CN104602313A - Environment-adapted routing method of wireless sensor network for maritime search and rescue - Google Patents

Abstract

The invention relates to an environment-adapted routing method of wireless sensor network of maritime search and rescue. The method is characterized in that the wireless sensor network formed by self-networking through a plurality of nodes includes a plurality of clusters formed by the nodes. The method comprises the steps of calculating the threshold of each node according to the node energy, the energy spacing and the channel quality during the cluster head election of each cluster; comparing the thresholds; selecting one node as the cluster head node by cycling; building route measurement adapted to the mobility of the nodes; selecting a reliable route chain according to the route measurement; creating a priority transfer node list; selecting the corresponding relay node through the cluster head node according to the priority in the transfer node list so as to transfer information to the outside. According to the method, the route measurement adapted to the environment can be created according to the environment change of the wireless sensor network; the characteristics of secondary clusters and opportunism routes are effectively utilized to adapt to the high-dynamics network, and therefore, the data transmission efficiency can be increased.

Description

A kind of environment self-adaption method for routing of maritime search and rescue wireless sense network

Technical field

The present invention relates to the wireless search and rescue technology in a kind of Marine case field and sensor network technology field, be specifically related to a kind of maritime search and rescue wireless sense network and environment self-adaption method for routing thereof.

Background technology

The present main search pattern of maritime search and rescue is all that the passive wait of searching target is searched and rescued, and such search and rescue mode cost is high, and efficiency is low, and searches and rescues target and can not be supplied to search and rescue side's effective status information.Wireless sensor network (Wireless Sensor Networks, WSN) technology is a kind of universal rapidly communication technology, sensor node self-organizing can form a multihop network, they can be collected the relevant information of search and rescue target in time and send to search and rescue side, greatly improve the search efficiency of search and rescue side.

Because WSN technology can be brought very large convenient to Jobs on the sea, therefore also start to popularize gradually in the application of maritime field, as marine oil overflow detection, maritime affairs monitoring etc.The advantage of WSN is low cost, low energy consumption, and volume is little, and can organize themselves into a communication network to collect useful information, is very suitable for being applied in maritime search and rescue field.

In the middle of at sea searching and rescuing during application WSN technology, the major issue solved is exactly ensure router efficiency as far as possible, router efficiency directly determines data transmission efficiency, namely under the prerequisite extending network lifecycle, as far as possible the effective information of all search and rescue targets can be collected and be sent to aggregation node, and be promptly sent search and rescue side.

The routing issue of WSN is not because of excessive power consumption under the prerequisite of death at guarantee node, select a less energy consumption, apart from short optimal path, make route tackle the uncontrollable change of network and can still have good robustness, and ensure that network life cycle maximizes, and network has higher communication efficiency.The research majority of present stage is static wireless sense network, and as forest fire protection, building monitoring, terrestrial reference detects etc., and in such applied environment, network is more stable, and sensor node does not have change in topology substantially, is therefore convenient to use, and network efficiency is higher.

Existing method for routing mainly divides two kinds, one is common prerequisite method for routing, namely the next-hop node of sensor selection problem is known, and because this route is used widely in every field, therefore most of route is all prerequisite routes, but at sea searching and rescuing, if still adopt prerequisite method for routing, due to the node dynamic topology change that marine particular surroundings causes, link-quality is poor, data transmission efficiency then can be caused lower, and node is dead because of constantly repeating transmitting energy to exhaust.Also having a kind of is the chance method for routing in recent years proposed, and this method for routing selects route after being, it can carry out data retransmission from the effective next-hop node of main separation, can improve forward efficiency.But the method extensibility is not strong, and is more suitable for small-scale wireless sensor network.Maritime search and rescue wireless sense network environment is poor, and node dynamic is strong, and node is not replaceable, therefore needs a kind of new method for routing ensure transfer rate and adapt to dynamic environment.

