Performance evaluation of acoustic network for

underwater autonomous vehicle in confined spaces

Siyuan Zheng, Xiuling Cao, F.Tong, Gangqiang Zhang, Yangze Dong
MOE Key Laboratory of Underwater Acoustic National Key Laboratory of Science and Technology on
Communication and Marine Information Technology, Underwater Acoustic Antagonizing,
Xiamen University, Xiamen, China Shanghai, China
ftong@xmu.edu.cn dongyangze@139.com

Abstract—As a kind of small size mobile platform, autonomous relay node, which perform the data transfer and information
underwater vehicles (AUV) are playing more and more collection respectively.
important role in different marine fields such as underwater As most of the current underwater acoustic networks are
monitoring and construction. While acoustic communication evaluated in terms of the communication and network
offer an effective communication link for AUV, under the
confined spaces in which small AUVs usually work, the
performance, few investigations has been reported on the
structured obstacles and closure or semi-closure barriers pose cross-media link supporting of small AUV in confined space.
serious challenges to communication link due to acoustic Therefore, performance evaluation the underwater acoustic
sheltering and significant reverberation. The underwater networks is performed in this paper for supporting cross-media
acoustic network is capable of improving acoustic connectivity of connectivity, in the presence of acoustic sheltering in confined
AUV via the distributed communication link of different nodes. space.
In this paper, an acoustic network is designed and evaluated for To be specific, connectivity supporting of small AUV is
communication support of AUV in a small lake with dense piers. equivalent to a cross-media data relay network as indicated in
While the presence of piers lead to obvious performance
Fig.1. Thus, from the viewpoint of AUV cross-media
degradation of point-to-point acoustic communication, the
experimental results verified the effectiveness of acoustic network
connectivity, the performance of AUV communication link in
for improving connectivity of AUV in confined space. confined space was evaluated based on an UWA(underwater
acoustic) communication network, in terms of its improvement
Keywords—AUV, Acoustic Network, Confined space with respect to the traditional point-to-point approach.

I. INTRODUCTION
In underwater confined space such as tunnel, harbor, cave
or pipeline, the small size AUVs are playing more and more
important role for various missions like detection,
search-rescue and picturing. While the acoustic
communication does provide an effective communication way
for AUV, the structured or un-structured barriers in underwater
confined space poses significant challenge to the reliable
point-to-point acoustic communication. As a result, the
application of small AUV in confined spaces is seriously Figure1. Illustration of the acoustic network for AUV link supporting
hindered.
In recent years underwater acoustic networks (UANs) have
a wide range of applications such as (i) underwater II. DESIGN OF THE ACOUSTIC NETWORK
environment monitoring, (ii) oceanic surveillance applications, A. Pyhsical layer
(iii) disaster prevention, (iv) offshore mining or drilling, (v)
As shown in Fig.2, the modem used in the proposed
tactical surveillance applications and (vi) assisted
network consists of a high performance float-point
navigation.[1-3]
micro-processor, a DAC(Digital to analog converter), an
Meanwhile, compared with the traditional point-to-point ADC(Analog to Digital Converter), an RS232 serial port, a
communication of underwater link, underwater acoustic power amplifier, a preamplifier, and a transceiver transducer.
network can provide a powerful support for the small AUV The micro-processor serves as the modem’s processing core.
with improved connectivity of distributed link. [4,5] Fig.3 shows the modem and transducers used in the
In view of the cross-media connectivity requirement of experiment.
small AUV in confined space, an experimental underwater We use DSSS(Direct Sequence Spread Spectrum)
acoustic networks is designed and tested. As indicated in Fig.1, modulation schemes at physical layer because it has good
the proposed underwater network consists of gateway node and anti-multipath and anti-jamming ability, can reduce the impact
of multipath effect and improve communication performance. For the design of MAC layer, we compared the features of
The DSSS modulation schemes are implemented on the same two popular MAC (Media Access Control) protocols, including
processor[6]. the ALOHA protocol and the MACA (multiple access collision
avoidance) protocol. Although ALOHA is low-complexity and
easy to implement, it may cause serious conflicts. Meanwhile
MACA can avoid data packet collision to some extent for the
handshake mechanism utilized. [8]

(a)

