- Email: [email protected]
Design of optical-acoustic hybrid underwater wireless sensor network
Author’s Accepted Manuscript Design of Optical-Acoustic Hybrid Underwater Wireless Sensor Network Jingjing Wang, Wei Shi, Lingwei Xu, Liya Zhou, Qiuna Niu, Juliu www.elsevier.com/locate/jnca
PII: DOI: Reference:
S1084-8045(17)30084-X http://dx.doi.org/10.1016/j.jnca.2017.02.014 YJNCA1870
To appear in: Journal of Network and Computer Applications Received date: 1 August 2016 Revised date: 24 January 2017 Accepted date: 21 February 2017 Cite this article as: Jingjing Wang, Wei Shi, Lingwei Xu, Liya Zhou, Qiuna Niu a n d Juliu, Design of Optical-Acoustic Hybrid Underwater Wireless Sensor N e t w o r k , Journal of Network and Computer Applications, http://dx.doi.org/10.1016/j.jnca.2017.02.014 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Design of Optical-Acoustic Hybrid Underwater Wireless Sensor Network Jingjing Wang1,2*, Wei Shi1, Lingwei Xu1, Liya Zhou1, Qiuna Niu1,3 and Juliu2 1
School of Information Science and Technology, Qingdao University of Science and Technology, Qingdao, Shandong Province, China 2 School of Information Science and Engineer, Shandong University, Jinan，Shandong Province, China 3 Shandong provincial Key Laboratory of Computer Network, Jinan, Shandong Province, China *
Corresponding author: Jingjing Wang, [email protected]
Abstract: High-speed underwater wireless transmission technologies are urgently demanded to transmit three-dimensional high-resolution observation data timely in ocean exploration. Currently, underwater acoustic sensor networks have obtained high transmission distance, but with low date rate, high power consumption, high cost and damage to marine mammals. Meanwhile, underwater optical communications can achieve high date rate, but with difficult networking and short communication distance. This paper proposed a kind of novel Optical-Acoustic hybrid Underwater Wireless Sensor Network (OA-UWSN) which employed optical communication for high-speed transmission at close range and employed acoustic communication for transmitting control commands and node localization. OA-UWSN accomplished long-distance optical transmission through directional optical communication and multi-hop transmission mechanism. The problem of difficult networking in optical communication was solved by designing the whole space optical repeater. With the characteristics of directional optical transmission, the Space Division Multiple Access (SDMA) technology was employed in Data Link Layer (DLL) to improve the data transmission efficiency. A routing protocol based on reverse route search was designed to improve the network life cycle and adapt to network topology which changed frequently. The paper proposed a solution for wireless transmission of real-time video and images in marine exploration and provided new methods for high-speed transmission of marine information detection.
Keywords: Optical-acoustic; Hybrid; Underwater sensor networks; Encoding modulation; Space division multiplexing; Multiuser access
1 / 24
1. Introduction The statistical data of U.S. National Oceanic and Atmospheric Administration in 2015 showed that 95% of the underwater world had not yet been detected [1]. The construction of underwater wireless sensor network (UWSN) and the provision of large scales of ocean exploration data with high spatial-temporal resolution, is of great significance for strengthening the utilization of ecological environment monitoring, ocean energy development, and marine scientific research. At present, there are three kinds of underwater wireless communication technology: underwater electromagnetic wave communication, underwater acoustic communication and underwater optical communication. The attenuation of electromagnetic wave in water is serious. For example, the attenuation of 2.4GHz band, which is used in Bluetooth technology and wireless LAN technology, is about 1695dB/m in the sea, the value is as high as 189dB/m even in pure fresh water [2]. The ultra low frequency electromagnetic wave can reduce the attenuation, while the system cost is high and the data transmission rate is low. At current, underwater radio communication system is only limited to shallow sea areas, and requires a huge antenna system. The acoustic communication is the most widely technology applied in underwater wireless communication over decades, but the speed of sound in water is only 1.5×103 meters per second, which possesses big difference, 5 orders of magnitude, with the speed of light in the water (3×108 m/s). In particular, factors such as the acoustic wave scattering, the inherent loss of transmission and return interference, make the development of underwater acoustic equipment very difficult. According to the report of the Natural Resources Defense Council, military sonar and other rising ocean noise is a serious impact on the life of dolphins, whales and other marine mammals, and even cause their hearing loss or death [3]. The most significant advantage of underwater acoustic communication is far transmission distance, but its propagation delay and signal attenuation is huge, communication bandwidth is low, multipath effect is serious and marine mammals may be harmed[4]. The underwater wireless optical communication is a kind of communication mode with light wave as the carrier of information. S. A. Sullian and S. Q. Dimtley et al in 1963 found in their study of propagation characteristics of light waves in the ocean that attenuation of 450 ~ 550nm in blue and green light in seawater is much smaller than the other bands of light [5], as shown in Figure 1. This significant physical discovery laid a solid foundation to solve long-term underwater target 2 / 24
detection, communication and the other essential problems.
Figure 1. Attenuation of light wave in seawater [4]
The performance comparison of three kinds of underwater wireless communication technology is shown as table 1. Wireless optical communication owns an overwhelming advantage in speed, power consumption, volume and other aspects, however, the short communication distance of wireless optical communication restricts the further development. This paper presented a kind of novel optical-acoustic hybrid underwater wireless sensor
network
which
maximized
the
advantages
of
underwater
optical
communication and acoustic communication. The optical communication was used to achieve high-speed transmission at close range and acoustic communication was used to transmit control command and node localization based on the advantages of long distance transmission. The rest of the paper is organized as follows. The research progress and development
of
underwater
optical
communication,
underwater
acoustic
communication and optical-acoustic hybrid UWSN are presented in turn in Section 2. Section 3 provides the solution of a new Optical-Acoustic Hybrid UWSN. The physical layer adaptive coding modulation schemes of the OA-UWSN are given in Section 4. The protocol of DLL and the routing protocol are proposed in Section 5 and Section 6 separately. In Section 7, we present simulation and numerical results of 3 / 24
the OA-UWSN and the traditional underwater acoustic sensor network. The summary is drawn in Section 8. Table 1. Comparison of three underwater wireless communication modes Transmission medium Parameter
Underwater acoustic communication
Underwater optical communication
One hundred meters
Several kilometers
Tens of meters to hundreds of meters
100bps
Several Kbps
Several Mbps
Transmit power Underwater transport rate Equipment volume
MW grade Near to speed of light Large
KW to tens of KW
Large
mW to W grade Near to speed of light Little
Equipment cost Confidentiality
high --
Interference immunity
Electromagnetic interference
high Bad Free from electromagnetic interference.
Low Good Free from electromagnetic interference.
Single node communication distance Communication bandwidth
Ultra low frequency long wave radio system
1500m/s
2. Background of underwater wireless communication Underwater optical communication technology in earlier stage is mainly used in the field of military. In 1977 and 1983, the U.S. Navy,
U.S. Defense Advanced
Research Project Agency and the former Soviet Union tested laser communication with submarine and communication by laser beam reflected through space orbit [6]-[8]. In 2005, Professor I. Vasilescu etc. at the Massachusetts Institute of Technology built the underwater optical network, they only spent 21 seconds to read 6.86 MB data through optical communication link, and spent 1.3 days through acoustic communication link at the speed of 480 bps. What’s more, the energy consumption of acoustic communication is dozens of times as much as optical communication [9]. In 2007 and 2008, the team of Professor J. Muth at the North Carolina state university in U.S. developed underwater optical communication systems respectively based on laser diode/PIN photodiode and light-emitting diodes/avalanche photodiode. They successfully realized the reliable transmission at the rate of 1 Mbps by blue-green in tank with a side of length 5m [10] [11]. Moreover, in 2010, J. A. Simpson and W. C. Cox etc. in the same team proposed 5 Mbps Optical Wireless Communication based on the RS error correction coding [12].
In 2011,
they also realize transmission at the rate level of Mbps by using laser and LED as 4 / 24
light source. Professor N. Farr etc. at Woods Hole Oceanographic Institution and Professor M. S. Alouini and his team at the King Abdullah University of Science and Technology have done a lot of advanced work on underwater optical communication channel modeling, communication tests under marine environment and other fields [13]-[16]. From 2013 to 2015, Dr. Yuhan Dong at Tsing Hua University considered parallel link and vertical link geometric models of the underwater optical link, built a new channel model, and applied digital method to describe the relationships between transmission power, link range, wind speed, bit error rate and water features, etc [17]-[19]. Global scientists have made important progress in the field of underwater optical communication, especially underwater optical communication rate in short range is up to Mbps, even for transmission distance of 100m in the limpid sea, communication rate still can reach more than Mbps. But currently most research results about underwater optical communication focus on the point to point communication scenarios, and less attention is paid to optical communication UWSN. In 2013, Dr. Z. U. Ahmad at the University of Warwick proposed underwater optical wireless sensor network [20]. He applied linear topology and used laser as light source. However, dynamic marine environment made it difficult for pointing at the laser communication receiver. In 1998, the U.S. Naval Postgraduate School (NPS) pioneers created an underwater wireless wide-area network of distributed sensors to develop Seaweb research [21]. In 2005, by means of an underwater acoustic communications network, the U.S. Persistent Littoral Undersea Surveillance Network (PLUSNet) put the effort into providing autonomous detection and tracking of quiet submarines [22]. In 2014, U.S. Navy and the Woods Hole Oceanographic Institution Cooperated to invest $35.54 million for upgrading the underwater monitoring network [23], which will be completed in 2019. Japanese researchers applied ARENA networks in 2002 [24]. The Marine science and technology project MAST- III of EU also carried out a series of underwater acoustic sensor network projects, such as LOTUS and ROBLINKS. The mentioned above major implementations promoted the rapid development of the underwater acoustic sensor network. And there are a lot of achievements about underwater acoustic communication involving clock synchronization in underwater wireless sensor network, node localization, energy consumption controlling mechanism, network protocols, etc.[25]-[35] However, the inherent disadvantages of the underwater acoustic communication, 5 / 24
such as big propagation delay, lower transmission rate and higher power consumption make this kind of network only can be used for small capacity data transmission, not suitable for transmitting large-capacity and high-speed data, such as video or image. Professor I. Vasilescu at MIT and his colleagues presented a simply-equipped optical and acoustic hybrid UWSN. In this network, the static nodes were used to measure and store visible change in the environment. An autonomous underwater vehicle (AUV) can locate the static nodes using optical location system. The AUV visited the static nodes periodically to upload the data by optical communications link. However, the lower cruising speed restricted the execution performance that made real time transmission difficult. Professor N. Farr proposed an integrated, underwater optical/acoustic communications system which took advantage of optical and acoustic communications, but the system aimed to transmit data from one node to another node and could not be consider as a hybrid UWSN [36]. No more optical and acoustic hybrid UWSN has not been found, maybe this is due to the application restrictions of UWSN, for example, short transmission distance and inconvenience networking[26].