Summary of the invention

The invention provides a kind of adaptive routing method of maritime search and rescue wireless sense network environment, solve the problem that maritime search and rescue wireless sense network interior joint router efficiency is low, can dynamic be adapted to higher, the network of link-quality instability, and reduce unnecessary energy ezpenditure.

In order to achieve the above object, technical scheme of the present invention is to provide a kind of adaptive routing method of maritime search and rescue wireless sense network environment:

In the wireless sense network formed by MANET by multiple node, comprise several bunches formed by these nodes at sea;

In each bunch of election of cluster head carried out, carry out the threshold value of each node in compute cluster based on node energy, internodal distance and channel quality and carry out threshold value and compare, selecting one of them as leader cluster node using round robin;

Set up the route adapted with the mobility of node to estimate, estimate according to this route and select reliable routing link, and the forward node list of setting up with priority, so that leader cluster node selects corresponding via node to come to a bunch outer forwarding information according to the priority in forward node list.

Preferably, elected leader cluster node is broadcasted one and is greeted packet, represents the threshold value of this leader cluster node with the threshold value of described greeting packet; Namely the threshold value also formation bunch of this greeting packet is preserved after neighbor node within the scope of this leader cluster node efficient communication receives; Further, when judging that the threshold value of this neighbor node itself is less than the threshold value of greeting packet, formation bunch again.

Preferably, by arranging timer to control a time interval, election of cluster head is re-started when this time interval arrives.

Preferably, leader cluster node sends packets of information as source node, in this packets of information containing this leader cluster node from bunch in the information that sends of the needs collected of each Nodes, the some of them node in bunch forwards this packets of information as some grades of via nodes.

Preferably, leader cluster node is broadcasted one and is greeted packet, and when listening to the node receiving this greeting packet and replying to the ACK that this leader cluster node feeds back according to priority, sends packets of information by the via node selected.

The random number that preferably, during election of cluster head, the node in bunch is arranged separately between (0,1), and random number is compared with threshold value T (n), the node making random number be less than threshold value T (n) becomes leader cluster node in epicycle, wherein:

$T\left(n\right)=\frac{\mathrm{\ρP}}{k-P\left[r\mathrm{mod}(k/P)\right]}$

Energy factor ρ:

$\mathrm{\ρ}=\frac{100}{L_{\mathrm{current}}{\∫}_0^{t}P\left(\mathrm{\τ}\right)\mathrm{d\τ}}[\left(r_s\mathrm{div}\frac{1}{p}\right)(1-\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }})+\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }}]$

L _currentfor present node is to leader cluster node) distance, P (τ) is moment τ channel loss, for the channel loss in time period t;

P is the probability that node becomes leader cluster node, and r is current election wheel number;

The label n ∈ G of node _round, G _roundfor not being elected to the node set of leader cluster node in the nearest wheel of 1/P continuously;

E _{n_current}for node present energy, E _{n_max}for the primary power of node;

K is bunch head number expected; r _sit is the round that node is not elected as leader cluster node continuously.

Preferably, the route of foundation estimate for:

$\begin{array}{c}{E}^{\′}\left(N_k\left(r\right)\right)=E\left(N_k\left(r\right)\right)+M\left(N_k\left(r\right)\right)\\ =\frac{{\mathrm{\Σ}}_{k=1}^r\left[d_{N_k}{p}_{N_k}{\mathrm{\Π}}_{n=0}^{k-1}\stackrel{\‾}{p_{N_n}}\right]}{E_{\mathrm{tx}}+r{E}_{\mathrm{rx}}}+\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C\end{array}$

Wherein, N _kr () is potential via node i ₁, i ₂... i _j, set, the priority of these potential via nodes is: i ₁>i ₂> ... >i _j; that the node that priority is higher sends failed probability;

E _tx, E _rxbe respectively node send message time and receipt message time energy consumption, r is the quantity of optional node in potential via node; N is the number of node;

The mobile cost that all nodes are total:

$M\left(N_k\left(r\right)\right)=\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C---\left(12\right)$

C is constant, it is the mobile cost of each node;

Wherein any one node j in the state of moment τ is:

$s\left(\mathrm{\τ}\right)={[x\left(\mathrm{\τ}\right),\stackrel{\·}{x}\left(\mathrm{\τ}\right),\stackrel{\·\·}{x}\left(\mathrm{\τ}\right),y\left(\mathrm{\τ}\right),\stackrel{\·}{y}\left(\mathrm{\τ}\right),\stackrel{\·\·}{y}\left(\mathrm{\τ}\right)]}^T---\left(9\right)$

S (τ) is the position of node j in cartesian coordinate system, the speed along x-axis and acceleration respectively; the speed along y-axis and acceleration respectively;

After time t, the position that this node j estimates at moment τ+t is:

$\hat{s}(\mathrm{\τ}+t)={[\hat{x}(\mathrm{\τ}+t),\widehat{\stackrel{\·}{x}}(\mathrm{\τ}+t),\widehat{\stackrel{\·\·}{x}}(\mathrm{\τ}+t),\hat{y}(\mathrm{\τ}+t),\widehat{\stackrel{\·}{y}}(\mathrm{\τ}+t),\widehat{\stackrel{\·\·}{y}}(\mathrm{\τ}+t)]}^T.$

The invention discloses a kind of adaptive routing method of maritime search and rescue wireless sense network environment, calculate threshold value by considering of energy distance etc., comparison node threshold value selects leader cluster node; Set up the new adaptability route being applicable to the environment of maritime search and rescue to estimate; Set up new forward node list; Priority according to forwarding list carries out nodes coordinating forwarding data.The route that the present invention can set up adaptive environment according to the environmental change of wireless sense network is estimated; Effectively utilize sub-clustering and opportunism route feature separately, the network that dynamic is higher can be adapted to, increase data transmission efficiency.

Maritime search and rescue wireless sense network of the present invention and environment self-adaption method for routing thereof are compared to the prior art, its advantage is, by introducing cluster-based techniques in chance route, can ensure that node can not be dead because of too much participation forward node list, can ensure that again source node finds suitable down hop path, improve router efficiency.

Accompanying drawing explanation

Fig. 1 is the configuration diagram of maritime search and rescue wireless sense network;

Fig. 2 is the structural representation of chance method for routing in the present invention;

Fig. 3 be the present invention compared with existing routing protocol, the schematic diagram of different jumping figure lower network throughput;

Fig. 4 be the present invention compared with existing routing protocol, along with node jumping figure increases the schematic diagram of data retransmission rate;

Fig. 5 be the present invention compared with existing routing protocol, along with several increase of giving out a contract for a project, the schematic diagram of the average energy consumption of network.

Embodiment

The present invention is applicable to maritime search and rescue wireless sense network, and as shown in Figure 1, this wireless sense network comprises: search and rescue terminal, be arranged at boats and ships, established a communications link by external satellite and the Internet; Aggregation node, is arranged at lifeboat raft etc., and this aggregation node is connected with search and rescue terminal (boats and ships) wireless telecommunications; Sensor node is arranged at life vest etc., and this sensor node real-time detection aggregation node is also connected with aggregation node wireless telecommunications.

Concrete, life vest or goods are equipped with wireless sensor node with ship, when wireless sense network is disposed, node primary power is identical, after overboard, node is met water and is opened, and self-organizing forms a multihop network, due to goods or persons falling in water due to marine stormy waves etc. impact entrained by sensor node may be relatively unstable; Relatively stable is lash ship, search and rescue boats and ships and search and rescue helicopter etc., thus in order to ensure that searching and rescuing target is arrived by the network coverage as far as possible, other sensor nodes of manually shedding search and rescue side will be added and front a kind of sensor node comes together to form wireless sense network.