Figure 2 Block chart of the hardware implementation of node

(b)
Figure 4 Small AUV used and lake trial network topology

To avoid the collision, MACA is adopted as the MAC

protocol. Before sending data, an RTS(reques-to-send) package
is transmitted from the initial node to check the status of target
node, receiving the CTS(clear-to-send) package from target
Figure 3 Figure of acoustic modem as physical layer node means that the channel is available and then the data
communication can be triggered.
Fixed routing is relatively simple and reliable for
TABLE I. PARAMETERS OF THE PHYSICAL LAYER small-scale networks in environments with little change. We
Carrier use the shortest path principle to establish routing.[9] In
Modulation DBPSK 16000Hz
frequency network initializing stage, each node exchanges message with
Bit per frame 192 Bit rate 63.49 its neighbor to classify it according to the hops of the sink
PN code M sequence Chip rate 4000
node.
In the process of network initialization and operation, there
Spread gain 63 Guard time 20ms
is a common conflict when more than one node receive the
message, they will send message at the same time. It will lead
Modulation parameter of the physical layer is presented in to serious collision of data packets. To prevent such collisions,
Table1. Sampling rate of the modem is 75ksps. The transducer nodes are designed to have random delay before they send and
bandwidth is 13-18kHz. listen to channels simultaneously.
In order to complete the connection between underwater When conflict occurs, we use binary exponential back-off
acoustic network and shore control center, the gateway node algorithm to retransmit to avoid the next conflict.
equipped an SV6500 long distance wireless data transmission
module. III. CONFIGURATION OF LAKE TRIAL
The field data was collected from Furong lake in the
B. MAC and routing layer Xiangan campus of Xiamen University. With the network
topology shown in Fig4(a), totally 6 nodes were adopted to successfully received. Specifically, the end to end delay of
form a underwater acoustic networks to support cross-media node 1 is defined as the time between the instant of node 1 data
acoustic link for small AUV as shown in Fig.4(b) in confined packet transmitting and that of its successful reception.
space. The length of the small AUV is 1.5m with a diameter of Shown in Fig.6 is the result of mean end to end delay of
18cm, which is equipped with an acoustic modem as data each node, one may notice that generally the average
transmitter[10]. Note that there is a pier between the node 4 connection time of the node 1 to the data transmitted by AUV
and the AUV, that acting as an obstacle to simulate the hostile is less than 80s.
channel of confined space in practical applications. The
distance of node1-2 and node1-3 is 200m, while that of node
4-2 and node4-3 is about 70m.
Consider the purpose of experimental network, node 5 is an
AUV equipped with an UWA modem acting as the small AUV
working in confined space. The AUV is designed to move
slowly while sending acoustic communication signal with
random time interval, thus simulating the AUV mission in
confined space. The mean interval of the data transmission is
180s. In our performance evaluation, successful demodulation
of the data packet transmitted by the AUV is used as the
performance metric for analysis.
Node 1 is the gateway node equipped with UWA modem
and wireless communication module, acting as the cross-media
transfer center between underwater sound and wireless RF. Figure 5 Throughput of underwater acoustic sensor network system
Node 2,3,4 are relay nodes, which are equipped with
acoustic modem to receive, demodulate and transmit the
communication packet. The demodulation data of each relay
node is sent to the Node 1 for accumulation via the underwater
acoustic networks, which is used for performance evaluation of
the AUV connectivity support.
Node 6 is a laptop connected to node 1 via the wireless
communication module, acting as a land based control center
for cross-media communication link. Used as the performance
indicator of connectivity support, the number of the total data
packet finally received by the node6 is collected for analysis
and comparison.

IV. EXPERIMENTAL RESULTS AND ANALYSIS

Figure 6 End-to-end delay of data packet
A. Results and analysis of network performance