3. The Architecture of OA-UWSN This paper proposed a kind of novel optical-acoustic hybrid underwater wireless sensor network. The architecture of the novel OA-UWSN is shown in Figure 2. It consists of sink nodes, mobile nodes, fixed nodes, AUV and control center.
Figure 2 Architecture of OA-UWSN
Fixed nodes are deployed on the seabed. Fixed nodes and mobile nodes contain 6 / 24
acoustic transmission component, three acoustic receiving components (for positioning to obtain X-axis, and Y-axis coordinates), optical transmitter and optical receiver, camera (for capturing video and images), pressure sensors (for obtaining the Z-axis depth) and other various sensors (for obtaining parameters of the marine environment, such as temperature, humidity, salinity, and cross-sectional flow).The mobile nodes are suspended in the sea. They are charge of data collection and transmission to the mobile nodes at the upper layer, and then layered spreading to sink node. The data is processed by the control center, responsible for issuing control commands (acoustic signal). The absolute position information of the mobile nodes and fixed nodes can be obtained using GPS. To ensure network robustness, all nodes use acoustic link to form a network. In our solution, the optical links provide high-speed wireless transmission with large amounts of data at close range (such as video and images); the acoustic links are used to transmit control commands and node location. When the optical communication coverage is not available, the acoustic links serve as backup links by transporting small-capacity data. When all optical links are unavailable among nodes and some video data are needed in emergency circumstance, an AUV can collect the data using optical communication. Taking into account small amount of data transmission in down links and big amount of transmission in up links, our solution utilizes tree topology, and network strategy of uploading data to control center over multiple hops. Figure 3 presents an exemplary of a three-tier topology. Purple, red and blue nodes are all mobile nodes. The nodes with same color stand for they are in the same tier.
Figure 3 Topology architecture of OA-UWSN
This solution gives physical layer adaptive optical-acoustic switching and 7 / 24
adaptive coding modulation scheme, data link layer protocol and network layer routing protocol of OA-UWSN.
3.1 OA-UWSN Physical Layer Adaptive Optical-Acoustic Switching and Adaptive Coding Modulation Scheme Given the complexity of underwater environment, water depth, horizontal and vertical directions, and the large underwater transmission distance difference between acoustic and light medium, the communication mode (acoustic or optical communication), coding and modulation strategy based on channel conditions should be selected adaptively in physical layer. The optical communication will be divided into high-speed, medium-speed, low-speed operating modes, and coding modulation schemes are different according to different operating modes. Combined with the DLL protocol and in accordance with measured Signal to Noise Rate (SNR) level, the communication mode will be determined adaptively.
3.2 OA-UWSN DLL Protocol Considering features of underwater acoustic and optical communications, this solution designs the data link layer protocol to address the problem of different types of nodes access and channels assignment. They are as follows: 3.2.1 Channel access mechanism Directional optical communication transmission characteristics are fully utilized, and according to the theory of spatial multiplexing, the Time Division Multiple Access (TDMA) technology is introduced to DLL protocol based on spatial multiplex. We study DLL access mechanism based on the Time Division Multiple Access (TDMA) joint Carrier Sense Multiple Access/Collision Avoidance (CSMA/CA), give a super-frame structure which consists of competitive and non-competitive parts, and achieve Media Access Control (MAC) protocol based on optical directional transmission. 3.2.2 Data transmission, reception and acknowledgment mechanism This solution establishes adaptive optical or acoustic communication selection mechanism, and uses channel state feedback information carried by request and response at the sender and the receiver in the link establishment process to track changes in the underwater channel. Underwater acoustic RTS/CTS handshaking protocol are utilized to determine optical-acoustic communication modes, which 8 / 24
adaptive transmission rate adjustment at different distances and different channel environments can be achieved.
3.3 OA-UWSN Network Layer Routing Protocol Considering the harsh underwater communication environment, and the three-dimensional mobility of network node affected by ocean currents, the paper studies how to make the routing protocols to adapt to the frequent changes of network topology. The solution proposes network routing protocol, which is suitable for underwater communication environment and optical-acoustic hybrid communication mechanism. The network routing protocol sets up the network for the mobile nodes, fixed nodes and sink nodes, and chooses the best path for data acquisition of each node. Aiming at the difficult battery replacement and weak processing capacity of underwater network node, this paper studies how to reduce energy consumption in the process of networking, and make a reasonable choice of the transmission path to improve the survival time of network.
4. Physical Layer Adaptive Coding Modulation Scheme of OA-UWSN According to the communication channel conditions, the received SNR is divided into four levels from high to low; each level corresponds to a range of SNR values. When the received SNR is the fourth level (SNR worst level), it indicates that the transmission distances of underwater channel or channel conditions are insufficient to support communication, acoustic communication should be used; when the received SNR is the other levels, optical communication should be used. The receiver sets channel coding and modulation scheme based on the level of received SNR, and feds back to the transmitter. Table 2 presents four kinds of coding modulation schemes for OA-UWSN.
9 / 24
Table 2 Two Communication Modes and Four Coding Modulation Schemes High-speed optical communication
Medium-speed optical communication
Low-speed optical communication
Acoustic communication
Channel coding
RS code
RS code
RS code
TURBO code
Modulation
QAM
PPM
OOK
PSK
0-30 meter
30-60 meter
60-100 meter
>100 meter
10Mbps
1-2Mbps
100kbps
8kbps
Communication distance Transmission rate
4.1 High-Speed Optical Communication Mode QAM modulation can achieve higher spectral efficiency by phase and amplitude joint control, and higher rate of data transmission within a defined frequency band. The error correction performance of RS code is moderate. When the code length is short, the RS code has a good performance. The hardware implementation of RS code is relatively simple, and the bit error rate is low with short-range communication. Considering the difficulty of system implementation and hardware resource consumption, optical communication has opted for RS codes as channel coding scheme. The high-speed optical communication mode chooses QAM modulation and RS channel coding. For underwater wireless high-speed communication, the symbol duration is short, and inter-symbol interference is severe. OFDM uses a parallel transmission which can increase the symbol duration at the same data rate conditions. The cyclic prefix is also used. The receiver can receive without inter-symbol interference by simple first-order channel equalization in frequency domain. So high-speed optical communication mode selects OFDM and QAM to use their advantages. System transmission rate is up to 10Mbps, communication distance is up to 30 meters. Figures 4 and 5 respectively present the block diagram of the transmitter and receiver of high-speed optical communication mode.
10 / 24
RS coding
Interweave
QAM
OFDM
modulation
modulation
Optical transmitter module
Figure 4 Block diagram of the transmitter of high-speed optical mode
Optical receiver antennas
OFDM
QAM
demodulation
demodulation
Deinterweave
RS
Information
decoding
flow
Figure 5 Block diagram of the receiver of high-speed optical mode
4.2 Medium-Speed Optical Communication Mode PPM modulation has high energy transfer efficiency and strong anti-interference ability. PPM modulation can reduce the average transmit power. So medium-speed optical communication mode selects the PPM modulation and RS channel coding. System transmission rate is up to 1-2Mbps, communication distance is up to 30-60 meters. Figures 6 and 7 respectively present the block diagram of the transmitter and receiver of medium-speed optical communication mode. Information flow
Interwe ave
Optical transmitter module
Figure 6 Block diagram of the transmitter of medium-speed optical mode
Interweav e
Optical transmitter module
Optical receiver antennas
Figure 7 Block diagram of the receiver of medium-speed optical mode
4.3 Low-Speed Optical Communication Mode OOK modulation is simple and classical technology, and is more suitable for low SNR and low transmission rate cases. So low-speed optical communication mode selects the OOK modulation and RS channel coding. System transmission rate is up to 100kbps, communication distance is up to 60-100 meters. Figures 8 and 9 respectively present the block diagram of the transmitter and receiver of low-speed optical communication mode. 11 / 24
RS
Information Informati RS coding flow on flow Figure 8 Block diagram of the transmitter of low-speed optical mode
Deinterwea ve
decodi ng
Optical transmitte r module Figure 9 Block diagram of the receiver of low-speed optical mode
4.4 Acoustic Communication Mode PSK modulation has high energy transfer efficiency and strong anti-interference ability. PSK modulation can increase communication distance. TURBO code is close to random code, has very good distance characteristics, and thus has a strong anti-interference and anti-attenuation ability. After interweaving, compared to the performance of encoded information, the performance of TURBO code in additive white Gaussian noise channel has a few dB gains. As long as the receiver can detect frequency interference, delete the signal and make an error correction decoding, it can achieve the performance that other codes are difficult to achieve. TURBO code is particularly suitable for harsh environments and long-distance communication. The PSK modulation and TURBO channel coding are used in acoustic communication mode. System transmission rate is up to 8kbps, communication distance can exceed 100 meters. Figures 10 and 11 respectively present the block diagram of the transmitter and receiver of acoustic communication mode. Optical transmitter module
Deinterwe ave
Optical receiver antennas
RS decoding
Figure 10 Block diagram of the transmitter of acoustic mode
Informati on flow
Information flow
TURBO coding
PSK modulatio n
Figure 11 Block diagram of the receiver of acoustic mode
12 / 24
Interweave
5. DLL Protocol of OA-UWSN 5.1 Establishment of DLL Channel Access Mechanism This project designs a DLL access procedure based on TDMA joint CSMA/CA, meanwhile, the spatial multiplexing technology are introduced based on TDMA. The optical beam transmission characteristics are efficiently used to improve the data transmission rate and system efficiency. PicoNet Coordinator (PNC) of OA-UWSN divides time into slots by utilizing superframes structure. Using a superframe as a transmission cycle, and its length can be set to a fixed value or a changed value. A superframe includes activation and inactive. PNC runs in low power sleep mode during the inactive. The structure of a superframe is shown as figure 12.
