CN101159689A - Mobile sensor network routing method based on region gradient updating - Google Patents

Abstract

The invention discloses a mobile sensor network routing method based on regional gradient updating. The method is helpful for the application and development of wireless sensor network technology in China, and will take important roles in environmental monitoring and forecasting, emergency treatment of natural disaster, scientific expedition and exploration, intelligent family, urban traffic, management of large-scale garage and warehouse, safety monitoring of large-scale industry park and other fields. The invention mainly solves the problems of the protocol in prior art, including large energy expenditure and large storage space. The method comprises a base station constructs regional gradient and performs periodical regional gradient updating, so as to synchronizing the time of the entire network. When a source node transmits data to a merge mode, the source node constructs a link going through the merge node, so as to conduct data transmission. The invention is based on wireless sensor network, increases the service life of wireless sensor network, and saves storage space of a sensor.

Description

Mobile sensor network routing method based on region gradient updating

Technical field

The present invention relates to a kind of in wireless sensor network the Route Selection between sensor node and the aggregation node, a kind of mobile sensor network routing method based on region gradient updating particularly is provided.

Background technology

Mobile sensor network is a kind of dynamic sensor network, and it is made of a large amount of dynamic pickup nodes and a small amount of static sensor node, and the dynamic pickup node moves in the network area at random, relies on Radio Link to communicate between the neighbors.Movable sensor node in the network is responsible for acquisition of information, and utilizes other sensor node to forward information to aggregation node.Method for routing is the information transmission path that is used for setting up from the source node to the aggregation node, need to select the intermediate node of forwarding information.All nodes on the transmission path will consume certain energy when transmitting data, how the effective Routing Protocol of design energy is one of key issue of sensor network always.

Because the characteristic that node moves at random in the mobile sensor network, existing sensor network Routing Protocol based on stationary nodes can't be applied directly in this network.At present, proposed several Routing Protocols, wherein comparatively be typically the disclosed a kind of mobile sensor network Routing Protocol of document " Impact of Mobility on Mobility-Assisted Information Diffusion (MAID) Protocols in Ad hoc Networks.USC Computer Science Department Technical Report " based on mobile sensor network.This agreement is a kind of data-centered Routing Protocol, node in the network in the process that moves with neighbor node information swap time that is run into, utilize two internodal Encounter Time length at interval to decide the size of time gradient value, each intranodal all to have an Encounter Time Grad table with other node.When node need send data to aggregation node, it in neighbors, seek one with the at interval the shortest node of aggregation node Encounter Time as its next-hop node, so repeatedly, till arriving aggregation node, the transmission path from the source node to the aggregation node is set up like this.

To set up convergence time shorter for designed Routing Protocol route in the document, but have following defective: network node needs in moving process that swap time, information was set up gradient continually, and energy expense is bigger; Each node need be preserved an Encounter Time Grad table with other node, need take a certain amount of storage area.

Summary of the invention

The object of the present invention is to provide a kind of mobile sensor network routing method based on region gradient updating.Based on wireless sensor network, increase wireless sensor network useful life, save the transducer memory space.

The present invention relates to sensor network is dense sensor network, promptly can guarantee mutual communication between the adjacent node in the network.Its network system that is suitable for is made up of transmitting base station, sensor node and aggregation node.

Transmitting base station: energy is sustainable to be replenished; The emission maximum radius can cover all sensor nodes in the network; Being responsible for whole network virtual subregion also is that the interior sensor node in each zone distributes Grad; When sensor node moved, it is the Grad of new sensor node more; Be responsible for the time synchronized of whole network.

Sensor node: finite energy and can not being replenished; Main being responsible for transmitted the information that senses to aggregation node.Sensor node receives only the information from the broadcast message of transmitting base station and the transmission of other sensor node; Transmit data to other sensor node; When some sensor nodes need send data to aggregation node, it was a source node just; Its position in network is unfixed.

Aggregation node: energy is sustainable to be replenished; Storage and computing capability all are better than sensor node; Connect sensor network and external network; Communication protocol between the translation-protocol stack; The data of collecting are dealt on the external network; Its position in network is fixed; Its Grad is 0.