In our investigation, the performance of underwater The Fig.7 provides the result of collision rate of each node,
acoustic networks and the traditional point-to-point way are which reveals that node 4 exhibits a much higher collision rate
analyzed and compared. As the node 1 and node 6 are compared with other nodes. Similarly the reason is the
connected with wireless link thus no error bit between, only the sheltering effect of piers.
performance of node 1 was investigated.
In our research, the throughput of the relay node is defined B.Results and analysis of AUV connectivity performance
as the ratio of the number of data packet successfully received As mentioned in previous section, the performance
to that totally transmitted. Specifically, throughput of node 1 is evaluation of the AUV connectivity supporting is based on the
defined as the number of data packet successfully received to number of the total receiving data packet that is originally
that totally transmitted by the node 5 (AUV). transmitted by the AUV.
As indicated in Fig.5, the throughput of all node generally Consider the hostile channel condition in confined space
exceed 60%, with node 2 and 3 exhibits a throughput of 100% caused by barriers or multipath, the advantage of acoustic
while node 4 achieves a relatively low throughput due to the network is that acoustic link can be achieved via different node
shading effect of piers as revealed in Fig.4. to improve the reliability of connectivity. With the underwater
The end to end delay is defined as the time delay between acoustic networks, the same data packet is received by several
the instant of data packet transmitting and that of its successful relay nodes and then transmitted to the same control node, thus
reception. The mean end to end delay is defined as the sum of the accumulative number of data packet received in node 1 can
end to end delay divided by the number of data packet be used to assess the connectivity supporting capability.
 Thus, while the data packets that transmitted by the AUV V. CONCLUSION
node and directly received by each relay node are countered as In view of the difficulties of connectivity supporting for
the performance metric of the traditional point-to-point small AUV in confined space, the hardware, software, network
acoustic link approach, the number of those data packet protocol of a cross-media UWA sensor network is provided for
receipted by the node 1 are countered for evaluating the enabling distributed link for small AUV. The experimental
networked sensing. results obtained from the lake trial are analyzed and discussed,
which verified the performance improvement of the
outperforms in AUV connectivity supporting.
ACKNOWLEDGMENT
The authors are grateful for the funding of the Natural
Science Foundation of China (No. 11274259 and No.
11574258), as well as the Open Funding of National Key
Laboratory of Science and Technology on Underwater
Acoustic Antagonizing (No. 614221401050217) in support of
the present research.
REFERENCES
[1] Dalhatu M , Mohammad A , Mahdi Z , et al. Game
Theory-Based Cooperation for Underwater Acoustic Sensor
Networks: Taxonomy, Review, Research Challenges and
Directions[J]. Sensors, 2018, 18(2):425-.
[2] Heidemann J , Stojanovic M , Zorzi M . Underwater sensor
Figure 7 Collision rate of data packet networks: applications, advances and challenges.[J].
Philosophical Transactions, 2012, 370(1958):158.
[3] Jindal H , Saxena S , Singh S . Challenges and issues in
underwater acoustics sensor networks: A review[C]//
International Conference on Parallel. IEEE, 2015.
[4] Manalang D , Delaney J R . [IEEE OCEANS 2016 MTS/IEEE
Monterey - Monterey, CA, USA (2016.9.19-2016.9.23)]
OCEANS 2016 MTS/IEEE Monterey - Axial seamount -
restless, wired and occupied: A conceptual overview of resident
AUV operations and technologies[J]. 2016:1-7.
[5] Dol H , Van Walree P . [IEEE OCEANS 2011 - SPAIN -
Santander, Spain (2011.06.6-2011.06.9)] OCEANS 2011 IEEE -
Spain - Underwater acoustic communication research at TNO —
Past and present[J]. 2011:1-6.
[6] Siyuan Zheng, Xiaoyang Wang, Weihua Jiang, et al. Lake trial
of an underwater acoustic cross-media network testbed [C]//
2017 IEEE International Conference on Signal Processing,
Communications and Computing (ICSPCC), 2017
[7] Chen K, Ma M, Cheng E, et al. A Survey on MAC Protocols for
Underwater Wireless Sensor Networks[J]. IEEE
Figure 8 The ratio of data packet received by network and individual node
Communications Surveys & Tutorials, 2014, 16(3):1433-1447.
[8] Ansari S, Gonzalez J P, Otero P, et al. Analysis of MAC
Strategies for Underwater Applications[J]. Wireless Personal
Shown in Fig.8 is the ratio of data packet reception by relay Communications, 2015, 85(2):359-376.
node (node 2,3,4) as well as the control node (node 1) to the [9] Ya-Qiong Z , Yun-Rui L . A routing protocol for wireless sensor
total data packet transmitted by AUV. Here it is used as an networks using K-means and Dijkstra algorithm[J]. International
indicator of the connectivity of AUV in confined space. We Journal of Advanced Media & Communication, 2016,
can see from Fig. 8 that, with the capability of underwater 6(2):109-121.
acoustic networks the node 1 exhibits the highest ratio of about [10] Q.Y. Tao, Y.H. Zhou, F.Tong, et al. , Evaluating acoustic
80%, while the sensor node 2,3,4 achieve lower ratio of 70%, communication performance of micro autonomous underwater
50% and 30% respectively. The experimental result of Fig.8 vehicles in confined spaces Front Inform Technol Electron Eng,
indicated that, AUV with cross-media connectivity supported 2018 19(8):1013-1023.
by underwater acoustic networks outperforms that by
traditional point-to-point acoustic link in the proposed scenario
of missions in confined space.