17
Underwater Wireless Communications: Modem Design and Implementation Yahong Rosa Zheng Department of Electrical & Computer Engineering Lehigh University, Bethlehem, PA, 18015 [email protected] The work is supported in part by USDoT, ONR and NSF

Underwater Wireless Communications: Modem Design and ...uwa.ua.edu/.../2/5/...communications_modem_design_and_implementation_1.…Underwater Wireless Communications: Modem Design and

  • Upload
    others

  • View
    10

  • Download
    0

Embed Size (px)

Citation preview

Underwater Wireless Communications: Modem

Design and Implementation

Yahong Rosa Zheng

Department of Electrical & Computer Engineering

Lehigh University, Bethlehem, PA, 18015

[email protected]

The work is supported in part by USDoT, ONR and NSF

Underwater Communication Means

2

❑Magneto-Inductive (MI) Communications:

Small bandwidth (<500 kHz),

Short range (~100 m)

❑Optical Beams: Large BW, short range: 20 m – 200 m

❑Sound Propagation (AComm):

Short range (<1 km): 300 kHz (HF)

Medium range (1-10 km): 10 - 100 kHz (MF)

Long range (1000 km): < 2 kHz (LF)

❑Radio Frequency (RF) does not work:

Limited BW (1MHz), very short range < 1 m

Acomm Multipath Channels

3

❑Small BW:

20 – 300 kHz

❑Multipath:

10 ms – 200 ms

❑Doppler:

5 – 20 Hz

❑Doppler ratio:

1e-4 ~ 1e-3

PHY-Layer Acomm Approaches

4

SC-FDE

❑MIMO Single-Carrier Time-Domain Turbo Equalization

❑MIMO Single-Carrier Frequency-Domain Turbo Equalization

❑MIMO OFDM with Turbo ICI Cancellation

Figures adopted from Falconer’s 1999 tutorial paper. Both are used for Cellular LTE.

5

Architecture of MIMO Transceiver

Y.R. Zheng, J. Wu, and C. Xiao, “Turbo Equalization for Underwater Acoustic Communications,” IEEE Commun. Mag., vol. 53, no. 11, pp. 79-

87, Nov. 2015.

Time-Domain Turbo Equalizers

6

Turbo Linear Equalizer

Turbo Soft-Decision Feedback Equalizer

7

Equalizer Implementation

B. Han, Z. Yang, and Y. R. Zheng, “FPGA Implementation of QR Decomposition for MIMO-OFDM Using Four CORDIC Cores,” IEEE ICC

2013, Budapest, Hungary, June 9-13, 2013. pp. 1-6.

Givens rotation for 4-by-4 matrix

QR decompensation

by 2D systolic array

8

FPGA Implementation of MIMO Transceiver

Channel decodingSoft decision

Channel Encoding

Interleaving

De-interleaving

DPD

DPD

PAD/A

A/D LNA

Switch

Interleaving

Mapping

Symbol

&

SubchannelizationPilot insertion

SubchannelizationPilot insertion

Lin

ear

Pre

cod

ing

CyclicPrefix

CyclicPrefix DUC

DUC CFR

CFR

Pilot ExtractionSubchannel

Pilot ExtractionSubchannel

Channel estimation

Channel estimation

So

ft d

ecis

ion

Sp

her

e d

eco

din

g RemoveCyclicPrefix

RemoveCyclicPrefix

PAD/A

A/D LNA

Switch

FrequencyCorrection

Timing &

FrequencyCorrection

Timing &

DDC

Baseband HF

DDC

Transmit

Bits

Bits

Ant1

Ant2

Baseband HF

Receive

❑MIMO Transceiver with Turbo Equalization and Sphere Decoding.

B. Han, Z. Yang, and Y. R. Zheng, “Efficient Implementation of an Iterative MIMO-OFDM Receiver Using MMSE Interference Cancelation,”

IET Commun., vol. 8, No. 7, pp. 990 – 999, May. 2014.

Altera FPGA Stratix III & Stratix IV

FFT1FFT2

RAM_ZinRAM_Sk

RAM_CE

Channel estimation

Frame extraction

output

MMSE-LE

E[Sk]

RAM_Esk

LE_H1LE_H2

LE_en_d2

Zin_wiremul_es_wiremul_es_oen Sym2bits

Ungroupsubcarrier

Bits_remove De-interleaver

LDPCDecoder

Bit_padding&Interleaver

Sk,

Sk_oen

Rx_bb_top_serial

Controllogics Bit2sym

Ungroupsubcarrier

ldpc_p2s

ldpc_s2p

Data input

Up-sampling

msg_I#msg_Q#

I#Q#

Data_I#Data_Q#

preamble_I#

preamble_Q#

fft_valid

H##N_I

H##N_Q

CE_valid

H##N_I_unsH##N_Q_uns

Es_wire

deintlv_o#x

deint_vld#x

Esk128

FFT#_realFFT#_imagFFT#_valid F

FT#_real_uns

FFT#_imag_uns

b2s_ob2s_vld

Zin_wire

pat_o#x

pat_oen#1

llr_e#xllr_en#

S2b_in#

s2b_in_en

bits#bits_oen#

dat_o#dat_oen#

zp_o#xzp_oen#1

zp_s#zp_sen#

Hard decision

Zin128

intlv_o#xint_vld#

zpp_o#x

zpp_oen#x

dat_Sk

wren_Sk

Sk_q

#: 1,2x: 1,2,3,4

dat_eon#

llr_p#xllr_pen#

rd_Zin

rden_Sk

rden_H

9

Results of ModelSim Output

10

11

TMS340C6713 DSK and Daughter Board

12

Hardware Implementation w/ TI MCU

TX

RX

MCU

Control

MI Communication System

13Version 3 MI system w/ 3-D antenna coils

MSP430

Underwater Internet of Things

14

Bridge Scour Monitoring

15

Rock shells weigh 50 – 200 lb

16

Gasconade River, July 2013

Rubidoux River, May 2018

Bridge Tests

Conclusion and Future Work

❑Two of the three means in underwater wireless communication:

❖Ultrasound (acoustic) communications (Acomm);

❖Magneto-Inductive (MI) Communications.

❑What are the most important factors: reliability, range, data rate,

power consumption, small form factor, cost?

❑Future work:

❖Hardware integration and field test;

❖Innovative network protocols and secure communication.

17