After wireless sense network is formed, sensor node collects the information of self goods or life entity, as vital sign, and geographical position etc., then by its information transmission to aggregation node, effective information is transferred to search and rescue terminal by aggregation node again.

The invention provides a kind of environment self-adaption method for routing being applicable to above-mentioned maritime search and rescue wireless sense network, be characterized in, the method is based on opportunism method for routing, sensor node in wireless sense network is carried out Uneven Cluster, adopt new route to estimate and set up forward node list, and produce forward node list priority, optimization data repeating process.

The process of the method, comprises:

Step 1, election of cluster head: by each node initializing, by compare threshold T (n), select leader cluster node at random to take turns the mode turned.

Step 2, selection via node: in order to reliable link can be selected, consider the mobility of node, set up new route and estimate.

Step 3, set up new repeating process: the priority listing node, carry out nodes coordinating data retransmission according to priority.

Below further illustrate specific embodiments of the invention:

election of cluster head

Example of the present invention is when wireless sense network is disposed, and node primary power is identical (this node can be the sensor node that goods or persons falling in water carry, and also can be shed afloat sensor node).LEACH (low energy consumption self adaptation clustering routing mechanism) is also most typical clustering route protocol the earliest, and agreement thought takes turns the mode turned to select leader cluster node at random, reduces network energy consumption and extend network lifetime.LEACH elects node when bunch head at random, ensures that energy is evenly distributed on the node in network.Random number between sensor node n stochastic generation one (0,1), and it is compared with threshold function table T (n); If the random number of node is less than threshold value, then this node becomes a bunch head in epicycle, and it is called a bunch head to other node broadcasts in effective communication distance.

Wherein:

$T\left(n\right)=\frac{P}{k-P\left[\mathrm{round}\mathrm{mod}(1/P)\right]},n\∈G_{\mathrm{round}}---\left(1\right)$

P is the probability becoming bunch head, the continuous round that round has been, G _roundfor not being elected to the node set of bunch head in nearest continuous 1/P wheel.

But in order to balance energy consumption, this method must based on two kinds of hypothesis: (1) all nodes primary power is equal; (2) consume energy when any node serves as bunch head equal.But, for the first time after cluster head random-selection, the first round starts sub-clustering and sub-clustering afterwards, bunch size, the energy that each residue energy of node and transmission data will consume, also have node all to start to occur change to the distance of base station, therefore, hypotheses (2) does not meet actual conditions.

For LEACH agreement, DCHS (deterministic cluster-head selection, accurate bunch of head selection algorithm) consider capacity factor, to make present energy and the low node of primary power ratio preferentially elected bunch head, the T (n) after being improved:

$T\left(n\right)=\frac{P}{k-P\left[r\mathrm{mod}(k/P)\right]}[\left(r_s\mathrm{div}\frac{1}{p}\right)(1-\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }})+\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }}]---\left(2\right)$

E _{n_current}represent node present energy, E _{n_max}represent the primary power of node.K represents bunch head number of expectation.R _sit is the round that node is not elected as leader cluster node continuously.R is current election wheel number.

In the method, we redefine a new parameter: energy factor ρ, so

$\mathrm{\ρ}=[\left(r_s\mathrm{div}\frac{1}{p}\right)(1-\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }})+\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }}]---\left(3\right)$

We redefine to consider the channel quality of adverse circumstances and unsteadiness, L _currentfor present node is to the distance of aggregation node (leader cluster node), P (τ) is moment τ channel loss, so be the channel loss in time period t.So the new threshold value of definition is

$\mathrm{\ρ}=\frac{100}{L_{\mathrm{current}}{\∫}_0^{t}P\left(\mathrm{\τ}\right)\mathrm{d\τ}}[\left(r_s\mathrm{div}\frac{1}{p}\right)(1-\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }})+\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }}]---\left(4\right)$

Combine with (2), be:

$T\left(n\right)=\frac{\mathrm{\ρP}}{k-P\left[r\mathrm{mod}(k/P)\right]}---\left(5\right)$

