Destination-Aware Focused Beam Routing (D-FBR): A routing protocol for underwater wireless sensor networks

Abstract

Delay Tolerant Networks enable data transmission in different applications, mainly in environments where communication infrastructure is missing. The existence of links is casual and even when links are created, they are short in time and unstable/unreliable in terms of connectivity. Thus, underwater network applications tend to be delay tolerant. In this paper, we propose a mobility-aware routing protocol for underwater wireless sensor networks. The protocol is based on Focused Beam Routing (FBR) protocol and considers nodes’ next destination location when making forwarding decisions. The routing protocol is called Destination-aware Focused Beam Routing (D-FBR) protocol and we compare its performance with FBR protocol for different FBR angles and different applications. We use Delivery Probability (DP), Average Number of Hops (ANH), Overhead Ratio (OR) and simulate our scenarios by The ONE simulator. Simulation results show that, for D-FBR protocol, when FBR angle is $30^{\circ}$ , OR and ANH is improved by $34.5 %$ and $23.1 %$ , respectively, while DP is $6.5 %$ lower, compared to FBR protocol. Both D-FBR and FBR outperform Epidemic Routing Protocol.

Keywords

Underwater Wireless Sensor Networks Focused Beam Routing UWSN FBR D-FBR ONE Simulator

1. Introduction

Underwater communications are shifting from military towards commercial applications [14]. Underwater Wireless Sensor Networks (UWSN) are becoming more and more popular in applications such as disaster detection/prevention, pollution monitoring in environmental systems, collection of scientific data in fields of biology and geology, mapping ocean floors in oceanography, coordinated navigation control and so on [18]. A typical application scenario includes several Autonomous Underwater Vehicles (AUVs), which are equipped with a wide range of sensors (environmental, chemical, navigation and so on). They explore the seabed, use their sensors to collect data and submit the sensed data up to the sea surface, where air-water interface ships, boats or buoys are located (see Fig. 1). Furthermore, these air-water interfaces might also use electromagnetic waves to send the aggregated data back to monitoring centers or data processing centers.

Fig. 1.

Underwater acoustic communications and aerial wireless communications.

The preferred communication media for underwater communications is acoustic waves in contrast to electromagnetic radio waves, which are widely used in the air. In fact, low frequency electromagnetic radio waves (30 Hz–300 Hz) can propagate for longer distances, but they require large antennae and high transmission power. Optical waves have better propagation, but they require directional coordination, which is almost impossible in underwater applications, where the devices are in constant movement [18].

Most of the research regarding the physical layer in underwater communications uses Phase Shift Keying (PSK) and Quadrature Amplitude Modulation (QAM) modulation techniques. With the increasing of processing capabilities of small devices, Orthogonal Frequency Division Modulation (OFDM) is also considered.

At the Medium Access Control (MAC) layer, Frequency Division Multiple Access (FDMA) does not work well with limited-band acoustic signals because they are affected by fading and multi-path [18]. Moreover, in acoustic communications there is a difference in delays between consecutive packets (jitter), which makes it very difficult to implement Time Division Multiple Access (TDMA) techniques. CDMA is so far the best solution, but newly designed techniques are yet to be developed, tested and implemented [1,25].

Because it is costly to implement network infrastructure in underwater environment, the network layer is mostly oriented towards adhoc architecture, where all participating nodes forward packets to other nodes, until the packets reach the intended destination. In fact, depending on the application, the adhoc architecture may consist of:

Real-time data transmission, where participating nodes, either already know where to forward the received data, or the can find out in a very short amount of time. Such amount of time is usually considered as “the real time”.

Delay-tolerant data transmission, where different techniques are used to increase packet delivery ratio (PDR) even when there are no available routes to the intended destination. Delay-tolerant data transmissions, us the store-carry-forward paradigm, to transmit data from one node to another, until the data reach the destination.

Thus, existing adhoc and Delay Tolerant Network (DTN) routing protocols need to be redesigned, in order to deal with unstable links and high delay variance in underwater environment.

In this paper, we introduce our proposed routing scheme, show its implementation in The One simulator and analyze its performance in terms of different evaluation parameters.

