Overview

Introduction

The Subsea Modem (SuM) is a project by SubSeaPulse Srl. It is a low cost, made in Italy, software-defined acoustic modem specifically tailored for underwater telecommunications. The device provides a series of robust modulations and coding schemes to optimise usage for any application, both research and industrial. Nonetheless it is suitable for low-power applications, allowing for battery-powered deployments. Its latest revisions also support real-time audio transmissions, constituting an effective acoustic intercom system.

The SuM is composed of a Raspberry Pi board, an HiFi sound card and the SuM-HAT, the first Raspberry Pi HAT engineered to transform the Raspberry into an analog front-end to pilot underwater acoustic transducers with great performance.

Operation principles

The SuM is a complete device, ready to be coupled with a transducer to stard communicating in the underwater realm. When employed for data transmission, the modem's software is designed to send and receive data through a TCP socket, so that it can be easily interfaced with any user-provided application. When used for audio transmission, the modem just needs to be connected over USB to an audio source and sink to operate.

Data communication

Upon receiving data from the user the modem will encode it into a modulated audio signal. This signal will be generated and amplified by the SuM's hardware in order to be then fed to the connected transducer. When not trasmitting, the modem will be listening for other signals captured by the transducer, ready to decode them and deliver their message to the user.

The device is built as a Software-Defined Modem (SDM), meaning that much freedom is left to the user in setting parameters and choosing modulation and coding schemes. In particular, the SuM supports out-of-the-box the modulation techniques below.

Modulation technique Type
Flexframe Single-carrier/OFDM, various modulations
JANUS Frequency-hopping, BFSK modulation
Analog audio (SSB, LSB, DSB) Analog frequency-modulation for audio streaming

Choosing single-carrier Flexframe allows the user to better specify what modulation and forward error correction scheme to use: below are listed the supported options.

Type Available parameters
Phase shift keying bpsk, qpsk, psk2, psk4, psk8, psk16, psk32, psk64, psk128, psk256
Differential phase shift keying dpsk2, dpsk4, dpsk8, dpsk16, dpsk32, dpsk64, dpsk128, dpsk256
Amplitude shift keying ask2, ask4, ask8, ask16, ask32, ask64, ask128, ask256
Quadrature amplitude qam4, qam8, qam16, qam32, qam64, qam128, qam256
Amplitude and phase shift keying apsk4, apsk8, apsk16, apsk32, apsk64, apsk128, apsk256
On-off keying ook

full list

Type Parameter
Hamming (7,4) h74
Hamming (7,4) plus parity bit h84
Hamming (12,8) h128
Repeat x3 rep3
Repeat x5 rep5
SEC-DED (72,64) secded
Convolutional K=7, df=10 cv27
Convolutional K=9, df=12 cv29
Convolutional K=9, df=18 cv39
Convolutional K=15, df<=57 cv615
Reed-Solomon m=8 rs

Real time audio communication

Recent SuM versions support, alongside the usual digital data transmission, analog transmission of a real-time vocal audio stream (300Hz - 3kHz bandwidth). This enables applications such as intercoms or remote-operated speakers. A push-to-talk button lets the user operate the modem as either a transmitter or a receiver, establishing a real time, half-duplex link over the underwater channel.