Communication and Sensing on One Waveform

Integrated Sensing and Communication (ISAC) — also called Joint Radar-Communication (JRC) — uses a single transmitted waveform to simultaneously deliver data to users and sense the physical environment (detect objects, measure range, estimate velocity). This dual-function approach eliminates the need for separate radar hardware, reduces spectrum usage, and enables new applications that neither system could achieve alone.

The ITU-R IMT-2030 framework (successor to IMT-2020/5G) formally includes sensing as a new capability pillar alongside communication, per ITU-R M.2160. 3GPP launched a Release 19 Study Item on ISAC (documented in RP-234069, "Study on Integrated Sensing and Communication") to evaluate sensing use cases, performance requirements, and air interface impacts.

OFDM Radar Fundamentals

NR already uses OFDM (Orthogonal Frequency-Division Multiplexing) as its downlink and uplink waveform. OFDM has a natural dual interpretation as a radar signal:

• Range information comes from the subcarrier domain (frequency)
• Velocity information comes from the OFDM symbol domain (time/Doppler)

Range Estimation

An OFDM waveform with N subcarriers at spacing Δf has a total bandwidth B = N · Δf. The range resolution is:

`

Δr = c / (2B)

`

where c = 3 × 10⁸ m/s. The maximum unambiguous range depends on the subcarrier spacing:

`

r_max = c / (2·Δf)

`

Velocity Estimation

Velocity is extracted from the Doppler phase shift across consecutive OFDM symbols. With M OFDM symbols over a coherent processing interval (CPI) of duration T_CPI = M · T_sym:

`

Δv = λ / (2·T_CPI)

v_max = λ / (4·T_sym)

`

where λ is the carrier wavelength and T_sym is the OFDM symbol duration (including cyclic prefix).

ISAC Parameter Summary

Sensing Topologies

ISAC systems operate in three distinct topologies, each with different capabilities and complexity:

Monostatic sensing (radar-like) provides the simplest geometry for range/velocity estimation but requires full-duplex or careful time-division to separate Tx and Rx at the same node. The self-interference cancellation requirement (typically >100 dB isolation) is a major engineering challenge addressed in TR 38.854 (study on full-duplex for NR). Bistatic sensing avoids self-interference entirely. The gNB transmits standard NR downlink, and a separate receiver (another gNB, a dedicated sensor, or even the UE itself) captures reflections from objects in the environment. The challenge is tight time and frequency synchronization between Tx and Rx nodes — typically within 1 ns and 1 Hz for sub-meter range accuracy.

Worked Example 1: ISAC Resolution at 28 GHz (FR2)

Calculate range and velocity resolution for a 28 GHz NR carrier with 400 MHz bandwidth, 120 kHz SCS, and a CPI of 64 OFDM symbols:

`

Range resolution:

Δr = c / (2·B) = 3×10⁸ / (2 × 400×10⁶) = 0.375 m

Maximum unambiguous range:

r_max = c / (2·Δf) = 3×10⁸ / (2 × 120×10³) = 1,250 m

Wavelength at 28 GHz:

λ = c/f = 3×10⁸ / 28×10⁹ = 10.71 mm

OFDM symbol duration (120 kHz SCS, normal CP):

T_sym = 1/(120×10³) + T_CP = 8.33 µs + 0.59 µs ≈ 8.92 µs

CPI duration:

T_CPI = 64 × 8.92 µs = 571 µs

Velocity resolution:

Δv = λ / (2·T_CPI) = 10.71×10⁻³ / (2 × 571×10⁻⁶) = 9.38 m/s = 33.8 km/h

Maximum unambiguous velocity:

v_max = λ / (4·T_sym) = 10.71×10⁻³ / (4 × 8.92×10⁻⁶) = 300.2 m/s = 1,081 km/h

`

The 0.375 m range resolution is sufficient for automotive collision avoidance. Velocity resolution of 33.8 km/h can be improved by extending the CPI (more symbols), at the cost of reduced maximum detectable acceleration.

