Menu
Latest
12 meetings · Prague, Jeju · 12–16 Oct
Reports
RAN1#126-bis · SA2#177

Glossary term · Physical Layer

TPMI

Transmitted Precoding Matrix Indicator

Physical Layer →

TPMI is an index signaled by the network to a device indicating the specific precoding matrix applied to a downlink transmission, enabling correct demodulation in closed-loop MIMO.

Introduced
Rel-8
Where
Radio Access Network › NG-RAN (5G)
Specifications
2 specs
TPMI Description Purpose What changed Family Neighbours Detected Changes Specifications

Description

The Transmitted Precoding Matrix Indicator (TPMI) is a critical downlink control information element in LTE and 5G NR MIMO operations. It is transmitted from the network (eNB in LTE, gNB in NR) to the User Equipment (UE) as part of the Downlink Control Information (DCI) on the Physical Downlink Control Channel (PDCCH). The TPMI value is an index that points to a specific precoding matrix or vector within a predefined codebook known to both the transmitter and receiver. This codebook contains a set of possible linear transformations (matrices) that can be applied to the layer-to-port mapping for transmit beamforming and spatial multiplexing. When the network schedules a downlink transmission for a UE, it selects a precoding matrix based on the UE's Channel State Information (CSI) feedback (which includes a Precoding Matrix Indicator, PMI, recommendation) and other scheduling algorithms. The applied matrix's index is then conveyed via the TPMI.

Upon receiving the DCI containing the TPMI, the UE uses this index to look up the corresponding precoding matrix W in its local codebook. The UE then applies this knowledge to its receiver processing chain. Specifically, it uses the matrix W to form hypotheses about the effective channel seen at the receiver, which is the product of the actual physical MIMO channel H and the precoding matrix W (i.e., H_eff = H * W). This allows the UE to perform accurate channel estimation for demodulation reference signals (DM-RS in NR, UE-RS in LTE) that are precoded with the same matrix as the data. Consequently, the UE can coherently demodulate the Physical Downlink Shared Channel (PDSCH) data. The TPMI is essential for closed-loop spatial multiplexing, especially for codebook-based precoding, as it resolves any ambiguity between what the UE recommended (via PMI) and what the gNB actually transmitted.

In terms of architecture, the TPMI is part of the dynamic link adaptation loop facilitated by the Physical Layer. The process involves: 1) UE measures the downlink channel and reports a recommended PMI and Rank Indicator (RI) via the PUCCH or PUSCH. 2) The gNB scheduler considers this recommendation, along with multi-user interference and traffic load, to select the final precoding matrix and transmission rank. 3) The gNB signals the chosen matrix via TPMI (and the transmission rank) in the DCI. 4) The UE decodes the DCI, retrieves the TPMI, and configures its receiver accordingly. This mechanism allows for fast, subframe-level adaptation of the precoding strategy to match time-varying channel conditions, maximizing throughput and reliability. In 5G NR, TPMI signaling is also used for uplink codebook-based transmission, where the gNB instructs the UE on which precoding matrix to use for its PUSCH transmission, showcasing its bidirectional role.

Purpose & Motivation

The TPMI was introduced to enable efficient and unambiguous implementation of closed-loop, codebook-based precoding for MIMO in LTE (from Release 8 onwards) and later carried forward into 5G NR. Prior to standardized closed-loop MIMO, systems relied on open-loop diversity or required complex blind detection techniques at the receiver to determine the applied precoding. The primary problem TPMI solves is the signaling of the transmitter's precoding choice to the receiver in a low-overhead manner. Without this explicit indicator, the UE would not know the exact spatial filter applied to its data, leading to severe performance degradation in demodulation, as the effective channel would be unknown.

Its creation was motivated by the drive to achieve higher spectral efficiency through multi-antenna techniques. By allowing the network to inform the UE of the precise precoding matrix, the UE can perform optimal receiver processing, such as Minimum Mean Square Error (MMSE) or Maximum Likelihood (ML) detection, tailored to the effective channel. This maximizes the signal-to-interference-plus-noise ratio (SINR) per layer. The TPMI, together with the PMI feedback, forms a tight feedback loop that adapts the beamforming pattern to the current channel state, focusing energy towards the intended UE and minimizing interference to others. This is fundamental for realizing the capacity gains promised by MIMO technology. In 5G NR, its purpose expanded to support more flexible antenna configurations, massive MIMO, and uplink MIMO, reflecting the evolution towards more dynamic and configurable spatial processing.

What changed in its specifications

The latest release-to-release comparison of specifications that define TPMI. Each summary describes the whole specification, not only this term.

