CN103973556A - Credible routing method of vehicular delay-tolerant network - Google Patents

Abstract

Disclosed is a credible routing method of a vehicular delay-tolerant network. The credible routing method includes adopting a movement model similarity evaluation scheme to transfer data to movement ranges of destination nodes according to similarity among movement models and local real-time information of vehicles, generating L copies by means of a copy equilibrium strategy after source nodes generate data packets, handing out in a cuttable manner, independently executing transfer of each copy until one of the copies reaches the destination nodes, dynamically adjusting the number of copies of different types of data packets, and realizing movement range oriented forwarding and dynamic multi-copies routing. The credible routing method aims to solve the problem that conventional methods are unsupportive to vehicular delay networking and is applicable to the conditions that starting and destination vehicles stay moving but routing cannot be determined in advance and information transmission is determined by relay vehicles skip by skip.

Description

A kind of credible routing method of vehicle-mounted time delay tolerant network

Technical field

The present invention relates to network route technology, the especially route technology of time delay tolerant network, specifically a kind of credible routing method of vehicle-mounted time delay tolerant network.

Background technology

In-vehicle networking is the wireless self-organization network that vehicle that a kind of dependence is provided with Wireless Telecom Equipment is realized transfer of data.In recent years, along with the development of wireless communication technology and vehicle GPS system, In-vehicle networking is widely used.For example: information of vehicle flowrate is broadcast to the vehicle on highway by In-vehicle networking, and driver can shift to an earlier date rerouting of road, avoids congested section; Traffic accident information is passed to nearest police office and first-aid centre, and accident can be processed timely; Near the information such as gas station, restaurant driver inquires about by In-vehicle networking, convenient trip.

By traditional MANET(mobile ad hoc network) Routing Protocol directly applies to In-vehicle networking, can not obtain gratifying performance.Its main cause is following 2 points: in (1) MANET, always between source and destination, seek a link transmission data that cost is minimum, the interruption of link is considered as to of short duration abnormal conditions.But in In-vehicle networking, the high mobility of vehicle causes the connection between vehicle often to be interrupted, and is difficult to ensure continual and steady connection.(2) in the case of the density of vehicle be lower, probably in isolated state, there is not the neighbours' vehicle that can communicate by letter in vehicle, causes abandoning of packet.

The intermittent connectivity that In-vehicle networking has and low-density make it more meet the time delay tolerant network feature of (delay-tolerant network is called for short DTN).Time delay tolerant network is a kind of existence between source node and destination node under the condition of end-to-end link, relies on storage forwarding asynchronous communication means to realize the wireless ad hoc network of exchanges data.

At present, the scheme that the thought that adopts time delay to tolerate both at home and abroad designs vehicle-mounted route mainly exists following several:

Scheme 1:Wu H, Fujimoto RM proposes MDDV agreement and Zhao J, the VADD Routing Protocol that Cao G proposes.The former utilizes vehicle to carry data to target area, and packet is spread to the vehicle in region.The latter is based on historical traffic flows information, and vehicle can be selected at crossing a suitable direction of transfer, and data are delivered to destination the most at last.But in these two kinds of methods, the destination of data or object region are all fixed, and cannot be applicable to the situation of mobile node.

Scheme 2:Huang HY, the research of Luo PE based on to Chinese Areas in Shanghai City taxi mobile alignment, introduces the concept of In-vehicle networking (SUVnet), and improves infecting route, proposes DAER Routing Protocol.But this scheme hypothesis node is known the current location of destination node, and this hypothesis cannot meet under certain condition.

Scheme 3:DingY, Wang C introduces the transmission of static infrastructure auxiliary data, in the time that vehicle finds that own moving direction deviates from the destination of packet, by package forward to fixation means, obtain packet continuation transmittance process by follow-up vehicle from fixation means.This method can ensure the accuracy that data are transmitted, but needs extra a large amount of infrastructure that increases.

