Introduction to FSK and 5G NR

Frequency Shift Keying (FSK) is one of the oldest and most reliable digital modulation techniques. It encodes data by shifting the carrier frequency between a set of discrete values, typically representing binary 0 and 1. FSK’s simplicity, resilience to amplitude variations, and low implementation cost have made it a mainstay in low-power, short-range applications such as Bluetooth Low Energy, RFID tags, and legacy paging systems. Its constant-envelope nature allows power amplifiers to operate efficiently, making it attractive for battery-constrained devices.

5G New Radio (NR), the global standard for fifth-generation cellular networks, takes a fundamentally different approach. It is built around Orthogonal Frequency Division Multiplexing (OFDM), which splits a wideband channel into many narrow, orthogonal subcarriers. OFDM, combined with advanced features like massive MIMO (multiple-input multiple-output), beamforming, and flexible numerology, delivers the high data rates, low latency, and spectral efficiency required for eMBB (enhanced mobile broadband), URLLC (ultra-reliable low-latency communications), and mMTC (massive machine-type communications).

The question of whether FSK can coexist with or be adapted to 5G NR is not merely academic. Network operators, chipset vendors, and IoT ecosystem players must decide whether to invest in backward-compatible modulation schemes or rely solely on OFDM-based variants. This article assesses the technical compatibility of FSK with 5G NR, identifies the fundamental barriers, and explores potential integration strategies for specific use cases.

Core Modulation Principles: FSK vs. OFDM

FSK Fundamentals

FSK modulates the carrier frequency according to the input bit stream. In its simplest binary form (BFSK), two frequencies represent 0 and 1. The frequency deviation from the center frequency determines the modulation index h. When h is less than 0.5, the spectrum is relatively compact but suffers from poor error performance; higher indices improve robustness but widen the occupied bandwidth. Continuous-phase FSK (CPFSK) variants such as GFSK (Gaussian-filtered FSK) smooth transitions to reduce out-of-band emissions, as seen in Bluetooth.

Frequency deviation is linked to data rate: to avoid intersymbol interference, the frequency spacing must be at least the bit rate for non-coherent detection. This creates an inherent trade-off between bandwidth and throughput. Even with advanced variants like multi-level FSK (MFSK), spectral efficiency remains modest compared to linear modulation schemes like QAM.

OFDM Architecture in 5G NR

5G NR uses OFDM with cyclic prefix (CP-OFDM) for both downlink and uplink (with optional DFT-spread OFDM for uplink to reduce PAPR). The subcarrier spacing is flexible: 15 kHz, 30 kHz, 60 kHz, 120 kHz, or 240 kHz, allowing adaptation to different frequency bands and deployment scenarios. Each subcarrier can be modulated independently with QPSK, 16QAM, 64QAM, or 256QAM (and up to 1024QAM in millimeter-wave bands). This orthogonality eliminates inter-carrier interference (ICI) as long as synchronization is maintained.

OFDM’s spectral efficiency arises from overlapping subcarriers without guard bands. For example, a 20 MHz channel with 15 kHz spacing can support over 1200 subcarriers, each carrying multiple bits per symbol. Combined with MIMO spatial multiplexing, OFDM achieves data rates exceeding 20 Gbps in ideal conditions. FSK, even with MFSK, cannot approach this density because its frequency spacing must be larger than the symbol rate to maintain orthogonal detection.

Robustness and Noise Performance

FSK excels in environments with severe amplitude fading or non-linear amplifier distortion because its constant envelope avoids amplitude-induced errors. This makes it ideal for low-cost transmitters where linearity is hard to guarantee. OFDM, with its high peak-to-average power ratio (PAPR), requires linear amplifiers and careful back-off, reducing power efficiency. In battery-operated IoT devices, FSK’s power advantage can be significant — often 3–6 dB better than OFDM for the same error rate.

However, OFDM can combat frequency-selective fading through adaptive bit loading and channel coding, whereas FSK’s wide frequency occupancy makes it more susceptible to narrowband interference. 5G NR also incorporates advanced forward error correction (LDPC and Polar codes) that OFDM exploits to close the link budget.

Compatibility Challenges in 5G NR Networks

Spectral Efficiency Mismatch

The most fundamental barrier is spectral efficiency. To support the massive data rates demanded by 5G, every hertz must carry many bits. FSK’s typical spectral efficiency is below 1 bit/s/Hz for binary modulation; even with 4-FSK, it reaches around 2 bits/s/Hz. 5G NR OFDM with 256QAM achieves over 8 bits/s/Hz per spatial stream. Deploying FSK within a 5G carrier would waste valuable spectrum that could otherwise serve hundreds of OFDM subcarriers.