For fear of the conflict of region gradient information signal and data communication, use broadcasting and two segregated channel of transfer of data in the sensor network.Wherein broadcast channel is that the base station is used for to network inner sensor node broadcasts area information; Data transmission channel is to be used for intercoming mutually between the network inner sensor node.Initiate to set up the region gradient line period region gradient updating of going forward side by side by the base station, and make whole network time synchronization.To when aggregation node sends data, initiate when source node, set up a link that leads to aggregation node, thereby finish transfer of data by source node.Its method is as follows:

(a) foundation of region gradient: be each the node distribution network unique identifying number (ID) in the network, after network node was disposed, whole network was in initial condition, and the node of base station and whole network carries out the synchronous of time.Then the base station is just periodically by the node broadcasts region gradient information of broadcast channel in network, and it broadcasts radius R _iBe increasing sequence, that is: R _i=ir, wherein: i={1,2,3...m}, r are the transmission radius of sensor node.

Node in the network just begins to intercept broadcast channel and data transmission channel after a time slot begins.In broadcast channel, when for the first time listening to the region gradient bag of base station broadcast, node just receives this packet and own Grad is set at the interior entrained Grad of bag.Can judge that according to the sequence numerical value in the bag this bag is the region gradient bag of base station which time slot broadcasting; Can know the gradient magnitude of node according to the Grad in the bag.At this moment, the region gradient of network node is set up.

(b) renewal of region gradient: because mobile sensor network is a kind of dynamic network, its internal sensor node is to move at random, therefore need the network area gradient of setting up be updated periodically.When node motion was regional to other, in the cycle, the Grad of node will be updated to the Grad of region at next base station broadcast, otherwise the Grad of node remains unchanged.

(c) foundation of route: this agreement belongs to on-demand routing protocol, and when source node need send data to aggregation node, it can be initiated route and set up process.Set up in the process in route, the principle that node is sought next-hop node is: gradient is greedy to be selected and energy perception selection.Generally speaking, node utilizes the greedy selection algorithm of gradient to carry out the search of next-hop node, and when node to be selected during greater than, node just starts energy perception selection algorithm and selects in node to be selected, aforesaid way carries out repeatedly, till finding aggregation node.After treating that route is set up successfully, aggregation node notification source node begins to send data.

Now the nodal information amount of overhead in this agreement and the MAID agreement is analyzed comparison.

If the network area area is D, sensor node quantity is N, and the transmission radius of node is r, and the average translational speed of node is v, then:

(1) network node density ρ is: $\mathrm{\ρ}=\frac{N}{D};$

The average neighbor node number N that meets when (2) node j moves with average speed v in time interval Δ t _iFor:

$N_i=\mathrm{S\ρ}=\frac{2\mathrm{rvN}}{D}\mathrm{\Δt}$

(3) node i is upgraded the number of times M of regional gradient information and is in time interval Δ t:

$M=\frac{\mathrm{\Δt}}{T_s}$

During route is set up,, get base station broadcast time slot T again in order to reduce the generation in route cavity _s=r/v, that is:

$M=\frac{\mathrm{\Δt}}{T_s}=\frac{v}{r}\mathrm{\Δt}$

(4) in the MAID agreement, node i is the average packets of information N that sends in time interval Δ t _S ^MAIDPackets of information N with average received _R ^MAIDBe respectively:

$N_S^{\mathrm{MAID}}=\frac{S}{{\mathrm{\&pi;r}}^2}=\frac{2\mathrm{rv\&Delta;t}}{{\mathrm{\&pi;r}}^2}=\frac{2v}{\mathrm{\&pi;r}}\mathrm{\&Delta;<figure-callout\; id="1"\; label="t"\; filenames="S2007101770244E00021.png"\; state="\{\{state\}\}">t</figure-callout>}$ ①

$N_R^{\mathrm{MAID}}=N_i=\frac{2\mathrm{rvN}}{D}\mathrm{\&Delta;t}$ ②

(5) in this agreement, node i is the average packets of information N that sends in time interval Δ t _S ^RGBPackets of information N with average received _R ^RGRBe respectively:

$N_S^{\mathrm{RGR}}=0$ ③

$N_R^{\mathrm{RGR}}=M=\frac{v}{r}\mathrm{\&Delta;<figure-callout\; id="4"\; label="t"\; filenames="S2007101770244E00041.png,S2007101770244E00061.png"\; state="\{\{state\}\}">t</figure-callout>}$ ④

Calculate N now _R ^MAID/ N _R ^RGB, will be 2. formula and 4. the formula substitution can get:

$\frac{N_R^{\mathrm{MAID}}}{N_R^{\mathrm{RGR}}}=\frac{2\mathrm{rvN}}{D}\mathrm{\&Delta;t}\&CenterDot;\frac{r}{\mathrm{v\&Delta;t}}=\frac{2r^2\mathrm{<figure-callout\; id="5"\; label="N\; D"\; filenames="S2007101770244E00041.png,S2007101770244E00061.png"\; state="\{\{state\}\}">N</figure-callout>}}{D}$ ⑤

Because network involved in the present invention is dense sensor network, so can divide following two kinds of situations that 5. formula is discussed:

First kind of situation is that N sensor node in the network is in best covering state, referring to Fig. 1.Wherein the area S of rectangle is:

S＝Nr·2r＝2r ²N

N the region area S ' that node covered is:

S′＝D

Then by Tu Kede: $\frac{S}{S^{\&prime;}}=\frac{2r^{2}N}{D}=1$

Promptly $\frac{N_R^{\mathrm{MAID}}}{N_R^{\mathrm{RGR}}}=\frac{2r^{2}N}{D}=1$ ⑥

Second kind of situation is that N sensor node in the network is in dense covering state, referring to Fig. 2.Wherein the area S of rectangle is:

S＝Nr·2r＝2r ²N

N the region area S ' that node covered is:

S′＝D

Then by Tu Kede: $\frac{S}{S^{\&prime;}}=\frac{2r^{2}N}{D}>1$

Promptly $\frac{N_R^{\mathrm{MAID}}}{N_R^{\mathrm{RGR}}}=\frac{2r^{2}N}{D}>1$ ⑦

Two kinds of situations being analyzed above comprehensive, by 6. formula and 7. formula can obtain following conclusion:

$\frac{N_R^{\mathrm{MAID}}}{N_R^{\mathrm{RGR}}}=\frac{2r^{2}N}{D}\&GreaterEqual;1$

Promptly $N_R^{\mathrm{MAID}}\&GreaterEqual;N_R^{\mathrm{RGR}}$ ⑧

By above analysis as can be known: in the identical time interval, sensor node lacks than the information packet quantity that receives in the MAID agreement in this agreement, and in this agreement node gradient set up and renewal process in broadcast message packet not, so this agreement is littler than the nodal information expense of MAID agreement.

The invention has the beneficial effects as follows: in the process of region gradient foundation and renewal, node in the network just receives the packets of information from base station broadcast, do not carry out the exchange of information between the neighbors, the node of Yi Donging reduces at unit interval internal information exchange capacity like this, and the energy consumption of node reduces; Intranodal just stores the Grad of oneself, does not store the Grad of other node, and node store information amount reduces, thereby increases the useful life of sensor network, the information stores space of saving sensor node.

Description of drawings

Fig. 1 is that sensor node of the present invention is in the best state diagram that covers, wherein: black color dots representative sensor node, the number of N representative sensor node, the transmission radius of r representative sensor node.

Fig. 2 is that sensor node of the present invention is in dense covering state diagram, wherein: black color dots representative sensor node, the number of N representative sensor node, the transmission radius of r representative sensor node.

Fig. 3 is a transmitting base station periodic broadcasting region gradient information schematic diagram of the present invention, wherein: black color dots representative sensor node, grey color dot is represented aggregation node, and a and b represent the length and the width of network area, R respectively ₁～R ₆Represent the different broadcasting radius of transmitting base station.

Fig. 4 is the present invention's base station transmitting power change curve in time in broadcast channel, wherein: t _bRepresent transmitting base station used time of region gradient bag of broadcasting, T _sRepresent broadcast cycle (time slot) time of base station, T _bThe used time of region gradient bag is taken turns in representative broadcasting one.

Fig. 5 and Fig. 6 are that node region Grad of the present invention upgrades schematic diagram, wherein: alphabetical A, B and the different virtual region of C representative, alphabetical a, b and c representative sensor node serial number, the Grad of alphabetical m, n and k representative sensor node.

Fig. 7, Fig. 8 and Fig. 9 are the greedy schematic diagrames of selecting of route gradient of the present invention, and wherein: grey color dot is represented source node, and other sensor node is its neighbor node, alphabetical d, f, g, h and j representative sensor node serial number.

Figure 10, Figure 11 and Figure 12 are that schematic diagram is selected in routing energy perception of the present invention, wherein: grey color dot is represented source node, other sensor node is its neighbor node, alphabetical x, y, z, u, v and w representative sensor node serial number, the current remaining value of ē representative sensor node.

Embodiment

Below in conjunction with the drawings and specific embodiments the present invention is elaborated.