Summary of the invention

The object of the invention is for the problems referred to above, a kind of credible routing method of vehicle-mounted time delay tolerant network is proposed, describe different mobility models are unified, and propose a kind of distortion and evaluate mechanism (MovementModel Similarity Evaluation Scheme, be called for short MMSE), forward dynamic many copies route (Movement Range Oriented Forwarding andDynamic Multi-copies Routing Protocol is called for short MROFDM) based on MMSE design is a kind of towards moving range.The method is utilized similarity between mobility model and the local real time information of vehicle, and data are forwarded in the moving range of destination node, adopts copy balance policy simultaneously, dynamically adjusts the copy number of different types of data bag.The present invention, compared with traditional many copies Routing Protocol, has good feasibility and adaptability under In-vehicle networking environment.

Technical scheme of the present invention is:

A kind of credible routing method of vehicle-mounted time delay tolerant network, this method is utilized similarity between mobility model and the local real time information of vehicle, data are forwarded to the moving range of destination node, adopt copy balance policy simultaneously, dynamically adjust the copy number of different types of data bag.

Of the present invention comprising the following steps:

The first step, source node adopts copy balance policy dynamically to increase copy number, adopts the value of the packet copy ex of following formula calculating source node increase, upgrades the value of the packet copy number bal of reduction simultaneously;

λ·ex=redu

bal=bal-ex≥0

Wherein, λ represents that node creates the speed of new uncertain UM packet; Redu represents packet reduction speed;

Second step, in the time that source node runs into arbitrary node at crossing, if what run into is destination node, directly completes the transmission of data;

The 3rd step, if what run into is non-destination node,, according to the similarity of the node that meets, carry out Data dissemination:

First, adopt following formula to calculate the source node M as mobility model ₁the node M of meeting with it ₂between similarity Sim (M ₁, M ₂);

Wherein, M ₁, M ₂represent two fixed model M ₁and M ₂, a represents M ₁, M ₂the same road segment of shiftable haulage line process, vertexs ₁represent source node M ₁the crossing number of process, vertexs ₂represent the node M of meeting with source node ₂the crossing number of process, rnd ₁represent source node M ₁random mobile probability, rnd ₂represent the node M of meeting with source node ₂random mobile probability;

Secondly, adopt following formula to calculate distribution number of copies;

Wherein, l represents the copy number of packet in source node model, l ₁represent the copy number that source node model oneself retains, Sim (Mn ₂, Mn _d) the adjacent nodal analysis method M of expression source node ₂with destination node model M _dsimilarity;

Use bal=bal+ (l-l ₁-l ₂) upgrade the value of bal; Use $\left\{\begin{array}{c}{\mathrm{redu}}_1=\frac{{\mathrm{\Δl}}_1}{t_1}\\ {\mathrm{redu}}_n={\mathrm{redu}}_{n-1}\·\frac{t_{n-1}}{t_{n-1}+t_n}+\frac{{\mathrm{\Δl}}_n}{{\mathrm{\Δt}}_n}\·\frac{t_n}{t_{n-1}+t_n}\end{array}\right.$ Upgrade reduction speed; Wherein Δ l _irepresent the copy number of the i time reduction, Δ t _irepresent the time of experience between the i-1 time reduction and the i time reduction, t _irepresent the timestamp of the i time reduction.

The 4th step, forwards.Use $\mathrm{MD}=\left\{\begin{array}{cc}\min \left\{\mathrm{dis}(v_d,v_i^m)\right\}& v_i^m\∈M.\mathrm{vertexs},\mathrm{type}\left(M\right)\≠\mathrm{UM}\\ \mathrm{undefined}& \mathrm{type}\left(M\right)=\mathrm{UM}\end{array}\right.$ Calculate destination v _dand the distance between mobility model.Wherein function is used for calculating destination node v _dand intermediate node between distance, MD represents v _dwith vertexs _mthe distance of nearest node in summit.

The movement that defines three class models and describe different vehicle in vehicle-mounted time delay tolerant network of the present invention, comprising:

Fixed model fixed model FM: vehicle moves according to fixing route;

Rule model regular model RM: certain statistical law is followed in the movement of vehicle; The place that vehicle often haunts and random mobile probability; The crossing that the former uses that vehicle often haunts is that fixed intersection represents, and the latter has reacted vehicle not at fixed intersection or the probability not moving to fixed intersection;

Ambiguous model uncertainty model UM: vehicle moves at random, does not possess stability factor.

