SignalEx: Linking environmental acoustics with the signaling schemes Michael Porter Ocean Sciences Division Science Applications International and Keyko McDonald, Paul Baxley, Joseph Rice Space and Naval Warfare Systems Center San Diego, CA
Mar 27, 2015
SignalEx: Linking environmental acoustics with the signaling schemes
Michael PorterOcean Sciences Division
Science Applications Internationaland
Keyko McDonald, Paul Baxley, Joseph Rice
Space and Naval Warfare Systems CenterSan Diego, CA
Outline
• Motivation Case study: Front engineering test
• SignalEx tests– Telesonar testbeds– Sites b, c, d– Measured channel impulse response– Predicted channel impulse response– Performance of a DPSK/DSSS system
SignalEx
The National Oceanographic Partnership Program (NOPP) FRONT system is being installed by a consortium led by U. Conn
• Cellular modems relay data to shore from the Montauk Point and Block Island USCG buoys
• ADCP sensor nodes with trawl-resistant bottom frame design
– Diver-free recovery: acoustic release holds ball floats, line packed in canister.
– Acoustic modem (azimuthal omnidirectional 409 transducer) at apex; all other components below its lower plane.
– Smooth outer surface to limit snaring of fishing gear.
9-km 9-km spacingspacing
340 342 344 346 3480
10
20
30
Win
d sp
eed
(mph
)
340 342 344 346 3480
50
100
150C
orre
ct re
cept
ions
(%) R = 0.5 km
R = 1.0 kmR = 1.5 kmR = 2.0 km
Upward refraction in FRONT-1 caused strong dependence on the sea-surface boundary
Upward refraction in FRONT-1 caused strong dependence on the sea-surface boundary
Year-day
BE
R=
0 (%
)S
/N (
dB
)W
ind
(kt
s)
1485 1490 1495 1500
0
10
20
30
40
50
Sound Speed (m/s)
Depth
(m
)D
epth
(m
)D
epth
(m
)
Range (m)
Sound speed (m/s)
SignalEx
Summary• The environment can have a big effect on
modem performance• These effects are not well understood• SignalEx program
– study a variety of modems in diverse environments– learn which work … and when– optimize modem parameters– develop and validate a channel model to provide a
predictive capability for modem performance
• Result: ’Smart Modem’ selects best operating mode for the channel
SignalEx
Modem schemes tested in SignalEx
type Method Analysis groupa Multi-frequency shift keyed
(MFSK)SAIC/SPAWARSSC
b Frequency-Hopped FSK (FH-MFSK)
Benthos
x Differential phase-shift keyed (DPSK)
Northeastern Univ.SAIC/SPAWARSSC
d N-QAM (BPSK, QPSK, 16-QAM)
Northeastern Univ., Delphi, NUWC, Benthos
e Pulse-Position Modulation (PPM)
SAIC/SPAWARSSC
g Orthogonal Frequency Division Multiplexing (OFDM)
Polytechnic Univ.
h Multi-Carrier Code Division Multiple Access (MC-CDMA)
Polytechnic Univ.
Mk-1, 1998-99Mk-1, 1998-99
Mk-1Mk-1
Mk-2, 2000-01Mk-2, 2000-01
Mk-1Mk-1Sublink’98Sublink’98
Telesonar
testbeds
Telesonar
testbeds
SignalEx
SignalEx waveform
LFM chirps(8-11 kHz)
7-tone comb
LFM chirps(8-16 kHz)
Type-a MFSK waveforms
Type-x DPSK waveforms
SignalEx spectrogram
LFM chirps(8-11 kHz)
7-tone comb
LFM chirps(8-16 kHz)
Type-a MFSK waveforms
Type-x DPSK waveforms
SignalEx 2000 experiment locations
SX-CSan Diego(in SubLink00)May 23-25, 2000 SX-B
New England Shelf(in ForeFront)April 17-20, 2000
SX-DBuzzard’s Bay(in SeaWeb00)August 10-11, 2000
SignalEx-B in ForeFRONT (New England Shelf)April 17-20, 2000
SignalEx-C in Sublink (Point Loma)May 23-25, 2000