Wherein ρ can obtain from equation (4).The random number that all nodes produce all can not be less than 0, otherwise cannot choose a bunch head.

cluster

Once a bunch head is determined, just broadcast " hello " packet is to find neighbor node for bunch head, and the maximum of T _ max=1 of we setting threshold T, neighbor node continues to monitor.If there is node to have received " hello " packet, namely preserve the threshold value (this threshold value represents the threshold value of bunch head) of " hello " packet.If the threshold value receiving the node of packet is less than the threshold value of " hello " packet, bunch again to be formed.And the node at set intervals in network can reset and form new bunch.

Energy balane adopts with drag: node a sends the data of k byte to another node b outside distance d, and energy consumption calculation mode is as follows:

$E_{T_r(k,d)}=E_{\mathrm{elec}\left(k\right)}+E_{\mathrm{amp}(k,d)}=\left\{\begin{array}{cc}k{E}_{\mathrm{elec}}+k{\mathrm{\ϵ}}_{\mathrm{fs}}{d}^2& d\≤d_0\\ k{E}_{\mathrm{elec}}+k{\mathrm{\ϵ}}_{\mathrm{mp}}{d}^4& d>d_0\end{array}\right.---\left(6\right)$

Wherein, E _{elec (k)}that the transtation mission circuit energy that sends k byte data and consume is (by expression emitter (transmit electronics, TE) and receiving circuit (receive electronics, RE) often sends or the power consumption of recruiting unit bit calculates); E _{amp (k, d)}it is the energy (energy that every bit unit of transfer square metre consumes being calculated by expression emission amplifier) that emission amplifier consumes; E _elecbe the energy that transtation mission circuit sends data consumes, depend on the factors such as digital code, filtering module and signal diffusion; ε _fsit is the specific energy consumption value of free space model emission amplifier; ε _mpit is the specific energy consumption value of multipath fading model emission amplifier; ε _fsd ²and ε _mpd ⁴parameter value is determined by the value of the data bit error rate in the distance and communication process of receiver; d ₀it is the distance value of a setting.

And node b is when receiving the message from node a, the energy consumption that its radio receiver produces is:

E _Rx(k)＝kE _elec(7)

The false code of algorithm 1 is as follows:

select via node and repeating process

The environment self-adaption method for routing of a kind of maritime search and rescue wireless sense network disclosed by the invention, the method, based on opportunism method for routing, is optimized it for the mobility of node and node scale.Chance method for routing fundamental mode is as follows, and if Fig. 2 is a directed graph G={N, L}, N are WSN interior joint number, and L is link.All nodes form a set.S is source node, and abcd is intermediate node, node for the purpose of D, and node abcd is in the effective communication distance of S, and these nodes are the potential via node of source node S.When S has data to send time, set up the list of forward node collection after the node abcd in its effective communication distance have received packet and according to priority sequentially send ACK response to the node S sending data.

Owing to considering energy, the factors such as distance, have scholar to propose a criterion:

$E\left(N_k\left(r\right)\right)=\frac{{\mathrm{\Σ}}_{k=1}^r\left[d_{N_k}{p}_{N_k}{\mathrm{\Π}}_{n=0}^{k-1}\stackrel{\‾}{p_{N_n}}\right]}{E_{\mathrm{tx}}+r{E}_{\mathrm{rx}}}---\left(8\right)$

Wherein, molecule is EPA (Expected Packet Advancement), N _kr () is potential via node i ₁, i ₂... i _j, d _nkset, the priority of these nodes is: i ₁>i ₂> ... >i _j.Wherein that the node that priority is higher sends failed probability, in denominator, E _tx, E _rxbe respectively the energy consumption of node when sending and receive, r is the quantity of optional node in potential via node.