Beneficial effect of the present invention:

The present invention has solved conventional method and has not supported the problem of vehicle-mounted time delay networking effectively, is applicable to source and destination vehicle all among motion, and routing procedure cannot pre-determine, by the situation of the transmission of relay vehicle hop-by-hop decision message.

Brief description of the drawings

Fig. 1 is the similarity schematic diagram between node motion model of the present invention.

Embodiment

Below in conjunction with drawings and Examples, the present invention is further illustrated.

As shown in Figure 1, a kind of credible routing method of vehicle-mounted time delay tolerant network, realizes towards moving range and forwards dynamic many copies vehicle-mounted time delay tolerant network routes (MROFDM) based on mobility model similarity evaluation mechanism (MMSE).

In the present invention, in In-vehicle networking, there are 3 kinds of important entities: vehicle, crossing and section, wherein the former is in the middle of moving, and then both are relatively fixing.Crossing and section have formed the static structure of network so, can be with figure G={V, and E} represents, and each crossing is a summit, i.e. and an element in V, if section connects two crossings, increases a limit so in E.In the process of travelling, there is not neighbours' vehicle in vehicle, only has two vehicles of working as to meet in the most of the time, message just may occur, and this meets the feature that time delay tolerant network chance connects.In most of the time, vehicle can carry message and move, and has increased the time delay that message is transmitted, and this embodies the essential characteristic of time delay tolerant network.But vehicle, as the node in network, is limited at motion on figure G, and this is that common time delay tolerant network does not have, and therefore the present invention is a special case that In-vehicle networking is considered as to time-delay network.

In the present invention, 3 class models are defined and have described the movement of different vehicle: (1) fixed model (fixed model FM), vehicle moves according to fixing route.The regular buses of such as public transport, enterprise etc. are stopped fixing website, meet FM mobility model.Use the crossing of process in mobile alignment to describe this mobility model.(2) rule model (regular model RM), certain statistical law is followed in the movement of vehicle, and for example people drive to travel to and fro between between family and unit.By this type of mobility model of 2 parametric descriptions: the place that vehicle often haunts and random mobile probability.The former uses the crossing (being called fixed intersection) that vehicle often haunts to represent, and the latter has reacted vehicle not at fixed intersection or the probability not moving to fixed intersection.(3) ambiguous model (uncertainty model UM), vehicle moves at random, does not possess stability factor.For example taxi determines destination according to passenger's demand, and each passenger's demand normally has nothing to do.In In-vehicle networking, mainly merge this 3 kinds of different mobility models, the unified mobility model that uses following structure description node of the present invention: M={vehicleid, rnd, vertexs[] }, wherein vehicleid has identified vehicle.Rnd represents the random mobile probability of vehicle, and FM model is fixed as 0, UM model and is fixed as 1, and RM model is actual random chance.Vertexs is an array, and vertexs[i] ∈ V, if FM model, represent so the crossing on the mobile alignment of vehicle, if RM model, represent the fixed intersection of vehicle, if UM model be sky.Can calculate the type of mobility model by rnd:

$\mathrm{type}\left(M\right)=\left\{\begin{array}{cc}\mathrm{FM}& \mathrm{RED}=0\\ \mathrm{RM}& 0<\mathrm{rnd}<l\\ \mathrm{UM}& \mathrm{rnd}=l\end{array}\right.$

Suppose mobility model M ₁={ vehicleid ₁, rnd ₁, vertexs ₁[] } and M ₂={ vehicleid ₂, rnd ₂, vertexs ₂[] }, and vehicleid ₁≠ vehicleid ₂, in the present invention, M ₁and M ₂there are following 5 kinds of relations.

(1) equivalence relation, two fixed models, follow same shiftable haulage line.

{type(M ₁)=type(M ₂)=FM,vertexs ₁=vertexs ₂}

For example vehicle on same public bus network.

(2) similarity relation, the shiftable haulage line of two fixed models is through identical section.