Worked Example 2: Sensing at 3.5 GHz (C-Band) vs 28 GHz

Compare ISAC performance between C-band (3.5 GHz, 100 MHz BW, 30 kHz SCS) and FR2 (28 GHz, 400 MHz BW, 120 kHz SCS):

`

C-Band (3.5 GHz) FR2 (28 GHz)

Bandwidth 100 MHz 400 MHz

Range resolution (Δr) 1.5 m 0.375 m

Max range (r_max) 5,000 m 1,250 m

Wavelength (λ) 85.7 mm 10.71 mm

Symbol duration 33.33 µs + CP 8.33 µs + CP

≈ 35.7 µs ≈ 8.92 µs

Velocity resolution (64 symbols CPI)

T_CPI 2.285 ms 0.571 ms

Δv 18.7 m/s 9.38 m/s

Δv (km/h) 67.4 km/h 33.8 km/h

Max velocity (v_max) 600 m/s 300 m/s

`

C-band sensing suits wide-area outdoor applications (traffic monitoring, weather). FR2 sensing suits precision short-range applications (V2X, gesture recognition, indoor positioning).

3GPP Release 19 ISAC Study

The 3GPP Release 19 study item (SI) on ISAC, documented in RP-234069 and progressing through RAN1 and RAN4 working groups, addresses:

Sensing Use Cases and Requirements

Air Interface Impact

The study evaluates whether existing NR waveforms and reference signals can support sensing without specification changes. Key findings:

• DL RS reuse: PRS (Positioning Reference Signal, per TS 38.211 Section 7.4.1.7) and CSI-RS can serve as sensing pilot signals
• Dedicated sensing RS: new reference signals may be needed for monostatic sensing with specific ambiguity function properties
• Duplex requirements: monostatic sensing at the gNB requires either full-duplex (per TR 38.854) or receive-only sensing gaps within the TDD frame
• Interference management: sensing signals from neighboring cells can create interference; coordination through Xn signaling is under study

Waveform Design Tradeoffs

The core ISAC waveform design problem is balancing communication optimality (maximize capacity) against sensing optimality (maximize detection probability and resolution):

Practical ISAC systems use hybrid approaches: standard NR data waveforms for communication, supplemented by sensing-optimized pilot sequences (e.g., Zadoff-Chu sequences with low autocorrelation sidelobes) inserted in dedicated OFDM symbols.

Real-World Research and Demonstrations

Qualcomm — Wi-Fi Sensing

While not cellular ISAC, Qualcomm's Wi-Fi Sensing technology (based on IEEE 802.11bf) demonstrates the ISAC principle at scale. Deployed in commercial Wi-Fi 7 chipsets, the system uses standard OFDM channel sounding to detect:

• Presence detection: >95% accuracy at <10 m range
• Room-level localization: <1.5 m accuracy using multiple APs
• Gesture recognition: 6 gestures classified with >90% accuracy at <3 m

The channel estimation data (CSI) is processed by on-chip ML models. Qualcomm has proposed extending this approach to NR for 3GPP ISAC, leveraging similar CSI-based sensing algorithms.

Nokia — Radar-Communications Prototype

Nokia Bell Labs demonstrated a 28 GHz ISAC prototype at Mobile World Congress 2024, using a standard NR FR2 waveform (400 MHz BW, 120 kHz SCS) for simultaneous data and radar:

• Data throughput: 1.5 Gbps DL to a connected UE
• Radar detection: vehicles detected at 200 m range with 0.4 m range accuracy
• Velocity measurement: ±1 m/s accuracy for targets at 50 km/h
• Angular resolution: 2° using a 256-element phased array

The prototype operated in bistatic mode, with separate Tx and Rx antenna panels on the same gNB. Nokia's roadmap targets commercial ISAC capability integrated into AirScale base stations by 2028.

Ericsson — ISAC for Smart Factories

Ericsson Research partnered with a European automotive manufacturer to trial ISAC in a factory environment using a 3.5 GHz (n78) private 5G network. The system detected:

• Forklift positions with <1 m accuracy using 100 MHz bandwidth
• Human presence in restricted zones (safety geofencing) with 99.7% detection rate
• Collision prediction between AGVs with <200 ms warning latency

The trial used existing NR DL signals (SSB and CSI-RS) for bistatic sensing, requiring no hardware modifications to the standard Ericsson RAN equipment. Signal processing was performed at the edge server using Ericsson's proprietary ISAC algorithms.

Challenges and Open Problems

Key Takeaway: ISAC unifies communication and radar sensing in a single NR waveform, enabling 6G networks to detect objects, measure velocities, and map environments while delivering data. OFDM naturally supports sensing — range resolution scales with bandwidth (Δr = c/2B), velocity resolution with coherent processing time. 3GPP's Rel-19 study item evaluates use cases from V2X to gesture recognition, while Nokia (28 GHz radar-comms prototype) and Ericsson (factory sensing on n78) demonstrate real-world feasibility. Expect normative ISAC specifications in Release 20, making sensing a built-in capability of every 6G base station.