Superframe #1
Quasi-omni beacon beacon in direction Quasi-omni direction #1
…
beacon in direction Quasi-omni direction #i
Superframe #2
Superframe #…
Contention Access Period CAP for direction# 1
…
CAP for direction# i
Superframe #m
Channel Time Allocation Period CTA1 CTA2 … CTAn
Inactive
Figure 12 A superframe structure of OA-UWSN
Activation is consist of three periods: Beacon, Contention Access Period (CAP), and Channel Time Allocation Period (CTAP). Beacon: The PNC broadcasts commands and access information by sending the beacon frame during acoustic communication. Directional optical communication mode is adopted in OA-UWSN, so the neighbor discovery is quite complex comparing with omnidirectional acoustic communication. To overcome this problem, a beacon period is divided into multiple time slots, and each time slot in turn send a beacon in different direction in order to make all the nodes access the network. Contention Access Period (CAP): CAP is a contention-based channel access period, and the equipment is connected to network by CSMA/CA scheme. CAP is mainly used to transmit data frames with low QoS. During this period, the strategy that the equipment completes the channel in all directions is adopted to ensure that all equipment can access the network successively. Channel Time Allocation Period (CTAP): PNC divides CTAP into several 13 / 24
channels Time Allocation (CTA). Then, CTA will be assigned to each link through TDMA. The length and position of CTA are decided by the channel sending request and the information is transmitted in CTA which has been established. In each allocated CTA, the equipment can alignment the data length on the premise of CTA time slot length is less than a given length. This period is mainly used to transmit data frames of high QoS, such as underwater video and image, etc. Inactive: When no data transmission needed, sensor nodes should be set to an inactive state to save power. Optical signal attenuates seriously in underwater condition, so no interference or only a little interference exist between two distant nodes. What’s more, the transmitter and the receiver beam transmission in underwater provides guarantee for directional spatial reuse. Different mobile nodes upload link would communicate in different CTA slots during CTAP if nodes work on TDMA because directional light are used to transmit data. If SDMA combined TDMA are adopted, different direction parallel transmission links with less interference at a time slot could be achieved which will increase the system capacity. As shown in figure 13, if mutual interference is very little between link 1, 2, 3 and 4, links can be arranged in the same CTA to parallel transmission which combining space division multiplexing with time division multiplexing can be achieved. The key to achieve high spatial multiplexing gain is designing rational links and time-slot allocation schemes, scheduling links with small interference into the same slot and meanwhile guaranteeing each link could be allocated to as many slots as possible. This paper maximizes the utilization rate of each CTA time slot on the premise of slot allocation fairness and improves the system capacity as much as possible. First, a Common Channel Interference (CCI) statistical table between links need to be established, and at the same time a link interference tolerance (LIT) table need to be set up which is used to calculate the upper limit of link interference under different transmission rate. In order to determine whether two or more links could coexist, the actual link interference in CCI table is compared with the interference tolerance of this link stored in LIT table. At the same time we maximize the utilization rate of each CTA time slot on the premise of slot allocation fairness to maximize system capacity of OA-UWSN.
14 / 24
Node 12
Node 11
4
nk
nk
Li
Li
3
Node 10
Node 8 Node 7
Node 6
Node 9
L
Node 1
k in
Node 4
1
Node 2
Lin
k 2
Node 5
Node 3
Figure 13 Diagram of parallel transmission links
5.2 Data Transmitting, Receiving and Acknowledgement mechanism According to channel conditions, the received SNR is divided into four level from high to low. The fourth level SNR (the worst level) indicates that the transmission distances or channel conditions are insufficient to support optical communication and acoustic communication should be chosen. Beam training is essential before optical communication. Mobile nodes are designed as a globate polyhedron with 20 transceiver plane (As shown in figure 14, the left one shows the planform.). Each plane contains the lens to achieve the congregation during optical transmission and reception. And every specified plane can transmit directional beam. Two-stage beam training mechanism are adopted which are sectors level (denoted as Si) training and precise beam level (denoted as Bi) training, as shown in figure 15. The mechanism of data transmitting, receiving and acknowledgement is shown in figure 16, which includes that selecting acoustic or optical communication mode according to channel conditions and SNR, turning on data transmission, switching communication mode and coding mode adaptively according to SNR before signal transmitting, and sending ACK when receiving data successfully, etc.
15 / 24
Figure 14 Design of mobile nodes
a) Sector beam
b) Precise beam
Figure 15 Two-dimensional beam model diagrams
16 / 24
Figure 16 Flow chart of the mechanism of data transmitting, receiving and acknowledgement
6. Network layer Routing Protocol of OA-UWSN In order to adapt the environments which network topology changes frequently,
reduce energy consumption of nodes used for the network routing and make full use 17 / 24
of high transmission characteristics of optical signal, this paper takes advantage of a reverse route scheme to search the optimal path and studies a routing protocol that temporary networks are set up by sink nodes periodically.
6.1 Networking Stage Firstly, search for nodes which are one hop away from sink nodes. Sink nodes find neighbor mobile nodes around by sending Hello packets by handshake agreement. Those mobile nodes have received Hello packets are called "the activated nodes". The receiving SNR, the residual energy of the sender and the depth of Z-axis are summed up to a weighted mean sorted by value and saved in the candidate neighbor table. Then the activated node selects a sink node with the biggest SNR as neighbor and returns an echo packet to it. So far, the route selection is completed. Take nodes in Fig. 17 as an example. Sink1, Sink2 and Sink3 send Hello packets around to search mobile nodes which can reach by one hop. The mobile node G receives the packets from Sink2 and Sink3 respectively. Then G makes a weighted average by the SNR of the transmission signal, the residual energy and the Z-axis depth, and sorts them by value. Finally, the results are saved in the candidate neighbor table. If Sink3 leads the list, it will be selected as a neighbor and G returns Echo packets to Sink3. In this period, the mobile node G uploads data to the neighbor Sink3 in priority.
Figure 17 Example of route searching Then find nodes which are multi hops away from sink nodes successively. The 18 / 24
activated mobile nodes start the next round search to find mobile nodes or fixed nodes which are two hops away from sink nodes. Each searched node makes a weighted average by the SNR of one or more groups of received Hello packets signals, the residual energy of the signal transmitter and the depth of Z-axis, and then restores it to its candidate neighbor table after ordering then by value. The node with the biggest weighted average is selected as the neighbor and the Echo response packet is returned to it. As a result, the neighbor will become the destination node to upload data. Take the scene in Fig. 17 for example. Node C has received the Hello packets from F 、 E and G. Make a comprehensive consideration of their channel communication quality, residual energy and the Z-axis depth. Suppose that E, F and G are included in the candidate neighbor table sequentially and C sends a Hello packet to E. In this cycle time, node C will give a priority to neighbor E to upload data. Repeat the cycle until all the nodes within the range of optical communication are connected in the network.
6.2 Data Transmission Stage All activated mobile nodes and activated fixed nodes can transmit data to their neighbor nodes. Data will be transferred layer upon layer via neighbor nodes and eventually be transported to sink node. Then, the node goes to sleep and waits for next upload. The longest duration of the data transfer can be set according to different circumstance.
6.3 Process of Special Circumstance If a node fails to establish a connection with other nodes within T2 time, acoustic communication mode will be started up and only sensor data will be transmitted. If a node cannot communicate with its current neighbor nodes due to the changes of the network topology, the node can select other nodes stored in the candidate neighbor table by making an attempt one by one until finding a new neighbor that can be contacted. If all the nodes in the candidate neighbor table cannot been communicated with, then acoustic model will be started and only small volume of data collected by sensors will be sent.
7. Simulation and numerical results In this section, we present simulation and numerical results for the OA-UWSN. The underwater acoustic (UWA) channel model and the related parameters in [37] are adopted when the acoustic communication mode of OA-UWSN is employed. [37] has presented numerical results demonstrating how the capacity is affected by the system 19 / 24
and channel parameters (i.e. link distance, the number, location, and power-delay profile (PDP) of significant taps) as well as environmental parameters (i.e. temperature, salinity, pressure, and depth). The underwater wireless optical communication channel models and the model parameters in paper [38] are adopted when optical communication mode of OA-UWSN is employed, including low-speed optical communication, medium-speed optical communication and high-speed optical communication. We consider a carrier frequency of 30 kHz for acoustic communication and 20W green/blue LEDs combined directional transmission for optical communication. The photomultiplier with a bandwidth of 140MHz is placed at the optical communication receiver. We assume that the temperature is T = 6°C and 200Mb video data is collected in a fixed node. The data is transmitted from a fixed node to a sink node. Two kinds of UWSN data transfer mode are simulated. One is classical acoustic transmission mode and another one is the self-adaptive multi-hop OA-UWSN mode proposed in this paper. The distance of each hop is 30 meters in multi-hop OA-UWSN mode. The distance from a fixed node to a sink node is assumed to be 30m, 60m and 90m separately. The classical acoustic transmission mode is used to transmit 30m, 60m and 90m separately by each time. The multi-hop OA-UWSN mode is used to transmit 30m, 60m and 90m separately by single hop, dual hop and triple hop. Assume that the underwater channels which correspond to pure sea, clear ocean, coastal and turbid harbor mode also correspond to high-speed optical communication, medium-speed optical
communication,
low-speed
optical
communication
and
acoustic
communication mode separately. The duration time of each channel environment is set to be 4s and the average transmission rate can be obtained after 200Mb data transmission is completed through four kinds of channels in turns, such as pure sea, clear ocean, coastal and turbid harbor channel in our simulation. We repeated the simulation about 1000 times and got the average transmission rate shown in table 3.