Different from other applied environments, the main task of maritime search and rescue is that information is sent to search and rescue person, and a lot of research is all focus on to reduce energy consumption to ensure network lifetime.Herein we propose one improve route estimate, in order to select reliable link, the mobility of node is taken into account by we.In the state of certain moment τ, we defined node j be:

$s\left(\mathrm{\τ}\right)={[x\left(\mathrm{\τ}\right),\stackrel{\·}{x}\left(\mathrm{\τ}\right),\stackrel{\·\·}{x}\left(\mathrm{\τ}\right),y\left(\mathrm{\τ}\right),\stackrel{\·}{y}\left(\mathrm{\τ}\right),\stackrel{\·\·}{y}\left(\mathrm{\τ}\right)]}^T---\left(9\right)$

Here, s (τ) is the position of node j in cartesian coordinate system, the speed along x-axis and acceleration respectively; the speed along y-axis and acceleration respectively.After time t, we can obtain the position that node j estimates at moment τ+t:

$\hat{s}(\mathrm{\τ}+t)={[\hat{x}(\mathrm{\τ}+t),\widehat{\stackrel{\·}{x}}(\mathrm{\τ}+t),\widehat{\stackrel{\·\·}{x}}(\mathrm{\τ}+t),\hat{y}(\mathrm{\τ}+t),\widehat{\stackrel{\·}{y}}(\mathrm{\τ}+t),\widehat{\stackrel{\·\·}{y}}(\mathrm{\τ}+t)]}^T.$

The mobile cost that we define each node is:

$m=\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C---\left(11\right)$

Note M is total mobile cost, so

$M\left(N_k\left(r\right)\right)=\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C---\left(12\right)$

C is constant, it is the mobile cost of each node.Estimate so new for:

$\begin{array}{c}{E}^{\′}\left(N_k\left(r\right)\right)=E\left(N_k\left(r\right)\right)+M\left(N_k\left(r\right)\right)\\ =\frac{{\mathrm{\Σ}}_{k=1}^r\left[d_{N_k}{p}_{N_k}{\mathrm{\Π}}_{n=0}^{k-1}\stackrel{\‾}{p_{N_n}}\right]}{E_{\mathrm{tx}}+r{E}_{\mathrm{rx}}}+\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C\end{array}$

The present invention is the self adaptation retransmission method being suitable for maritime environment.In traditional routing, if certain nodes listen have sent ACK response to even higher level of node, it can abandon the packet that it receives.If source node does not receive ACK response within a period of time, it is retransmission data packet may then.For example of the present invention, when certain node has data to need to send, it can calculate oneself corresponding threshold value, then lists the forward node list with priority.But in the presence of a harsh environment, if once have node will send information just set up node listing, then can cause congested, some node also may be caused to participate in list or forwarding and depleted of energy frequently simultaneously.So when there being node will send information, message is sent to a bunch head by it, bunch head to collect bunch in member node information after formed a packets of information (bundle), outside re-sending to bunch.Then leader cluster node sends this packets of information (bundle) as source node.Consider the mobility of node, if we are late for renewal bunch, some members can due to mobile disengaging bunch, and therefore arrange a TCP timer here, re-starts election of cluster head when envoy presses by it.

The false code of algorithm 2 source node is as follows:

The false code of algorithm 3 both candidate nodes is as follows:

The process efficiently utilizes the advantage of opportunism route, the forwarding data of network node maximal efficiency can be made, and avoid part of nodes and excessively participate in list and forwarding and dead.Therefore this method has better adaptability to maritime environment, compares common chance route more effective.

Adaptive routing method below by the problem model set up the present invention and invention carries out sunykatuib analysis, introduces execution mode and the advantage in actual applications thereof of example of the present invention further.This sunykatuib analysis can adopt emulation platform to carry out.

This method utilizes NS2 platform to carry out emulation platform to carry out Computer Simulation.Optimum configurations is as follows: in the square shaped sensor region of 2000m*2000m, and random dispersion 100 nodes.The bandwidth of each node is 1MB.Emulation adopts this research Routing Protocol (showing in the drawings for EA-COR), E-adaptive and ECBR-MWSN.

