Acoustic & Hydroacoustic Telemetry

Why Acoustic? The RF Problem in Water

Radio waves attenuate dramatically in seawater. The conductivity of seawater (approximately 4 S/m at 20°C) causes exponential absorption of electromagnetic energy. At 1 GHz, the skin depth in seawater is only about 0.2 cm — meaning RF signals are absorbed within centimeters. Even at ELF (3–30 Hz), the practical range is limited to a few thousand kilometers, and data rates drop to characters per minute.

Sound, by contrast, propagates through water with remarkably low attenuation. The speed of sound in seawater is approximately 1500 m/s — five orders of magnitude slower than the speed of light (3×10⁸ m/s) in vacuum. This enormous difference in propagation velocity makes sound waves practical for long-range communication where electromagnetic waves cannot survive. Acoustic attenuation in seawater is typically 0.01–0.1 dB/km at frequencies below 10 kHz, allowing communication over tens of kilometers with modest transmit power.

Underwater Channel Characteristics

The underwater acoustic channel is among the most challenging communication environments known. Unlike radio channels, the acoustic channel exhibits characteristics that defy conventional assumptions:

• Extreme multipath: Sound reflects off the sea surface, bottom, and any underwater structures. A single transmission may arrive at the receiver via 10+ distinct paths, with path differences spanning hundreds of meters. The delay spread can reach 10–100 ms in shallow water — comparable to the entire symbol period at practical data rates.
• Doppler spread: Surface waves, transmitter/receiver motion, and current-driven scatterers create Doppler shifts of 2–5 Hz even at acoustic carrier frequencies of 10–30 kHz. This is a Doppler spread that is significant relative to the signal bandwidth.
• Frequency-dependent attenuation: Acoustic absorption increases with the square of frequency (approximately α ∝ f²), creating a hard upper bound on usable bandwidth. At 10 kHz, total absorption is about 1 dB/km; at 100 kHz, it rises to 10 dB/km; at 1 MHz, it exceeds 100 dB/km.
• Slow propagation: The 1500 m/s speed of sound introduces significant latency. A signal from a source 1.5 km away arrives 1 second later — unacceptable for many real-time applications but tolerable for telemetry and monitoring.
• Variable sound speed profile: Temperature, salinity, and pressure create a layered sound speed profile (SSP) that bends (refracts) acoustic rays, creating shadow zones, convergence zones, and surface ducts that dramatically alter propagation.

Acoustic Frequency Bands

Acoustic telemetry systems operate across several frequency bands, each offering a different tradeoff between range, data rate, and transducer size. The choice of frequency band is the single most important design decision in any underwater acoustic system.

VLF Acoustic (10–30 Hz)

The lowest acoustic frequencies are used for extreme-range applications such as submarine communication, seismic monitoring, and deep-ocean acoustic tomography. At these frequencies, absorption is negligible (less than 0.001 dB/km), enabling communication over thousands of kilometers through the SOFAR channel. However, the available bandwidth is minuscule — typically less than 10 Hz — limiting data rates to a few bits per second. Transducers are large (meters in diameter) and tuned to narrow frequency bands. The US Navy's SOSUS (Sound Surveillance System) operated primarily in the 10–30 Hz band for passive detection of submarine noise signatures.

MF Acoustic (30–300 kHz)

The mid-frequency acoustic band is the workhorse of commercial and research underwater telemetry. Systems operating at 30–300 kHz offer a practical compromise: absorption ranges from 0.1 to 10 dB/km, enabling communication over distances of 1–10 km, while available bandwidth (tens of kHz) supports data rates from 100 bits/sec to 100 kbit/s. Transducers are compact enough to mount on AUVs and ROVs. Most commercial acoustic modems — including those from Telesonics, LinkQuest, and LinkQuest — operate in this band. Applications include AUV command and control, sensor network telemetry, diver communication, and subsea instrumentation.

HF Acoustic (300 kHz–1 MHz)

High-frequency acoustic systems sacrifice range for bandwidth. At 500 kHz, absorption exceeds 50 dB/km, limiting range to a few hundred meters. However, the available bandwidth (100+ kHz) enables data rates exceeding 1 Mbit/s over short distances. HF acoustic is used for high-bandwidth applications: real-time video from AUVs, high-resolution sonar imaging, and short-range acoustic LANs connecting underwater sensor clusters. The range limitation is a fundamental tradeoff — at 1 MHz, useful communication range drops to tens of meters, making HF acoustic essentially a short-range wireless technology.