Table 3 Average Transmission Rate of the OA-UWSN and UWA Transmission Distance
30m
60m
90m
2.87Mbps
1.42Mbps
360Kbps
178Kbps
0.96Mbps 86Kbps
Mode OA-UWSN UWA
20 / 24
8. Summary This paper proposed a solution of OA-UWSN which maximized the advantages of underwater optical and acoustic communication. High-speed transmission could be achieved within short communication distance by means of optical communication. Transmitting control commands and node localization were achieved via acoustic communication based on the advantages of long distance transmission. The noise of the OA-UWSN was effectively reduced with no interference between acoustic and optical signals. At the mean time less use of acoustic signals could lower the sound harm to marine lives.
ACKNOWLEDGEMENTS The authors would like to thank the referees and editors for providing very helpful comments and suggestions. This project was the National Natural Science Foundation of China (No. 61671261, No. 61304222), the Key Research and Development Program of Shandong Provice (No.2016GGX101007) and the Open Research Fund from Shandong provincial Key Laboratory of Computer Network (No. SDKLCN-2015-04).
Reference ［1］ [Online]http://www.noaa.gov/educational-resources ［2］ W. Wei, N. N. Chen, X. Zhang, J. H. Rao and W. B. Wang, Research and survey of wireless optical communication for Underwater Wireless Sensor Networks[J], Sensor world, 2011, 17(3): 6-12 ［3］ [Online]http://www.nrdc.org/ ［4］ G.D. Yao, Z.G. Jin, and Y.S. Su, An environment-friendly spectrum decision strategy for underwater acoustic networks[J], Journal of Network and Computer Applications, 2016, 73: 82-93 ［5］ [Online]http://www.physics.umd.edu/grt/taj/104a/104anotessupps.html,
T.
Jacobson,
“Phys104,” 2008 ［6］ M. Callaham, Submarine communications[J], IEEE Magazine, 1981, 19(11):16-25 ［7］ J. Puschell, R. Giannaris, and L.Stotts, The autonomous data optical relay experiment: first two way laser communication between an aircraft and submarine[C], Proc. National Telesystems Conference (NTC’92), Washington, DC, May. 1992, 1427-1430 ［8］ T. Wiener and S.Karp, The role of blue/green laser systems in strategic submarine communications[J], IEEE Translations on Communications, 1980, 28(9): 1602-1607 ［9］ I. Vasilescu, K. Kotay, D. Rus, M. Dunbabin, and P. Corke, Data Collection, Storage, 21 / 24
and Retrieval with an Underwater Sensor Network[C]. Proceedings of the 3rd International Conference on Embedded Networked Sensor Systems, 2005, 154-165 ［10］ W. C. Cox, A 1 Mbps Underwater Communication System Using a 405 nm Laser Diode and Photomultiplier tube [D]. North Carolina State University: M.S. Thesis, 2008 ［11］ J. A. Simpson. A 1 Mbps Underwater Communications System Using LEDs and Photodiodes with Signal Processing Capability [D]. North Carolina State University: M.S. Thesis, 2007 ［12］ J. A. Simpson, W. C. Cox, J. R. Krier, B. Cochenour, B. Hughes, and J. F. Muth, 5 Mbps Optical Wireless Communication with Error Correction Coding for Underwater Sensor Nodes[C]. IEEE Pro. OCEANS 2010, Seattle, WA, 2010, 1-4 ［13］ C. Pontbriand, N. Farr, J. Ware, J. Preisig, and H. Popenoe, Diffuse high-bandwidth optical communications, Proc. MTS/IEEE OCEANS (OCEANS’08), Quebec City, QC, Canada, Sep. 2008, 1-4 ［14］ N. Farr, J. Ware, C. Pontbriand, and M. Tivey, Demonstration of wireless data harvesting from a subsea node using a ship of opportunity[C], Proc. MTS/IEEE OCEANS (OCEANS’13), San Diego, CA, Sep. 2013, 1-5 ［15］ C.P. Li, K. H. Park, and M. S. Alouini, On the Use of a Direct Radiative Transfer Equation Solver for Path Loss Calculation in Underwater Optical Wireless Channels [J], IEEE Wireless Communication Letters, 2015, 4(5):561-564 ［16］ H. M. Oubei, C. P. Li, K. H. Park, T. K. Ng, M. S. Alouini, and B. S. Ooi, 2.3 Gbit/s underwater wireless optical communications using directly modulated 520 nm laser diode, Optics express, 2015, 23(16): 20743-20748 ［17］ Y. Dong, S. Tang, X. Zhang, Effect of Random Sea Surface on Downlink Underwater Wireless Optical Communications [J], IEEE Journal on Selected Areas in Communications, 2013, 17(11): 2164-2167 ［18］ H. Zhang, and Y. Dong, Impulse response modeling for general underwater wireless optical MIMO links[J], IEEE Communications Magazine, 2016, 54(2): 56-61 ［19］ H. Zhang, and Y. Dong, General Stochastic Channel Model and Performance Evaluation for
Underwater
Wireless
Optical
Links[J],
IEEE
Transactions
on
Wireless
Communications, 2016, 15(2): 1162-1173 ［20］ Z. U. Ahmad, Underwater optical wireless sensor network [D], University of Warwick， 2013 ［21］
Naval
Postgraduate
School
USA,
http://www.nps.edu/About/Nev>s/NPS-Pioneers-Seaweb-Underwaler-Sensor-Networks.h tm 22 / 24
［22］ M. Grund, L. Freitag, and J. Preisig, The PLUSNet underwater communications system: acoustic telemetry for undersea surveillance[C], Proceedings of IEEE OCEANS, 2006, 1-5 ［23］ United States Department of Defense(DOD), USA, http://www.defense.gov/contracts /contract.aspx?contractid=5409 ［24］ J. Kasahara, Y. Shirasaki, and K. Asakawa, Scientific application of ARENA networks[C], Proceedings of International Workshop on SUSCRT, 2003, 272-275 ［25］ A.P. Das, and S.M. Thampi, Fault-resilient localization for underwater sensor networks[J], Ad Hoc Networks, 2016, 55(2): 132–142 ［26］ Z. Zeng, H. Zhang, Y. Dong, and J. Cheng, A Survey of Underwater Wireless Optical Communication,
IEEE
Communications
Surveys
&
Tutorials,
2016,
DOI:
10.1109/COMST.2016.2618841 ［27］ N. Nusrat, K. Gour, K. Joarder, PRADD: Apathreliability-aware data delivery protocol for underwater acoustic sensor networks[J], Journal of Network and Computer Applications, 2016, 75:385-397 ［28］ N.Z. Zenia, M. Aseeri, M.R. Ahmed, Z.I. Chowdhury, and M.S. Kaiser, Energy-efficiency and reliability in MAC and routing protocols for underwater wireless sensor network: A survey[J], Journal of Network and Computer Applications, 2016, 71: 72-85 ［29］ X. Guo, F. X. Ge, S. L. Ren, and L. H. Guo, Eigenanalysis-based adaptive beamforming using worst-case performance optimization[J], Acta Acustica, 2015, 40(2):187-197 ［30］ J. L. Wang, Y. H. Liu, Y. He, W. Dong, and M. Li, QoF: Towards Comprehensive Path Quality Measurement in Wireless Sensor Networks[J], IEEE Transactions on Parallel and Distributed Systems, 2014, 25(4): 1003-1013 ［31］ Z. Liao, D. Li, and J. Chen, Joint Bandwidth Optimization and Media Access Control for Multihop Underwater Acoustic Sensor Networks[J], IEEE Sensors Journal, 2015, 15(8): 4292-4304 ［32］ L. F. Liu, N. S. Zhang, Y. Liu, Topology control models and solutions for signal irregularity in mobile underwater wireless sensor networks[J], Journal of Network and Computer Applications, 2015, 51: 68-90 ［33］ F. Qu, Z. Wang, L. Yang, and Z. Wu, A journey toward modeling and resolving doppler in underwater acoustic communications[J], IEEE Communications Magazine, 2016, 54(2): 49-55 ［34］ S.F. Yan, Optimal design of modal beamformers for circular arrays[J], The Journal of the Acoustical Society of America, 2015, 138(4):2140-2151 ［35］ J. F. Jiang, G. J. Han, H. Guo, L. Shu, J. J. Rodrigues, Geographic multipath routing 23 / 24
based on geospatial division in duty-cycled underwater wireless sensor networks[J], Journal of Network and Computer Applications, 2016, 59: 4-13 ［36］ N. Farr, A. Bowen, J. Ware, C. Pontbriand, and M. Tivey, An integrated, underwater optical/acoustic communications system[C], Proc. IEEE/MTS OCEANS, Sydney, NSW, Australia, May 2010, pp. 1–6. ［37］ H. Nouri, M. Uysal, E. Panayirce, and H. Senol, Information theoretical performance limits of single-carrier underwater acoustic systems[J], IET Communications, 2014, 8(15): 2599-2610 ［38］ C. Gabriel, M. A. Khalighi, S. Bourennane, and P. Leon, Monte-Carlo-based channel characterization for underwater optical communication systems [J]. IEEE/OSA Journal of Optical Communications & Networking, 2013, 5(1):1-12.