{type(M ₁)=type(M ₂)=FM,vertexs ₁[i]=vertexs ₂[j],

vertexs ₁[i+1]=vertexs ₂[j+1],···vertexs ₁[i+a]=vertexs ₂[j+a]a≥1,i≥0,j≥0}

For example two public bus networks have identical website, and the vehicle on these two circuits can be through one section of overlapping section so.

(3) independence, there is not common factor in fixed intersection or the shiftable haulage line of two rule models or a rule model and a fixed model.

{ type (M ₁)=type (M ₂)=RM or type (M ₁)=FM, type (M ₂)=RM or

type(M ₁)=RM,type(M ₂)=FM,vertexs ₁⌒vertexs ₂=φ}

For example two private vehicles haunt about different places, and their range of movement is relatively independent, therefore be called independence.

(4) parallel relation, there is not common factor in the shiftable haulage line of two fixed models.

{type(M ₁)=type(M ₂)=FM,vertexs ₁⌒vertexs ₂=φ}

For example two vehicles on two public bus networks that do not have a same site, they can not meet substantially, are similar to two parallel lines that converge at infinity in same plane, therefore be called parallel relation.

(5) default relationship, does not belong to above 4 kinds of situations, is all classified as default relationship.

In the present invention, the process of MROFDM is: source node is produced L copy and carried out cuttable distribution by copy balance policy after producing packet, and each copy is independently carried out repeating process afterwards, until one of them copy arrives destination node.

In the present invention, certain vertex v _ddistance MD between ∈ V and mobility model M, is described as:

$\mathrm{MD}=\left\{\begin{array}{cc}\min \left\{\mathrm{dis}(v_d,v_i^m)\right\}& v_i^m\&Element;M.\mathrm{vertexs},\mathrm{type}\left(M\right)\&NotEqual;\mathrm{UM}\\ \mathrm{undefined}& \mathrm{type}\left(M\right)=\mathrm{UM}\end{array}\right.$

Wherein dis (v ₁, v ₂) function is used for calculating vertex v ₁and v ₂between distance, MD represents v _dwith vertexs _mthe most subapical distance in summit.

In the present invention, repeating process is: as two node n ₁and n ₂after meeting, for n ₁packet p in buffering area _i, its destination node is n _d, copy number is the situation use copy balance policy that 1(copy number is greater than 1), if n ₂be more suitable for being p _ivia node, n ₁by package forward to n ₂, n ₂packet in buffering area is as the same.Therefore the process forwarding is real in finding the process of suitable via node.

In the present invention, use MMSE to evaluate respectively n _dwith n ₁mobility model and n _dwith n ₂the similarity of mobility model, holds p by the node that similarity is high _iif similarity is identical, according to the real time information of node, calculate respectively n ₁and n ₂current destination and n _dmobility model between distance, by distance short node hold p _i.If distance is undefined, continue to hold p by original node _i.

In the present invention, utilized respectively distortion and real-time destination forwarding strategy in package forward process, its object is just to allow packet send towards the moving range of destination node, therefore be referred to as the forwarding towards moving range.

In the present invention, according to the mobility model of packet rs destination node, packet is divided into 3 classes: FM packet, RM packet and UM packet.

In the present invention, copy balance policy is described as: in the distribution phase of copy, ensure under the prerequisite of transmission success rate, and the number of copies of reducing dynamically FM packet, and the number of copies of reduction is supplemented to the packet to UM.In the present invention, distribution procedure is described as: the L of a packet copy is passed to L-1 different node by source node, and oneself retains a copy, and each node refusal repeatedly receives the copy of same packet.As two node n ₁and n ₂after meeting, traditional copy distribution implementation is: for node n ₁packet p in buffering area _i, its destination node is n _d, copy number is l>1, and n ₂in buffering area, there is no p _icopy, n so ₁by l ₂individual copy sends to n ₂, oneself retains l ₁individual, need to determine l ₁with l ₂value, make l ₁+ l ₂=l, n ₂packet in buffering area is as the same.

