CN104618979A - Adaptive partition routing method based on cross aiding - Google Patents

Abstract

The invention discloses an adaptive partition routing method based on cross aiding. The method is that an adaptive dynamic partition model is utilized to adapt the quickly-varied VANETs network; the cross transferring method is improved based on the constant cross information; a small-delay and high-density route is selected to improve the arrival rate of a data packet; a neighbor list is divided into the list of the neighbors in the same direction and the list of the neighbor in the opposite directions according to the neighbor node moving direction information, and the priority of the neighbor nodes in the same direction is superior to that of the neighbor nodes in the opposite directions, and therefore, the data can be transferred in the optimal route direction, and as a result, the delay is reduced. The method can be widely applied to the field of wireless network.

Description

A kind of adaptive partition method for routing auxiliary based on crossing

Technical field

The present invention relates to field of wireless, especially a kind of adaptive partition method for routing auxiliary based on crossing.

Background technology

Vehicular ad hoc network (VANETs, Vehicular Ad-hoc Networks) be wireless distributed structure, multi-hop, the continuous outdoor communication network changed, its basic thought is: in certain communication range, mutually can exchange the data of the information such as the speed of a motor vehicle, position each other and onboard sensor perception between vehicle, set up the mobile network of a self-organizing automatically.The single-hop communication scope of node only has hundreds of rice to a km, and each node (vehicle) is not only a transceiver, simultaneously or a router, therefore adopts the mode of multi-hop data retransmission to farther vehicle.Due to the movable self-organization character that VANETs is extremely special, there is following distinguishing feature:

(1) node high-speed mobile change.Due to node (vehicle) high-speed mobile, cause the topology of network to change frequently, make the route lifetime set up in network topology short;

(2) unsteadiness of inter-node communication.By the high-speed mobile of node and the impact of actual scene environment, the link establishment of communication often has provisional;

(3) in network, between section, the density situation of node differs.In actual scene, the node density under urban environment on section is comparatively large, and the node on section, countryside is then more sparse;

(4) movement of node is by the restriction of road, can only move on road.

(5) node (vehicle) is mounted with GPS device, in conjunction with electronic chart, make routing policy realize more convenient.

Therefore, character due to its uniqueness of VANETs: topological frequent variations, node height dynamic mobile, the link maintenance time is very short, determine traditional MANET agreement to be not suitable for being applied in vehicular ad hoc network, particularly in suburb, node density is low, when sparse, the performance of traditional MANET agreement more can't bear.

At present, the Routing Protocol used in vehicular ad hoc network roughly can be divided into 3 classes: based on route (TBR, the topology-based routing) agreement of topology; Location-based route (PBR, position-based routing) agreement; Based on route (MBR, the map-based routing) agreement of map.And these single Routing Protocol algorithms, well can not adapt to actual traffic scene, can not according to the forwarding strategy of the real-time change packet of traffic conditions.

Summary of the invention

In order to solve the problems of the technologies described above, the object of the invention is: provide a kind of assist based on crossing high packet delivery ratio, low delay adaptive partition method for routing.

The technical solution adopted in the present invention is: a kind of adaptive partition method for routing auxiliary based on crossing, comprises the following steps:

A, judge whether present node position is crossing and sets up crossing list, the list of described crossing to be used in memory node moving process the intersection information of process;

Node in B, movement, according to the directional information of neighbor node movement, generates neighbor list and is also divided into neighbor list and reverse neighbor list in the same way, and calculate in the list of crossing, the neighbor node in the same way between crossing and reverse neighbor node number;

Whether C, forward node are in section according to the directional information calculating neighbor node of host node and neighbor node; If so, the section then performing step e forwards; If not, then the crossing performing step D forwards;

D, adopt dynamic self-adapting partition model to select optimal path, from neighbor list, then select to move with selected direction, crossing the consistent and node that distance objective crossing is nearest as next-hop node, then perform the section forwarding of step e;

E, employing section forward, until packet is forwarded to intersection, then perform step D.

Further, the data at current crossing, one-level crossing and secondary crossing are preserved in the crossing list in described steps A.

Further, in described steps A and step C, forward node according to the direction signs value of the directional information of host node and neighbor node computing node respectively, and is worth the position of neighbor node to be in section or crossing according to the direction signs value difference of host node and neighbor node.