Modulation Schemes for Underwater Channels

The underwater acoustic channel demands modulation techniques that are robust against extreme multipath, Doppler spread, and time-varying attenuation. Several modulation approaches have been developed and refined for underwater use:

Frequency-Shift Keying (FSK)

Non-coherent FSK is the most widely deployed modulation in commercial underwater modems. The transmitter alternates between two (or more) discrete frequencies to represent binary 0s and 1s. Because FSK does not require phase coherence at the receiver, it is inherently robust against Doppler-induced phase fluctuations. The tradeoff is low spectral efficiency — typically 0.5–1 bit/s/Hz — but this is acceptable when the channel provides only a few kilohertz of usable bandwidth. M-ary FSK (using 4, 8, or 16 frequency tones) can increase data rate at the cost of increased receiver complexity and reduced noise immunity.

Phase-Shift Keying (PSK)

Coherent PSK (BPSK, QPSK) offers higher spectral efficiency (1–2 bit/s/Hz) than FSK, but requires accurate carrier phase estimation at the receiver. In the underwater channel, Doppler spread and multipath cause rapid phase fluctuations that make carrier recovery challenging. Adaptive equalizers (decision-feedback or linear) are essential for PSK underwater systems. The combination of DFE (Decision Feedback Equalizer) with PSK modulation has been the standard for high-rate underwater telemetry since the 1990s. Doppler compensation using resampling or frequency tracking loops must precede equalization.

Orthogonal Frequency-Division Multiplexing (OFDM)

OFDM has been adapted from terrestrial wireless to the underwater acoustic domain, though significant modifications are required. The long multipath delay spread of the underwater channel (10–100 ms) necessitates long OFDM symbols with correspondingly narrow subcarrier spacing. Doppler shifts must be compensated using a combination of carrier frequency offset (CFO) estimation and Doppler resampling algorithms. When properly implemented, OFDM achieves high spectral efficiency and naturally handles frequency-selective fading. Recent research demonstrations have achieved OFDM data rates exceeding 50 kbit/s over distances of 1 km in shallow water.

Spread Spectrum

Direct-sequence spread spectrum (DSSS) and time-hopping spread spectrum (THSS) are used in applications requiring multiple access, low probability of intercept, or interference rejection. The spreading gain provides resilience against narrowband interference and multipath fading. Undersea sensor networks using CDMA (Code Division Multiple Access) can share the acoustic channel among dozens of sensors, each using a unique spreading code. The tradeoff is reduced data rate — spreading factors of 10–100 are typical, reducing effective throughput by an equivalent factor.

Applications

Submarine Communication (ELF/VLF)

Nuclear-powered submarines on patrol are among the most challenging communication targets. They cannot use RF antennas at depth, and they must receive messages without revealing their position. The solution is ELF (3–30 Hz) and VLF (3–30 kHz) one-way acoustic/electromagnetic communication. The US Navy's Project Sanguine (later Seafarer) operated at 76 Hz to communicate with submarines through the Earth itself. Modern submarine communication uses ELF/VLF as a "wake-up call" — a short pre-formatted message that instructs the submarine to surface or deploy a buoy for higher-bandwidth communication via satellite.

• ELF (3–30 Hz): Penetrates to full ocean depth, data rate ~1 character per minute, requires submarine trailing wire antenna 1–2 km long
• VLF (3–30 kHz): Penetrates to 100–200 m patrol depth, data rate 10–200 bits/sec, submarine uses near-surface towed antenna
• Built-in redundancy: Messages are repeated many times to overcome fading; submarines may receive incomplete messages and request retransmission at scheduled intervals

AUV and ROV Telemetry

Autonomous Underwater Vehicles (AUVs) and Remotely Operated Vehicles (ROVs) require real-time or near-real-time telemetry for mission control, status monitoring, and data offload. AUVs typically use acoustic modems at 20–40 kHz for command and control during launch and recovery, and higher frequencies (100+ kHz) for mid-mission data transfers. ROVs connected by fiber-optic tethers do not require acoustic telemetry for primary communication, but use acoustic links for diver-ROV coordination and short-range positioning. Data rates for AUV telemetry range from 100 bits/sec (low-power sensor data) to 100 kbit/s (compressed imagery over short range).

