Even though Underwater Wireless Sensor Networks (UWSNs) share some common properties with terrestrial sensor networks, such as limited resources at the nodes, acoustic waves have to be used for communications and the way they propagate in water makes a significant difference from conventional WSNs. Unlike the air, water as a propagation medium is highly asymmetric, so that the setup of any two-way communication mechanism necessitates dedicated design. The major challenges to designing UWSNs are the long propagation delays, low bandwidths and the interference coming from nodes that could be far away in the network. Solutions for terrestrial WSNs cannot be directly deployed in UWSNs, and new protocol stacks are required for both single-hop and multi-hop communications.
The SUNSET Framework
In order to perform real-life testing, we have developed a new testing architecture based on the well know network simulator ns-2, to test new protocols for underwater sensor networks in realistic scenarios.
The proposed framework, named SUNSET , for Sapienza University Networking framework for underwater Simulation, Emulation and real-life Testing, aims at providing a powerful networking solution kit for sensor network researchers. The basic idea concerns re-using code written for simulations to perform tests in real-life. This should make it easier for the networking community to design and test protocol stacks, since many of them have been already written and simulated through ns-2 (which is written in C and C++). Researchers can first perform preliminary simulations to validate their protocols. They can then port their ns-2 code on real devices (we have used Gumstix devices for this purpose, which have OpenEmbedded Linux as Operating System), interfacing it with real hardware for data transmissions, to evaluate the protocol performance in a real testbed.
Although the proposed framework is general enough to be used for emulation and real-life testing of any kind of networks (radio, optical, underwater, etc.), it has been widely and extensively validated in the underwater environment. Different commercial acoustic modems have been used at the price of providing new drivers for the interaction between ns-2 and the modem. Moreover, it has been successfully integrated and interfaces with environmental underwater sensors (for temperature and CO2 and methane concentrations) and the navigation system of Autonomous Underwater Vehicles (AUVs).
More details can be found HERE
Research on MAC Protocols
We have dedicated particular attention to the study and design of Medium Access Control (MAC) protocols that would account for the many challenges imposed from acoustic signal propagation.
A new CSMA-based MAC solution has been proposed, termed PDAP for Propagation Delay-Aware Protocol , that aims at maximizing the bandwidth usage also enabling interleaved, yet reliable communications between different pairs of nodes. A performance evaluation comparison of the proposed solution with a selection of underwater MAC protocols (DACAP, T-Lohi, APCAP, CSMA, slotted CSMA) as representative of different channel access techniques has been conducted in order to evaluate their pros and cons. Moreover, different possible solutions have been studied to improve network performance counteracting aspects of acoustic underwater communication:
- Optimal packet size selection in multi-hop underwater acoustic network , . We have seen that packet size is a fundamental parameter whose proper selection has significant impact on throughput, latency, and energy consumption.
- We have investigated data and control channels multiplexing for random channel access based on RTS/CTS
exchange , with the objective to
eliminate the collisions between data and control packets, which have been shown to cause remarkable loss in throughput.
Results show that multiplexing offers improvements in throughput efficiency and on energy consumption.
- Packet fragmentation and selective repeat ARQ .
The selection of the optimal number of fragments results in improved performance in terms of throughput efficiency,
energy per bit consumption, packet latency per meter and route length, i.e., the benefits of reducing control/data collisions
effects through fragmentation outweigh the higher overhead that fragmentation itself imposes.
Detailed description: Show
Research on Cross-layer Protocols
MAC and routing protocols face new design challenges with respect to their terrestrial counterpart, in that channel access mechanisms and routing techniques are highly affected by a link quality that varies so differently from wireless radio links. Most of the solutions proposed for UWSN protocols at the MAC and routing layers address the problem of channel access and multi-hop routing separately. Recently, however, it has been shown that cross-layer techniques can impact protocol performance positively, especially in networks with limited resources and/or deployed in challenging environments, like UWSNs.
A new cross layer routing protocol for underwater wireless sensor networks has been proposed and investigated. The solution, termed CARP for Channel-aware Routing Protocol , exploits link quality information for cross layer relay determination. Nodes are selected as relays if they have a (recent) history of successful transmissions to their neighbors. CARP combines link quality with simple topology information (hop count), thus being able to route around connectivity voids and shadow zones. The protocol is also designed to take advantage of power control for selecting robust links.