The remainder of the paper is organized as follows. A state of the art of Routing in UWSN is given in Section 2. In Section 3, we describe some DTN routing protocols for UWSN and discuss their challenges. In Section 4, we introduce our proposed D-FBR and its implementation in The ONE simulator. We describe our simulation settings in underwater scenario in Section 5. In Section 6, we present and analyze the simulation results. We conclude with conclusions and future works, in Section 7.

2. Related work

DTNs are enabling various networking applications in different fields. Work have been done towards using DTN in order to enable vehicular communication in VANET [5,7,10]. Other applications of DTN consider underwater environment. In [12], the authors present an underwater framework, which uses DTN to enable communication between AUVs. They also discuss various complex application scenarios for control operations involving single and multiple AUVs. Other works focus on acoustic signal generation in underwater environment.

The authors in [6] have designed an acoustic repeater device which extends the range of underwater communications. They have conducted tests in both an on-campus confined tank and open-air river waters. The tests show that the acoustic repeater can retransmit the acoustic signals with a relatively small delay of around 30 ms. In [17], the authors present a performance analysis of a laser communication system. The experiments are conducted in real underwater environment in the Baltic Sea. A 80 MB video file was transmitted in real time without errors in an experimental distance of 2 m. The maximum data rate achieved during the transmission was 21 Mbps.

In [27], authors construct an underwater acoustic network in a small lake with a lot of piers. They evaluate network connectivity for AUVs in a small lake with piers. They claim that the communication is effective and that the presence of piers and other obstacles decrease the performance.

Many wireless ad hoc geographical routing protocols use positioning and other geographical data to make forwarding decisions. While it is easy to get positioning data from gnss satellite system, in underwater environment, global positioning knowledge is not an easy problem. Extensive research have been done in localization and tracking in underwater environment.

In order to estimate the velocity of underwater sensor velocity, the authors of [24], use a method based on mean value and finite-difference approximation. The proposed method derives the sensor velocity values from multiple sensors incorporated in underwater sensors. In [22], a Underwater acoustic localization module is designed and implemented. The module uses multiple signal classification from four installed hydrophones. The authors in [11] utilize a group of autonomous surface vehicles to design an underwater acoustic source localization strategy. In order to minimize resources, the strategy consists of two phases: target search and target drive. They claim the strategy is effective and prove their claims by presenting simulation results. Sonar images are used in [23] to propose a dynamic target tracking method based on model predictive control algorithm. Sonar images, are firstly used to identify and track the target. Then, a Kalman filter is used to predict moving target’s direction and velocity. The authors claim that the proposed method is accurate in tracing moving targets and they prove their claims by simulation results. In [21], an autonomous surface vehicle design and implementation is presented. The authors aim to utilize the vehicle as a “satellite” to support Internet of Underwater Things applications. The vehicle also supports diving operations (diver following, monitoring, assistance, and improved safety of the diver), navigation operations and networking operations (data forwarding, localisation) when used in multiple numbers.

Aiming for an energy-aware environment [16], several approaches to routing have been proposed to increase efficiency and lower resource utilization in underwater communications. In [3] a cross-layer channel-aware routing protocol for underwater wireless sensor networks is presented. Link quality, hop count and successful transmission history are used to make forwarding decisions. The authors evaluate the performance of the proposed routing protocol by ns-2 simulation and show that it outperforms FBR and Depth Based Routing (DBR) routing protocols.

In [26], the authors propose a routing method that utilizes number of interactions and contact duration between nodes in order to make routing decisions. The authors claim that the proposed method improves overhead cost and achieves shorter delivery delay while not decreasing its performance. The claims are supported by simulation results.

The authors of [4] propose two routing protocols for ocean vessel ad hoc networks. The paper proposes a probabilistic forwarding decision process based on the number of encounters between vessels. As shown from the simulation results the novel approaches minimize overhead while keeping high rated of packet delivery.

In [2], the authors propose an underwater pragmatic routing approach, which uses packet reverberation mechanism in order to decrease energy consumption. The consider both node condition and link quality in the forwarding decision process. They compare the performance of the proposed protocol with other routing protocols by simulations (ns-2), and show that it has a better performance in terms of packet delivery rate, delay, energy consumption, network lifespan, and network throughput.

3. Underwater wireless sensor network routing protocols

Fig. 2.

Underwater routing protocol paradigms.