Oceanographic Sensor Networks (ARGO Floats)

The Argo program maintains over 4000 profiling floats distributed across the world's oceans. Each float descends to 2000 m, drifts with deep currents, then ascends while measuring temperature and salinity. At the surface, the float transmits its data to satellites via Iridium. However, Argo floats also use underwater acoustic telemetry for internal communication: command interrogation from surface ships, inter-float ranging for position fixes, and data relay from seafloor instruments to surface floats. The acoustic subsystem operates at 6–12 kHz with data rates of 80–500 bits/sec, sufficient for the small data packets required by Argo instruments.

Seismic Monitoring and Ocean Bottom Seismometers

Ocean bottom seismometers (OBS) record earthquakes and seismic reflections from beneath the seafloor. Data must be retrieved either by physically recovering the instrument (costly and slow) or by acoustic telemetry to a surface relay. Acoustic telemetry of seismic data is limited by the enormous data volumes — a single 3-component seismometer sampling at 200 Hz generates ~1 Mbit/day. Compressed or event-triggered transmission over acoustic links (1–10 kbit/s) can transmit seismic event data in near-real-time, while bulk data offload requires physical recovery or high-rate short-range acoustic transfer during ship passes.

Diver Communication

Military and commercial divers require voice and data communication. Underwater acoustic communication systems for divers operate at 20–40 kHz with ranges of 100–500 m. The US Navy's AN/UQC-1 diver communication system and its successors provide encrypted voice communication between divers and surface support. Modern diver comms systems integrate acoustic ranging for diver tracking, enabling surface operators to monitor diver positions in real-time while maintaining voice links.

Commercial Acoustic Modem Systems

Several manufacturers produce commercial acoustic modems for underwater telemetry. These systems vary in frequency, data rate, range, power consumption, and form factor. The choice of modem depends on the specific application requirements:

Telesonics (Teledyne Marine)

• LinkQuest SoundLink series: Mid-frequency modems (20–300 kHz), data rates up to 315 kbit/s, ranges from 100 m to 10 km depending on frequency and conditions
• Wideband modems: Adaptive frequency-hopping OFDM, automatically selects optimal bandwidth and carrier frequency based on channel conditions
• Low-power modes: Sensor network modems operating at microwatt power levels, enabling battery life measured in years

LinkQuest

• UWM series: Universal acoustic modems supporting FSK, PSK, and OFDM modulation with automatic rate adaptation
• Data rates: 70 bit/s to 153 kbit/s depending on range and channel conditions
• Network topology: Point-to-point, point-to-multipoint, and mesh network configurations

Data Rate vs. Range Tradeoff

The fundamental tradeoff in underwater acoustics is range vs. data rate. This is governed by the physics of absorption and the available signal-to- noise ratio (SNR):

• 10 km range: Maximum data rate approximately 1–10 kbit/s at 10–20 kHz carrier frequency, limited by absorption and multipath
• 1 km range: 10–100 kbit/s achievable at 20–50 kHz carrier, with adaptive equalization
• 100 m range: 100 kbit/s to 1 Mbit/s possible at 100–500 kHz carrier, sufficient for compressed video
• 10 m range: Multiple Mbit/s at 500 kHz–1 MHz, used for short-range AUV docking and high-rate data transfer

These rates assume typical shallow-water conditions. Deep-water channels with less multipath can achieve somewhat higher rates, while highly reverberant environments (harbors, under ice) may degrade performance significantly.

Channel Modeling and Propagation

Understanding underwater acoustic propagation requires sophisticated modeling tools that account for the ocean environment. Unlike RF channel models, acoustic models must handle the full complexity of 3D ocean acoustics including variable sound speed, bathymetry, and surface/bottom interactions.

Sound Speed Profile (SSP)

The speed of sound in seawater depends on three variables: temperature, salinity, and pressure (depth). The empirical formula (Mackenzie, 1981) is:

c = 1448.96 + 4.591T − 0.05304T² + 0.0002374T³ + 1.340(S − 35) + 0.01630D + 1.675×10⁻⁷D²

where T is temperature (°C), S is salinity (ppt), and D is depth (m). Typical values: ~1500 m/s at the surface in temperate waters, increasing to ~1550 m/s at 4000 m depth due to pressure. The resulting sound speed profile varies with latitude, season, and time of day, creating the complex refractive environment that governs acoustic propagation.

