PUSCH — Physical uplink shared channel — the mirror of PDSCH
Channels — what carries what
Where it sits
What it is
What it is. The uplink mirror of PDSCH. TS 38.211 clause 6.3.1, procedures in TS 38.214 clause 6, coding in TS 38.212 clause 6.2. PDSCH §3.3 lists the differences; this section gives them their clauses.
The same chain, almost exactly. CRC, LDPC base-graph selection, segmentation, rate matching and concatenation are the same clauses of TS 38.212 used with UL-SCH substituted for DL-SCH. Scrambling, modulation, layer mapping and resource mapping mirror TS 38.211 clause 7.3.1. If you have read Parts II and III of the PDSCH note, you have read most of PUSCH.
Difference 1 — a second waveform. PUSCH may use DFT-spread OFDM (clause 6.3.1.4), which lowers the peak-to-average ratio at the cost of the frequency-selective scheduling that CP-OFDM allows. It exists because it is the device's amplifier that pays (frame structure §9.2).
Difference 2 — control rides on it. When a device owes an acknowledgement or a CSI report and also has a grant, the control information is multiplexed into PUSCH rather than sent separately (TS 38.213 clause 9.3). One transmission is cheaper than two, and the device's power budget is the scarce thing.
Difference 3 — precoding is not transparent. PDSCH §6.4 shows the downlink antenna-port mapping being an identity because the base station hides its precoder. On the uplink the network tells the device which precoder to use, from a codebook — so the uplink kept the codebook that the downlink abandoned.
Read on
This concept was first written up in ref-channels, which reads the whole group as one argument.
Before this concept, the hierarchy says to learn the following — the full chain, in order: