1 of 19

Quantum and Quantum-Inspired Computing

for Large-Scale NOMA-MIMO Wireless Networks

Interns: Jeffrey Tang, Alex Markley

Advisors: Minsung Kim, Byungjun Kim

1

Week 7

2 of 19

Implementation of ParaMax: Quantum-Inspired Maximum Likelihood Detection

2

Goal

101001110010101...

N_t

N_r

3 of 19

Implementation of ParaMax: Quantum-Inspired Maximum Likelihood Detection

3

Goal

101001110010101...

N_t

N_r

4 of 19

Implementation of ParaMax: Quantum-Inspired Maximum Likelihood Detection

4

Goal

5 of 19

Implementation of ParaMax: Quantum-Inspired Maximum Likelihood Detection

5

Goal

6 of 19

6

Progress

This week:

Simultaneous transmission from multiple Tx
Verifying timing synchronization
Alternative MIMO detection algos

Next week:

Solution for timing sync issues

Post-processing or USRP config debugging

MIMO Python integration on hardware scheme

7 of 19

Under ideal cases, no interference. Realistically, what is transmitted is not what is received.

7

MIMO Detection

8 of 19

Work backwards by finding the data that was most likely sent given the channel state and recv’d data. This is an optimization problem now.

8

MIMO Detection

9 of 19

9

10 of 19

10

11 of 19

Sphere Decoding (equiv to ML) performs DFS tree pruning. FlexCore uses different PEs to traverse each path concurrently and selects the best candidates

11

FlexCore

MIMO Detection

Pre-processing tree example

3x3 MIMO with QPSK, pick the top 4 p-vectors

Special Thanks to Brian Zhang for the slides on FlexCore

12 of 19

Sample possible x values and start to settle into a minima if we find one.

12

ParaMax

MIMO Detection

Best found

13 of 19

Add multiple parallel annealers, one that stays near the best solution so far and one that explores for a better minima.

13

ParaMax

MIMO Detection

14 of 19

14

15 of 19

15

16 of 19

16

Preamble

Payload

17 of 19

17

observed: 50 samps

expected: 50

observed: ~400 samps

expected: 100

18 of 19

18

19 of 19

Thank you!

Any Questions?

19