Further, described step D is specially: adopt dynamic self-adapting partition model to take host node as initial point, host node is x-axis to the vector of destination node, set up a new relative coordinate axle and direct graph with weight, with current crossing for the center of circle, current crossing is the dynamic picture circle of radius-adaptive to secondary crossing distance farthest, and this border circular areas is divided into three different priority regions, dynamic calculating from current crossing to selected priority regions in the optimal path at crossing, to select from neighbor list and the consistent and node that distance objective crossing is nearest is moved as next-hop node in selected direction, crossing, then the section performing step e forwards.

Further, when performing section forwarding in described step e, calculate local optimum and reverse local optimum situation in the same way simultaneously.

Further, the described situation of local optimum is in the same way that neighbor node distance objective crossing is minimum in neighbor list in the same way distance is greater than self situation to the distance at target crossing.

Further, described reverse local optimum situation is that the distance that in reverse neighbor list, neighbor node distance objective crossing is minimum is greater than the situation that self arrives the distance at target crossing.

Further, in described step D, do not obtain next-hop node according to dynamic self-adapting partition model, then present node not forwarding data bag, and by present node with data on new section, then perform step e section forward.

The invention has the beneficial effects as follows: the present invention is by the dynamic partition model of self adaptation, the fast-changing VANETs network of better adaptation, utilize constant intersection information, improve crossing retransmission method, select an arrival rate that time delay is little, density is high path improves packet; And according to the directional information of neighbor node movement, neighbor list is divided into neighbor list and reverse neighbor list in the same way, and the priority of neighbor node, higher than reverse neighbor node, ensure that the direction forwarding of data retransmission along optimal path, reduces time delay in the same way.

Accompanying drawing explanation

Fig. 1 is the flow chart of steps of adaptive partition of the present invention method for routing;

Fig. 2 is the inventive method interior joint moving direction mark schematic diagram;

Fig. 3 is partition model schematic diagram of the present invention;

Fig. 4 is that the reference axis of Fig. 3 rebuilds schematic diagram;

Fig. 5 is the present invention's three kinds of local optimum situation schematic diagrames.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described further:

With reference to Fig. 1, a kind of adaptive partition method for routing auxiliary based on crossing, comprises the following steps:

A, judge whether present node position is crossing and sets up crossing list, the list of described crossing to be used in memory node moving process the intersection information of process;

Present node often arrives a crossing all to be needed to recalculate once;

Node in B, movement, according to the directional information of neighbor node movement, generates neighbor list and is also divided into neighbor list and reverse neighbor list in the same way, and calculate in the list of crossing, the neighbor node in the same way between crossing and reverse neighbor node number;

Whether C, forward node are in section according to the directional information calculating neighbor node of host node and neighbor node; If so, the section then performing step e forwards; If not, then the crossing performing step D forwards;

D, adopt dynamic self-adapting partition model to select optimal path, from neighbor list, then select to move with selected direction, crossing the consistent and node that distance objective crossing is nearest as next-hop node, then perform the section forwarding of step e;

E, employing section forward, until packet is forwarded to intersection, then perform step D.

Be further used as preferred embodiment, in the crossing list in described steps A, preserve the data at current crossing, one-level crossing and secondary crossing.

Be further used as preferred embodiment, in described steps A and step C, forward node according to the direction signs value of the directional information of host node and neighbor node computing node respectively, and is worth the position of neighbor node to be in section or crossing according to the direction signs value difference of host node and neighbor node.

With reference to Fig. 2, concrete determination methods is described: the angle of the moving direction of node is θ _i, specific as follows:

If then direction signs value is 0; If then direction signs value is 4;

If then direction signs value is 1; If then direction signs value is 5;

If then direction signs value is 2; If then direction signs value is 6;

If then direction signs value is 3; If then direction signs value is 7.

The direction signs value difference value of host node and neighbor node, if there is difference to be not equal to 0 or be not equal to 4, then this host node is in crossing.Through crossing, now the coordinate of vehicle is labeled as the coordinate information of intersection, and this intersection information comprises: the unique identifier at crossing, i.e. the coordinate information at crossing.Node is preserved in this intersection information to oneself crossing list, and broadcasts intersection information, and the renewal of crossing list upgrades along with neighbor list.Through the node at same crossing, the up-to-date crossing in its crossing list is identical.