In UWSNs, the store-carry-forward paradigm is popular among routing protocols, hence Epidemic Routing (ER), which is visualized in Fig. 2(a), is the main forwarding technique. Here, the forwarding area is defined from the communicating distance. One problem with ER protocol is that the packets are copied to every possible relay node creating a huge overhead in the network, which also impacts the network lifetime due to battery depletion. In fact in some applications where nodes are scarce, the increased number of copies helps to increase the delivery probability, but in general, decreasing the overhead is preferred for UWSN, because UWSN nodes also have limited storing resources.

The objective of routing protocols for UWSN in the majority of applications is to forward data towards the data collectors in the water surface. Decreasing the number of copied packets, forcing nodes to strictly forward packets only to nodes that are closer to the surface has proven beneficial in terms of performance, for these applications [15,19,20,28]. This category of routing protocols is called Depth-Based Routing (DBR) and is shown in Fig. 2(b).

In order to furthermore decrease the overhead and energy consumption, nodes that use Focused Beam Routing (FBR), limit their forwarding area in the direction of data collector, and define it by a relatively small angle θ, as shown in Fig. 2(c). Sender Node S will forward available messages only to nodes 1, 5 and 6, which are located inside the FBR area. A more detailed description of FBR, can be found in [13].

4. Proposed Destination-aware FBR (D-FBR) protocol and its implementation

Fig. 3.

Destination FBR (D-FBR) overview.

The original FBR protocol does forwarding decisions based on the actual position of the nodes. However, it does not consider the mobility of the nodes. We introduce a mobility-aware version of FBR, called Destination-aware FBR (D-FBR), where each forwarding decision is done based on the next destination of a potential relay node, as shown in Fig. 3. When the original FBR is used, Sender Node S would forward messages to node 5, but not to node 4, based on their actual location. When D-FBR is used instead, because the next destination of node 4 is inside FBR area, and the next destination node 5 is outside FBR area, the data will be forwarded only to node 4. By doing this, we want to achieve similar delivery probability values, while minimizing the overhead in the network.

While there are a lot of routing protocols proposed and implemented around the world, we could not find an open implementation of FBR that could be easily editable and verifiable in different scenarios. Thus, we implemented FBR and D-FBR in the well-known The One simulator and compared their performances in different scenarios. The following assumptions are made, in order to simplify the implementation:

The environment is considered as 2-dimensional: horizontal dimension represents the width of a water environment and the vertical axis represents its depth. We believe that the behaviour of our proposed protocol might have a slightly different performance in 3-dimensional environment, mainly because each node should calculate FBR cone, instead of FBR angle.

Every node knows every other nodes’ location in every moment, which in fact is very difficult to achieve in reality without a well-established infrastructure.

Participating node is mobile and moves based on the Random Waypoint Mobility Model, as implemented in The ONE. We wanted to use random mobility patterns, in order to evaluate the performance of our protocol for continuously moving sensors and AUVs, without defining a specific application.

There is only one data collector node, and all transmissions are directed towards this node, which is located in the top-center of our 2D environment.

Then, as shown in Fig. 3, whenever a participating node (node O) receives a new message or contacts a new node, the following happens.

Directional angle towards the data collector is calculated in degrees.

FBR angle θ defines the transmission area on both sides of line towards the collector. The transmission area is confined by these lines.

Communication distance confines the transmission area, which in the best case scenario is an arc.

Define FBR area (shaded area)], which makes the transmissions directed towards the buoy.

Forwarding Decision is made for every candidate, based on their next destination coordinates. If a candidate’s next destination is outside FBR area, messages are not forwarded, and vice versa.

5. Simulation environment

In order to analyse the proposed D-FBR protocol and compare its performance to that of FBR protocol, we conducted simulations in different scenarios, and set the environment as shown in Table 1. We used The ONE Simulator [8], which has already implemented the ER protocol and the store-carry-forward mechanism. We limited the buffer size of UWSN nodes, in order to see the effect of overhead packets.

