Underwater Acoustic Network Multicarrier Communication And Cross-layer Design

Underwater acoustic networks(UANs)is an effective way to explore the oceanic resources and obtain the environment parameters.Underwater acoustic channels have the characteristics of limited bandwidth,low propagation speed,severe attenuation and fast time variation.Also the underwater equipment is powered by battery which is hard to exchange which limits the performance of the UANs.How to increase the throughput and extend the life of the networks is an urgent problem needed to solve.Lots of mature techniques of terrestrial networks can't be directly used in UANs due to the unique and complex characteristic of the underwater acoustic channels.The strong multipath interference and the complex noise lead to a high bit error rate(BER)of physical layer which consequently leading to a low throughput of the underwater networks.So how to realize robust and high data rate data transmission is an important factor to increase the throughput.In addition,the underwater sound speed is very low(1500ms/s)which leads to high data transmission latency.It lowers the channel utilization severely.How to design high efficiency physical layer data transmission and MAC protocol are two important aspects to increase the throughput of the network.The traditional layered networks didn't consider the influences of the channel and the mutual influence of each layer.So the network can't fully utilize the channel resource and optimize the resource allocation.Based on the traditonal layered network structure,the cross-layer design regards several or all layers as a whole to optimize and control the performance.It sufficiently considers the interference and strong correlation of each layer which could effectively utilize the channel resource and optimize and promote the network performance.The UAN is normally divided into five layers including application layer,transport layer,network layer,data link layer and physical layer.The bottom layer protocols and the communication techniques are the important factors to effect the network's performance.The paper focuses on the physical layer(PHY)and the data link layer(mainly MAC layer)design and try to improve the throughput through PHY/MAC cross-layer design.Physical layer adopts high data rate OFDM modulation.Underwater acoustic channel is a main factor to influence the performances of the physical and MAC layer.Familiar with the characteristic of underwater acoustic channel and model the channel could effective instruct the system simulation and the performance analysis.So,firstly,the underwater acoustic channel modeling and real channel analysis were researched in the paper.Physical layer,as the bottom layer of the network,is responsible for data transmitting and receiving.Its performance has great influence to the network.So how to design high speed and robust physical layer communication is an important research aspect in the paper.MAC layer mainly controls the access of the nodes to the wireless channel which is also an important factor to network's performance.How to efficiently utilize the channel resource for long propagation delay,hence to increase the throughput is another research interest.Based on the robust physical layer data transmission and high efficiency MAC protocol,through PHY/MAC cross-layer design to optimize and utilize the channel resource,the UANs performance could be greatly enhanced.Underwater acoustic channel is an important factor to affect underwater acoustic communication and network performance.To fully know about the underwater channel characteristic is helpful to receiver design and top level resources allocation.In the paper,the characteristics of shallow underwater channels of different areas were analyzed.The amplitudes,delays,Doppler and the time correlation et.al.were analyzed in time and frequency domain which provides real data support for underwater acoustic channel modeling.A time-varying underwater channel modeling method was proposed.The channel fading is divided into large-scale variation and small-scale variation.The random variations caused by scattering are considered in the model.The cause of Doppler shift is modeled as stable motion and random motion which is more closely to the real underwater channel.Aim to solve the problem of OFDM inter-symbol interference caused by strong multipath interference,a time reversal signal processing based on the sparse channel estimation was proposed.According to the sparsity of underwater channel,two sparse channel estimation methods were proposed:matching pursuit(MP)and basis de-noising pursuit(BPDN).Compare to traditional channel estimation methods,sparse channel estimation has higher estimation accuracy.Time reversal mirror(TRM)could realize the compression in the time domain.Apply TRM technique to OFDM demodulation could compress the channel,reduce the ISI and pilot subcarriers so as to increase the data rate.In order to reduce the non-minimal phase interference caused by time reversal processing,cyclic postfix is needed to be added in the OFDM symbol.In order to remove the interference caused by the probe signal in passive time reversal processing,virtual time reversal processing was proposed.The sparse channel estimation is used in the virtual time reversal processing to enhance the processing gain.The control frame of underwater acoustic sensor networks is usually transmitted in a low data rate which occupies the channel for a long time.Orthogonal multicarrier M-ary cyclic shift keying(CSK)spread spectrum technique was proposed to increase the data rate for control frame transmission.The data is firstly modulated by M-ary and CSK spread spectrum and then modulated on multiple parallel subcarriers.A composite sequence was used to suppress the inter-carrier interference to improve the system performance.This modulation algorithm has the advantage of spread spectrum such as resisting multipath interference and attenuation and also has a higher channel utilization.It could effectively shorten the length of control frame and subsequently increase the transmission efficiency.Aim at solving the low throughput of the UANs,a PHY/MAC cross-layer design is proposed.Through exchanging and sharing the information between two layers,the PHY and MAC layer could adaptively adjust the modulation parameters and the resource allocation scheme which could effectively utilize the channel resource so as to increase the throughput of the network as well as to decrease the average end-to-end delay.As the requirement of cross-layer design,an adaptive OFDM modulation and coding scheme and improved CSMA/CA protocol were proposed.Adaptive modulation and coding scheme adaptively adjusts the modulation level,coding rate and frequency diversity order to adapt to different channels.The receiving node feeds back the channel state information to the transmitting node which is used to choose the appropriate modulation parameters.Improved CSMA/CA detects the channel state by overhearing the data frame using virtual carrier sensing instead of handshaking which reduces the overhead and both increases the throughput and decreases the data transmission delay.