As 3G director of Spreadtrum Communications Inc, Dr. Kang Yi is the man at the wheel for TD-SCDMA project. Under his lead, Spreadtrum finished developing 1st TD-SCDMA baseband IC within only one year around the world. Dr. Kang has more than ten-years practical experiences in R&D and project management. He ever participated IC design for DSP core, high-performance CPU, visual phone, wireless communications etc. Dr. Kang also has following successful cases: 1st DMB-T transmitting IC for terrestrial digital TV broadcasting, 1st G-Ethernet firewall with network processor. Dr. Kang has abundant background knowledge in digital signal processing, high-speed CPU, telecommunications system and he ever worked for some silicon valley companies such as Sun Microsystems etc. Dr. Kang received his EE bachelor, master degree from TsingHua University and doctor degree from University of Illinois at Urbana Champaign.
-------------------------------------------------------------------------------------------------------------------------------

Q&A on Line:
I: About TD-SCDMA Chip


· Q: Dr. Kang, what’s the difference between the chips from Spreadtrum and T3G ? thanks.
A: All the chips from Spreadtrum, T3G and CYIT are used for baseband processing.
Spreadtrum’s chip has denser integration. Others are chipsets while ours is single-chip solution with full functions of digital basband, analog baseband and power management.


· Q: Spreadtrum’s chip is baseband IC ? RF IC ? or both ?
A: It is baseband IC, not including RF part.


· Q: Would you please give me a general introduction of SC8800 ?
A: SC8800 is 1st TD-SCDMA IC developed by Spreadtrum. We got successful prototyping in this May and passed IC test in this July.


· Q: How about the reliability of Spreadtrum chip ?
A: In last IC test, SC8800 can maintain stable voice call for long time. It’s reliability also has been tested.


· Q: Someone thinks although SC8800 is a milestone for China 3G, perhaps it is not a mature technology. Dr. Kang, What’s your opinion about such comments ?
A: As you said, “China Chip” is a milestone for China 3G. TD-SCDMA has been proved as a mature technology. “China Chip” is fully based on TD-SCDMA standard. So it should be one mature solution.


· Q: Dr. Kang, is it possible for one type of chip to occupy dominating TD-SCDMA market just like Qualcomm has done in CDMA field ? If not, how to resolve interoperability question in short time ? It seems in WCDMA system this problem has not been cracked yet for several years.
A: I don’t think there will be another Qualcomm in TD-SCDMA market, neither for network, terminal or IC suppliers. All the network and terminal vendors need cooperate to resolve interoperability problem.

· Q: Hello Dr. Kang, I want to know whether TD-SCDMA protocol stack is also available on this chip ?
A: Yes. We are doing this work.

· Q: Does protocol stack has been finalized ? Is it developed via SDL ?
A: Generally protocol developing all are done by SDL.

· Q: Does Spreadtrum also develop protocol testing set ? I heard TD-SCDMA Industry Alliance is doing this. By the way, Whether TD-SCDMA IC suppliers also get governmental fund support ?
A: Spreadtrum will support TD-SCDMA Alliance to finish protocol testing set. To solid TD-SCDMA industry, government supports both network and IC suppliers . Spreadtrum also received fund support.

· Q: It is reported that “China Chip” can support both R99 and R4. TD-SCDMA is standardized since R4 version. Would you please give me a explanation ?
A: Here I want to clarify we support R4 only.

· Q: Dr. Kang, Has Spreadtrum participated recent 3G trial ? How about the testing result ?
A: Until now Spreadtrum didn’t participate 3G trial. We plan to join the trial in near future.

· Q: “China Chip” surely is very great. Compared with other traditional IC suppliers, what ‘s Spreadtrum’s unique advantages can guarantee product compatibility and stability ?
A: Spreadtrum is a new company. We can challenge foreign IC giants by our leading technology and good China market understanding.