24 / 24
PII: DOI: Reference:
S1084-8045(17)30084-X http://dx.doi.org/10.1016/j.jnca.2017.02.014 YJNCA1870
To appear in: Journal of Network and Computer Applications Received date: 1 August 2016 Revised date: 24 January 2017 Accepted date: 21 February 2017 Cite this article as: Jingjing Wang, Wei Shi, Lingwei Xu, Liya Zhou, Qiuna Niu a n d Juliu, Design of Optical-Acoustic Hybrid Underwater Wireless Sensor N e t w o r k , Journal of Network and Computer Applications, http://dx.doi.org/10.1016/j.jnca.2017.02.014 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Design of Optical-Acoustic Hybrid Underwater Wireless Sensor Network Jingjing Wang1,2*, Wei Shi1, Lingwei Xu1, Liya Zhou1, Qiuna Niu1,3 and Juliu2 1
School of Information Science and Technology, Qingdao University of Science and Technology, Qingdao, Shandong Province, China 2 School of Information Science and Engineer, Shandong University, Jinan，Shandong Province, China 3 Shandong provincial Key Laboratory of Computer Network, Jinan, Shandong Province, China *
Corresponding author: Jingjing Wang, [email protected]
Abstract: High-speed underwater wireless transmission technologies are urgently demanded to transmit three-dimensional high-resolution observation data timely in ocean exploration. Currently, underwater acoustic sensor networks have obtained high transmission distance, but with low date rate, high power consumption, high cost and damage to marine mammals. Meanwhile, underwater optical communications can achieve high date rate, but with difficult networking and short communication distance. This paper proposed a kind of novel Optical-Acoustic hybrid Underwater Wireless Sensor Network (OA-UWSN) which employed optical communication for high-speed transmission at close range and employed acoustic communication for transmitting control commands and node localization. OA-UWSN accomplished long-distance optical transmission through directional optical communication and multi-hop transmission mechanism. The problem of difficult networking in optical communication was solved by designing the whole space optical repeater. With the characteristics of directional optical transmission, the Space Division Multiple Access (SDMA) technology was employed in Data Link Layer (DLL) to improve the data transmission efficiency. A routing protocol based on reverse route search was designed to improve the network life cycle and adapt to network topology which changed frequently. The paper proposed a solution for wireless transmission of real-time video and images in marine exploration and provided new methods for high-speed transmission of marine information detection.
Keywords: Optical-acoustic; Hybrid; Underwater sensor networks; Encoding modulation; Space division multiplexing; Multiuser access
1 / 24
1. Introduction The statistical data of U.S. National Oceanic and Atmospheric Administration in 2015 showed that 95% of the underwater world had not yet been detected [1]. The construction of underwater wireless sensor network (UWSN) and the provision of large scales of ocean exploration data with high spatial-temporal resolution, is of great significance for strengthening the utilization of ecological environment monitoring, ocean energy development, and marine scientific research. At present, there are three kinds of underwater wireless communication technology: underwater electromagnetic wave communication, underwater acoustic communication and underwater optical communication. The attenuation of electromagnetic wave in water is serious. For example, the attenuation of 2.4GHz band, which is used in Bluetooth technology and wireless LAN technology, is about 1695dB/m in the sea, the value is as high as 189dB/m even in pure fresh water [2]. The ultra low frequency electromagnetic wave can reduce the attenuation, while the system cost is high and the data transmission rate is low. At current, underwater radio communication system is only limited to shallow sea areas, and requires a huge antenna system. The acoustic communication is the most widely technology applied in underwater wireless communication over decades, but the speed of sound in water is only 1.5×103 meters per second, which possesses big difference, 5 orders of magnitude, with the speed of light in the water (3×108 m/s). In particular, factors such as the acoustic wave scattering, the inherent loss of transmission and return interference, make the development of underwater acoustic equipment very difficult. According to the report of the Natural Resources Defense Council, military sonar and other rising ocean noise is a serious impact on the life of dolphins, whales and other marine mammals, and even cause their hearing loss or death [3]. The most significant advantage of underwater acoustic communication is far transmission distance, but its propagation delay and signal attenuation is huge, communication bandwidth is low, multipath effect is serious and marine mammals may be harmed[4]. The underwater wireless optical communication is a kind of communication mode with light wave as the carrier of information. S. A. Sullian and S. Q. Dimtley et al in 1963 found in their study of propagation characteristics of light waves in the ocean that attenuation of 450 ~ 550nm in blue and green light in seawater is much smaller than the other bands of light [5], as shown in Figure 1. This significant physical discovery laid a solid foundation to solve long-term underwater target 2 / 24
detection, communication and the other essential problems.
Figure 1. Attenuation of light wave in seawater [4]
The performance comparison of three kinds of underwater wireless communication technology is shown as table 1. Wireless optical communication owns an overwhelming advantage in speed, power consumption, volume and other aspects, however, the short communication distance of wireless optical communication restricts the further development. This paper presented a kind of novel optical-acoustic hybrid underwater wireless sensor
network
which
maximized
the
advantages
of
underwater
optical
communication and acoustic communication. The optical communication was used to achieve high-speed transmission at close range and acoustic communication was used to transmit control command and node localization based on the advantages of long distance transmission. The rest of the paper is organized as follows. The research progress and development
of
underwater
optical
communication,
underwater
acoustic
communication and optical-acoustic hybrid UWSN are presented in turn in Section 2. Section 3 provides the solution of a new Optical-Acoustic Hybrid UWSN. The physical layer adaptive coding modulation schemes of the OA-UWSN are given in Section 4. The protocol of DLL and the routing protocol are proposed in Section 5 and Section 6 separately. In Section 7, we present simulation and numerical results of 3 / 24
the OA-UWSN and the traditional underwater acoustic sensor network. The summary is drawn in Section 8. Table 1. Comparison of three underwater wireless communication modes Transmission medium Parameter
Underwater acoustic communication
Underwater optical communication
One hundred meters
Several kilometers
Tens of meters to hundreds of meters
100bps
Several Kbps
Several Mbps
Transmit power Underwater transport rate Equipment volume
MW grade Near to speed of light Large
KW to tens of KW
Large
mW to W grade Near to speed of light Little
Equipment cost Confidentiality
high --
Interference immunity
Electromagnetic interference
high Bad Free from electromagnetic interference.
Low Good Free from electromagnetic interference.
Single node communication distance Communication bandwidth
Ultra low frequency long wave radio system
1500m/s
2. Background of underwater wireless communication Underwater optical communication technology in earlier stage is mainly used in the field of military. In 1977 and 1983, the U.S. Navy,
U.S. Defense Advanced
Research Project Agency and the former Soviet Union tested laser communication with submarine and communication by laser beam reflected through space orbit [6]-[8]. In 2005, Professor I. Vasilescu etc. at the Massachusetts Institute of Technology built the underwater optical network, they only spent 21 seconds to read 6.86 MB data through optical communication link, and spent 1.3 days through acoustic communication link at the speed of 480 bps. What’s more, the energy consumption of acoustic communication is dozens of times as much as optical communication [9]. In 2007 and 2008, the team of Professor J. Muth at the North Carolina state university in U.S. developed underwater optical communication systems respectively based on laser diode/PIN photodiode and light-emitting diodes/avalanche photodiode. They successfully realized the reliable transmission at the rate of 1 Mbps by blue-green in tank with a side of length 5m [10] [11]. Moreover, in 2010, J. A. Simpson and W. C. Cox etc. in the same team proposed 5 Mbps Optical Wireless Communication based on the RS error correction coding [12].
In 2011,
they also realize transmission at the rate level of Mbps by using laser and LED as 4 / 24
light source. Professor N. Farr etc. at Woods Hole Oceanographic Institution and Professor M. S. Alouini and his team at the King Abdullah University of Science and Technology have done a lot of advanced work on underwater optical communication channel modeling, communication tests under marine environment and other fields [13]-[16]. From 2013 to 2015, Dr. Yuhan Dong at Tsing Hua University considered parallel link and vertical link geometric models of the underwater optical link, built a new channel model, and applied digital method to describe the relationships between transmission power, link range, wind speed, bit error rate and water features, etc [17]-[19]. Global scientists have made important progress in the field of underwater optical communication, especially underwater optical communication rate in short range is up to Mbps, even for transmission distance of 100m in the limpid sea, communication rate still can reach more than Mbps. But currently most research results about underwater optical communication focus on the point to point communication scenarios, and less attention is paid to optical communication UWSN. In 2013, Dr. Z. U. Ahmad at the University of Warwick proposed underwater optical wireless sensor network [20]. He applied linear topology and used laser as light source. However, dynamic marine environment made it difficult for pointing at the laser communication receiver. In 1998, the U.S. Naval Postgraduate School (NPS) pioneers created an underwater wireless wide-area network of distributed sensors to develop Seaweb research [21]. In 2005, by means of an underwater acoustic communications network, the U.S. Persistent Littoral Undersea Surveillance Network (PLUSNet) put the effort into providing autonomous detection and tracking of quiet submarines [22]. In 2014, U.S. Navy and the Woods Hole Oceanographic Institution Cooperated to invest $35.54 million for upgrading the underwater monitoring network [23], which will be completed in 2019. Japanese researchers applied ARENA networks in 2002 [24]. The Marine science and technology project MAST- III of EU also carried out a series of underwater acoustic sensor network projects, such as LOTUS and ROBLINKS. The mentioned above major implementations promoted the rapid development of the underwater acoustic sensor network. And there are a lot of achievements about underwater acoustic communication involving clock synchronization in underwater wireless sensor network, node localization, energy consumption controlling mechanism, network protocols, etc.[25]-[35] However, the inherent disadvantages of the underwater acoustic communication, 5 / 24
such as big propagation delay, lower transmission rate and higher power consumption make this kind of network only can be used for small capacity data transmission, not suitable for transmitting large-capacity and high-speed data, such as video or image. Professor I. Vasilescu at MIT and his colleagues presented a simply-equipped optical and acoustic hybrid UWSN. In this network, the static nodes were used to measure and store visible change in the environment. An autonomous underwater vehicle (AUV) can locate the static nodes using optical location system. The AUV visited the static nodes periodically to upload the data by optical communications link. However, the lower cruising speed restricted the execution performance that made real time transmission difficult. Professor N. Farr proposed an integrated, underwater optical/acoustic communications system which took advantage of optical and acoustic communications, but the system aimed to transmit data from one node to another node and could not be consider as a hybrid UWSN [36]. No more optical and acoustic hybrid UWSN has not been found, maybe this is due to the application restrictions of UWSN, for example, short transmission distance and inconvenience networking[26].
