Integrated Sensing and Communications (ISAC)
Traditional wireless systems face a fundamental limitation, they are designed for communication. Adding sensing capabilities typically adds overhead like dedicating spectrum resources to special sensing signals, implementing separate processing chains, and coordinating complex time-frequency resource splits between communication and sensing functions. The result is compromised performance for both.
Cohere’s Pulsone Technology, powered by the Zak-OTFS waveform, solves this problem via native operation in the Delay-Doppler domain, the same domain used by radar systems. Unlike OFDM, which operates in time-frequency space and requires overhead to add sensing, Zak-OTFS inherently processes signals in the radar domain. The Delay dimension captures multipath and distance information while the Doppler dimension captures velocity and motion—exactly what radar systems need.
This creates “data-derived sensing” where communication signals themselves provide sensing capability with zero spectral efficiency overhead. The same waveform simultaneously delivers communications and environmental awareness.
Demonstrated results prove the advantage. Cohere’s real-time neural receiver implementation—showcased at NVIDIA GTC Government Conference in October 2025—achieved 4x better resolution in target detection, tracked 4x more simultaneous targets, detected objects one-quarter the size of OFDM-based systems, and maintained superior link budget characteristics. All while using communication data signals without any dedicated sensing overhead.
“Last fall at the GPU Technology Conference (GTC) DC, we worked jointly with Cohere and Duke to demonstrate the industry’s first online, real-time neural receiver for integrated sensing and communications (ISAC) on NVIDIA’s Jetson AGX Orin to show what’s possible with Zak-OTFS,” said Dr. Lingjia Liu, IEEE Fellow, Andrew J. Young Professor of Electrical and Computer Engineering and Co-Director of Wireless at Virginia Tech. “The computational complexity of the introduced neural receiver for Zak-OTFS-based ISAC is extremely low, without the need for offline training to enable real-time operation. Together, we can show the industry the ability for AI/ML to work seamlessly across waveforms, protocols, and redefine what is possible for next-generation wireless networks.”
The system converges within a single symbol time, adapts in real-time to changing conditions, and requires no offline training datasets. This makes Zak-OTFS the only waveform that provides integrated sensing and communications with zero capacity sacrifice—transforming ISAC from a theoretical concept requiring painful compromises into a practical capability that enhances both functions simultaneously.
Critical Applications Driving ISAC Adoption
Integrated Sensing and Communications represents a paradigm shift in wireless network architecture. Rather than building separate systems for data transmission and environmental awareness, ISAC unifies these functions into a single platform. This convergence matters because the spectrum, hardware, and infrastructure costs of deploying dual systems are becoming unsustainable as networks densify and applications demand both connectivity and situational awareness.
ISAC implementation on OFDM waveforms certainly works, the fundamental physics of using communication signals for sensing applies regardless of modulation scheme. However, mission-first communications for national defense, homeland security, and critical infrastructure protection cannot accept “workable” when superior alternatives exist. When metropolitan areas generate millions of potential sensing events per minute—aircraft transits, UAV movements, ground vehicle traffic, maritime activity, the challenge of correlating these events into coherent tracks and actionable intelligence remains identical whether processing OFDM or OTFS returns. Advanced signal processing, AI/ML algorithms, and track correlation engines consume the same computational resources regardless of input waveform. The concern shifts from processing capacity to detection sensitivity and resolution at the sensing layer itself.
Mission-First Sensing Demands Maximizing what Enters the Processing Pipeline
If downstream computational costs remain constant across waveforms, operational priority dictates deploying the waveform with highest detection sensitivity and finest resolution. OTFS’s demonstrated 4x improvement in target resolution, 4x increase in simultaneous target capacity, and ability to detect objects one-quarter the size of OFDM-based systems directly translates to detecting threats earlier, tracking more targets simultaneously, and identifying smaller objects than an OFDM-based ISAC would miss entirely. For defense applications, where detecting a small UAV five minutes earlier or tracking twice as many threats simultaneously can mean mission success versus failure, choosing the inferior sensing waveform to preserve OFDM ecosystem compatibility represents an unacceptable compromise. National security applications cannot optimize for backward compatibility when threat detection capability hangs in the balance.
Critical infrastructure monitoring becomes viable at scale through ISAC. Airport perimeter security can detect intrusions while maintaining communications across the facility. Utilities can monitor transmission lines for ice loading or vegetation encroachment while coordinating grid operations. Maritime applications combine vessel tracking with ship-to-shore communications.