· Q: For unknown reasons, in some reports roadmap is calculated from Sample directly to Commercial Phone. Chip Quantity Shipment is missing. What’s your opinion?
A: For IC suppliers, the most important is chip quality, i.e. the procedure from R&D to commercial product. IC manufacturing with large quantity depends on customers’ order. When the chip reaches commercial level, manufacturing cycle is only 2~3 months. We say available time is next June means the quality of this chip reaches commercial level at that time.

· Q: Dr. Kang, from your understanding, are there any other attractions for foreign companies except for China is a big market ? Regarding R&D capability, what’s the difference between China and North American ? How about human power cost ? Is just for this reason, Spreadtrum develops IC in China ? If the design house is in North American, what are the difficulties ?
A: China is not only a huge market but also a center of person with ability. This is another attractions for foreign companies. Although domestic IC design is not so good as North American, China has excellent IC, software and system engineers. In China, the cost of developer surely is lower than North American. If TD-SCDMA IC design is located in North American, how to co-debug and test with network system·

· Q: So far how many times has Spreadtrum joined field test ? How about the successful probability· Does Spreadtrum is also developing relative communications software ?
A: We have not carried out field testing. Relative communications software developing is ongoing.

· Q: Dr. Kang, one base question, what’ mean the relative communications software ? high level applications ? Telecommunications service realized on high speed PHY channels ? In next June, are both voice and data services available ?
A: So-called communications software means something related with PHY and protocol layers. Data service will also be available in next year.

· Q: For TD-SCDMA chip, how about cost control ? How about power consumption ?
A: Cost and power consumption of the phone designed based on our chip will be comparative or lower than WCDMA terminal.

· Q: Does Spreadtrum also involve in IC manufacturing ? or by out-sourcing ?
A: Spreadtrum is FABLESS, a IC design company. It has no manufacturing.

· Q: Currently it is said four types of TD-SCDMA chip are available. Has the standardization started ?
A: I heard 4 to 5 suppliers are developing TD-SCDMA IC, not 4 types of chip available. Standardization has been formed from the beginning.

II Cooperation in Industry Chain

· Q: It is said Amoi has adopted Spreadtrum’s TD-SCDMA chip. May I know how Amoi to resolve software and other problems ?
A: Amoi is developing high layer software by itself. For TD-SCDMA fast going ahead, we hope more terminal suppliers join TD-SCDMA family to guarantee TD-SCDMA terminal available in next year. We expect more cooperations with other terminal suppliers.

· Q: Hello Dr. Kang, do you think there are some bottlenecks at lower end of TD-SCDMA industry chain ?
A: I think there are two bottle-neck problems: RF solution is very limited (Just one supplier); available testing instrument is very scarce.

· Q: It is said, Mr. Lou (vice Minister of MII) together with other leaders inspected 3G trials of CNC and CRC. Are they satisfied with the status of current 3G trials ?
A: About this event, please turn to MTnet for more information.

· Q: Dr. Kang, one key point for successful 3G is a solid industry chain. As a member of TD-SCDMA Alliance, does Spreadtrum has some detailed plan· What role does Spreadtrum wants to play in it ?
A: The target of TD-SCDMA Alliance is to form a perfect industry chain. As a member Spreadtrum is working hard for this target.

· Q: I think it is easily to realize interoperability within domestic vendors. I wonder how about the interoperability between terminal chip and NodeB chip ? Spreadtrum has achieved successful call with DaTang’s NodeB. How about with NodeBs from other suppliers ? Only HuaWei and DaTang as you mentioned ? Who are NodeB chip providers ?
A: DaTang, ZTE, HuaWei and PuTian, all can provide NodeBs. We’ll realize full interoperability with their products.

· Q: In your “China Chip” exhibition , I narrowly recognized one RF IC provider. Is he the only RF provider ? It was reported, COMMIT will also provide RF solution. Has COMMIT succeeded in this ?
A: Yes, Maxium is only one of RF IC supplier. I don’t know whether COMMIT will has RF product. I hope so.

III. Terminal

· Q: DaTang declared they can provide 3G phone with cost under 2000RMB. I wonder if terminal based your TD-SCMDA chip, how about the cost ?
A: Our solution is based on single chip ASIC which can realize the lowest cost. Cost should be lower than 2000RMB.