3. The Architecture of OA-UWSN This paper proposed a kind of novel optical-acoustic hybrid underwater wireless sensor network. The architecture of the novel OA-UWSN is shown in Figure 2. It consists of sink nodes, mobile nodes, fixed nodes, AUV and control center.
Figure 2 Architecture of OA-UWSN
Fixed nodes are deployed on the seabed. Fixed nodes and mobile nodes contain 6 / 24
acoustic transmission component, three acoustic receiving components (for positioning to obtain X-axis, and Y-axis coordinates), optical transmitter and optical receiver, camera (for capturing video and images), pressure sensors (for obtaining the Z-axis depth) and other various sensors (for obtaining parameters of the marine environment, such as temperature, humidity, salinity, and cross-sectional flow).The mobile nodes are suspended in the sea. They are charge of data collection and transmission to the mobile nodes at the upper layer, and then layered spreading to sink node. The data is processed by the control center, responsible for issuing control commands (acoustic signal). The absolute position information of the mobile nodes and fixed nodes can be obtained using GPS. To ensure network robustness, all nodes use acoustic link to form a network. In our solution, the optical links provide high-speed wireless transmission with large amounts of data at close range (such as video and images); the acoustic links are used to transmit control commands and node location. When the optical communication coverage is not available, the acoustic links serve as backup links by transporting small-capacity data. When all optical links are unavailable among nodes and some video data are needed in emergency circumstance, an AUV can collect the data using optical communication. Taking into account small amount of data transmission in down links and big amount of transmission in up links, our solution utilizes tree topology, and network strategy of uploading data to control center over multiple hops. Figure 3 presents an exemplary of a three-tier topology. Purple, red and blue nodes are all mobile nodes. The nodes with same color stand for they are in the same tier.
Figure 3 Topology architecture of OA-UWSN
This solution gives physical layer adaptive optical-acoustic switching and 7 / 24
adaptive coding modulation scheme, data link layer protocol and network layer routing protocol of OA-UWSN.
3.1 OA-UWSN Physical Layer Adaptive Optical-Acoustic Switching and Adaptive Coding Modulation Scheme Given the complexity of underwater environment, water depth, horizontal and vertical directions, and the large underwater transmission distance difference between acoustic and light medium, the communication mode (acoustic or optical communication), coding and modulation strategy based on channel conditions should be selected adaptively in physical layer. The optical communication will be divided into high-speed, medium-speed, low-speed operating modes, and coding modulation schemes are different according to different operating modes. Combined with the DLL protocol and in accordance with measured Signal to Noise Rate (SNR) level, the communication mode will be determined adaptively.
3.2 OA-UWSN DLL Protocol Considering features of underwater acoustic and optical communications, this solution designs the data link layer protocol to address the problem of different types of nodes access and channels assignment. They are as follows: 3.2.1 Channel access mechanism Directional optical communication transmission characteristics are fully utilized, and according to the theory of spatial multiplexing, the Time Division Multiple Access (TDMA) technology is introduced to DLL protocol based on spatial multiplex. We study DLL access mechanism based on the Time Division Multiple Access (TDMA) joint Carrier Sense Multiple Access/Collision Avoidance (CSMA/CA), give a super-frame structure which consists of competitive and non-competitive parts, and achieve Media Access Control (MAC) protocol based on optical directional transmission. 3.2.2 Data transmission, reception and acknowledgment mechanism This solution establishes adaptive optical or acoustic communication selection mechanism, and uses channel state feedback information carried by request and response at the sender and the receiver in the link establishment process to track changes in the underwater channel. Underwater acoustic RTS/CTS handshaking protocol are utilized to determine optical-acoustic communication modes, which 8 / 24
adaptive transmission rate adjustment at different distances and different channel environments can be achieved.
3.3 OA-UWSN Network Layer Routing Protocol Considering the harsh underwater communication environment, and the three-dimensional mobility of network node affected by ocean currents, the paper studies how to make the routing protocols to adapt to the frequent changes of network topology. The solution proposes network routing protocol, which is suitable for underwater communication environment and optical-acoustic hybrid communication mechanism. The network routing protocol sets up the network for the mobile nodes, fixed nodes and sink nodes, and chooses the best path for data acquisition of each node. Aiming at the difficult battery replacement and weak processing capacity of underwater network node, this paper studies how to reduce energy consumption in the process of networking, and make a reasonable choice of the transmission path to improve the survival time of network.
4. Physical Layer Adaptive Coding Modulation Scheme of OA-UWSN According to the communication channel conditions, the received SNR is divided into four levels from high to low; each level corresponds to a range of SNR values. When the received SNR is the fourth level (SNR worst level), it indicates that the transmission distances of underwater channel or channel conditions are insufficient to support communication, acoustic communication should be used; when the received SNR is the other levels, optical communication should be used. The receiver sets channel coding and modulation scheme based on the level of received SNR, and feds back to the transmitter. Table 2 presents four kinds of coding modulation schemes for OA-UWSN.
9 / 24
Table 2 Two Communication Modes and Four Coding Modulation Schemes High-speed optical communication
Medium-speed optical communication
Low-speed optical communication
Acoustic communication
Channel coding
RS code
RS code
RS code
TURBO code
Modulation
QAM
PPM
OOK
PSK
0-30 meter
30-60 meter
60-100 meter
>100 meter
10Mbps
1-2Mbps
100kbps
8kbps
Communication distance Transmission rate
4.1 High-Speed Optical Communication Mode QAM modulation can achieve higher spectral efficiency by phase and amplitude joint control, and higher rate of data transmission within a defined frequency band. The error correction performance of RS code is moderate. When the code length is short, the RS code has a good performance. The hardware implementation of RS code is relatively simple, and the bit error rate is low with short-range communication. Considering the difficulty of system implementation and hardware resource consumption, optical communication has opted for RS codes as channel coding scheme. The high-speed optical communication mode chooses QAM modulation and RS channel coding. For underwater wireless high-speed communication, the symbol duration is short, and inter-symbol interference is severe. OFDM uses a parallel transmission which can increase the symbol duration at the same data rate conditions. The cyclic prefix is also used. The receiver can receive without inter-symbol interference by simple first-order channel equalization in frequency domain. So high-speed optical communication mode selects OFDM and QAM to use their advantages. System transmission rate is up to 10Mbps, communication distance is up to 30 meters. Figures 4 and 5 respectively present the block diagram of the transmitter and receiver of high-speed optical communication mode.
10 / 24
RS coding
Interweave
QAM
OFDM
modulation
modulation
Optical transmitter module
Figure 4 Block diagram of the transmitter of high-speed optical mode
Optical receiver antennas
OFDM
QAM
demodulation
demodulation
Deinterweave
RS
Information
decoding
flow
Figure 5 Block diagram of the receiver of high-speed optical mode
4.2 Medium-Speed Optical Communication Mode PPM modulation has high energy transfer efficiency and strong anti-interference ability. PPM modulation can reduce the average transmit power. So medium-speed optical communication mode selects the PPM modulation and RS channel coding. System transmission rate is up to 1-2Mbps, communication distance is up to 30-60 meters. Figures 6 and 7 respectively present the block diagram of the transmitter and receiver of medium-speed optical communication mode. Information flow
Interwe ave
Optical transmitter module
Figure 6 Block diagram of the transmitter of medium-speed optical mode
Interweav e
Optical transmitter module
Optical receiver antennas
Figure 7 Block diagram of the receiver of medium-speed optical mode
4.3 Low-Speed Optical Communication Mode OOK modulation is simple and classical technology, and is more suitable for low SNR and low transmission rate cases. So low-speed optical communication mode selects the OOK modulation and RS channel coding. System transmission rate is up to 100kbps, communication distance is up to 60-100 meters. Figures 8 and 9 respectively present the block diagram of the transmitter and receiver of low-speed optical communication mode. 11 / 24
RS
Information Informati RS coding flow on flow Figure 8 Block diagram of the transmitter of low-speed optical mode
Deinterwea ve
decodi ng
Optical transmitte r module Figure 9 Block diagram of the receiver of low-speed optical mode
4.4 Acoustic Communication Mode PSK modulation has high energy transfer efficiency and strong anti-interference ability. PSK modulation can increase communication distance. TURBO code is close to random code, has very good distance characteristics, and thus has a strong anti-interference and anti-attenuation ability. After interweaving, compared to the performance of encoded information, the performance of TURBO code in additive white Gaussian noise channel has a few dB gains. As long as the receiver can detect frequency interference, delete the signal and make an error correction decoding, it can achieve the performance that other codes are difficult to achieve. TURBO code is particularly suitable for harsh environments and long-distance communication. The PSK modulation and TURBO channel coding are used in acoustic communication mode. System transmission rate is up to 8kbps, communication distance can exceed 100 meters. Figures 10 and 11 respectively present the block diagram of the transmitter and receiver of acoustic communication mode. Optical transmitter module
Deinterwe ave
Optical receiver antennas
RS decoding
Figure 10 Block diagram of the transmitter of acoustic mode
Informati on flow
Information flow
TURBO coding
PSK modulatio n
Figure 11 Block diagram of the receiver of acoustic mode
12 / 24
Interweave
5. DLL Protocol of OA-UWSN 5.1 Establishment of DLL Channel Access Mechanism This project designs a DLL access procedure based on TDMA joint CSMA/CA, meanwhile, the spatial multiplexing technology are introduced based on TDMA. The optical beam transmission characteristics are efficiently used to improve the data transmission rate and system efficiency. PicoNet Coordinator (PNC) of OA-UWSN divides time into slots by utilizing superframes structure. Using a superframe as a transmission cycle, and its length can be set to a fixed value or a changed value. A superframe includes activation and inactive. PNC runs in low power sleep mode during the inactive. The structure of a superframe is shown as figure 12.