The performance of CARP has been evaluated through ns2-based simulations, and compared to the performance of two previously proposed routing protocols, namely, FBR and DBR. The obtained results show that CARP robust relay selection mechanism enables it to achieve throughput efficiency that is up to twice the throughput of FBR and almost three times that of DBR. CARP also obtains remarkable performance improvements over FBR and DBR with respect to end-to-end packet latency and energy consumption. The simulations results proves that including link quality explicitly into relay selection is key to obtain superior throughput efficiency, end-to-end latency and energy consumption.
In April 2012, first investigation of CARP performance running at sea experiments have been conducted at the NATO Centre for Maritime Research & Experimentation (CMRE), formerly NURC. Five static nodes and two mobile nodes have been used.
More details can be found HERE.
Applications and Experiments
CO2 ocean monitoring systems
The geological Carbon Capture and Storage (CCS) technique consists of capturing CO2 from power and industrial activities and storing it in deep geological reservoirs in order to prevent large quantities of CO2 from being released into the atmosphere. CCS has the potential to contribute over 15% of the needed reductions in CO2 emissions by 2050, while at the same time providing society with a "bridge" between fossil fuels and renewable energy supplies.
Current CO2 ocean monitoring systems are typically based on one of two different approaches:
- Deployment of underwater nodes that record data during the monitoring mission, and then recover the instruments to retrieve the data.
- Cabling of the underwater nodes to a surface station in order to collect the data on-line and in real-time.
Detailed description: Show
AUV navigation control
When Autonomous Underwater Vehicles (AUVs) are used, there is usually no real-time on-line control on the vehicle once it is underwater, or the control on the vehicle is really limited. A mission file, which contains all the tasks the vehicle has to perform, is usually preloaded before the vehicle goes underwater and the mission starts. Real-time and on-line interaction with the AUV is usually possible when the vehicle is cabled or when it is at the sea surface, using a radio link, if radio communication is supported.
To overcome these limitations we have integrated and interfaced our framework SUNSET with the AUV control system. As we have done for the acoustic modems, we have designed and implemented a communication interface to interact with the control system of the AUV. The AUV can be locally or remotely instructed to perform a specific action, e.g. to navigate to a given location, to start the detection of some event, to deliver navigation data or environmental measurements to a central station, etc. To remotely control the AUV operations, requests/commands are sent to the AUVs through acoustic links, using the networking capabilities of our framework. The user can select the desired protocol stack to be used: MAC, routing, etc. The framework is responsible to create the acoustic packet at the transmitter and to deliver it to the selected destination. Once the acoustic packet is received the framework will extract the payload containing the request/command and provide it to the AUV control system in order to perform the requested action.
SUNSET has already been interfaced with two different mobile vehicles: Folaga (produced by Graaltech) and MARES (produced by the Oceansys group).
More details can be found HERE.
ASV navigation control
Autonomous Surface Vehicles (ASVs) usually support both radio and acoustic communication. Radio communication is generally used to control the vehicle operation in a real-time and on-line way, while acoustic communication is considered for a basic interaction with nodes deployed underwater. However, it could be possible that in some scenarios radio connection is not properly working or it could be not available (ASV moving out of the radio communication range of the control station, signal reflection on the sea surface making the radio link not stable, etc.) Therefore, acoustic communication can be used as a backup line to control the vehicle operations (providing longer communication range even though with lower bit rates). Moreover, a smart and more efficient use of acoustic communication capabilities allows to create heterogeneous networks where ASVs, AUVs and static nodes (moored or at the surface) can cooperate together to perform collaborative tasks.
For this purpose, as previously done for the MARES AUV vehicle, we have integrated and interfaced SUNSET with the control system of the ASV developed by the Oceansys group. Acoustic commands can be sent in a real-time and on-line way to remotely instruct the ASV on the action to perform (in a single hop or multi hop way) and to let the vehicle acoustically cooperate with the other nodes deployed in the network. The ASV can be locally or remotely instructed to perform a specific action, e.g. to navigate to a given location, to start the detection of some event, to deliver navigation data or environmental measurements to a central station, etc. SUNSET communication and networking capabilities are used to allow the interaction and data exchange among the different nodes in the network. For the communication at the surface radio links can also be used, if available. The user can select the desired protocol stack to be used (MAC, routing, etc.) according to the application requirements or the network conditions. SUNSET is responsible to create the acoustic packet at the transmitter and to deliver such packet to the intended destination. Once the acoustic packet is received the framework will extract the payload containing the request/command and provide the message to the ASV control system in order to perform the requested action.