· Q: Now everyone is talking TD-SCDMA standard and “China Chip”. Current dual-mode terminal is transitional or final commercial version ? How about the market for dual-mode terminal ?
A: Our dual-mode terminal is final commercial version for GSM and TD-SCDMA hybrid network.

IV. TD-SCDMA commercial procedure

· Q: Currently voice call is the mainstream mobile service. Does TD-SCDMA has enough charm to attract operators ? Is there some gap between WCDMA and TD-SCDMA because the latter’s little operational experience ?
A: TD-SCDMA efficiently support asymmetric data service. On the other hand because smart antenna is adopted, it can be used in high density area. This feature also benefit voice service and operator is interested in it.

· Q: I think operators will achieve full coverage and voice service via GSM network. Now there are no specific TD-SCDMA 3G application suppliers. Whether there is a roadmap for TD-SCDMA 3G applications ?
A: 3G applications is not TD-SCDMA specific problem. The other two standards also have it. It need system and terminal suppliers work together to resolve this question.

· Q: Does Spreadtrum also have WCDMA product ? Which one is your preference, W or TD ?
A: Personally I think TD-SCDMA will has more bright future in China.

· Q: 3G will coexists with 2G network in future. Different standard of WCDMA, CDMA2000 and TD-SCDMA also coexist. Does Spreadtrum has plan to develop tri-mode or multi-mode terminal ? What about the status of multi-mode terminal ?
A: SC8800 can support GPRS/GSM/TD-SCDMA. We also think WCDMA&TD-SCDMA dual-mode terminal has nice marketing chance.

· Q: Dr. Kang, according to current condition, there is no mature 3G application in next year running on commercial 3G terminal. If only voice service, can TD-SCDMA alone network be possible ? I think WCDMA & TD-SCMDA hybrid network will harm TD-SCDMA because other two standard have more mature 3G applications, especially for CDMA2000 1x EV DO.
A: 3G application has no relation with radio access methods. I’m sure WCDMA 3G application can also be supported by TD-SCDMA.

· Q: Dr. Kang, what’s the biggest obstacle to block TD-SCDMA commercial procedure ?
A: I think is TD-SCDMA has no enough field testing under full network environment.

· Q: What’s the biggest difficulty for TD-SCDMA industry ? How to resolve it ? What’s your expectation from government support ? Thanks!
A: My own opinion is TD-SCDMA has too limited fund compared to CDMA , WCDMA industry. On the other hand, I admit investment should be from market. If market has enough requirement investors will dig into this industry. For government, our expectation is policy support. I’m sure if market requirement and policy are available investment is everywhere.

· Q: I’m afraid it is difficult to migrate WCDMA 3G application to TD-SCDMA network because of different specification understanding. This is just why there is a bottle-neck problem in terminal interoperability testing. When different TD-SCDMA IC providers face the same application, they will have different understanding. So TD-SCDMA also has interoperability problem. Subscribers can’t enjoy TD-SCDMA application freely unless all the terminals use the same chips from one supplier. Is my understanding is correct ?
A: Firstly thank you for so deep thinking. If DaTang, HuaWei and ZTE can open their system, TD-SCDMA interoperability will not be a big problem. TD-SCDMA can achieve this relatively easily. Any how they all are domestic vendors unlike WCDMA.

· Q: Dr. Kang, how to evaluate TD-SCDMA marketing requirement for the purpose of financing ?
A: I have no financial background knowledge. But I’m sure TD-SCDMA has huge marketing requirement.

· Q: Dr. Kang, I think TD-SCDMA drops a lot behind CDMA2000 and WCDMA. If three 3G licenses will be released in next year, for TD-SCDMA how to deal with this condition ?
A: What TD-SCDMA can only do is to enhance its capability and catch up the other two standards.