Get in energy model emulation experiment: d ₀=75m; ε _fs=13pJ/bit/m ²; ε _mp=0.0013pJ/bit/m ⁴; E _fusion=5nJ/bit; E _elec=50nJ/bit;

Bunch radius r=20m, data package size is 500byte, and broadcast data packet size is 25byte, and node primary power is 0.5J, often takes turns Data Collection frequency n=10.

Three kinds of agreements are contrasted from following three aspects:

(1) throughput.

(2) packet forward rate, namely routing protocol data is surrounded by effect efficiency of transmission.

(3) efficiency.

Fig. 3 is that the network throughput under different jumping figure compares.Number of always giving out a contract for a project is 200, and a packets of information (bundle) of a bunch of head is inner 20 bags, as can be seen from Figure 3, along with the increase of jumping figure, the bag number be forwarded also increases, and network throughput will reduce, but when we compare same number of hops, this method is than high both other.Because this method simplifies the repeating process of existing chance route, can more effective management node.

As can be seen from Figure 4 the forward rate of ECBR-MWSN and this method are all increase in whole process, but possibly cannot ensure due to the channel quality selected, and ECBR-MWSN is all the time lower than this method.Time node number is 40-120, E-adaptive and this method due to can be autonomous afterwards best next-hop node, so all higher than ECBR-MWSN.Along with the increase of number of nodes, network size expands, and effective transmission rate reduces, and causes the efficiency of E-adaptive entirety to reduce.

From Fig. 5, we can see, in energy consumption, this method performance is best.The forwarding strategy of this method take into account energy consumption and Link State simultaneously, and by contrast, ECBR-MWSN can not successfully manage poor link-quality, result in forward rate low, and E-adaptive cannot effective management node, when node popularization, more energy will be consumed.

In sum, the route that the present invention can set up adaptive environment according to the environmental change of wireless sense network is estimated; Effectively utilize sub-clustering and opportunism route feature separately, the network that dynamic is higher can be adapted to, increase data transmission efficiency.

Although content of the present invention has done detailed introduction by above preferred embodiment, will be appreciated that above-mentioned description should not be considered to limitation of the present invention.After those skilled in the art have read foregoing, for multiple amendment of the present invention and substitute will be all apparent.Therefore, protection scope of the present invention should be limited to the appended claims.

Claims (7)

1. an adaptive routing method for maritime search and rescue wireless sense network environment, is characterized in that,

In the wireless sense network formed by MANET by multiple node, comprise several bunches formed by these nodes at sea;

In each bunch of election of cluster head carried out, carry out the threshold value of each node in compute cluster based on node energy, internodal distance and channel quality and carry out threshold value and compare, selecting one of them as leader cluster node using round robin;

Set up the route adapted with the mobility of node to estimate, estimate according to this route and select reliable routing link, and the forward node list of setting up with priority, so that leader cluster node selects corresponding via node to come to a bunch outer forwarding information according to the priority in forward node list.

2. adaptive routing method as claimed in claim 1, is characterized in that,

Elected leader cluster node is broadcasted one and is greeted packet, represents the threshold value of this leader cluster node with the threshold value of described greeting packet; Namely the threshold value also formation bunch of this greeting packet is preserved after neighbor node within the scope of this leader cluster node efficient communication receives; Further, when judging that the threshold value of this neighbor node itself is less than the threshold value of greeting packet, formation bunch again.

3. adaptive routing method as claimed in claim 1, is characterized in that,

By arranging timer to control a time interval, re-start election of cluster head when this time interval arrives.

4. the adaptive routing method as described in claim 1 or 3, is characterized in that,

Leader cluster node sends packets of information as source node, in this packets of information containing this leader cluster node from bunch in the information that sends of the needs collected of each Nodes, the some of them node in bunch forwards this packets of information as some grades of via nodes.