One might consider using FSK only in bandwidth-limited segments, but 5G NR’s resource grid is designed for contiguous OFDM symbol allocation. Inserting an FSK waveform would create guard bands that reduce overall spectral utilization. The 3GPP specifications do not currently define an FSK-based waveform for any 5G NR channel.

Interference with OFDM Subcarriers

FSK signals are not orthogonal to OFDM subcarriers. The frequency shift jumps cause wideband side lobes that fall into adjacent OFDM subcarriers, creating inter-carrier interference. Even with Gaussian filtering, the out-of-band emissions from an FSK transmitter can exceed the stringent spectral mask requirements for 5G NR base stations and user equipment. 5G NR uses a guard band and a resource block structure; any non-OFDM transmission would require careful coordination, likely at the cost of capacity.

Conversely, OFDM signals appear as noise to an FSK receiver. The aggregate OFDM power across many subcarriers can desensitize the FSK front-end, especially if the FSK receiver lacks narrowband filtering. This makes co-existence within the same carrier impractical without dynamic spectral sharing techniques.

Hardware and Radio Front-End Constraints

5G NR user equipment (UE) and base stations are optimized for OFDM. The RF chains incorporate linear power amplifiers, wideband analog-to-digital converters (ADCs), and digital predistortion to handle OFDM’s high PAPR. These same components can support FSK, but the reverse is not true: an FSK-specific radio would need significant add-ons to support OFDM. For dual-mode operation, the cost and complexity increase.

Moreover, the synchronization requirements differ. OFDM relies on precise time and frequency synchronization to maintain orthogonality; 5G NR achieves this through primary and secondary synchronization signals (PSS/SSS). FSK receivers often use non-coherent detection, which is more tolerant of frequency offsets but less efficient. Merging the two chains in a single device requires either separate baseband processors or a configurable digital front-end.

Latency and Synchronization Issues

URLLC services require sub-millisecond latencies. FSK, with its longer symbol duration for a given data rate, may introduce additional latency compared to OFDM with short subcarrier spacing. For example, a 15 kHz OFDM symbol lasts 66.7 µs, while a 100 kbps BFSK signal with 100 kHz deviation might have a symbol period exceeding 10 µs, but the overall packet duration could be longer due to lower data rates. Synchronization and channel estimation overhead in hybrid systems could further increase latency.

Power Consumption Trade-offs

While FSK transmitters can consume less power than OFDM transmitters for short bursts (due to constant-envelope amplifiers), 5G NR has introduced power-saving features like discontinuous reception (DRX), sleep modes, and wake-up signals. An FSK-based wake-up receiver (WUR) is actually one potential application (see section on integration strategies). However, for continuous data transmission, OFDM with efficient coding can actually achieve better energy per bit at high throughput.

Potential Integration Strategies for FSK in 5G Systems

Hybrid Modulation Schemes

A pragmatic approach is to blend FSK with OFDM in a single waveform. For instance, an OFDM subcarrier could employ FSK as an inner modulation, while the outer structure remains OFDM. This is sometimes called FSK-OFDM or frequency shift keying on subcarriers. In practice, this reduces to MFSK on each subcarrier, which yields lower spectral efficiency than QAM but retains constant-envelope properties for selected subcarriers. The 3GPP has considered variations like π/2-BPSK for low-PAPR uplink transmissions in LTE-M and NB-IoT, but those are phase modulations, not frequency shifts.

Another hybrid is to use FSK for control channels where robustness is prioritized over data rate. For example, the 5G NR physical uplink control channel (PUCCH) uses BPSK and QPSK; substituting a narrowband FSK could improve link budget for cell-edge devices. However, such changes would require standardization and backward-compatible signaling.

Software-Defined Radio and Flexible Waveforms

Software-defined radio (SDR) platforms enable dynamic waveform selection based on channel conditions and service requirements. A 5G base station equipped with SDR could, in principle, allocate a small resource block for FSK transmissions to serve legacy IoT devices or to test novel concepts. The Open Radio Access Network (O-RAN) architecture supports modular baseband processing, allowing custom slots for non-OFDM waveforms. This is especially relevant for private 5G networks in industrial environments where ultra-reliable low-rate links for sensor networks may coexist with high-throughput eMBB.

The key challenge is real-time waveform switching. 5G NR’s frame structure (10 ms radio frame divided into slots and mini-slots) must be respected. A dedicated time slot or frequency band could be reserved for FSK, but this reduces OFDM resource availability. SDR also requires programmable accelerators (FPGAs, GPUs) to handle the diverse modulation and demodulation chains, increasing hardware cost.

