An underwater acoustic network simulation methodology is presented that is based on lookup tables (LUTs) with physical-layer error ratios. These LUTs are prepared with a validated replay channel simulator, which is here driven by channel measurements from the Kauai Acomms MURI 2011 (KAM11) experiment. Three physical-layer candidates are considered: a coherent singlecarrier scheme, a coherent multicarrier scheme, and an incoherent scheme. The three modulation schemes are operated at a fixed message size and at four different data rates. Low rates are more robust to noise, interference, and channel dispersion, but also more prone to collisions in a network with busy traffic, because the packets are longer. Error statistics for colliding packets have been measured and are incorporated in the network simulator as collision LUTs. Example simulations are presented for a reduced flooding protocol with or without retransmissions. The results demonstrate how retransmissions pose a tradeoff between performance at high and low traffic load. The best network performance is obtained if the highest data rate is selected that yields a reasonably well-connected network. Collision avoidance is more important than the extended connectivity offered by low-rate signaling. On the other hand, at a given bit rate, a physical-layer scheme with extended connectivity but relatively weak links is shown to outperform a scheme with stronger links over fewer connections.

No abstract

No abstract

This paper presents results from year-long passive acoustic recordings in the seasonally ice-covered ocean north of Spitsbergen (Svalbard archipelago) from July 2016 to June 2017. Two moorings were deployed by FFI, each with an AMAR Ultra Deep acoustic recorder (JASCO Applied Sciences) equipped with a single hydrophone. The moorings were deployed and retrieved during open-water conditions and remained during periods of partially to near fully ice-covered conditions. We present results from analysis of the acoustic data for ambient noise spectra and statistics, and discuss characteristics of the spectra in relation to environmental factors including ice cover, wind and ocean waves. Seasonal noise spectra are compared with historic measurements from the eastern Arctic. Components of the sound field including transients due to ice, marine mammals, and anthropogenic noise will also be discussed.

This paper presents analysis of a one-year (2018–2019) recording of ambient noise (40–2000 Hz) at a seasonally ice-covered location on the continental slope between the Svalbard archipelago and the Nansen Basin, northeast Atlantic Arctic. Time series of ambient noise show highest correlations with ice concentration and wind speed. A log-wind speed regression model is fitted to spectral noise data for three categories of ice concentration. Wind-speed dependence decreases with increasing ice concentration and increases with frequency, except at high ice concentration. Periodicity in noise during the ice-covered season is related to the M2 and M4 tidal current constituents.