As the world’s appetite for high-speed wireless communication surges, the challenge is no longer simply building networks, it is learning to use spectrum intelligently. Nigerian telecommunications expert Odinaka-Olisa James Okonkwo, a senior officer at Phase3 Telecom Limited, Abuja, co-authored an important early study into the Recursive One-Sided Hypothesis Testing (ROHT) technique, an under-explored algorithm with major implications for future spectrum efficiency. Today, with global networks strained by unprecedented demand, his research is strikingly timely. In this interview, he discusses the insights his work revealed and why cognitive radio may be the key to the next generation of adaptive, reliable connectivity.

You presented research on the performance of the ROHT technique in cognitive radio. For a general audience, what problem were you trying to solve?

Every device that connects to the internet; your phone, your laptop, even IoT sensors, depends on access to the radio spectrum. The problem is that the spectrum is getting crowded. Cognitive radio was proposed as a solution. It allows devices to sense which parts of the spectrum are free and use them without interfering with licensed users.

But cognitive radio only works if devices can detect signals reliably, especially in very noisy environments. Our research looked at how one of the key detection algorithms, the ROHT technique, behaves under different signal-to-noise conditions. Essentially: How low can the signal be before the system starts failing?

Why is this important now, especially for countries like Nigeria?

Nigeria is moving toward large-scale digital transformation; smart agriculture, online education, fintech, remote healthcare. All of these rely on wireless connectivity. But building new spectrum infrastructure is expensive.

Cognitive radio offers a way to unlock unused frequencies and expand internet access without building new towers or buying new spectrum. It’s incredibly important for rural communities and growing cities.

So understanding the limits of detection algorithms helps us design more robust systems for our environment.

What exactly is the ROHT algorithm, in simple terms?

Imagine you’re in a crowded room trying to hear someone whisper. The ROHT algorithm behaves like a smart listener. It continuously studies the noise around it, recalculates what “normal noise” looks like, and then tries to detect if there is an actual signal hidden in that noise.

It adapts its threshold dynamically, meaning it doesn’t use the same standard for quiet and noisy environments. That adaptability is what makes it attractive for cognitive radio.

And what did your research discover about its performance?

We discovered that the ROHT algorithm performs very well when the signal is reasonably strong; say, at 10 dB or 5 dB signal-to-noise ratio. But once the signal drops below 3 dB, the algorithm struggles to detect it reliably.

In simple terms: below a certain signal level, ROHT can’t tell the difference between noise and an actual transmission. That has implications for real-world deployment, because in many wireless environments in urban areas, busy networks often operate in low signal-to-noise conditions.

What surprised you most about the results?

We expected performance to decline gradually as the signal weakened. But what we saw instead was a sort of “cliff.” The algorithm was very stable at higher SNR levels, then suddenly at 1 dB, it almost entirely failed to detect the signal even though it maintained a good false-alarm rate.

That means it doesn’t misinterpret noise as a signal, but it can miss signals altogether. This finding is important because it identifies the exact point where ROHT stops being reliable.

How can your findings be used today?

 There are two key uses: 

1.Guiding system design. Engineers now know that ROHT-based energy detectors must be optimized if they’re expected to operate in very noisy environments.

2. Inspiring more robust algorithms. Our study showed where the limitations are; now researchers can focus on improving performance below 3 dB using adaptive optimization or hybrid detection methods.

This kind of foundational research helps prevent expensive deployment mistakes in real networks.

You’ve since moved into data analytics, AI, and intelligent systems. How does this early work connect to what you’re doing now?

It taught me two important things. One, systems don’t fail randomly, they fail predictably, and you can model those limits. Two, Intelligence isn’t just about data; it’s about how systems respond to uncertainty.

Those exact principles define the future of engineering and data science. Whether it’s telecom networks, energy grids or healthcare systems, the goal is the same: predict, adapt, and optimize.

That early research shaped my interest in intelligent infrastructure, algorithmic optimization and real-time decision systems.

What do you think the future of cognitive radio looks like for Nigeria and Africa?

The future is bright. Africa has the advantage of not being locked into legacy systems. If we adopt intelligent spectrum management supported by cognitive radio, we can massively expand broadband access without massive capital expenditure.

It’s also crucial for emerging technologies: drones, remote farming sensors, telemedicine, and smart logistics. All of these need a flexible spectrum.

Ultimately, cognitive radio gives us a way to democratize connectivity, especially in underserved communities.

What excites you most about where this research could go?

I’m excited about merging cognitive radio with machine learning. Algorithms that can learn from the environment, adjust their thresholds in real time and operate in extremely low-SNR conditions, that’s the future.

If we can get there, we unlock reliable, low-cost, wide-coverage internet access for millions. That’s the real impact.

Final thoughts for policymakers or industry leaders reading this?

Invest in spectrum research, support local talent and encourage innovation in wireless technology. The next generation of digital infrastructure will not be built by hardware alone, but by intelligent systems that understand their environment.

If Nigeria wants to lead in digital transformation, spectrum intelligence must be a national priority.