SUNSET has already been interfaced with the Zarco and Gama ASVs (produced by the Oceansys group).
More details can be found HERE.
Improvements to existing simulators
Improvement to the ns2-Miracle network simulator and WOSS underwater acoustic channel model to make the simulations more realistic and adherent to environmental changes as well as to the behavior of underwater nodes.
In the original version, the ns2-MIRACLE platform was assuming an over-simplistic packet interference model, thus impacting on the correct computation of the bit/symbol/packet error rates.
When two or more packet receptions were overlapping at some node, the ns2-MIRACLE network simulator was not considering the different packet chunks where the overlapping was actually occurring, but it was averaging the interference of the interfering packets throughout the entire packet and assuming such interference constant for SINR computation. In this way, a single under-estimated interference value is considered over the entire packet instead of different higher interference values over the exact packet chunks where the packets overlapping was occurring.
An improved packet interference model has been implemented, within the CLAM project, by the UWSN Group together with the SIGNET Lab of the University of Padova and made available to the community in the latest version of ns2-MIRACLE. In the improved packet interference model, if the interference level varies throughout the duration of a packet, the typically different error rates that result are separately accounted for in the computation of the overall packet error rate. Therefore, the error probability is now computed over all separate chunks having the same interference level (Figure). Chunk-per-chunk packet error rate computation is more precise in the presence of bursty interference (e.g., short signaling packets).
|Figure: Improved packet interference model.|
Underwater networks may embrace long periods of (simulated) time. Environmental features such as the temperature of the various water layers and the resulting sound speed profile may change during long periods of time, and should be tracked in order to provide realistic simulation results. First versions of WOSS simulator only allowed to specify one sound speed profile at the beginning of the simulation, keeping this environmental parameter constant for the whole simulation time.
In order to model the environment and its changes with better accuracy, the UWSN Group together with the SIGNET Lab has implemented an option to change sound speed profile over time in WOSS .
In particular, the user can specify a set of SSPs and the time where they have to be used. At time t0, the first SSP has to be used. This SSP should be substituted by the second SSP at time t1, by the third SSP at time t2 and so on. After the last SSP has been specified, WOSS automatically wraps around and restarts from the first SSP provided. Therefore, if the user has specified that the provided SSPs span the time of one day and the simulation time covers several days, the same sequence of SSP is repeated automatically for every simulated day.
These improvements have been made available to the community at large in the new WOSS version 1.3, freely downloadable from the WOSS webpage.
We have designed analytical models to benchmark routing and scheduling solutions for multi-hop UWSNs . It produces the optimum solution for small to medium scale underwater networks. Moreover, scalable, centralized heuristics have been investigated, which combine approximate analytical models and scheduling heuristics, and are able to generate solutions close to the optimum. The overall result is a powerful tool to derive benchmark results (upper bounds) for underwater protocol performance and to understand the trade-offs and performance limits of such systems. The routing/scheduling problem has been formulated into an Integer Linear Programming (ILP) model which yields the joint optimal routing and scheduling which maximizes the network throughput and minimizes the energy consumption.
Differently from most of prior research in the area, we consider an accurate interference model where we account not only for the presence of other single transmitters but also for the combination of multiple transmissions that can overlap during the reception at the packet destination. The resulting solution can only scale to a few tens of nodes. This is in any case larger than current deployments but demands for more scalable heuristic solutions to understand the performance trade-offs of future UWSNs.
We have therefore proposed a heuristic solution which scales well with the problem size. Since the complexity of the optimal strategy stems from the need to jointly optimize routing and scheduling, in the heuristic we use the "divide et impera" principle: First we compute the routing; then, with the routing fixed, we schedule packets transmissions.
Our experiments show that, for network sizes where the optimum can be derived, the proposed heuristic produces results very close to the optimal ones (we observed a throughput degradation of 7% on average) at a fraction of time. Moreover, the proposed heuristic is able to scale to larger networks.
The proposed solution can therefore be used as guideline and benchmark for the design and evaluation of distributed protocol stacks.
