8 Integrated Sensing and Communication for IoT

8.1 Start With the Story

Radio Remi is helping a factory send control messages across a busy hall. The same radio waves also bounce from a moving trolley. The team could use those echoes to add a warning zone, but sensing symbols consume time, bandwidth, power, and compute. Remi asks for one schedule that keeps the control link useful while producing a modest range update. The team must show both results, not hide one behind the other.

8.2 Overview

Integrated sensing and communication, or ISAC, reuses parts of a radio system for both data exchange and environmental measurement. Sharing may include waveform, spectrum, antennas, clocks, compute, or scheduling; it does not mean every communication packet becomes a complete radar. Known OFDM symbols can expose a channel estimate. Transforming that estimate across subcarriers produces delay or range structure, while changes across repeated symbols can expose Doppler. The design still needs calibration, geometry, clutter handling, privacy controls, and a communications service-level target.

8.3 Learning Objectives

By the end of this chapter, you will be able to:

• Explain how a known OFDM transmission can also support range and motion estimation.
• Compare communication throughput with sensing bandwidth, update rate, and ambiguity.
• Distinguish research and standards work from sensing capabilities available to an IoT designer today.

8.4 Reuse an OFDM frame

An OFDM transmitter places known or decodable symbols $X[k]$ on many subcarriers. The receiver observes $Y[k]$. Where $X[k]$ is known and non-zero, the channel estimate is $H[k]=Y[k]/X[k]$. An inverse Fourier transform across subcarriers turns frequency variation into a delay profile. A calibrated delay can map to range, although clock offsets and one-way versus round-trip geometry must be handled explicitly.

Several OFDM symbols add a slow-time axis. A target moving toward or away from the radio changes phase across that axis, which can support a Doppler estimate. Wider occupied bandwidth improves the ability to separate nearby delays; longer coherent observation improves Doppler resolution but slows the result and demands more stable timing. Data symbols, pilots, guard intervals, retransmissions, beams, and sensing bursts all compete for finite resources.

8.5 Read the shared resource grid

Figure 8.1 shows how one OFDM frame can carry both payload and sensing evidence without pretending the resources are free.

In Figure 8.1, begin with the time-frequency grid. Blue cells carry payload, teal cells are known pilots, and amber cells reserve a denser sensing burst. The middle plot transforms the estimated channel into two delay peaks; close echoes could merge if their delay bins are too wide. The final curve is a design boundary, not a universal law: more sensing time improves update opportunities while leaving fewer resource elements for payload. A chosen operating point must report both link and sensing metrics.

8.6 Balance resolution, update rate, and throughput

For a simple monostatic range model, delay resolution is roughly $\Delta R\approx c/(2B)$, where $B$ is usable sensing bandwidth. More bandwidth can separate closer reflectors, but spectrum availability, radio front ends, channel occupancy, and regulations constrain it. More repeated symbols can improve Doppler estimation, yet they occupy airtime and may increase latency or energy use. Power aimed at stronger echoes is not automatically power available for reliable communication at every receiver.

Build a Pareto view instead of one combined score. Report payload throughput, packet delay, packet loss, range-bin spacing, update interval, false alarms, missed detections, energy, and compute load separately. A point is dominated when another schedule is no worse on every required metric and better on at least one. The application chooses among the remaining points using a stated priority, such as a factory exclusion-zone update deadline.

8.7 Understand current and future use

An IoT designer today can use Wi-Fi CSI from supported hardware, communication-channel measurements, positioning features, or a dedicated radar beside the network. Those are practical building blocks, but their APIs and evidence differ by chipset and deployment. 3GPP Release 19 includes a study on integrated sensing and communication, while the ITU IMT-2030 framework names ISAC as a future usage scenario. A study, framework, prototype, or vendor demonstration is not the same as an interoperable sensing service available on every 5G-Advanced network.

Candidate use cases include drone or object detection, factory monitoring, transport safety, mapping, and occupancy. Each one needs its own range, field of view, update rate, false-alarm cost, and privacy boundary. Sensing Privacy and Consent applies because a shared waveform can still produce observations about people and spaces.

8.8 Decision and Trade-offs

Reuse a communication waveform when shared hardware and spectrum materially reduce cost or latency and when the network can expose the required measurements. Keep dedicated radar when sensing availability, geometry, calibration, or safety assurance must remain independent of packet load. A hybrid system may schedule quiet sensing intervals or fuse both paths, but it must show which subsystem owns each decision and failure.

8.9 Practice the Method

The linked JupyterLite lab creates a known OFDM frame, applies a synthetic two-path channel, estimates $H[k]$, and transforms it into a range profile. You will change occupied bandwidth and the share of symbols reserved for sensing, then compare range-bin spacing with remaining payload throughput. The result card records one justified operating point and one rejected alternative.

8.10 Check Your Reasoning

8.11 Summary

ISAC shares radio resources, but it does not erase engineering trade-offs. Known OFDM symbols support a channel estimate, frequency structure supports a delay profile, and repeated symbols can support Doppler. Every design must state what is shared, what is calibrated, and how communication and sensing outcomes were measured together.

• $H[k]=Y[k]/X[k]$ is useful only where the transmitted symbol is known and non-zero.
• Bandwidth, coherent time, airtime, power, and compute connect sensing quality to communications cost.
• Current components and research roadmaps must not be presented as universal deployed 5G-A or 6G capability.

8.12 Sources and Boundaries

• 3GPP TR 22.837 records the Release 19 study on integrated sensing and communication.
• The ITU IMT-2030 programme identifies ISAC among future IMT usage scenarios and capabilities.