Superframe #1
Quasi-omni beacon beacon in direction Quasi-omni direction #1
…
beacon in direction Quasi-omni direction #i
Superframe #2
Superframe #…
Contention Access Period CAP for direction# 1
…
CAP for direction# i
Superframe #m
Channel Time Allocation Period CTA1 CTA2 … CTAn
Inactive
Figure 12 A superframe structure of OA-UWSN
Activation is consist of three periods: Beacon, Contention Access Period (CAP), and Channel Time Allocation Period (CTAP). Beacon: The PNC broadcasts commands and access information by sending the beacon frame during acoustic communication. Directional optical communication mode is adopted in OA-UWSN, so the neighbor discovery is quite complex comparing with omnidirectional acoustic communication. To overcome this problem, a beacon period is divided into multiple time slots, and each time slot in turn send a beacon in different direction in order to make all the nodes access the network. Contention Access Period (CAP): CAP is a contention-based channel access period, and the equipment is connected to network by CSMA/CA scheme. CAP is mainly used to transmit data frames with low QoS. During this period, the strategy that the equipment completes the channel in all directions is adopted to ensure that all equipment can access the network successively. Channel Time Allocation Period (CTAP): PNC divides CTAP into several 13 / 24
channels Time Allocation (CTA). Then, CTA will be assigned to each link through TDMA. The length and position of CTA are decided by the channel sending request and the information is transmitted in CTA which has been established. In each allocated CTA, the equipment can alignment the data length on the premise of CTA time slot length is less than a given length. This period is mainly used to transmit data frames of high QoS, such as underwater video and image, etc. Inactive: When no data transmission needed, sensor nodes should be set to an inactive state to save power. Optical signal attenuates seriously in underwater condition, so no interference or only a little interference exist between two distant nodes. What’s more, the transmitter and the receiver beam transmission in underwater provides guarantee for directional spatial reuse. Different mobile nodes upload link would communicate in different CTA slots during CTAP if nodes work on TDMA because directional light are used to transmit data. If SDMA combined TDMA are adopted, different direction parallel transmission links with less interference at a time slot could be achieved which will increase the system capacity. As shown in figure 13, if mutual interference is very little between link 1, 2, 3 and 4, links can be arranged in the same CTA to parallel transmission which combining space division multiplexing with time division multiplexing can be achieved. The key to achieve high spatial multiplexing gain is designing rational links and time-slot allocation schemes, scheduling links with small interference into the same slot and meanwhile guaranteeing each link could be allocated to as many slots as possible. This paper maximizes the utilization rate of each CTA time slot on the premise of slot allocation fairness and improves the system capacity as much as possible. First, a Common Channel Interference (CCI) statistical table between links need to be established, and at the same time a link interference tolerance (LIT) table need to be set up which is used to calculate the upper limit of link interference under different transmission rate. In order to determine whether two or more links could coexist, the actual link interference in CCI table is compared with the interference tolerance of this link stored in LIT table. At the same time we maximize the utilization rate of each CTA time slot on the premise of slot allocation fairness to maximize system capacity of OA-UWSN.
14 / 24
Node 12
Node 11
4
nk
nk
Li
Li
3
Node 10
Node 8 Node 7
Node 6
Node 9
L
Node 1
k in
Node 4
1
Node 2
Lin
k 2
Node 5
Node 3
Figure 13 Diagram of parallel transmission links
5.2 Data Transmitting, Receiving and Acknowledgement mechanism According to channel conditions, the received SNR is divided into four level from high to low. The fourth level SNR (the worst level) indicates that the transmission distances or channel conditions are insufficient to support optical communication and acoustic communication should be chosen. Beam training is essential before optical communication. Mobile nodes are designed as a globate polyhedron with 20 transceiver plane (As shown in figure 14, the left one shows the planform.). Each plane contains the lens to achieve the congregation during optical transmission and reception. And every specified plane can transmit directional beam. Two-stage beam training mechanism are adopted which are sectors level (denoted as Si) training and precise beam level (denoted as Bi) training, as shown in figure 15. The mechanism of data transmitting, receiving and acknowledgement is shown in figure 16, which includes that selecting acoustic or optical communication mode according to channel conditions and SNR, turning on data transmission, switching communication mode and coding mode adaptively according to SNR before signal transmitting, and sending ACK when receiving data successfully, etc.
15 / 24
Figure 14 Design of mobile nodes
a) Sector beam
b) Precise beam
Figure 15 Two-dimensional beam model diagrams
16 / 24
Figure 16 Flow chart of the mechanism of data transmitting, receiving and acknowledgement
6. Network layer Routing Protocol of OA-UWSN In order to adapt the environments which network topology changes frequently,
reduce energy consumption of nodes used for the network routing and make full use 17 / 24
of high transmission characteristics of optical signal, this paper takes advantage of a reverse route scheme to search the optimal path and studies a routing protocol that temporary networks are set up by sink nodes periodically.
6.1 Networking Stage Firstly, search for nodes which are one hop away from sink nodes. Sink nodes find neighbor mobile nodes around by sending Hello packets by handshake agreement. Those mobile nodes have received Hello packets are called "the activated nodes". The receiving SNR, the residual energy of the sender and the depth of Z-axis are summed up to a weighted mean sorted by value and saved in the candidate neighbor table. Then the activated node selects a sink node with the biggest SNR as neighbor and returns an echo packet to it. So far, the route selection is completed. Take nodes in Fig. 17 as an example. Sink1, Sink2 and Sink3 send Hello packets around to search mobile nodes which can reach by one hop. The mobile node G receives the packets from Sink2 and Sink3 respectively. Then G makes a weighted average by the SNR of the transmission signal, the residual energy and the Z-axis depth, and sorts them by value. Finally, the results are saved in the candidate neighbor table. If Sink3 leads the list, it will be selected as a neighbor and G returns Echo packets to Sink3. In this period, the mobile node G uploads data to the neighbor Sink3 in priority.
Figure 17 Example of route searching Then find nodes which are multi hops away from sink nodes successively. The 18 / 24
activated mobile nodes start the next round search to find mobile nodes or fixed nodes which are two hops away from sink nodes. Each searched node makes a weighted average by the SNR of one or more groups of received Hello packets signals, the residual energy of the signal transmitter and the depth of Z-axis, and then restores it to its candidate neighbor table after ordering then by value. The node with the biggest weighted average is selected as the neighbor and the Echo response packet is returned to it. As a result, the neighbor will become the destination node to upload data. Take the scene in Fig. 17 for example. Node C has received the Hello packets from F 、 E and G. Make a comprehensive consideration of their channel communication quality, residual energy and the Z-axis depth. Suppose that E, F and G are included in the candidate neighbor table sequentially and C sends a Hello packet to E. In this cycle time, node C will give a priority to neighbor E to upload data. Repeat the cycle until all the nodes within the range of optical communication are connected in the network.
6.2 Data Transmission Stage All activated mobile nodes and activated fixed nodes can transmit data to their neighbor nodes. Data will be transferred layer upon layer via neighbor nodes and eventually be transported to sink node. Then, the node goes to sleep and waits for next upload. The longest duration of the data transfer can be set according to different circumstance.
6.3 Process of Special Circumstance If a node fails to establish a connection with other nodes within T2 time, acoustic communication mode will be started up and only sensor data will be transmitted. If a node cannot communicate with its current neighbor nodes due to the changes of the network topology, the node can select other nodes stored in the candidate neighbor table by making an attempt one by one until finding a new neighbor that can be contacted. If all the nodes in the candidate neighbor table cannot been communicated with, then acoustic model will be started and only small volume of data collected by sensors will be sent.
7. Simulation and numerical results In this section, we present simulation and numerical results for the OA-UWSN. The underwater acoustic (UWA) channel model and the related parameters in [37] are adopted when the acoustic communication mode of OA-UWSN is employed. [37] has presented numerical results demonstrating how the capacity is affected by the system 19 / 24
and channel parameters (i.e. link distance, the number, location, and power-delay profile (PDP) of significant taps) as well as environmental parameters (i.e. temperature, salinity, pressure, and depth). The underwater wireless optical communication channel models and the model parameters in paper [38] are adopted when optical communication mode of OA-UWSN is employed, including low-speed optical communication, medium-speed optical communication and high-speed optical communication. We consider a carrier frequency of 30 kHz for acoustic communication and 20W green/blue LEDs combined directional transmission for optical communication. The photomultiplier with a bandwidth of 140MHz is placed at the optical communication receiver. We assume that the temperature is T = 6°C and 200Mb video data is collected in a fixed node. The data is transmitted from a fixed node to a sink node. Two kinds of UWSN data transfer mode are simulated. One is classical acoustic transmission mode and another one is the self-adaptive multi-hop OA-UWSN mode proposed in this paper. The distance of each hop is 30 meters in multi-hop OA-UWSN mode. The distance from a fixed node to a sink node is assumed to be 30m, 60m and 90m separately. The classical acoustic transmission mode is used to transmit 30m, 60m and 90m separately by each time. The multi-hop OA-UWSN mode is used to transmit 30m, 60m and 90m separately by single hop, dual hop and triple hop. Assume that the underwater channels which correspond to pure sea, clear ocean, coastal and turbid harbor mode also correspond to high-speed optical communication, medium-speed optical
communication,
low-speed
optical
communication
and
acoustic
communication mode separately. The duration time of each channel environment is set to be 4s and the average transmission rate can be obtained after 200Mb data transmission is completed through four kinds of channels in turns, such as pure sea, clear ocean, coastal and turbid harbor channel in our simulation. We repeated the simulation about 1000 times and got the average transmission rate shown in table 3.