Table 1
Simulation parameters’ settings

Parameters Value

Movement model Random waypoint

Simulation time 5400 s (1.5 h)

Simulation area 500 m × 500 m

Message size Small: 180 kB–220 kB

Medium: 400 kB–600 kB

Large: 1300 kB–1700 kB

FBR’s Angle (θ) $30^{\circ}$ , $45^{\circ}$ , $90^{\circ}$

Parameters	Value
Movement model	Random waypoint
Simulation time	5400 s (1.5 h)
Simulation area	500 m × 500 m
Message size	Small: 180 kB–220 kB
Medium: 400 kB–600 kB
Large: 1300 kB–1700 kB
FBR’s Angle (θ)	$30^{\circ}$ , $45^{\circ}$ , $90^{\circ}$

Table 2

Participating nodes in simulations

Parameters	Collector nodes	Sensor nodes	Relay nodes (AUVs)
Number of nodes	1	50	10
Mobility	Static at (250,0)	Random Waypoint Mobility Model
Moving speed	0	0.5–1.5 m/s	2–5 m/s
Buffer size	Unlimited	10 MB	300 MB
Interface	High Speed Interface (HSI)	Low Speed Interface (LSI)	HSI and LSI
Transmission speed	10 Mbps	250 kbps	10 Mbps and 250 kbps
Transmission distance	100 m	50 m	100 m and 50 m

Fig. 4.

Simulation scenario.

Fig. 5.

Overhead ratio for 3 different message sizes, comparing FBR and D-FBR for different FBR angles.

We used different message size for data generation, while generating the same amount of data throughout the simulation. As shown in Table 1, for message size Small, the message size is generated randomly with values in 180 kB–220 kB interval. These small messages are transmitted approximately every 4 s, totalling a generation datarate of around 3 MBytes/minute. In order to keep the datarate fixed, Medium messages are generated every 10 s and Large messages are generated every 15 s. Because underwater sensors use LSI at 250 kbps, Large messages require more time to be transmitted from one node to another, and occupy more buffer space when being carried around the simulation area.

In Table 2, the settings of each type of nodes used in our simulations is summarized. As shown in Fig. 4, a single data collector is located at the top-center (250 m, 0 m) of our simulation area (500 m × 500 m). Then we added 40 underwater sensor nodes, which are usually small devices attached to marine animals or floating in the water. These sensor nodes generate data in 3 different data patterns: Small message, Medium message and Large message. They are equipped with Low Speed Interface (LSI), in order to forward their sensed data to each-other and the relay nodes. Relay nodes are more powerful devices equipped with extra resources and they act as data mules for underwater sensors. In our case, they are equipped with 2 interfaces; one LSI interface in order to get data from senors and a High Speed Interface (HSI) in order to send data to the data collector. The data collector node receives messages only from relay nodes, by using its HSI.

6. Simulation results

We present our simulation results for Overhead Ratio (OR), Average Number of Hops (ANH) and Delivery Probability (DP), in Figs 5, 6 and 7, respectively. We note that, when FBR angle is $180^{\circ}$ , both protocols act as ER protocol. This is confirmed by showing exactly the same performance when FBR angle is $180^{\circ}$ .

Fig. 6.

Average number of hops for 3 different message sizes, comparing FBR and D-FBR for different FBR angles.

In Fig. 5, we notice that when generated messages are small (∼200 kB), OR is smaller for all FBR angles and both protocols. We assume that smaller messages can be forwarded faster and stored easier in node’s buffers. Moreover, it can be seen that D-FBR has lower overhead for smaller FBR angles ( $30^{\circ}$ and $45^{\circ}$ ). When we analyze Fig. 6, we can see that the ANH is smaller for smaller message sizes and smaller FBR angles. D-FBR shows better performance than FBR, mainly for smaller FBR angles.

Fig. 7.

Delivery probability for 3 different message sizes, comparing FBR and D-FBR for different FBR angles.

By achieving lower overhead and smaller number of hops for each communication, we claim that our proposed D-FBR shows better performance than FBR, mainly for smaller FBR angles. From Fig. 7, we can observe that DP for D-FBR is slightly smaller, but the improvement in overhead and ANH are more significant. D-FBR shows the best performance compared to FBR, for medium message sizes and $30^{\circ}$ FBR angle. In this case, OR is $34.5 %$ lower and ANH is $23.1 %$ lower, but DP is only $6.5 %$ lower.

D-FBR shows great performance in terms of decreasing resource usage, with slightly lower delivery ratio. Both protocols outperform ER protocol for OR, ANH and DP. When sensors send smaller (but frequent) messages in the network, all protocols show lower DPs, but better OR and ANH.

