Patent Application

App. No. 18/736,370

Abstract

Claims (37)

1 . A method for Parallel-Processing Multi-Antenna Calibration (PPMAC) for massive Multi-Input Multi-Output (mMIMO) antennas in a 5G New Radio (NR) next generation NodeB (gNB) system, the method comprising the steps of:

providing a plurality of digital front-end units (DFEs), each of the plurality of DFEs includes an antenna calibration block having an internal memory and a plurality of associated antennas used in a 5G NR gNB, each antenna connected to a port;

connecting each of the plurality of DFEs in parallel with each other;

each of the antenna calibration blocks injecting a Zadoff-Chu (ZC) sequence from the calibration blocks internal memory into respective waveforms in a calibration feedback network to form network-affected ZC sequences;

the calibration feedback network returning the network-affected ZC sequences to a designated receiver port;

capturing the network-affected ZC sequences with the calibration block in one or more of the plurality of DFEs after a programmed delay; and

determining phase offsets between antenna ports in both the transmit (Tx) and receive (Rx) directions based on the network-affected ZC sequences.

2 . The method of claim 1 , further comprising the steps of:

applying a Hadamard code on top of the network-affected ZC sequences; and

simultaneously injecting the network-affected ZC sequences to a subset of the plurality of associated antennas with each DFE.

3 . The method of claim 2 , wherein the Hadamard code that is applied on the parallel injecting multi-antennas converts the network-affected ZC sequences to P orthogonal ZC-Hadamard sequences that are decodable and separable from the captured ZC-Hadamard sequence.

4 . The method of claim 3 , wherein

the plurality of DFEs comprises a first, second, third and fourth DFE; and

the plurality of associated antennas with each antenna calibration block is eight;

wherein the plurality of antennas are grouped into eight sets with one antenna per DFE in each set for a total of four antennas per set.

5 . The method of claim 4 , wherein each of the DFEs comprises a field programmable gate array (FPGA).

6 . The method of claim 4 , wherein calibration of all antenna sets is performed periodically in both the Tx and Rx directions.

7 . The method of claim 6 , wherein the calibration blocks perform the injection and capture during a Time Division Duplex (TDD) guard period between Tx and Rx simultaneously per antenna set in the Tx or Rx direction.

8 . The method of claim 7 , wherein the plurality of DFEs are coupled to a fronthaul field programmable gate array (FH FPGA), which is adapted to send and receive data and compensated values over an inter-FPGA Ethernet connection to the plurality of DFEs.

9 . The method of claim 8 ,

wherein the FH FPGA and each of the first, second, third and fourth DFE comprises a GPIO3 in, and a GPIO3 out;

wherein the FH FPGA GPIO3 out is connected to each of the first, second, third and fourth DFE GPIO3 in;

wherein the FH FPGA GPIO3 in is connected to the first DFE GPIO3 out; and

wherein each of the second, third and fourth DFE GPIO3 out are not connected.

10 . The method of claim 9 , wherein calibration is managed by the FH FPGA such that, prior to each calibration cycle, each DFE will receive a message relating to: the Carrier Components (CC) to be calibrated, a local port number for the DFE, and whether the DFE should inject, capture or both.

11 . The method of claim 7 , wherein each DFE applies the same network-affected ZC sequence to each antenna set.

12 . The method of claim 11 , wherein the Hadamard code comprises a 4×4 Hadamard matrix that is applied simultaneously to each antenna set.

13 . The method of claim 12 , further comprising the steps of:

converting the network-affected ZC sequences for each antenna set to Time-Domain; and

saving Time-Domain converted network-affected ZC sequences in corresponding AC Inject/Capture blocks.

14 . The method of claim 13 , further comprising the step of:

applying a smoothing function (ƒ) on an output of each antenna calibration channel estimation, to reduce a noise component contribution and increase quality and accuracy of the channel estimation of each antenna in a set.

15 . The method of claim 14 , wherein the smoothing function if a curve-fitting algorithm or a comb filter.

16 . The method of claim 15 , wherein the curve-fitting algorithm is a linear curve-fitting that linearizes channel estimation into a first-order estimation that only includes an average and slope information either for each part of a segmented bandwidth or for the entire bandwidth one part.

17 . The method of claim 15 , wherein a length of the comb filter determines an accuracy level of tracking an original channel estimation versus noise attenuation.

18 . The method of claim 17 , wherein the comb filter has a filter length of 32 taps.

19 . The method of claim 1 , wherein the antennas are aligned and equalized in terms of gain, delay, and phase for both Tx and Rx.

Assignments (13)