Key Propagation Effects

• Thermocline: The layer of rapid temperature decrease (typically 50–200 m depth) causes strong downward refraction, creating a shadow zone below the thermocline where direct-path signals are attenuated. The thermocline acts as an acoustic lens, focusing energy downward and reducing communication range for surface-to-deep links.
• Surface duct: In conditions where the surface mixed layer has a positive sound speed gradient (due to pressure), sound rays are refracted upward and reflected off the surface, creating a surface duct. Signals trapped in this duct propagate long distances with minimal loss, enabling long-range communication at moderate frequencies.
• SOFAR channel: The SOFAR (Sound Fixing and Ranging) channel exists at the depth of minimum sound speed (typically 600–1200 m in temperate waters). Rays launched near this depth are trapped by refraction, propagating thousands of kilometers with spreading loss only — no bottom or surface interaction. This is the ultra-long-range acoustic propagation channel used by ocean acoustic tomography and SOSUS.
• Convergence zones: In deep water, rays refracted downward by the thermocline curve back upward due to the pressure gradient, focusing at the surface approximately 50–70 km from the source. These convergence zones provide periodic high-intensity signal regions that can be exploited for long-range communication.
• Bottom reflections: In shallow water, the dominant propagation mode is repeated surface-bottom reflection. The bottom interaction introduces loss and phase changes that depend on bottom type (sand, mud, rock). Soft bottoms (mud, silt) absorb more energy than hard bottoms (rock, coral), affecting both range and signal quality.

Bellhop Ray Tracing

The Bellhop model (developed at SIO, now widely used) is the standard tool for predicting underwater acoustic propagation. It solves the acoustic wave equation using ray theory with Gaussian beam tracking, which avoids the caustic singularities of pure ray theory. The model takes as input the sound speed profile, bathymetry, surface and bottom properties, and source/receiver geometry. Outputs include transmission loss, impulse response, arrival angle, and eigenray identification. Bellhop runs in 2D (range-dependent) and 3D modes and is used to design and evaluate acoustic communication links, sonar systems, and underwater sensor networks.

• Input parameters: SSP, bathymetry (range-dependent), bottom geoacoustic properties (density, sound speed, attenuation), surface roughness, source depth/position, frequency
• Output products: Transmission loss vs. range/depth, impulse response (multipath structure), arrival time, angle of arrival, eigenray identification
• Applications: Link budget analysis, modem performance prediction, optimal source/receiver placement, communication system design

Shallow Water vs. Deep Water Propagation

The character of acoustic propagation changes fundamentally between shallow water (depths less than 200 m) and deep water (depths greater than 1000 m):

• Shallow water: Strong multipath from surface and bottom reflections, modal dispersion, depth-limited propagation, high reverberation. Communication is difficult due to the dense multipath (delay spread 10–100 ms). Range is limited to a few kilometers at useful data rates.
• Deep water: Less multipath (refraction dominates over reflection), longer range (SOFAR channel enables thousands of km), convergence zones create periodic signal enhancement. However, the deep-water channel can still exhibit significant Doppler and time variability due to internal waves and ocean dynamics.

Current Research Frontiers

Underwater acoustic communication remains an active area of research, driven by the growing demand for ocean observation, autonomous vehicles, and subsea industry. Key research areas include:

• MIMO acoustic systems: Using multiple transducers and hydrophones to exploit spatial diversity and increase data rates without increasing bandwidth
• Adaptive modulation and coding: Real-time channel estimation to select optimal modulation scheme, coding rate, and bandwidth allocation
• Machine learning for equalization: Neural network-based channel equalizers that adapt to rapidly changing channel conditions without explicit channel estimation
• Underwater optical-acoustic hybrids: Combining acoustic ranging and signaling with optical data links (blue-green laser) for short-range high-bandwidth communication
• Cooperative and networked acoustics: Multi-hop acoustic networks with relay nodes, network coding, and distributed MIMO to extend range and increase aggregate throughput
• Low-power acoustic IoT: Miniaturized, ultra-low-power acoustic modems for dense underwater sensor networks, enabling continuous ocean monitoring at unprecedented scale