TS 36.212Rel-18 → Rel-19

Rel-19 adds NTN IoT Phase 3 support to TS 36.212 for LTE. It adds OCC enabled/disabled and OCC sequence index fields to the DCI Format N0 description.

Full comparison in TS 36.212 →

Technology family

MIMO and beam management · Radio Access Network › NG-RAN (5G)

  1. R99 Transmit diversity, TSTD
  2. Rel-7 MIMO
  3. Rel-8 CSI, DM-RS, MU-MIMO, PMI, RI, SDMA, SRS, SU-MIMO, TPMI
  4. Rel-10 CSI-RS, NZP CSI-RS
  5. Rel-11 CoMP, CSI-IM, QCL, UL CLTD, UL MIMO
  6. Rel-13 CRI, FD-MIMO
  7. Rel-15 Beam Level Mobility, Beam management, BFD, BFD-RS, BFR, PT-RS, TCI, ZP CSI-RS
  8. Rel-18 CJT

Also in this family (introduction release not pinned to a single defining specification): Beamforming

Members and relations were proposed independently by several models and kept only where at least two agree; quotes are verified verbatim against the specifications. Release = first approved version of the defining specification.

Release Timeline

Detected Changes Across Releases

verified against 3GPP sources

Documented changes in 6 releases, each verified against the text of 3GPP specifications or approved change requests and shown with the exact wording. The release timeline above shows the same releases. A release without an entry has no documented change for this term.

Rel-8 introduced

TPMI is specified for the first time in 36212, as a precoding information field in DCI formats indicating which codebook index is used for closed-loop spatial multiplexing and single-layer transmission.

  • “Note that TPMI indicates which codebook index is used in Table 6.3.4.2.3-1 or Table 6.3.4.2.3-2 of [2].” TS 36.212 v8.8.0
  • “TPMI information for precoding – number of bits as specified in Table 5.3.3.1.3A-1” TS 36.212 v8.8.0
  • “TPMI Transmitted Precoding Matrix Indicator” TS 36.212 v8.8.0
Rel-13 specification text

The TPMI information field is now stated to be present only when PDSCH is configured with TM6 for DCI formats mapped onto the UE-specific search space given by the C-RNTI.

  • “This field is present only if PDSCH transmission is configured with TM6 for DCI formats scheduling PDSCH which are mapped onto the UE specific search space” TS 36.212 v13.11.0
Rel-15 specification text

NR PUSCH codebook-based transmission specified for the first time: the UE derives its precoder from SRI, TPMI and rank, and TPMI also determines the codebook subset and UL PT-RS ports.

  • “the UE determines its PUSCH transmission precoder based on SRI, TPMI and the transmission rank” TS 38.214 v15.16.0
  • “the actual number of UL PT-RS port(s) is determined based on TPMI and/or number of layers” TS 38.214 v15.16.0
  • “the UE determines its codebook subsets based on TPMI and upon the reception of higher layer parameter codebookSubset” TS 38.214 v15.16.0
Rel-16 specification text

TPMI-based codebook PUSCH precoding was extended to the new DCI format 0_2, with a separate codebookSubsetDCI-0-2 parameter, and TPMI indication for UL full power transmission was clarified.

  • “the UE determines its PUSCH transmission precoder based on SRI, TPMI and the transmission rank” TS 38.214 v16.17.0
  • “the UE determines its codebook subsets based on TPMI and upon the reception of higher layer parameter codebookSubset in pusch-Config” TS 38.214 v16.17.0
  • “Clarification of TPMI indication for UL full power transmission” TS 38.212 v16.15.0
Rel-17 specification text

For codebook-based PUSCH, one or two TPMIs can now be indicated by DCI and applied per SRS resource set to PUSCH repetitions, with PT-RS port derivation updated for multiple TPMIs.

  • “one or two SRI(s), and one or two TPMI(s) are given by the DCI fields of two SRS resource indicator” TS 38.214 v17.16.0
  • “The TPMI is used to indicate the precoder to be applied over the layers {0…ν-1}” TS 38.214 v17.16.0
Rel-19 specification text

TPMI-based codebook subset determination and UL PT-RS port determination are extended to codebook-based UL transmission with three antenna ports, using the 'nonCoherent' subset.

  • “For codebook based transmission with three antenna ports, the UE determines its codebook subsets based on TPMI(s)” TS 38.214 v19.5.0
  • “or non-coherent codebook-based UL Transmission with 3 antenna ports” TS 38.214 v19.5.0

Defining Specifications

3GPP specifications that define or reference TPMI, with the latest known release. Sourced from the 3GPP document catalog — see methodology.

SpecificationTitleRelease
TS 36.212 vj30 E-UTRA Physical Layer Procedures Rel-19 read
TR 38.889 vg00 NR-based access to unlicensed spectrum study Rel-16 read