Periodic Reporting for period 1 - MONUSEN (MONtenegrin center for Underwater SEnsor Networks)
Periodo di rendicontazione: 2022-06-01 al 2023-08-31
In the realm of sustainable collaboration, which forms our second objective, UoM researchers have actively collaborated with partner institutions to execute three targeted research actions focused on the design of efficient and secure USNs, for fixed and dynamic topologies. The outcomes of these collaborative research endeavors include the development of an innovative chaos-based modulation scheme, adaptive MAC protocol solutions that effectively manage the complexity of underwater acoustic channels, a blockchain-assisted authentication system for Underwater Acoustic Sensor Networks (UASNs), and novel cooperative guidance control protocols for unmanned surface vehicles. To empirically validate these research findings, we have carried out three joint experimental campaigns. Our published work, so far, includes two Q1 journal articles and six peer-reviewed conference papers.
The third project objective centers on enhancing UoM's research management capacities. To this end, a dedicated research management unit has been established within UoM's Faculty of Electrical Engineering. The effectiveness of this unit has been further elevated through targeted capacity-building measures, including two staff exchanges, three expert-visits, and two specialized research management training sessions led by the leading partners.
For our fourth objective, focused on widening UoM's scientific influence, we’ve engaged in various networking activities. These include co-organization of two research-industry workshops (EU-funded projects in Marine Robotics and Applications, EMRA), two specialized summer schools. These events have served not only to broaden our research collaborations but also to strengthen links with industry and potential end-users of USN technology.
Adaptive MAC Protocol Based on Reinforcement Learning
In a significant departure from existing approaches, our project developed an adaptive Medium Access Control (MAC) protocol that employs reinforcement learning, treating sensor nodes as asynchronous learning agents. This enables the nodes to continuously optimize their transmission strategies through trial-and-error, eliminating the need for gathering detailed network information. The nodes' transmission strategies involve a combination of time-slot and transmission-offset selection, resulting in a system that not only enhances network resilience against changing topologies and environmental conditions but also outperforms existing reinforcement-based solutions in this space.
Advanced Waveforms for Underwater Communication
Novel waveforms based on chaotic signals were developed, addressing key challenges in underwater communication. Unlike traditional modulation techniques, these novel waveforms demonstrate exceptional resilience to adverse conditions, including low signal-to-noise ratios and severe multipath distortions. Their unique noise-similar spectrum offers the added advantage of minimizing ecological impact on marine life. Importantly, the efficacy of these waveforms has been rigorously validated in real-world conditions through a series of North Sea trials. These tests confirmed a 100% success rate in message detection and decoding across substantial distances—up to 5km—solidifying their potential as a robust and eco-friendly solution for underwater communications.
Authentication methods for mobile acoustic USNs
The project has engineered an authentication method underpinned by blockchain technology, specifically tailored for large-scale acoustic USNs featuring mobile sensor nodes. The blockchain network is formed from multiple surface gateways that act as cluster heads in the network. Acting as a decentralized ledger, the blockchain securely stores network metadata and offers distributed certificate authority services, thereby eliminating vulnerability points and fortifying network resilience. Additionally, the blockchain framework provides a robust defense against replay attacks by storing comprehensive communication histories between all network nodes. The method is geared towards facilitating fast and lightweight authentication processes, an essential feature given that mobile nodes frequently switch between surface gateways due to natural or intentional mobility patterns.
Multi-Vehicle Cooperation in UASNs
A distributed control protocol was developed to solve the output tracking problem in multi-vehicle USNs. The proposed protocol enables follower vehicles to track or maintain a specific distance (thus forming the desired formation) from a reference trajectory generated by the leader vehicle. Unlike existing protocols, the proposed solution does not require the followers to share controller internal information. This significantly reduces communication costs and allows the protocol to be implemented using only relative distance sensors.
Sideslip Compensation in USV Guidance
An improved Line of Sight (LOS) guidance law was proposed to tackle the problem of sideslip in USVs caused by drift forces like ocean currents, wind, and waves. This approach estimates the sideslip angle concurrently with the cross-track error, using an augmented extended Kalman filter. The result is a more effective and accurate guidance system for USVs.
The advancements made have the potential to redefine the state-of-the-art in USNs, with far-reaching impacts across diverse applications like environmental monitoring, natural resource exploration, and national defense.