Inferensys

Glossary

Phase Tracking Reference Signal (PTRS)

A 5G NR reference signal designed to track and compensate for phase noise introduced by local oscillators at high carrier frequencies, such as millimeter wave bands.
Stylish WeWork-like workspace with hot desks and document wall, professional searching through enterprise knowledge base on a mounted ultrawide display, warm industrial pendants overhead.
5G NR PHYSICAL LAYER

What is Phase Tracking Reference Signal (PTRS)?

A specialized 5G New Radio reference signal designed to compensate for phase noise, a critical impairment at millimeter wave frequencies.

A Phase Tracking Reference Signal (PTRS) is a 5G NR pilot signal specifically designed to enable the receiver to estimate and compensate for phase noise introduced by local oscillator imperfections. This impairment is particularly severe at the high carrier frequencies used in millimeter wave (mmWave) bands (FR2), where it causes a common phase error and inter-carrier interference that degrades the error vector magnitude of high-order QAM constellations.

PTRS is configured in the time domain with a density that adapts to the scheduled modulation and coding scheme (MCS) and bandwidth. It is associated with a specific Demodulation Reference Signal (DMRS) port and is transmitted only in the resource blocks allocated for a Physical Downlink Shared Channel (PDSCH) or Physical Uplink Shared Channel (PUSCH) transmission, minimizing overhead while tracking phase rotation across OFDM symbols.

PHASE NOISE COMPENSATION

Key Characteristics of PTRS

The Phase Tracking Reference Signal (PTRS) is a 5G NR-specific pilot signal designed to mitigate the devastating effects of phase noise on high-order modulation schemes at millimeter-wave frequencies. Its characteristics are tailored to the oscillator quality and scheduled bandwidth.

01

Phase Noise Compensation

PTRS is the primary tool for common phase error (CPE) correction. Local oscillator imperfections at high carrier frequencies (e.g., FR2, >24 GHz) cause random phase rotations that rotate the entire received constellation. The receiver estimates this common rotation using known PTRS symbols and applies an inverse phase de-rotation to every subcarrier within the OFDM symbol, preventing catastrophic bit error rate floors in 64QAM and 256QAM transmissions.

02

Time-Domain Density (L_ptrs)

The time density of PTRS is configurable to match the channel's coherence time and the oscillator's phase noise profile. The standard defines four levels:

  • L_ptrs = 1: Present in every OFDM symbol (highest density, for severe phase noise).
  • L_ptrs = 2: Present in every 2nd symbol.
  • L_ptrs = 4: Present in every 4th symbol.
  • L_ptrs = 0: PTRS is not present (sufficient for low-order modulation or low carrier frequencies).
03

Frequency-Domain Density (K_ptrs)

The frequency density defines how many subcarriers within a scheduled resource block carry PTRS. It is tied to the scheduled bandwidth to balance estimation accuracy against overhead:

  • K_ptrs = 2: One PTRS subcarrier every 2 RBs (for bandwidths < 4 RBs).
  • K_ptrs = 4: One PTRS subcarrier every 4 RBs (for bandwidths ≥ 4 RBs). This sparse frequency allocation is sufficient because phase noise is highly correlated across frequency, unlike the channel response estimated by DMRS.
04

Association with DMRS

PTRS is always associated with a specific DMRS port. The PTRS port is a function of the associated DMRS port index. The PTRS sequence is derived from the same pseudo-random sequence generator used for DMRS, but it is scrambled with a different n_SCID (scrambling identity) to ensure orthogonality. The power of PTRS is also scaled relative to the associated DMRS to maintain a consistent energy per resource element (EPRE) ratio.

05

PTRS Sequence Generation

The PTRS sequence is a Gold sequence (length-31). The initialization of the sequence generator depends on:

  • The slot number and OFDM symbol index within a radio frame.
  • The cell ID (N_ID).
  • The higher-layer parameter n_RNTI (Radio Network Temporary Identifier). This ensures that the PTRS is pseudo-random and unique per cell and per UE allocation, minimizing inter-cell interference.
06

Resource Element Mapping

PTRS is mapped to resource elements in the physical resource blocks scheduled for the PDSCH (downlink) or PUSCH (uplink). The mapping avoids collision with DMRS, CSI-RS, and other critical signals. The specific subcarrier index within the scheduled bandwidth is determined by the DMRS port association and the frequency density configuration, ensuring a deterministic and known pattern for the receiver.

PHASE NOISE COMPENSATION

Frequently Asked Questions

Common questions about the Phase Tracking Reference Signal (PTRS) and its role in maintaining demodulation integrity at millimeter wave frequencies in 5G NR networks.

A Phase Tracking Reference Signal (PTRS) is a UE-specific reference signal introduced in 3GPP Release 15 for 5G NR that enables the receiver to estimate and compensate for phase noise generated by local oscillators. Unlike the Demodulation Reference Signal (DMRS), which provides a baseline channel estimate, PTRS is a sparse pilot signal distributed in the time domain specifically to track the rapid, sample-to-sample phase rotation caused by oscillator imperfections. This is critical at millimeter wave frequencies (FR2, above 24 GHz) , where phase noise power increases proportionally with carrier frequency. PTRS is configured per scheduled user and is only present when higher modulation orders (e.g., 64QAM, 256QAM) or high-rank MIMO transmissions are scheduled, as these are most susceptible to phase noise degradation. The signal is mapped to specific resource elements within the scheduled physical resource blocks, with its time density configurable based on the subcarrier spacing and modulation and coding scheme (MCS).

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.