each channel's bandwidth is 5MHz, then there're only 3 channels.

    However, what limits the amount of users per channel if they can all together transmit over the same channel's bandsiwth at the same time?
     
    Last edited: Sep 26, 2010
  8. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    EngInto, what are you not understanding here? You said you understand and the image you posted above tells you how it relates why are you asking the same questions again when you've already got the answers?
     
  9. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,
    It's because I don't think we covered that point.

    The image I posted shows that if users use smaller bandwidth, then the 3.84MHz is not as bussy as the 1MHz, which both are centred at the 800MHz carrier frequency.

    But what does it have to do with the code length?

    I don't see any relation to the code length in our whole discussion here.
     
  10. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    I'm not even sure what you mean by codelength I don't know that much about CDMA I was basing my answers on your statement that code length is proportional to datarate.
     
  11. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,

    At the first port I quoted an article which said that longer code are transmitted in low data rate.

    Longer code = larger length code.

    I'm trying to see what is behind this - longer code <-> lower data rate.
     
  12. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    Then you need to refer to the original article, so I have something to go by, like I said I don't know that much about WCDMA standards. The encoding for CDMA is not simple, especially considering you didn't reference the article that explains the exact methods that you're looking at.
     
  13. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,
    You're right.

    Here is the article.
    Please take a look at page seven - upper right corner.
     
  14. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    That's going to take me a while to digest =)
     
  15. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Oh, I thank you so much for your patience :)

    I must say, it's so interesting.

    I started studying it a week ago, and ever since I can't stop reading about it.
    It's a good stuff ;)
     
  16. sceadwian

    New Member

    Jun 1, 2009
    499
    37
    If you're still confused try explain better and I'll try to do the same =) I'm as patiente as your willingness to continue questioning is, if you hit hit a dead end with explanations, try different words or different ways of looking at the problem as you see it, might give me enough perspective to figure out how to answer well.
     
  17. EngIntoHW

    Thread Starter Member

    Apr 24, 2010
    128
    0
    Hi,
    Thank you so much for your patience, it's very kind of you.

    I was just entering here to correct myself.

    I confused between data rate (rate of encoder input which encodes the user's data) and chips rate (which is constantly 3.84MHz at WCDMA).

    According to articles, when data rate at input is high, a low spreading factor (also called processing gain) is used, which is what i'm trying to understand - the reason that they use low SF (spread factor) when data rate at input is high.
     
Loading...