Those Years When I Was Forced to Top the Forbes List

Chapter 262 Technical Reconciliation (A purely transitional chapter, you can skip it)

The small conference room on the 7th floor was stuffy and windowless. The moment the light was turned on, the air conditioner turned on at full blast with a beep.

Nearly twenty people entered, filling the room to capacity. They stood around the circular conference table. Zhang Hao spread the documents and laptop he had brought on the table and, before the other party could speak, hurriedly said:

"We need your channel switching algorithm!"

The atmosphere froze for a moment, and Yu Chengdong, who was arranging for someone to pour water, also turned to look at Zhang Hao.

Chen Xuebing, who hadn't originally planned to participate, walked from the doorway to the table upon seeing this and said with a smile:
"I have already discussed the conditions with Mr. Ren. He said he can fully cooperate. For all the research projects proposed, Huawei, Spreadtrum, and Chongqing University of Posts and Telecommunications will mutually authorize the final research results."

"This involves each other's technologies and patents. Without this license, no one can use the final technology, right?"

"Today, we are all project team leaders. Mr. Wu from Spreadtrum and Mr. Li from Chongqing University of Posts and Telecommunications are also here. They are all professional technicians. The one who spoke just now is Dr. Zhang Hao, who returned from the United States. Tech people are always straightforward. Since Mr. Ren is not here, I will preside over the meeting. Is that alright?"

"With so many people, let's skip the introductions and just ask each other questions in an orderly fashion. Whoever is in charge of a project should answer the question. Let's make the most of our time and get things done quickly."

Upon hearing this, Yu Chengdong laughed: "President Chen, you're making it sound like we're unprofessional. Fine, then describe your TD module first. Does Datang Card work on the physical layer or the protocol stack?"

Wu Ping immediately laid out a chip architecture diagram with L2 (L1-L4) security level, which showed some standardized interfaces.

"At the physical layer, we have changed the way solid-state antennas are used, adopting analog 8-antenna beamforming and using Spreadtrum chips for digital pre-equalization to compensate for beam distortion, thus circumventing their calibration patents."

"As for the protocol stack, we and Chongqing University of Posts and Telecommunications are different. Chongqing University of Posts and Telecommunications develops its own protocol stack. They have no experience in GSM and are a little behind in the dual-mode field. However, they optimize HSDPA efficiency through dynamic time slot allocation, with a measured peak of 2.8Mbps, and circumvent Datang's fixed pilot synchronization patent."

"Our solution is an independently designed alternative based on reverse engineering, with several optimizations."

"However, neither of us can bypass their switching technology."

Basic patents are not so easy to bypass. Spreadtrum managed to bypass the physical layer, but it couldn't bypass the communication protocol.

After listening, Yu Chengdong didn't give a direct answer. Instead, he rested his chin on his hand and asked, "Isn't your self-designed system more likely to infringe on patents? Or are you using Chongqing University of Posts and Telecommunications' protocol stack as a base?"

“It still works,” Zhang Hao said.

"The RF-baseband joint debugging interface provided by Spreadtrum can help Chongqing University of Posts and Telecommunications (CUP) dynamically adjust equalizer parameters to adapt to the calibrated channel characteristics, resulting in a 20% reduction in the measured bit error rate."

"Spreadtrum's hardware-accelerated HARQ engine reduces retransmission processing latency from 5ms to 2ms. Combined with the adaptive scheduling algorithm of Chongqing University of Posts and Telecommunications' protocol stack, the actual throughput of HSDPA is increased to 85% of the theoretical peak."

Furthermore, the joint work of Spreadtrum's Time-Domain Equalizer (TEQ) and Chongqing University of Posts and Telecommunications' Midamble channel estimation can improve the multipath delay extension tolerance from 4μs to 8μs in dense urban scenarios.

Dr. Zhang's words, based on his understanding of the protocol stack technologies of both parties, especially the 2ms latency of the hard acceleration engine and the I/Q calibration technology, are all parts that laid the foundation for their independent development.

The Huawei staff also glanced at them slightly.

The conversation turned.

"In addition, Chongqing University of Posts and Telecommunications' protocol stack HSDPA uses dynamic pilots and non-standard Gray coding to circumvent Datang's fixed pilot patent, which requires an additional 3dB of signal-to-noise ratio for 16QAM demodulation. EVM is controlled at 8% through I/Q imbalance calibration of Spreadtrum RF chip... However, due to constellation diagram rotation to circumvent the patent, the actual peak rate is about 30% lower than the theoretical value."

It's still a protocol stack.