7. Conclusions and future work

In this paper, we proposed an Destination-aware FBR (D-FBR), which considers the next destination of each candidate before forwarding messages. We implemented D-FBR in The ONE Simulator, and compared its performances with FBR and ER, two well-known routing protocols for DTNs. From the simulation results, we draw the following conclusions.

D-FBR shows better performance than FBR for small FBR angles.

Both D-FBR and FBR outperform ER in almost all scenarios.

When sensors send smaller (but frequent) messages, all protocols show lower DPs, but better OR and ANH.

We believe that, D-FBR needs further improvements, in order to increase its predictability. We are continuously improving and testing D-FBR, in order to furthermore decrease resource usage (OR and ANH), while keeping an acceptable performance in message delivery. In this work, we made various assumptions, which in real life might not be easily implemented. In the future, we would like to add realistic conditions to our simulations, including real mobility traces [9] and real data applications.

Conflict of interest

The author has no conflict of interest to report.

References

I.F.

Akyildiz,

Pompili and

Melodia, Underwater acoustic sensor networks: Research challenges, Ad Hoc Networks 3(3) (2005), 257–279. doi:10.1016/j.adhoc.2005.01.004.

Ashraf,

Gao,

Chen,

Naeem,

Ahmad and

Ahmed, Underwater pragmatic routing approach through packet reverberation mechanism, IEEE Access 8 (2020), 163091–163114. doi:10.1109/ACCESS.2020.3022565.

Basagni,

Petrioli,

Petroccia and

Spaccini, Channel-aware routing for underwater wireless networks, in: 2012 Oceans, Yeosu, 2012, pp. 1–9. doi:10.1109/OCEANS-Yeosu.2012.6263538.

Bing,

Jiang and

Hong, Exploiting social network characteristics for efficient routing in ocean vessel ad hoc networks, in: 2019 IEEE 38th International Performance Computing and Communications Conference (IPCCC), 2019, pp. 1–8. doi:10.1109/IPCCC47392.2019.8958733.

Bylykbashi,

Spaho,

Barolli and

Xhafa, Routing in a many-to-one communication scenario in a realistic VDTN, International Journal of High Speed Networks 24(2) (2018), 107–118. doi:10.3233/JHS-180584.

Chen,

Hsiao,

Chuang and

Hu, Establishment of a robust underwater acoustic repeater with applications in underwater acoustic transmission and detection, in: OCEANS 2015, Genova, 2015, pp. 1–5. doi:10.1109/OCEANS-Genova.2015.7271770.

Cuka,

Shinko,

Spaho,

Oda,

Ikeda and

Barolli, A simulation system based on ONE and SUMO simulators: Performance evaluation of different vehicular DTN routing protocols, International Journal of High Speed Networks 23(1) (2017), 59–66. doi:10.3233/JHS-170557.

Keränen,

Ott and

Kärkkäinen, The ONE simulator for DTN protocol evaluation, in: Proceedings of the 2nd International Conference on Simulation Tools and Techniques, Rome, Italy, 2009.

Kim,

Kotz and

Kim, Extracting a mobility model from real user traces, in: Proceedings IEEE INFOCOM 2006. 25th IEEE International Conference on Computer Communications, 2006, pp. 1–13. doi:10.1109/INFOCOM.2006.173.

10.

S.S.

Magdum,

Sharma,

S.M.

Kala,

A.A.

Franklin and

B.R.

Tamma, Evaluating DTN routing schemes for application in vehicular networks, in: 2019 11th International Conference on Communication Systems & Networks (COMSNETS), 2019, pp. 771–776. doi:10.1109/COMSNETS.2019.8711206.

11.

M.H.A.

Majid and

M.R.

Arshad, Underwater acoustic source localization strategy by a group of autonomous surface vehicles, in: 2016 IEEE International Conference on Underwater System Technology: Theory and Applications (USYS), 2016, pp. 26–31. doi:10.1109/USYS.2016.7893941.

12.

E.R.B.

Marques

et al., AUV control and communication using underwater acoustic networks, in: OCEANS 2007, Europe, 2007, pp. 1–6. doi:10.1109/OCEANSE.2007.4302469.

13.