Table 3 Average Transmission Rate of the OA-UWSN and UWA Transmission Distance
30m
60m
90m
2.87Mbps
1.42Mbps
360Kbps
178Kbps
0.96Mbps 86Kbps
Mode OA-UWSN UWA
20 / 24
8. Summary This paper proposed a solution of OA-UWSN which maximized the advantages of underwater optical and acoustic communication. High-speed transmission could be achieved within short communication distance by means of optical communication. Transmitting control commands and node localization were achieved via acoustic communication based on the advantages of long distance transmission. The noise of the OA-UWSN was effectively reduced with no interference between acoustic and optical signals. At the mean time less use of acoustic signals could lower the sound harm to marine lives.
ACKNOWLEDGEMENTS The authors would like to thank the referees and editors for providing very helpful comments and suggestions. This project was the National Natural Science Foundation of China (No. 61671261, No. 61304222), the Key Research and Development Program of Shandong Provice (No.2016GGX101007) and the Open Research Fund from Shandong provincial Key Laboratory of Computer Network (No. SDKLCN-2015-04).
Reference ［1］ [Online]http://www.noaa.gov/educational-resources ［2］ W. Wei, N. N. Chen, X. Zhang, J. H. Rao and W. B. Wang, Research and survey of wireless optical communication for Underwater Wireless Sensor Networks[J], Sensor world, 2011, 17(3): 6-12 ［3］ [Online]http://www.nrdc.org/ ［4］ G.D. Yao, Z.G. Jin, and Y.S. Su, An environment-friendly spectrum decision strategy for underwater acoustic networks[J], Journal of Network and Computer Applications, 2016, 73: 82-93 ［5］ [Online]http://www.physics.umd.edu/grt/taj/104a/104anotessupps.html,
T.
Jacobson,
“Phys104,” 2008 ［6］ M. Callaham, Submarine communications[J], IEEE Magazine, 1981, 19(11):16-25 ［7］ J. Puschell, R. Giannaris, and L.Stotts, The autonomous data optical relay experiment: first two way laser communication between an aircraft and submarine[C], Proc. National Telesystems Conference (NTC’92), Washington, DC, May. 1992, 1427-1430 ［8］ T. Wiener and S.Karp, The role of blue/green laser systems in strategic submarine communications[J], IEEE Translations on Communications, 1980, 28(9): 1602-1607 ［9］ I. Vasilescu, K. Kotay, D. Rus, M. Dunbabin, and P. Corke, Data Collection, Storage, 21 / 24
and Retrieval with an Underwater Sensor Network[C]. Proceedings of the 3rd International Conference on Embedded Networked Sensor Systems, 2005, 154-165 ［10］ W. C. Cox, A 1 Mbps Underwater Communication System Using a 405 nm Laser Diode and Photomultiplier tube [D]. North Carolina State University: M.S. Thesis, 2008 ［11］ J. A. Simpson. A 1 Mbps Underwater Communications System Using LEDs and Photodiodes with Signal Processing Capability [D]. North Carolina State University: M.S. Thesis, 2007 ［12］ J. A. Simpson, W. C. Cox, J. R. Krier, B. Cochenour, B. Hughes, and J. F. Muth, 5 Mbps Optical Wireless Communication with Error Correction Coding for Underwater Sensor Nodes[C]. IEEE Pro. OCEANS 2010, Seattle, WA, 2010, 1-4 ［13］ C. Pontbriand, N. Farr, J. Ware, J. Preisig, and H. Popenoe, Diffuse high-bandwidth optical communications, Proc. MTS/IEEE OCEANS (OCEANS’08), Quebec City, QC, Canada, Sep. 2008, 1-4 ［14］ N. Farr, J. Ware, C. Pontbriand, and M. Tivey, Demonstration of wireless data harvesting from a subsea node using a ship of opportunity[C], Proc. MTS/IEEE OCEANS (OCEANS’13), San Diego, CA, Sep. 2013, 1-5 ［15］ C.P. Li, K. H. Park, and M. S. Alouini, On the Use of a Direct Radiative Transfer Equation Solver for Path Loss Calculation in Underwater Optical Wireless Channels [J], IEEE Wireless Communication Letters, 2015, 4(5):561-564 ［16］ H. M. Oubei, C. P. Li, K. H. Park, T. K. Ng, M. S. Alouini, and B. S. Ooi, 2.3 Gbit/s underwater wireless optical communications using directly modulated 520 nm laser diode, Optics express, 2015, 23(16): 20743-20748 ［17］ Y. Dong, S. Tang, X. Zhang, Effect of Random Sea Surface on Downlink Underwater Wireless Optical Communications [J], IEEE Journal on Selected Areas in Communications, 2013, 17(11): 2164-2167 ［18］ H. Zhang, and Y. Dong, Impulse response modeling for general underwater wireless optical MIMO links[J], IEEE Communications Magazine, 2016, 54(2): 56-61 ［19］ H. Zhang, and Y. Dong, General Stochastic Channel Model and Performance Evaluation for
Underwater
Wireless
Optical
Links[J],
IEEE
Transactions
on
Wireless
Communications, 2016, 15(2): 1162-1173 ［20］ Z. U. Ahmad, Underwater optical wireless sensor network [D], University of Warwick， 2013 ［21］
Naval
Postgraduate
School
USA,
http://www.nps.edu/About/Nev>s/NPS-Pioneers-Seaweb-Underwaler-Sensor-Networks.h tm 22 / 24
［22］ M. Grund, L. Freitag, and J. Preisig, The PLUSNet underwater communications system: acoustic telemetry for undersea surveillance[C], Proceedings of IEEE OCEANS, 2006, 1-5 ［23］ United States Department of Defense(DOD), USA, http://www.defense.gov/contracts /contract.aspx?contractid=5409 ［24］ J. Kasahara, Y. Shirasaki, and K. Asakawa, Scientific application of ARENA networks[C], Proceedings of International Workshop on SUSCRT, 2003, 272-275 ［25］ A.P. Das, and S.M. Thampi, Fault-resilient localization for underwater sensor networks[J], Ad Hoc Networks, 2016, 55(2): 132–142 ［26］ Z. Zeng, H. Zhang, Y. Dong, and J. Cheng, A Survey of Underwater Wireless Optical Communication,
IEEE
Communications
Surveys
&
Tutorials,
2016,
DOI:
10.1109/COMST.2016.2618841 ［27］ N. Nusrat, K. Gour, K. Joarder, PRADD: Apathreliability-aware data delivery protocol for underwater acoustic sensor networks[J], Journal of Network and Computer Applications, 2016, 75:385-397 ［28］ N.Z. Zenia, M. Aseeri, M.R. Ahmed, Z.I. Chowdhury, and M.S. Kaiser, Energy-efficiency and reliability in MAC and routing protocols for underwater wireless sensor network: A survey[J], Journal of Network and Computer Applications, 2016, 71: 72-85 ［29］ X. Guo, F. X. Ge, S. L. Ren, and L. H. Guo, Eigenanalysis-based adaptive beamforming using worst-case performance optimization[J], Acta Acustica, 2015, 40(2):187-197 ［30］ J. L. Wang, Y. H. Liu, Y. He, W. Dong, and M. Li, QoF: Towards Comprehensive Path Quality Measurement in Wireless Sensor Networks[J], IEEE Transactions on Parallel and Distributed Systems, 2014, 25(4): 1003-1013 ［31］ Z. Liao, D. Li, and J. Chen, Joint Bandwidth Optimization and Media Access Control for Multihop Underwater Acoustic Sensor Networks[J], IEEE Sensors Journal, 2015, 15(8): 4292-4304 ［32］ L. F. Liu, N. S. Zhang, Y. Liu, Topology control models and solutions for signal irregularity in mobile underwater wireless sensor networks[J], Journal of Network and Computer Applications, 2015, 51: 68-90 ［33］ F. Qu, Z. Wang, L. Yang, and Z. Wu, A journey toward modeling and resolving doppler in underwater acoustic communications[J], IEEE Communications Magazine, 2016, 54(2): 49-55 ［34］ S.F. Yan, Optimal design of modal beamformers for circular arrays[J], The Journal of the Acoustical Society of America, 2015, 138(4):2140-2151 ［35］ J. F. Jiang, G. J. Han, H. Guo, L. Shu, J. J. Rodrigues, Geographic multipath routing 23 / 24
based on geospatial division in duty-cycled underwater wireless sensor networks[J], Journal of Network and Computer Applications, 2016, 59: 4-13 ［36］ N. Farr, A. Bowen, J. Ware, C. Pontbriand, and M. Tivey, An integrated, underwater optical/acoustic communications system[C], Proc. IEEE/MTS OCEANS, Sydney, NSW, Australia, May 2010, pp. 1–6. ［37］ H. Nouri, M. Uysal, E. Panayirce, and H. Senol, Information theoretical performance limits of single-carrier underwater acoustic systems[J], IET Communications, 2014, 8(15): 2599-2610 ［38］ C. Gabriel, M. A. Khalighi, S. Bourennane, and P. Leon, Monte-Carlo-based channel characterization for underwater optical communication systems [J]. IEEE/OSA Journal of Optical Communications & Networking, 2013, 5(1):1-12.
24 / 24