Chiara Petrioli and Roberto Petroccia, "SUNSET: Simulation, Emulation and Real-life Testing of Underwater Wireless Sensor Networks," in Proceedings of IEEE UComms 2012, (Sestri Levante, Italy), IEEE Oceanic Engineering Society, September, 12-14 2012. - [BIB] - [PDF]
C. Petrioli, R. Petroccia, and M. Stojanovic, "A comparative performance evaluation of MAC protocols for underwater sensor networks," in Proceedings of MTS/IEEE OCEANS 2008, (Quebec City), MTS/IEEE Oceanic Engineering Society, September 15-18 2008. - [BIB] - [PDF]
S. Basagni, C. Petrioli, and R. Petroccia, and M. Stojanovic, "Optimized Packet Size Selection in Underwater WSN Communications," IEEE Journal of Oceanic Engineering, Volume 37, Issue 3, pp. 321-337, July 2012. - [BIB] - [PDF]
S. Basagni, C. Petrioli, R. Petroccia, and M. Stojanovic "Choosing the Packet Size in Multi-hop Underwater Networks," in Proceedings of MTS/IEEE OCEANS 2010, (Sydney, Australia), IEEE Oceanic Engineering Society, May, 24-27 2010. - [BIB] - [PDF]
S. Basagni, C. Petrioli, R. Petroccia, and M. Stojanovic, "Multiplexing Data and Control Channels in Random Access Underwater Networks," in Proceedings of MTS/IEEE OCEANS 2009, (Biloxi, Mississippi, USA, October), October, 26-29 2009. - [BIB] - [PDF]
S. Basagni, C. Petrioli, R. Petroccia, and M. Stojanovic "Optimizing Network Performance through Packet Fragmentation in Multi-hop Underwater Communications," in Proceedings of MTS/IEEE OCEANS 2010, (Sydney, Australia), IEEE Oceanic Engineering Society, May, 24-27 2010. - [BIB] - [PDF]
C. Petrioli, and R. Petroccia, and J. Potter, "Performance evaluation of underwater MAC protocols: From simulation to at-sea testing," in Proceedings of IEEE/OES OCEANS 2011, (Santander, Spain), IEEE Oceanic Engineering Society, June, 6-9 2011. - [BIB] - [PDF]
S. Basagni, C. Petrioli, R. Petroccia, and D. Spaccini, "Channel-aware Routing for Underwater Wireless Networks," in Proceedings of MTS/IEEE OCEANS 2012, (Yeosu, Korea), IEEE Oceanic Engineering Society, May, 21-24 2012. - [BIB] - [PDF]
A. Annunziatellis, S. Graziani, S. Lombardi, C. Petrioli and R. Petroccia, "CO2Net: A marine monitoring system for CO2 leakage detection," in Proceedings of MTS/IEEE OCEANS 2012, (Yeosu, Korea), IEEE Oceanic Engineering Society, May, 21-24 2012. - [BIB] - [PDF]
S. Azad, P. Casari, C. Petrioli, R. Petroccia, and M. Zorzi, "On the impact of the environment on MAC and routing in shallow water scenarios," in Proceedings of MTS/IEEE OCEANS 2011, (Santander, Spain), IEEE Oceanic Engineering Society, June, 6-9 2011. - [BIB] - [PDF]
C. Petrioli, R. Petroccia, J. Shusta, and L. Freitag, "From underwater simulation to at-sea testing using the ns-2 network simulator," in Proceedings of MTS/IEEE OCEANS 2011, (Santander, Spain), IEEE Oceanic Engineering Society, June, 6-9 2011. - [BIB] - [PDF]
L. Badia, M. Mastrogiovanni, C. Petrioli, S. Stefanakos, and M. Zorzi, "An optimization framework for joint sensor deployment, link scheduling and routing in underwater sensor networks," SIGMOBILE Mob. Comput. Commun. Rev. 11, 4 (October 2007), 44-56 (selected paper from ACM WUWNET 2006). - [BIB] - [PDF]
C.Petrioli, R. Petroccia and D. Spaccini. "SUNSET version 2.0: Enhanced Framework for Simulation, Emulation and Real-life Testing of Underwater Wireless Sensor Networks," in Proceedings of ACM WUWNet 2013, (Kaohsiung, Taiwan), ACM, November 11--13, 2013. - [PDF]