Miquel Jornet,

Stojanovic and

Zorzi, Focused beam routing protocol for underwater acoustic networks, in: Proceedings of the Third ACM International Workshop on Underwater Networks (WuWNeT’08). Association for Computing Machinery, New York, NY, USA, pp. 75–82.

14.

Potter,

Alves,

Green,

Zappa,

Nissen and

McCoy, The JANUS underwater communications standard, in: 2014 Underwater Communications and Networking (UComms), Sestri Levante, 2014, pp. 1–4.

15.

Qiang,

Shigeyasu and

Ch.

Chen, A new DTN routing strategies ensuring high message delivery ratio while keeping low power consumption, Internet of Things 17 (2022). doi:10.1016/j.iot.2021.100463.

16.

Ricciardi,

Careglio,

Santos-Boada,

Solé-Pareta,

Fiore and

Palmieri, Towards an energy-aware Internet: Modeling a cross-layer optimization approach, Telecommunication Systems 52(2) (2013), 1247–1268.

17.

Scholz, Laser based underwater communication experiments in the Baltic Sea, in: 2018 Fourth Underwater Communications and Networking Conference (UComms), 2018, pp. 1–3. doi:10.1109/UComms.2018.8493174.

18.

Stojanovic, Underwater Acoustic Communication, Wiley Encyclopedia of Electrical and Electronics Engineering, 1999. doi:10.1002/047134608X.W5411.

19.

Thangarajan,

Siva and

Boopalachakaravarthy, Adaptive energy efficient routing protocol with extended lifetime in underwater sensor networks, in: 2013 International Conference on Communication Systems and Network Technologies, 2013, pp. 322–326. doi:10.1109/CSNT.2013.74.

20.

Trevathan,

Schmidtke,

Read,

Sharp and

Sattar, An IoT General-Purpose Sensor Board for Enabling Remote Aquatic Environmental Monitoring, Internet of Things 16 (2021). doi:10.1016/j.iot.2021.100429.

21.

Vasilijevic,

Nad and

Miskovic, Autonomous surface vehicles as positioning and communications satellites for the marine operational environment – step toward Internet of underwater things, in: 2018 IEEE 8th International Conference on Underwater System Technology: Theory and Applications (USYS), 2018, pp. 1–5. doi:10.1109/USYS.2018.8778993.

22.

Vivek and

Vadakkepat, Multiple signal classification (MUSIC) based underwater acoustic localization module (UALM) for AUV, in: 2015 IEEE Underwater Technology (UT), 2015, pp. 1–4. doi:10.1109/UT.2015.7108288.

23.

Wei,

Zhu and

Chu, Underwater dynamic target tracking of autonomous underwater vehicle based on MPC algorithm, in: 2018 IEEE 8th International Conference on Underwater System Technology: Theory and Applications (USYS), 2018, pp. 1–5. doi:10.1109/USYS.2018.8779215.

24.

Yang,

Sun,

Hong and

Zhou, Research of directivity of underwater acoustic velocity gradient sensors, in: 2013 Symposium on Piezoelectricity, Acoustic Waves, and Device Applications 2013, pp. 1–4. doi:10.1109/SPAWDA.2013.6841143.

25.

Yin,

Du,

Yang and

Zhou, Space-division multiple access for CDMA multiuser underwater acoustic communications, Journal of Systems Engineering and Electronics 26(6) (2015), 1184–1190. doi:10.1109/JSEE.2015.00129.

26.

Yoon,

Kim,

Lee and

Jang, An enhanced friendship-based routing scheme exploiting regularity in an opportunistic network, in: 2016 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData), 2016, pp. 51–57. doi:10.1109/iThings-GreenCom-CPSCom-SmartData.2016.36.

27.

Zheng,

Cao,

Tong,

Zhang and

Dong, Performance evaluation of acoustic network for underwater autonomous vehicle in confined spaces, in: 2018 IEEE 8th International Conference on Underwater System Technology: Theory and Applications (USYS), 2018, pp. 1–4. doi:10.1109/USYS.2018.8778974.

28.

Zou,

Lin and

Sun, A cluster-based adaptive routing algorithm for underwater acoustic sensor networks, in: 2019 International Conference on Intelligent Computing, Automation and Systems (ICICAS), 2019, pp. 302–310. doi:10.1109/ICICAS48597.2019.00072.