HSUPA Q&A

[ritgail]

Lades I would like to through these questions and hope it will create a good thread for EUL::p

In what release HSUPA was introduced? Who are target for HSUPA and what are the Handsets categories?

3GPP introduced HSUPA taking what in consideration so that a huge uplink data would be supported?

What is the spreading factors used in HSUPA?What are the modulations? What is TBF sizing and it depends on what?


What is the maximum throughput in HSUPA? How it is calcaulated? What is the role of i-Q modulation? Is it digital modulation or a RF modulation?

Code Tree utilisation in HSUPA max throughput?

What is the maximum no HRAQ process are there in 2ms TTI and 10ms TTI?

Who schedules the TTI allocation ans which layer does the job?

How SIR target is fixed for uplink in HSUPA? What is the general concept of SIR target fixing in R99?

......to be continued

[moslyi0891]

3GPP introduced HSUPA taking what in consideration so that a huge uplink data would be supported?

The hardware must support 2SF2 + SF4 and TTI 2 ms as well it is feature

What is the spreading factors used in HSUPA?What are the modulations?



SF2 SF4 SF8

[ritgail]

I-Q modulations comes into picture and code tree utilisation violation can be said...



3GPP introduced HSUPA taking what in consideration so that a huge uplink data would be supported?

The hardware must support 2SF2 + SF4 and TTI 2 ms as well it is feature

What is the spreading factors used in HSUPA?What are the modulations?



SF2 SF4 SF8

[moslyi0891]

Higher order modulationsare not used in HSUPA as they need more energy per bit (unlike in HSDPA where complexmodulations are supported).

[moslyi0891]

3GPP introduced HSUPA taking what in consideration so that a huge uplink data would be supported?

The error rate is really being the most important considaration that s why using TTI 2 ms for faster link adaptation, H-ARQfor making re-transmissions more effective, faster scheduling, faster DCH set up and higherorder modulations.

Code Tree utilisation in HSUPA max throughput?
Code tree utilization must take into considration the SF2 , SF4 mainly are being used and there will be high utilization

Maximum through put will be 5,6 Mbps as we are using 2*2 (2SF2(2*1056Kbps)) + 2*4 (2SF4(2*850Kbps))

What is the maximum no HRAQ process are there in 2ms TTI and 10ms TTI?
HybridAutomaticRepeatreQuestorHARQinHSUPAissimilartothatofHSDPAexceptthatin the case of HSUPA it is fully synchronous and with IR (Incremental Redundancy) makingre-transmissions more effective. Also, the buffer is maintained by the base station rather thanby the mobile. IR and CC are present in the HSUPA as in the HSDPA (explained before inthis chapter). The process is synchronous in the HSUPA. All the timing sequences are definedwith respect to 2 ms and 10 ms (4 HARQ processes) and these tell which HARQ process isbeing used. The number of these processes is not required to be configured

Who schedules the TTI allocation ans which layer does the job?

As in HSDPA, the scheduling is moved to the base station in HSUPA as well. But unlike inHSDPA where all of the power can be re-directed to a single user, in HSUPA resources aredistributed evenly across all users, that is the approach is a dedicated one unlike in HSDPAwhereasharedapproachisused.TheschedulerintheHSUPAsystemdoesthetaskofallocatingresourcesofexistinguserswhenadmittingnewusersintothenetwork.Ithasinformationaboutthe interference in the uplink direction and is faster in controlling interference as comparedto the scheduler in the RNC. The scheduling in HSUPA has better spectral efficiency (L1HARQ) and is faster (base station-based scheduling). Three physical channels are used forthis purpose: E-DPCCH in uplink with E-AGCH and E-RGCH in the downlink direction.There are two scheduling methods, long-term grants and short-term grants. Long-term grantsa reissued tomany mobiles that can sendd at a simultaneously,there by th egrantsareincreasingor decreasing according to the load of the cell and is done in the code domain. Short-termgrants allow multiplexing mobiles in the time domain. Scrambling and channelization codesarenotsharedbetweenvariousmobilesinordertoallowthemultiplexinguplinktransmissionsof many mobiles

[ritgail]

Higher order modulationsare not used in HSUPA as they need more energy per bit (unlike in HSDPA where complexmodulations are supported).


in HSUPA there is no need of SIR target on BLER value..as it wont be there..so RNC decides SIR target depending upon no of retransmission in the uplink.

Whereas in R99 it is fixed from the BLER value by RNC in the outer loop..inner loop continues till next SIR is obtained..