The information at three crossings is only preserved in crossing list: current crossing, one-level crossing, secondary crossing.Described current crossing is a up-to-date crossing of preserving in the list of crossing; Described one-level crossing is the neighbours crossing at current crossing; Described secondary crossing is the neighbours crossing at one-level crossing.

Be further used as preferred embodiment, described step D is specially: adopt dynamic self-adapting partition model to take host node as initial point, host node is x-axis to the vector of destination node, set up a new relative coordinate axle and direct graph with weight G (V, E), with current crossing for the center of circle, current crossing is the dynamic picture circle of radius-adaptive to secondary crossing distance farthest, and this border circular areas is divided into three different priority regions, dynamic calculating from current crossing to selected priority regions in the optimal path at crossing, to select from neighbor list and the consistent and node that distance objective crossing is nearest is moved as next-hop node in selected direction, crossing, then the section performing step e forwards.

Below the dynamic self-adapting partition model in step D is illustrated: according to the node motion direction signs schematic diagram of Fig. 2, the direction signs value difference value of host node and neighbor node is 0 (in the same way) or value when being 4 (oppositely) (this node n is to up-to-date crossing I), forward node is according to the intersection information receiving neighbours and send over, renewal intersection information is J → P → I, and the intersection information that forward node receives has following four kinds of situations:

(1) node through P crossing does not also upgrade neighbours' intersection information of intersection information: K _a→ J _b→ P;

(2) node through I crossing does not also upgrade neighbours' intersection information of intersection information: M _c→ N _d→ I;

(3) neighbours' intersection information of intersection information has been upgraded toward the node of crossing I movement through P crossing: J _b→ P → I;

(4) neighbours' intersection information of intersection information has been upgraded toward the node of crossing P movement through I crossing: N _d→ I → P.

D (I, J) is calculated respectively, d (I, K according to above-mentioned four kinds of situations _a), d (I, J _b), d (I, M _c), d (I, N _d), and obtain current crossing to secondary crossing distance d (I) farthest:

D (I)=max{d (I, J), d (I, K _a), d (I, J _b), d (I, M _c), d (I, N _d) formula (1)

Wherein: a=1,2,3 ..., t, b=1,2,3 ..., w, c=1,2,3 ..., u, d=1,2,3 ..., v, and t, w, u, v represent the crossing quantity meeting above-mentioned intersection information respectively.

Now, with crossing I for the center of circle, d (I) draws a circle for radius, builds crossing preference pattern, as shown in Figure 3, supposes d (I)=d (I, J).

If d is (I, D) > d (I) (for the purpose of D node), the straight line that can obtain node D and crossing I and circle Ο (I) intersect at E, F 2 point, and cross D two tangent lines tangent with circle Ο (I), point of contact is respectively A, B, connects IA, IB.

In the reference axis of rebuilding, as shown in Figure 4, all crossings (one-level crossing and secondary crossing) and the angle between crossing I and destination node D three in circle is calculated:

${\mathrm{\ω}}_i=\mathrm{ac}\cos \left(\frac{\stackrel{\→}{\mathrm{ID}}\×\stackrel{\→}{{\mathrm{II}}_i}}{\left|\stackrel{\→}{\mathrm{ID}}\right|\·\left|\stackrel{\→}{{\mathrm{II}}_i}\right|}\right)$ Formula (2)

$\mathrm{\ω}=\mathrm{ac}\cos \left(\frac{d(I,A)}{d(I,D)}\right)$ Formula (3)

According to right hand principle, carry out prioritization by formula (2) and formula (3), specific as follows:

(1)-ω≤ω _iduring≤ω, the crossing priority level on this interval (as AIB fan-shaped in Fig. 3) is the highest;

(2) ω < ω _iduring≤π, the crossing priority level on this interval (as AIF fan-shaped in Fig. 3) is placed in the middle;

(3)-π≤ω _iduring <-ω, the crossing priority level on this interval (as FIB fan-shaped in Fig. 3) is minimum.

