Rebirth: Starting from Lighting Up the Tech Tree

Chapter 51 Review

On March 15th, Han Lu passed her thesis defense.

She came to the company to complete her onboarding procedures the day after graduating from Huada Business School—her official position: Business Director of 402 Technology. Zuo Cheng offered her a monthly salary of 12,000 yuan, second only to himself in 402's salary system.

"Isn't that expensive?" Zhang Lei asked Zuo Cheng privately. "She just graduated. Twelve thousand yuan could hire a business manager with three years of experience in Huaxia City."

"Do you think a business manager with three years of experience could have stood up and read out an 'or clause' at last year's bidding qualification review meeting?" Zuo Cheng asked rhetorically.

Zhang Lei thought for a moment and then remained silent.

The first thing Han Lu did after joining the company was to review all the terms of the contracts currently in effect for Project 402—including the Sky Dome Phase II, LoRa Smart Agriculture, and the after-sales obligations for the three completed projects. She made a list, specifying the acceptance standards and liability clauses for each deliverable, and then posted the list on the office wall.

"From now on, everyone will be able to see the contract risks for every project," she told Zuo Cheng. "You're responsible for the technical aspects, but if any business issues arise, come to me."

Zuo Cheng glanced at the table. The densely packed words were written neatly, and the table lines were drawn by hand with a ruler—Han Lu didn't like using electronic versions on the wall, saying, "Paper has a stronger visual impact, forcing people to look at it."

Lu Mingyuan arrived on March 20th.

It wasn't a conference call; he came in person.

The Sky Dome Business Unit has scheduled mid-quarter reviews for all partners in the second phase, with 402 being the first. Lu Mingyuan brought two technical review experts and notified Zuo Cheng a day in advance to prepare a technical progress report.

Zuo Cheng spent one night summarizing the development progress of the past forty-five days—the architecture design of the multi-star pipeline scheduler was 85% complete, the adaptive parameter sharing engine was 70% complete, the beam cooperative controller was 90% complete (Tang Xu's hierarchical accuracy scheme has been verified by simulation), and the spectrum sensing front-end was 60% complete.

The average progress of the four core modules is 76%. Half of the ninety days have passed, and we are ahead of schedule.

But Zuo Cheng knew that 60 percent of the spectrum sensing front end was a hidden danger.

This module is responsible for sensing the electromagnetic environment around the terminal in real time, identifying available spectrum resources, and avoiding interference. The technical challenge lies in the fact that the spectrum environment of low-Earth orbit satellites is dynamic. In the few minutes that a satellite passes overhead, Doppler shift, atmospheric attenuation, and interference from adjacent beams all change rapidly. Traditional spectrum sensing algorithms are based on static models and cannot adapt to this highly dynamic scenario.

Zuo Cheng has been thinking about this problem for two weeks and has tried three solutions, but none of them have been effective.

Upon arrival, Lu Mingyuan skipped the pleasantries and went straight into the technical review.

After reviewing the progress reports of the four modules, he rated the first three modules as "meeting expectations." But when he got to the spectrum sensing front end, he stopped and looked at it for a long time.

"Sixty percent." He looked up at Zuo Cheng. "The other three modules are all above seventy, this one is dragging us down. What's the reason?"

"The optimal solution for real-time sensing algorithms in high-dynamic spectrum environments has not yet been found," Zuo Cheng stated frankly. "Traditional methods only achieve a 73% accuracy rate in spectrum identification during satellite transits, which falls short of our target of 90%."

A reviewer next to Lu Mingyuan flipped through the proposal documents: "Have you considered using compressed sensing? Under the assumption of spectral sparsity, compressed sensing can achieve fast spectral reconstruction with fewer sampling points."

"We've tried it," Zuo Cheng said. "Compressed sensing works very well in static scenarios, but the assumption of spectral sparsity for low-Earth orbit satellites doesn't hold true—when multiple satellites are in the field of view simultaneously, the spectrum occupancy is dense, not sparse. We've tested it; with 120 satellites running in the entire network, the peak spectral occupancy rate within the terminal's visible arc can reach over 60%, far exceeding the applicable range of sparsity."

The expert glanced at the data in his hand, nodded, and didn't ask any further questions.

Lu Mingyuan looked at Zuo Cheng, his tone calm but serious: "Zuo Cheng, spectrum sensing is the eyes of the entire platform. Without it, pipeline scheduling and beam coordination are like blind men directing traffic. I suggest you raise the priority of this module to the highest level."

"clear."

Before leaving after the review, Lu Mingyuan privately told Zuo Cheng, "I'm very satisfied with the overall progress. The spectrum sensing issue needs to be resolved as soon as possible—the technical committee will be holding a mid-term report meeting next month, and Zhou will attend."

Zhou Henian is about to hear another report. This means that the issue of spectrum sensing is not just a technical challenge, but a crucial test of whether 402 can continue to gain trust within the Sky Dome system.

After seeing Lu Mingyuan off, Zuo Cheng returned to his workstation and stared at the architecture diagram of the spectrum sensing front end for a long time.

High dynamism, non-sparseness, and real-time performance—these three constraints coexist, effectively eliminating most traditional solutions.

He needs a new approach.

That evening, he called Yu Ying, not to talk about feelings, but to talk about technology.

"Kongkong, in your research on low-Earth orbit satellite anti-jamming, does it involve dynamic spectrum sensing?"

Yu Ying paused for two seconds: "Yes. One research direction my supervisor recently asked me to explore is using cyclic stationary feature detection for satellite signal spectrum sensing. This method doesn't rely on the assumption of spectral sparsity, but rather utilizes the inherent periodic stationary characteristics of communication signals to distinguish between signals and noise."

Zuo Cheng's attention was instantly drawn.

"Cyclic stationary feature detection," he repeated. "How does this method perform in real-time in highly dynamic scenes?"

"Theoretically, it's possible to achieve millisecond-level performance, but the computational load is quite large. I've read several papers, and they say it can't run in real-time on general-purpose processors." Yu Ying paused for a moment, "But if you're doing hardware acceleration on an embedded platform—you understand that better than I do."

Zuo Cheng didn't speak. His mind was rapidly connecting two things—the cyclic stationary feature detection that Yu Ying had mentioned, and his own "embedded intelligent channel processing" fusion blade.

One is the algorithm, and the other is the engineering implementation capability.

If the core algorithm of cyclic stationary detection is extracted and customized for the hardware characteristics of embedded platforms—such as pipelined parallelism, fixed-point arithmetic, and lookup tables to replace complex mathematical operations—it is entirely possible to complete real-time spectrum sensing in milliseconds.

"Kongkong, do you have the code for a cycle stationarity detection algorithm?"

"I have some simulation code from my advisor's research group, written in MATLAB. I can send it to you for reference, but you can't use it directly due to issues of academic ownership."

"I don't need the code itself, just an understanding of the algorithm's mathematical core. I'll write the engineering implementation myself."

"No problem. I'll send it to you tonight."

After hanging up the phone, Zuo Cheng leaned back in his chair and closed his eyes.

The breakthrough in spectrum sensing may lie here.

The tech tree trembled slightly deep within his consciousness—not as a mission prompt, but as a more subtle signal, like leaves gently swaying in the wind.

Whenever he stands on the verge of a technological breakthrough, the tech tree reacts in this way.

He opened his eyes and glanced at the time. It was 11 p.m.

We'll start working on this tomorrow.