PDA

View Full Version : Question TTI from 10 ms to 2ms



_lisi4king
2012-01-16, 11:24 PM
Good day, UMTS guru.

Can you explain me, for example in UL, if we change TTI from 10 ms to 2ms we have throughput increaseng from 2 Mbps to 5. I want to know, why? As i understand, TTI this is time when resousre sceduling for UE. How this time infuent on throughpu?

Thanks for responces.

jan74
2012-01-17, 01:34 AM
Its because you can then send the SRB on the HSUPA channel, freeing up the codes that can be allocated in such a way that you can get 2×SF2 + 2×SF4, allowing you up to 5.8Mbps.

_lisi4king
2012-01-17, 01:39 AM
So, in case of 10 ms i can't use SRB on the HSUPA? And in case of 2ms (HSDPA) i can use SRB on HSDPA?

jan74
2012-01-17, 01:55 AM
So, in case of 10 ms i can't use SRB on the HSUPA? And in case of 2ms (HSDPA) i can use SRB on HSDPA?

Yes, with 10ms, UL SRB will be on the DCH, not the HSUPA channel.

_lisi4king
2012-01-17, 02:01 AM
Yes, with 10ms, UL SRB will be on the DCH, not the HSUPA channel.

What about 10->2 ms in DL?

wolverine
2012-01-17, 02:47 AM
Yes, with 10ms, UL SRB will be on the DCH, not the HSUPA channel.

This actually is not correct. E******* configure the SRB on E-DCH even for the 10ms TTI.

In answer to the original question, it is not the transition from 10ms to 2ms that gives the gain in throughput, but rather the allowed increase in number of parallel E-PDCH and the lower SF they can use.

_lisi4king
2012-01-17, 03:05 AM
This actually is not correct. E******* configure the SRB on E-DCH even for the 10ms TTI.

In answer to the original question, it is not the transition from 10ms to 2ms that gives the gain in throughput, but rather the allowed increase in number of parallel E-PDCH and the lower SF they can use.

Let me clarify: it doen't matter 10ms or 2ms, throughput can be the same? If yes, it shold be correct for UL and DL?

jan74
2012-01-17, 06:00 AM
This actually is not correct. E******* configure the SRB on E-DCH even for the 10ms TTI.

In answer to the original question, it is not the transition from 10ms to 2ms that gives the gain in throughput, but rather the allowed increase in number of parallel E-PDCH and the lower SF they can use.

Wolverine, maybe i didnt make it clear enough. The SRB can be on HSUPA for 10ms but for 2ms, it is only on HSUPA at least on NSN. Regardless, as i also mentioned, freeing up the extra codes allows the SF2 to be used. Thus allowing higher throughput. With respect to 10ms to 2ms DL, i dont understand. All HSDPA TTIs are 2ms regardless.

wolverine
2012-01-17, 06:17 AM
23793
Let me clarify: it doen't matter 10ms or 2ms, throughput can be the same? If yes, it shold be correct for UL and DL?

See extract from the specs above. As you can see the specs only allow the higher throughputs on the smaller TTI. One Transport block fits more bits on the 2ms TTI as opposed to the 10ms TTI due to less coding. As already answered in the DL on HSDPA only the 2ms TTI is defined by the specs.

Ahmed_attar
2012-01-17, 06:32 AM
TTI, Transmission Time Interval, is a parameter in UMTS (and other digital telecommunication networks) related to encapsulation of data from higher layers into frames for transmission on the radio link Layer. TTI refers to the duration of a transmission on the radio link. The TTI is related to the size of the data blocks passed from the higher network layers to the radio link layer.
To combat errors due to fading and interference on the r (http://en.wikipedia.org/w/index.php?title=Radio_link&action=edit&redlink=1)adiolink data is divided at the transmitter into blocks and then the bits within a block are e (http://en.wikipedia.org/wiki/Encoder)ncoded and interleaved. The length of time required to transmit one such block determines the TTI. At the receiver all bits from a given block must be received before they can be deinterleaved and decoded. Having decoded the bits the receiver can estimate the BER. And because the shortest decodable transmission is one TTI the shortest period over which BER can be estimated is also one TTI. Thus in networks with link adaption techniques based on the estimated BER the shortest interval between reports of the estimated performance, which are used to adapt to the conditions on the link, is at least one TTI. In order to be able to adapt quickly to the changing conditions in the radio link a communications system must have shorter TTIs. In order to benefit more from the effect of interleaving and to increase the efficiency of error-correction and compression techniques a system must, in general, have longer TTIs. These two contradicting requirements determine the choice of the TTI.
In UMTS Release '99 the shortest TTI is 10 ms (http://en.wikipedia.org/wiki/Millisecond) and can be 20 ms, 40 ms, or 80 ms. In UMTS Release-5 the TTI for HSDPA (http://en.wikipedia.org/wiki/HSDPA) is reduced to 2ms. This provides the advantage of faster response to link conditions and allows the system to quickly schedule transmissions to mobiles which temporarily enjoy better than usual link conditions. As a result the system most of the time transmits data over links which are better than the average conditions, because of this the bit rates in the system most of the time are higher than what the average conditions would allow. This leads to increase in system capacity.

hope this could help

atul22
2012-01-17, 03:35 PM
Hi,
As per my knowledge, in 10 ms TTI it is transmitted 32 kbps and it is consumed 1 CE.
In 2 ms TTI it is transmtted 160 kbps and consumed 8CE.
In 2ms TTI ,for max throughput 2X2+2X4 SF . so SF2 it transmitted 1.92mbps and SF4 it transmitted 960 kbps .it will get around 5.7 mbps .
and 96 CE is consumed to get this much speed.

BR//
atul

_lisi4king
2012-01-17, 03:48 PM
Hi,
As per my knowledge, in 10 ms TTI it is transmitted 32 kbps and it is consumed 1 CE.
In 2 ms TTI it is transmtted 160 kbps and consumed 8CE.
In 2ms TTI ,for max throughput 2X2+2X4 SF . so SF2 it transmitted 1.92mbps and SF4 it transmitted 960 kbps .it will get around 5.7 mbps .
and 96 CE is consumed to get this much speed.

BR//
atul

Thank for your responce, but my main qestion was not about HSDPA or HSUPA, main qestion is how does TTI influent on throughput (if influent :))

dholama
2012-01-17, 06:12 PM
with 2ms TTI you can respond faster to Link conditions. So if you have a link which is good you can get higher throughput due to quick allocation. Hence we could get better than average bit rates and hence gain in throughput.

_lisi4king
2012-01-17, 07:37 PM
with 2ms TTI you can respond faster to Link conditions. So if you have a link which is good you can get higher throughput due to quick allocation. Hence we could get better than average bit rates and hence gain in throughput.

So i'm i rigth, if we are in laboratory and in excelent conditions, possible to say, that it doesn't matter which TTI (2 ms or 10 ms) we use?

oswe
2012-01-19, 11:07 AM
Thank for your responce, but my main qestion was not about HSDPA or HSUPA, main qestion is how does TTI influent on throughput (if influent :))

it is like they´ve explained before... that it frees up codes to be used...

firstmaxim
2012-01-19, 12:45 PM
Another way of looking at it is that you more HARQ instances that can be defined with 2ms TTI compared to a 10ms TTI.