5. adaptive routing method as claimed in claim 4, is characterized in that,

Leader cluster node is broadcasted one and is greeted packet, and when listening to the node receiving this greeting packet and replying to the ACK that this leader cluster node feeds back according to priority, sends packets of information by the via node selected.

6. adaptive routing method as claimed in claim 1, is characterized in that,

The random number that during election of cluster head, the node in bunch is arranged separately between (0,1), and random number is compared with threshold value T (n), the node making random number be less than threshold value T (n) becomes leader cluster node in epicycle, wherein:

$T\left(n\right)=\frac{\mathrm{\ρP}}{k-P\left[r\mathrm{mod}(k/P)\right]}$

Energy factor ρ:

$\mathrm{\ρ}=\frac{100}{L_{\mathrm{current}}{\∫}_0^{t}P\left(\mathrm{\τ}\right)\mathrm{d\τ}}[\left(r_s\mathrm{div}\frac{1}{p}\right)(1-\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }})+\frac{E_{n\_\mathrm{current}}}{E_{n\_\max }}]$

L _currentfor present node is to the distance of leader cluster node, P (τ) is moment τ channel loss, for the channel loss in time period t;

P is the probability that node becomes leader cluster node, and r is current election wheel number;

The label n ∈ G of node _round, G _roundfor not being elected to the node set of leader cluster node in the nearest wheel of 1/P continuously;

E _{n_current}for node present energy, E _{n_max}for the primary power of node;

K is bunch head number expected; r _sit is the round that node is not elected as leader cluster node continuously.

7. adaptive routing method as claimed in claim 1, is characterized in that,

Set up route estimate for:

$\begin{array}{c}{E}^{\′}\left(N_k\left(r\right)\right)=E\left(N_k\left(t\right)\right)+M\left(N_k\left(r\right)\right)\\ =\frac{{\mathrm{\Σ}}_{k=1}^r\left[d_{N_k}{p}_{N_k}{\mathrm{\Π}}_{n=0}^{k-1}\stackrel{\‾}{p_{N_n}}\right]}{E_{\mathrm{tx}}+r{E}_{\mathrm{rx}}}+\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C\end{array}$

Wherein, N _kr () is potential relaying jointpoint set, the priority of these potential via nodes is: i ₁>i ₂> ... >i _j; that the node that priority is higher sends failed probability;

E _tx, E _rxbe respectively node send message time and receipt message time energy consumption, r is the quantity of optional node in potential via node; N is the number of node;

The mobile cost that all nodes are total:

$M\left(N_k\left(r\right)\right)=\frac{1}{N}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}\frac{1}{|\hat{s}(\mathrm{\τ}+t)-\hat{s}\left(\mathrm{\τ}\right)|}*C---\left(12\right)$

C is constant, it is the mobile cost of each node;

Wherein any one node j in the state of moment τ is:

$s\left(\mathrm{\τ}\right)={[x\left(\mathrm{\τ}\right),\stackrel{.}{x}\left(\mathrm{\τ}\right),\stackrel{..}{x}\left(\mathrm{\τ}\right),y\left(\mathrm{\τ}\right),\stackrel{.}{y}\left(\mathrm{\τ}\right),\stackrel{..}{y}\left(\mathrm{\τ}\right)]}^T---\left(9\right)$

S (τ) is the position of node j in cartesian coordinate system, the speed along x-axis and acceleration respectively; the speed along y-axis and acceleration respectively;

After time t, the position that this node j estimates at moment τ+t is:

$\hat{s}(\mathrm{\τ}+t)={[\hat{x}(\mathrm{\τ}+t),\widehat{\stackrel{.}{x}}(\mathrm{\τ}+t),\widehat{\stackrel{..}{x}}(\mathrm{\τ}+t),\hat{y}(\mathrm{\τ}+t),\widehat{\stackrel{.}{y}}(\mathrm{\τ}+t),\widehat{\stackrel{..}{y}}(\mathrm{\τ}+t)]}^T.$