Targeted Applications: IoT, M2M, and Low-Power Devices

The most compelling use case for FSK in a 5G context is massive IoT (mMTC). 5G NR Lite (RedCap) devices already reduce complexity by supporting lower bandwidths (e.g., 20 MHz) and fewer MIMO layers. Going further, an FSK-based narrowband channel could be overlaid on the 5G carrier for the simplest IoT sensors that transmit a few bytes per day.

3GPP Release 17 introduced Narrowband IoT (NB-IoT) and LTE-M within the 5G framework, but those use a modified OFDM (single-tone or multi-tone QPSK). They achieve similar power efficiency as FSK through single-tone transmissions with narrow bandwidth. Nevertheless, true FSK could offer even lower peak current draw because of the constant-envelope amplifier. Some vendors have proposed an FSK-based NB-IoT extension, but it has not been adopted.

Example: Wake-Up Receivers (WUR)

5G NR devices can incorporate a low-power wake-up receiver that monitors for an FSK trigger signal while the main OFDM transceiver is in deep sleep. The WUR uses very low power (microamps) and can wake the main radio only when needed. This is a hybrid approach where FSK serves as an always-on, low-rate control channel. The 3GPP has studied WUR enhancements for power saving in Release 18. In such systems, the FSK waveform coexists with OFDM by operating in a separate narrowband channel (e.g., 5 MHz dedicated to wake-up signals).

5G NR also operates in unlicensed bands (NR-U) and in millimeter-wave. In these environments, FSK’s wide bandwidth can be tolerated more easily. For example, in the 60 GHz band (802.11ad/ay), single-carrier modulations like FSK might offer better robustness against phase noise than OFDM. However, the high data rate targets push most designers toward OFDM-like waveforms.

For wireless backhaul, where fixed links connect small cells, FSK could be used for low-rate telemetry and control, while the main data uses OFDM. This is a niche application, but it avoids the need for a separate radio interface.

Future Research Directions and Standardization Efforts

Role of 3GPP

The 3GPP is the primary body defining 5G NR standards. So far, only OFDM-based waveforms have been specified for NR. Any introduction of FSK would require a new study item and work item, likely under the banner of “new waveform” for specific deployment scenarios. A key document is 3GPP TS 38.211, which defines physical channels and modulation. Adding an FSK option would affect all layers — from resource mapping to coding. Given the maturity of Release 17/18, it is unlikely that FSK will be added as a general-purpose modulation, but future releases (e.g., Release 19 or 20) could consider it for extreme IoT.

Advanced Digital Signal Processing

Modern DSP can mitigate some of FSK’s disadvantages. Advanced algorithms such as frequency-domain equalization for FSK (FD-FSK) can improve spectral efficiency by allowing closer spacing of tones without increasing error rate. Interference cancellation techniques (successive interference cancellation, or SIC) could enable FSK and OFDM to share spectrum if the receiver can subtract the FSK signal. Research papers (see example) have demonstrated that with careful design, an FSK signal can be embedded within an OFDM guard band or even overlaid on the data subcarriers using low-rate coding.

Cognitive Radio and Dynamic Spectrum Access

In a cognitive radio framework, a 5G base station could detect unused portions of the spectrum and assign them for FSK transmissions. This is similar to licensed shared access (LSA) but on a fine time-frequency granularity. The base station would schedule an FSK resource block that does not conflict with active OFDM allocations. Machine learning algorithms can predict traffic patterns and allocate the right modulation per device. This is still experimental, but it aligns with the vision of a flexible 5G air interface.

Conclusion

FSK and 5G NR OFDM represent fundamentally different design philosophies. FSK prioritizes simplicity, robustness to amplifier nonlinearities, and low power consumption at the expense of spectral efficiency and data rate. 5G NR OFDM maximizes throughput and flexibility through complex multi-carrier processing. Direct substitution of FSK for OFDM in a 5G carrier is impractical due to spectral inefficiency, interference issues, and hardware mismatch.

However, FSK can play a supporting role in 5G ecosystems. Potential integration paths include hybrid waveforms (FSK-OFDM), software-defined radios that dynamically allocate narrowband FSK channels, wake-up receivers for power saving, and specialized IoT overlays in unlicensed spectrum. These approaches leverage FSK’s strengths without compromising 5G’s core performance metrics. Standardization bodies like 3GPP are unlikely to adopt FSK as a primary waveform, but they may consider it for specific enhancements in future releases.

For network operators and device manufacturers, the prudent strategy is to monitor advancements in flexible waveforms and low-power techniques while continuing to rely on OFDM for mainstream 5G services. For niche IoT applications that demand extreme battery life and low data rates, dedicated FSK radios (e.g., Bluetooth, Zigbee, or proprietary sub-GHz systems) remain a viable complement to 5G NR. The long-term vision is a heterogeneous air interface where FSK and OFDM coexist through intelligent resource management, each serving the use case it handles best.