An environment-aware Q-learning-based trust evaluation scheme inUnderwater Acoustic Sensor Networks (UASNs)

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F04274644%3A_____%2F26%3A%230001279" target="_blank" >RIV/04274644:_____/26:#0001279 - isvavai.cz</a>

Result on the web

<a href="https://www.sciencedirect.com/science/article/abs/pii/S0920548925001163?via%3Dihub" target="_blank" >https://www.sciencedirect.com/science/article/abs/pii/S0920548925001163?via%3Dihub</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.csi.2025.104087" target="_blank" >10.1016/j.csi.2025.104087</a>

Result language

angličtina

Original language name

An environment-aware Q-learning-based trust evaluation scheme inUnderwater Acoustic Sensor Networks (UASNs)

Original language description

Underwater acoustic sensor networks (UASNs) have diverse applications in military and civilian domains but are vulnerable to various security threats due to their broadcast nature and challenging underwater environment. Trust mechanisms have emerged as effective solutions to enhance security and reliability in UASNs. However, existing trust models often lack efficient trust update mechanisms that can manage inevitable dynamic fluctuations in the underwater environment and various potential attacks. In this paper, an environment-aware Q-learning-based trust evaluation (EAQTE) scheme is presented in UASNs. EAQTE incorporates environmental features such as communication channel quality and node stability into the trust computation. Communication quality is assessed based on the variance in successful packet transmission probability, while node stability is measured through movement similarity. Each node collects three types of trust evidence — energy-based, data-based, and communication-based — by interacting with neighboring nodes. Energy-based evidence includes residual energy, current energy change rate, and the similarity of energy change sequences to normal patterns. Data-based evidence evaluates the consistency of collected data, and communication-based evidence considers successful and unsuccessful interactions. EAQTE uses a Q-learning algorithm with three trust states (belief, disbelief, uncertainty) to dynamically adapt trust levels. Simulation results demonstrate that EAQTE improves detection accuracy by 7.01% compared to TUMRL, ARTMM, and TMC based on simulation time. However, under attack mode switching scenarios, EAQTE’s detection accuracy is approximately 2.86% lower than TUMRL. Additionally, EAQTE reduces the false alarm rate by 19.65% relative to TUMRL when node speed varies, and by 11.8% compared to TUMRL under different node densities. Furthermore, EAQTE achieves higher energy efficiency and improves it by 5.19% over TUMRL when the percentage of compromised nodes increases, and by approximately 5.66% across varying node densities. These results indicate that EAQTE effectively balances adaptability, accuracy, and energy consumption in challenging underwater environments.

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2026

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Computer Standards & Interfaces

ISSN

0920-5489

e-ISSN

—

Volume of the periodical

96

Issue of the periodical within the volume

March

Country of publishing house

NL - THE KINGDOM OF THE NETHERLANDS

Number of pages

16

Pages from-to

1-16

UT code for WoS article

001607241600001

EID of the result in the Scopus database

2-s2.0-105022209394