Concrete enforcement is as follows: this agreement is used broadcasting and two segregated channel of transfer of data for fear of the conflict of region gradient information signal and data communication in the sensor network.Wherein broadcast channel is that the base station is used for to network inner sensor node broadcasts area information; Data transmission channel is to be used for intercoming mutually between the network inner sensor node.This agreement is made up of following three parts:

(a) foundation of region gradient: after network node was disposed, whole network was in initial condition, and the node of base station and whole network carries out the synchronous of time.Then the base station just periodically by broadcast channel to network node broadcast area gradient information, the broadcasting radius R _tBe time dependent increasing sequence, that is: R _i=ir, wherein: i={1,2,3...m}, r are the transmission radius of sensor node.Referring to Fig. 3, suppose that network coverage is a rectangle, its length of side is a and b, covers for making the base station broadcast radius reach the whole network, requires the maximum broadcasting in base station radius R _mGreater than the longest distance L between 2 of the network edges, again in the rectangle the longest 2 be diagonal apart from line segment, that is: $L=\sqrt{a^2+b^2},$ And

Then:

Wherein

It is right to represent

Round up.

Wherein the communication format of region gradient bag is:

Heading

Broadcast flag

Sequence number

Grad

Heading comprises message authentication codes, is used for reducing the biography code check.

Broadcast flag is the identification number of this packet.

Sequence number is to distinguish different time slots.If twice of base station with same emission radius broadcast area gradient bag, claims that then this time interval is a broadcast cycle, i.e. a time slot.

Grad is the relative distance value that is assigned to the base station for the node in each zone.

Referring to Fig. 4, in broadcast channel, the used time of region gradient bag of base station broadcast is t _b, then to take turns the used time of region gradient bag be mt in its broadcasting one _b, that is: T _b=mt _b, wherein m be the base station in a number of broadcast times of taking turns in the broadcasting, just broadcast the number of radius; Broadcast cycle (time slot) time of base station is T _s, when a time slot began, the base station just began mode with the incremental broadcast radius to Web broadcast region gradient bag, takes turns off-the-air up to this.

Node in the network just begins to intercept broadcast channel and data transmission channel after a time slot begins.In broadcast channel, when for the first time listening to the region gradient bag of base station broadcast, node just receives this packet and own Grad is set at the interior entrained Grad of bag.Can judge that according to the sequence numerical value in the bag this bag is the region gradient bag of base station which time slot broadcasting; Can know the gradient magnitude of node according to the Grad in the bag.In this time slot, node is just closed broadcast channel after receiving the region gradient bag, continues data transmission channel is intercepted, up to T _bAfter time, node enters sleep state.At this moment, the region gradient of network node is set up.

(b) renewal of region gradient: because mobile sensor network is a kind of dynamic network, its internal sensor node is to move at random, therefore need the network area gradient of setting up be updated periodically.When node motion was regional to other, the Grad of node will be updated to this regional Grad, otherwise the Grad of node remains unchanged.Referring to Fig. 5 and Fig. 6, network is divided into three virtual regions, that is: A district, B district and C district, and wherein: the region gradient value in A district is m; The region gradient value in B district is n; The region gradient value in C district is k.Among Fig. 5, at t ₀Constantly, the node a in the A district is to the motion of B district, and the node c in the B district moves in the one's respective area, and the node b in the C district also moves to the B district; Among Fig. 6, at t ₁Constantly, node a, node b and node c all are in the B district, when base station during at next round broadcast area gradient bag, the region gradient value m of node a is updated to n, the region gradient value k of node b also is updated to n, and the region gradient value of node c remains unchanged, and still is n, the process that the node region Grad upgrades more than has been described after node location changes.

(c) foundation of route: this agreement belongs to on-demand routing protocol, and when source node need send data to aggregation node, it can be initiated route and set up process.Set up in the process in route, the principle that node is sought next-hop node is: the gradient greediness ^[8]Select and energy perception selection.Generally speaking, node utilizes the greedy selection algorithm of gradient to carry out the search of next-hop node, when node to be selected during greater than one, node just starts energy perception selection algorithm and selects in node to be selected, according to aforesaid way, till finding aggregation node, the information path from the source node to the aggregation node has been set up and has been finished like this.

Introduce the greedy selection algorithm of gradient and the energy perception selection algorithm of Route Selection below.