· Q: Members of TD-SCDMA Alliance include network vendors, IC suppliers. It seems system software suppliers and ISP,ICP are missing. Does this mean TD-SCDMA will adopt current software platform, such as CDMA2000’s BREW ?
A: Regarding member information, please turn to TD-SCDMA Alliance. I think software platform for TD-SCDMA is in a preliminary phase. I want to clarify that BREW is not a open platform.

· Q: Dr. Kang, compared with WCDMA , TD-SCDMA has more leading features, such as smart antenna, Baton handover etc. More complexity will lead to decrease of stability. Does Spreadtrum meet such stability problems ? Now I’m caring when can I use fully Chinese TD-SCDMA mobile phone ?
A: I think TD-SCDMA terminal has no more technical complexity than WCDMA terminal. As to when you can use fully Chinese TD-SCDMA mobile phone, it depends on when 3G license will be released.

· Q: Dr. Kang, from your understanding, what’s the deficiency for TD-SCDMA technology ? Or what are the technical questions we can’t conquer before commercial launch ?
A: I don’t think there is any deficiency for TD-SCDMA technology. TD-SCDMA must pass field testing before its commercial operation.

· Q: As you mentioned, Spreadrum’s TD-SCDMA chip will be available in next June. We still need to wait for a real commercial mobile phone for another one year. Is my understanding correct ?
A: In next June, not only chip but also communication software are available.

· Q: There is little foreign risky investment in TD-SCDMA industry. To attract foreign investment, what should be done for enterprise and government ?
A: If TD-SCDMA has enough marketing requirement of foreign risky investment surely gives it more support. Currently there already are such investment serving TD-SCDMA, although they are not too much. In fact, both COMMIT and T3G consist of some famous foreign members.

· Q: Dr. Kang, does TD-SCDMA specification has been frozen ? I heard there still are many CRs in WCDMA specification.
A: All the 3GPP specifications are on the improved way.

· Q: Would you please give me a optimistic estimation on TD-SCDMA commercial time schedule ?
A: It should be in next year.

· Q: How TD-SCDMA supports simultaneous video streaming and voice call ? In other words, If protocol layer wants to support simultaneous real-time multiple service, what RTOS should do ?
A: TD-SCDMA can support mixed multiple service, such as data + voice. How to configure RTOS depends on detailed realization method. For example, we can use one RTOS to support TD-SCDMA protocol while another OS processes video stream playing.

· Q: You mean using two RTOS ? If use multiple RTOS, multiple CPUs have to be used. If traffic is big, how to realize real time communication between each CPU ?
A: Just like WCDMA, our protocol stack fully complies with TD-SCDMA specification. If we support real time service of 384kbps, this traffic is not so much for communication between two CPUs.

· Q: Does TD-SCDMA commercial procedure will be affected for continuous 3GPP standard evolution ?
A: No. Changes of standard version will not affect TD-SCDMA commercial time schedule.

· Q: Dr. Kang, The difference between LCR and WCDMA is only air interface and L3 signaling. Is it true ?
A: LCR and WCDMA has the same L3 and different air interface.

· Q: Dr. Kang, my job is focusing on RTOS. What’s requirement for RTOS if I want to support TD-SCDMA protocol stack and 3G applications(such as necessary functions, technical data etc.) ?
A: If you want to support protocol stack, RTOS needs good real-time capability. Simple RTOS can’t support complex 3G applications. It is not easily for RTOS to support both TD-SCDMA protocol and 3G applications. The key point is when RTOS processes 3G applications, it should not lose its real-time ability.

· Q: Is it possible for TD-SCDMA to cover full national area in GSM+TD-SCDMA scenario ? I think TD-SCDMA technology can’t meet this requirement although it has been improved somewhere.
A: I think it is very possible. Why do you think TD-SCDMA can’t provide large area coverage· Smart antenna can concentrate power to a certain direction which fit well rural coverage usage.

· Q: It is difficult for current multi-mode chip to be changed to single-mode. Is it possible for pure TD-SCDMA network to come into use (even if it gets own 3G license) ?
A: To subtract functions is very easily.

Read more...