SignalEx-D (Buzzard’s Bay)August 10-11, 2000
Ray/beam trace and incoherent TL
CTD Record
Predicted Impulse Response
SX-B (New England Shelf)impulse response
SignalEx
SX-C (San Diego)Eigenrays and impulse response
SX-D (Buzzard’s Bay) impulse response
Drift 1: 0-3.5 km
Drift 2: 2.2-4.2 km
Drift 3: 1.4-3.8 km
Type-x (DPSK) bit error rates
Range = 5 km Range = 7 km
SignalEx
MFSK BER
DSSS/DPSK (type-x)
Transmitter•½ rate, contraint length 7 convolutional coder•(interleaver)•Gold sequence for spreading (4000 chips/sec)•BPSK on I/Q channels (QPSK out) (12 kHz carrier)•Shaping filter
Receiver•RAKE receiver, variable number of taps (or sparse)•Delay-locked loop•Viterbi decoder, 35 stage lookback
(John Proakis/Ethem Sozer Delphi/NEU)
DSSS/DPSK (type-x) bit errors in SX-B
400-bit transmissions8 kHz bandwidth
SignalEx
Transmission 10 bps 50 bps 100 bps 200 bps 400 bps1 0 3 13 400 1582 0 0 3 46 1303 0 0 400 25 1264 0 0 1 10 605 0 0 0 0 306 0 0 0 4 747 0 0 0 0 158 0 0 0 15 1599 0 0 0 15 127
10 0 37 24 28 7311 0 400 205 400 207
DSSS/DPSK bit errors in SX-C64-bit transmissions; 3 kHz bandwidth
SignalEx
50 bps 100 bps 150 bps 300 bps 500 bps0 0 1 17 150 0 0 1 140 0 0 32 170 0 0 8 210 0 0 2 200 0 0 6 210 0 1 3 240 0 1 8 170 0 0 7 210 0 0 9 250 0 0 1 160 0 1 5 210 0 0 5 180 0 0 2 140 0 0 4 200 0 0 7 160 0 37 11 220 0 0 0 160 0 0 8 210 0 0 9 240 0 0 6 21
37 0 0 5 260 0 38 5 240 0 1 9 230 0 0 7 210 0 0 4 220 0 0 6 180 0 0 2 220 0 0 5 20
50 bps 100 bps 150 bps 300 bps 500 bps0 0 2 23 640 0 0 8 190 0 0 12 240 1 1 10 220 0 1 6 200 0 0 11 180 0 0 2 140 0 0 8 150 0 0 12 110 3 14 14 230 0 1 8 270 0 64 11 240 64 3 16 280 0 64 12 26
38 0 0 5 180 0 0 7 180 0 0 4 160 0 0 4 150 0 1 9 160 0 0 8 160 0 0 9 230 0 0 4 130 0 0 6 170 0 0 1 130 0 0 18 220 0 1 11 220 0 1 9 190 1 0 15 210 2 3 22 26
R=3 km R=5 km
DSSS/DPSK bit errors in SX-D (Drift 1)
500-bit transmissions; 8 kHz bandwidth SignalEx
Channel errors Convolutional coding
20 bps 50 bps 100 bps 200 bps 400 bps 20 bps 50 bps 100 bps 200 bps 400 bps7 0 0 11 46 0 0 0 0 387 1 0 5 41 0 0 0 0 230 0 0 1 28 0 0 0 0 00 0 0 0 23 0 0 0 0 8
242 1 5 1 32 121 0 0 0 050 4 0 0 24 36 0 0 0 25 0 0 64 62 0 0 0 34 577 0 0 0 60 0 0 0 0 361 0 0 1 46 0 0 0 0 158 0 2 0 47 0 0 0 0 286 0 0 2 27 4 0 0 0 0
123 0 7 30 32 93 0 0 13 3128 24 26 17 47 4 9 15 9 101 0 0 0 40 0 0 0 0 53 0 0 1 19 0 0 0 0 01 0 0 0 26 0 0 0 0 0
42 0 65 3 86 8 0 55 0 37
DSSS/DPSK bit errors in SX-D (Drift 3)
500-bit transmissions; 8 kHz bandwidth SignalEx
Channel errors Convolutional coding20 bps 50 bps 100 bps 200 bps 400 bps 20 bps 50 bps 100 bps 200 bps 400 bps
0 0 0 3 116 0 0 0 0 1170 0 254 7 43 0 0 125 0 122 0 0 10 62 0 0 0 0 940 0 0 3 56 0 0 0 0 270 0 0 1 83 0 0 0 0 591 0 0 5 42 0 0 0 0 110 0 0 0 66 0 0 0 0 470 0 0 6 145 0 0 0 0 910 0 0 0 22 0 0 0 0 00 0 0 1 38 0 0 0 0 130 0 0 6 33 0 0 0 0 00 0 0 3 25 0 0 0 0 00 0 0 0 41 0 0 0 0 60 0 0 0 31 0 0 0 0 40 0 0 2 48 0 0 0 0 350 0 0 3 33 0 0 0 0 120 0 0 0 30 0 0 0 0 40 0 0 1 193 0 0 0 0 1110 0 0 1 59 0 0 0 0 490 0 0 0 38 0 0 0 0 110 0 0 0 46 0 0 0 0 44
Summary
•Channel impulse response is well-predicted by classical multipath picture
•Type-x multi-access DPSK performs reliably at 100 bps in all cases tested to date (ranges from 0-7 km)
•Further SignalEx analysis will provide common-platform comparisons between many signaling schemes