The greedy selection algorithm of gradient: by base station periodic broadcasting region gradient packets of information, the node in the network has been assigned with Grad, and it just can select next-hop node according to the greedy selection algorithm of gradient.Referring to Fig. 7, Fig. 8 and Fig. 9, its Grad of digitized representation on the node, source node d has data need send to aggregation node among Fig. 7, and it is broadcasted routing request packet (REQ) to neighbor node and initiates route requests, and neighbor node h, j, f and g have all received the routing request packet (REQ) of node d; Neighbor node h, j, f and g pass through the Grad (G of self among Fig. 8 _h, G _j, G _fAnd G _g) with the Grad (G of node d _d) carry out matching ratio, obtain: G _j, G _fAnd G _gAll be not more than G _d, but G _hGreater than G _d, so neighbor node j, f and g meet the condition of node to be selected, they answer a route response bag (RES) to node d respectively, and neighbor node h is learning that oneself not meeting condition to be selected just enters sleep state; After node x receives the route response bag (RES) of neighbor node j, f and g transmission among Fig. 9, with G _j, G _fAnd G _gCompare, obtain: G _jMinimum, it just selects neighbor node j as next-hop node then, sends a route affirmation bag (ACK) to node j and confirms, at this moment, sets up to the route of node j from node d.

Energy perception selection algorithm: when utilizing the greedy selection algorithm of gradient to carry out the selection of next-hop node, if during more than one of node to be selected, at this moment route initiation node just starts energy perception selection algorithm, and the node that dump energy of selection is maximum in node to be selected is as its next-hop node.Referring to Figure 10, Figure 11 and Figure 12, its Grad of digitized representation on the node, ē is the current remaining value of node, source node x broadcasts routing request packet (REQ) to neighbor node and initiates route requests among Figure 10, and neighbor node u, v, w, y and z have all received the routing request packet (REQ) of node x; Neighbor node u, v, w, y and z pass through the Grad (G of self among Figure 11 _u, G _v, G _w, G _yAnd G _z) with the Grad (G of node x _x) carry out matching ratio, obtain: G _v, G _w, G _yAnd G _zAll be not more than G _x, but G _uGreater than G _xTherefore neighbor node v, w, y and z meet the condition of node to be selected, they answer a route response bag (RES) to node x respectively, wherein wrap the current remaining value ē that internal information comprises this node, and neighbor node u is learning that oneself not meeting condition to be selected just enters sleep state; After node x receives the route response bag (RES) of neighbor node v, w, y and z transmission among Figure 12, with G _v, G _w, G _yAnd G _zCompare, obtain: G _yAnd G _zMinimum, it is with regard to the residual energy value ē of comparison node y and z, wherein: ē then _z＞ē _y, final node x selects neighbor node z as next-hop node, sends a route affirmation bag (ACK) to node z and confirms.After treating that route is set up successfully, aggregation node notification source node begins to send data.

Claims (1)

1. mobile sensor network routing method based on region gradient updating, it is characterized in that: network system is made up of transmitting base station, sensor node and aggregation node; Use broadcast channel and two segregated channel of data transmission channel in the sensor network; (a) set up by transmitting base station initiated area gradient, be each the node distribution network unique identifying number (ID) in the network, after network node is disposed, whole network is in initial condition, the node of base station and whole network carries out the synchronous of time, then the base station is just periodically by the node broadcasts region gradient information of broadcast channel in network, and it broadcasts radius R _iBe increasing sequence, that is: R _i=ir, wherein: i={1,2,3...m}, r is the transmission radius of sensor node, node in the network is after a time slot begins, just begin to intercept broadcast channel and data transmission channel, in broadcast channel, when for the first time listening to the region gradient bag of base station broadcast, node just receives this packet and own Grad is set at the interior entrained Grad of bag, judge that according to the sequence numerical value in the bag this bag is the region gradient bag of base station which time slot broadcasting, know the gradient magnitude of node according to wrapping interior Grad, at this moment, the region gradient of network node is set up; (b) renewal of region gradient: when node motion was regional to other, in the cycle, the Grad of node will be updated to the Grad of region at next base station broadcast, otherwise the Grad of node remains unchanged; (c) foundation of route: when source node need send data to aggregation node, it can be initiated route and set up process, set up in the process in route, the principle that node is sought next-hop node is that greedy selection of gradient and energy perception are selected, node utilizes the greedy selection algorithm of gradient to carry out the search of next-hop node, and when node to be selected during greater than, node just starts energy perception selection algorithm and selects in node to be selected, aforesaid way carries out repeatedly, till finding aggregation node.

Images