What step D adopted is crossing forwarding, wherein:

1, the weight computing in section

If d (I, D) > d (I), namely destination node is outside circle, utilizes weights formula:

$V_{\mathrm{ij}}=\mathrm{\&alpha;}\frac{l_{\mathrm{ij}}}{\stackrel{\&OverBar;}{v_{\mathrm{ij}}}}+\mathrm{\&beta;}(1-\frac{d_{\mathrm{ij}}}{d(I,D)+d_2\left(I\right)})$ Formula (4)

Obtain the weights size V on each section i → j _ij.Wherein, l _ijrepresent the length in section between crossing i and adjacent intersection j, i.e. l _ij=| i → j|, represent the upper node average speed of section i → j, d _ij=d (j, D).Therefore l is had _ij=l _ji,

If d (I, D)≤d (I), namely destination node is in circle, then revise a parameter of formula (4), as follows, calculates the weights in section:

$V_{\mathrm{ij}}=\mathrm{\&alpha;}\frac{l_{\mathrm{ij}}}{\stackrel{\&OverBar;}{v_{\mathrm{ij}}}}+\mathrm{\&beta;}(1-\frac{d_{\mathrm{ij}}}{2\&times;d_2\left(I\right)})$ Formula (5)

2, crossing is selected

$U_{\mathrm{ij}}=\underset{k\&Element;N\left(j\right)}{\mathrm{\&Sigma;}}{\mathrm{\&sigma;}}_{\mathrm{ij}}\&times;V_{\mathrm{jk}}$ Formula (6)

Wherein, σ _ijrepresent the standard variance of the node number on the i → j of section, and σ _ij=σ _ji, N (j) represents the one-level crossing set of crossing j, V _ijrepresent the weights on i → j section.

3, σ _ijcalculating

I → j section has the probability of k many mobile nodes obey Poisson distribution:

$P(X=k)=\frac{{\left({\mathrm{ul}}_{\mathrm{ij}}\right)}^k}{k!}{e}^{{-\mathrm{ul}}_{\mathrm{ij}}}$ Formula (7)

Wherein ev represents the expectation of the upper speed of section i → j, and λ represents the average arrival rate of node on section, l _ijrepresent the distance in i → j section.So obtain

At the forward node at crossing place, design of graphics 0, V is all crossings in circle, and E is the line between crossing, U _ijfor the weights on i → j direction, to direct graph with weight G (V, E), finding with I is start crossing, I _ifor terminating crossing (I _icrossing in selected priority regions) an optimal path.Suppose that obtaining possible optimal solution is minimum, then next crossing that packet should be selected is Q, direction I → Q.

Obtaining optimal solution is U _iQ, forward node should in its neighbor list optional m igration direction consistent with I → Q, and the nearest node of distance crossing Q is as down hop.

Be further used as preferred embodiment, if such neighbor node does not exist, then packet does not forward, node with data mobile through crossing on a new section, on new section, adopt section forwarding strategy to forward.

Be further used as preferred embodiment, when performing section forwarding in described step e, calculate local optimum and reverse local optimum situation in the same way simultaneously.

Be further used as preferred embodiment, the described situation of local optimum is in the same way that neighbor node distance objective crossing is minimum in neighbor list in the same way distance is greater than self situation to the distance at target crossing.

Be further used as preferred embodiment, described reverse local optimum situation is that the distance that in reverse neighbor list, neighbor node distance objective crossing is minimum is greater than the situation that self arrives the distance at target crossing.

What step e adopted is section forwarding, supposes now through crossing forwarding strategy algorithm, and obtain optimum crossing forwarding and represent current crossing for I → J, I, J represents next crossing routing direction of selection.In the forwarding strategy of crossing, the node motion direction of the down hop of forwarder selection must along I → J direction.

Situation 1: if now the moving direction of forward node is I → J direction

If neighbor list is not empty in the same way, then obtains the distance at each nodal distance J crossing in neighbor list in the same way, and obtain node corresponding to minimum range, that is:

M _s={ m _q| min [d (m _q, J)] } q=1,2,3 ..., p formula (8)

D (m _s, J) and≤d (n, J) formula (9)

If d is (m _s, J) and≤d (n, J), then select m _scorresponding node m _qas down hop.Wherein, n represents forward node, and d (n, J) represents the distance of node n to crossing J, m _qrepresent q node in neighbor list in the same way, p represents the quantity of neighbor node in the same way of forward node n.

If d is (m _s, J) > d (n, J) or in the same way neighbor list be empty, this kind of situation for local optimum problem in the same way as shown in Figure 5: when packet is forwarded to node L, there is local optimum situation in the same way in L.Now from reverse neighbor list, calculate the optimum node m in reverse neighbor list according to formula (8) and formula (9) _r.

