【Editorial Discussion】

A TDD slot format on paper defines symbols and a guard. On a real radio, the time that is actually clean between the end of a downlink burst and a usable uplink is set by PA ramp, filter delay, switching transients, and sometimes the remote radio’s framing. If you only check that the configuration matches the standard format, you can still fail a coexistence or a UL sensitivity test.

1) Measure the RF envelope at the antenna reference, not only the baseband timeline. The last useful downlink energy, the switch transient, and the first usable uplink symbol should be visible on the same capture. A format that “has enough guard symbols” can still leave residual TX energy in the first UL symbol.

2) Ramp and bias sequencing belong in the budget. A PA or switch that turns off slowly eats guard. Temperature and output power change that tail. Record the condition under which the guard was verified, not only a lab mid-power case.

3) The uplink receiver’s settling is a separate item. Even if the transmitter is quiet, the RX path (filters, AGC, DC tracking) may not be ready at the first UL symbol. Sensitivity or EVM on that symbol should be checked explicitly if the standard or the product claims it is usable.

4) Keep the slot format, the measured envelope, and the claim you make to a higher layer in one note. A protocol stack that assumes a clean symbol you have not actually cleared will look like a “PHY bug.”

Open question: in your last TDD guard failure, was the leftover energy on the TX tail, the switch, or the RX not being ready? Do you require an envelope capture at the antenna plane as part of format sign-off? Please share what you measure besides the slot-format table.