The protocol stack of Chongqing University of Posts and Telecommunications has been modified. Although it avoids patents by using non-standard coding, it also requires a higher signal-to-noise ratio for demodulation.

This problem is a major headache even for Zhang Hao.

Yu Chengdong then turned to the person next to him: "Shunmao, can you solve this problem?"

"Yes, our base stations support dynamic pilot density adjustment. By adding or removing pilot symbols according to channel conditions, we should be able to reduce spectral efficiency loss while bypassing constellation rotation technology. Using an adaptive pilot interpolation algorithm, we can compress the additional 3dB signal-to-noise ratio required for 16QAM demodulation to 1.5dB."

Zhang Shunmao, the base station system architect, said with a smile, "Actually, we have also applied a real-time I/Q imbalance compensation algorithm in the base station radio frequency unit. Our NodeB dynamically increases or decreases the pilot density from 1/4 to 1/8 based on channel coherence time prediction. Combined with the SVD interpolation algorithm to reduce signal-to-noise ratio loss, the EVM can be controlled below 3%, which can support 64QAM modulation and improve the terminal throughput by 15%."

EVM（实际信号和理想信号的误差幅度）8%，3%，高下立现。    周光平咝了一声：“EVM 3%的话，64QAM的峰值速率应该能从2.8mbps提到3.2mbps了吧？”

Yu Chengdong smiled and nodded, then looked at Chen Xuebing: "President Chen, it seems we are the more professional ones."

Chen Xuebing keeps his eyes down and his mind at ease; when he needs something from someone, he remains silent.

He knew why Ren Zhengfei had transferred this person from Europe to work with them.

Yu Chengdong, according to the path he took in his previous life, was supposed to be transferred to the terminal business department as president. He was probably sent here to cultivate a good relationship with me in advance.

But this guy's mouth just can't stay still.

You think you're so tough? I'll make things difficult for you someday.

Seeing that the "powerful patron" Ren had mentioned had fallen silent, Yu Chengdong chuckled dryly and began to bring up the most valuable part:

"Mr. Xu, what's your opinion on the patent switching issue?"

Huawei's management style is modeled after IBM's, which means there are many presidents. The young Mr. Xu, who was from some department of the wireless product line, said: "Our distributed base station handover algorithm has passed the test of the Portuguese operator. However, Mr. Ren instructed that our Indian research institute transmits English technical documents through a VPN channel. You should also set up your servers overseas to avoid Datang finding evidence."

"Then let's put the server in Singapore! If Datang wants to investigate, let them deal with the Singapore courts first!"

After Chen Xuebing finished speaking in a cold voice, he walked outside with his hands behind his back.

...

The technical reconciliation lasted all night.

Chen Xuebing took a short nap in the office next door for five or six hours.

It was late at night when he woke up and found that there was a lot of noise outside, and the people seemed to be getting more and more energetic as they talked.

"Two carriers? TD's time slot structure is far more complex than WCDMA's. Multi-carrier synchronization is a huge problem! We have scheduling algorithms that you can't use!"

"Don't you have joint time slot alignment technology? The base station embeds inter-carrier time difference compensation information in the downlink synchronization time slot, and the terminal uses this for precise synchronization."

"Spreadtrum's next-generation chips can also integrate multi-carrier digital front-ends and support dynamic carrier switching!"

"Actually, TD-SCDMA's biggest advantage is its smart antenna. Datang's patents are mainly focused on fixed beamforming, but if we do adaptive beam tracking..."

"Since Huawei is already conducting preliminary research on 8-antenna MIMO, if it can be combined with dynamic beamforming, it can theoretically improve the spectrum efficiency by 3 times."

"The number of antennas on the terminal side is limited, and 8-antenna MIMO is not practical for mobile phones."

"It doesn't necessarily have to be 8 antennas. We can do virtual MIMO! The base station uses multiple antennas, and the terminal uses 2 antennas. We can improve the gain through space-time coding."

"Chongqing University of Posts and Telecommunications' protocol stack can optimize the uplink sounding reference signal, allowing the base station to estimate the channel state more accurately."

"Wait, if we combine multi-carrier, MIMO, and dynamic beamforming... wouldn't that be considered the 3GPP LTE 4G standard?" Wu Ping's voice came through.

There was a two-second silence inside.

“That’s right… According to the research roadmap proposed by 3GPP in 2004, if we start accumulating these technologies now, we will definitely be able to gain some initiative in the future. However, the technical roadmap is not yet clear. If we go and do this preliminary research, we may suffer great losses.”

Upon hearing about 4G, Chen Xuebing perked up and rushed into the conference room.

(End of this chapter)