If there is multiple neighbor node to meet simultaneously, then random selecting one is as down hop.

If d is (m _s, J) and > d (n, J) and d (m _rj) > d (n, J), then occur complete local optimum problem, as shown in Figure 5, now forward node N is just in complete local optimum situation, it will carry packet and move together, do not forward yet not packet loss, until arrive employing crossing, crossing forwarding strategy forwarding data bag, or forwarding data bag when finding best down hop.

Situation 2: in 1, packet may be forwarded to the node in reverse neighbor list, the direction of now forward node movement is contrary with I → J direction, path of selection, then pay the utmost attention to the reverse neighbor list of this forward node, find the node m nearest apart from next crossing in lists according to formula (8) and formula (9) _qif, d (m _q, J) and≤d (n, J), then select q node as down hop; Otherwise there is reverse local optimum situation, as shown in Figure 5, forward node M is in this kind of situation, now in its in the same way neighbor list, find the node nearest apart from next crossing, distance is here less to the distance d (n, J) at next crossing than forward node, otherwise there is complete local optimum situation, now node can only carry packet and moves, until run into best down hop or arrive crossing, is just forwarded by packet.

In reverse local optimum situation, its reverse neighbor list is empty scarcely, because the packet that now forward node receives forwards from its reverse neighbor node.

In whole data forwarding process, until packet is forwarded to crossing, then changes into and adopt crossing to forward.

More than that better enforcement of the present invention is illustrated, but the invention is not limited to described embodiment, those of ordinary skill in the art can also make all equivalents or replacement under the prerequisite without prejudice to spirit of the present invention, and these equivalent distortion or replacement are all included in the application's claim limited range.

Claims (8)

1., based on the adaptive partition method for routing that crossing is auxiliary, it is characterized in that: comprise the following steps:

A, judge whether present node position is crossing and sets up crossing list, the list of described crossing to be used in memory node moving process the intersection information of process;

Node in B, movement, according to the directional information of neighbor node movement, generates neighbor list and is also divided into neighbor list and reverse neighbor list in the same way, and calculate in the list of crossing, the neighbor node in the same way between crossing and reverse neighbor node number;

Whether C, forward node are in section according to the directional information calculating neighbor node of host node and neighbor node; If so, the section then performing step e forwards; If not, then the crossing performing step D forwards;

D, adopt dynamic self-adapting partition model to select optimal path, from neighbor list, then select to move with selected direction, crossing the consistent and node that distance objective crossing is nearest as next-hop node, then perform the section forwarding of step e;

E, employing section forward, until packet is forwarded to intersection, then perform step D.

2. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 1, is characterized in that: the data of preserving current crossing, one-level crossing and secondary crossing in the crossing list in described steps A.

3. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 1, it is characterized in that: in described steps A and step C, forward node according to the direction signs value of the directional information of host node and neighbor node computing node respectively, and is worth the position of neighbor node to be in section or crossing according to the direction signs value difference of host node and neighbor node.

4. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 1, it is characterized in that: described step D is specially: adopt dynamic self-adapting partition model to take host node as initial point, host node is x-axis to the vector of destination node, sets up a new relative coordinate axle and direct graph with weight with current crossing for the center of circle, current crossing is the dynamic picture circle of radius-adaptive to secondary crossing distance farthest, and this border circular areas is divided into three different priority regions, dynamic calculating from current crossing to selected priority regions in the optimal path at crossing, select to move the consistent and node that distance objective crossing is nearest as next-hop node with selected direction, crossing from neighbor list, then perform the section forwarding of step e.

5. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 1, is characterized in that: when performing section forwarding in described step e, calculate local optimum and reverse local optimum situation in the same way simultaneously.

6. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 5, is characterized in that: the described situation of local optimum is in the same way that neighbor node distance objective crossing is minimum in neighbor list in the same way distance is greater than self situation to the distance at target crossing.

7. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 5, is characterized in that: described reverse local optimum situation is that the distance that in reverse neighbor list, neighbor node distance objective crossing is minimum is greater than the situation that self arrives the distance at target crossing.

8. a kind of adaptive partition method for routing auxiliary based on crossing according to claim 1, it is characterized in that: in described step D, next-hop node is not obtained according to dynamic self-adapting partition model, then present node not forwarding data bag, and by present node with data on new section, then perform step e section forward.