In the present invention, if n ₂with n _dthe similarity of mobility model is greater than 0, and they have the higher overview of meeting so, can be by n ₂be considered as p _isuitable via node.At definite l ₁with l ₂when value, allow l ₁+ l ₂≤ l, realizes the object of dynamically reducing copy and ensureing transmission success rate.In the present invention, l ₁with l ₂value is described as:

In the present invention, this dynamic reduction distribution policy, in the time finding suitable via node, according to the similarity between via node and destination node mobility model, the dynamically number of reduction copy in the time not running into suitable via node, uses binary distribution policy to distribute fast copy simultaneously.Each node uses the number of copies of variable bal accumulative total reduction: bal=bal+ (l-l ₁-l ₂).

In the present invention, in the time that node creates new UM packet, if the value of its bal is greater than 0, can be suitably its supplementary ex copy, the value of simultaneously corresponding minimizing bal.The generation speed of node data bag is λ ₁, wherein UM packet accounts for γ, and the generation rate of UM packet is λ=λ so ₁γ.Suppose that dynamically reducing speed is redu, in order to keep the balance between reduction speed and the supplementary speed of copy, ex need to meet:

λ·ex=redu

bal=bal-ex≥0

In the present invention, node uses the mode of recursion to calculate redu:

$\left\{\begin{array}{c}{\mathrm{redu}}_1=\frac{{\mathrm{\&Delta;l}}_1}{t_1}\\ {\mathrm{redu}}_n={\mathrm{redu}}_{n-1}\&CenterDot;\frac{t_{n-1}}{t_{n-1}+t_n}+\frac{{\mathrm{\&Delta;l}}_n}{{\mathrm{\&Delta;t}}_n}\&CenterDot;\frac{t_n}{t_{n-1}+t_n}\end{array}\right.$

Wherein Δ l _irepresent the copy number of the i time reduction, Δ t _irepresent the time of experience between the i-1 time reduction and the i time reduction, t _irepresent the timestamp of the i time reduction.In the time calculating reduction speed, stab the weight of conduct service time, can ensure the real-time of estimation, take into account historical data simultaneously.Node is renewal reduction speed accumulative total bal value after each distribution, after each generation UM packet, calculates ex value, determines copy number and reduces bal value.

The part that the present invention does not relate to all prior art that maybe can adopt same as the prior art is realized.

Claims (3)

1. the credible routing method of a vehicle-mounted time delay tolerant network, it is characterized in that this method utilizes similarity between mobility model and the local real time information of vehicle, data are forwarded to the moving range of destination node, adopt copy balance policy simultaneously, dynamically adjust the copy number of different types of data bag.

2. the credible routing method of vehicle-mounted time delay tolerant network according to claim 1, is characterized in that it comprises the following steps:

The first step, source node adopts copy balance policy dynamically to increase copy number, adopts the value of the packet copy ex of following formula calculating source node increase, upgrades the value of the packet copy number bal of reduction simultaneously;

λ·ex=redu

bal=bal-ex≥0

Wherein, λ represents that node creates the speed of new uncertain UM packet; Redu represents packet reduction speed;

Second step, in the time that source node runs into arbitrary node at crossing, if what run into is destination node, directly completes the transmission of data;

The 3rd step, if what run into is non-destination node,, according to the similarity of the node that meets, carry out Data dissemination:

First, adopt following formula to calculate the source node M as mobility model ₁the node M of meeting with it ₂between similarity Sim (M ₁, M ₂);

Wherein, M ₁, M ₂represent two fixed model M ₁and M ₂, a represents M ₁, M ₂the same road segment of shiftable haulage line process, vertexs ₁represent source node M ₁the crossing number of process, vertexs ₂represent the node M of meeting with source node ₂the crossing number of process, rnd ₁represent source node M ₁random mobile probability, rnd ₂represent the node M of meeting with source node ₂random mobile probability;

Secondly, adopt following formula to calculate distribution number of copies;

Wherein, l represents the copy number of packet in source node model, l ₁represent the copy number that source node model oneself retains, Sim (Mn ₂, Mn _d) the adjacent nodal analysis method M of expression source node ₂with destination node model M _dsimilarity;

Use bal=bal+ (l-l ₁-l2) upgrade the value of dzl; Use $\left\{\begin{array}{c}{\mathrm{redu}}_1=\frac{{\mathrm{\&Delta;l}}_1}{t_1}\\ {\mathrm{redu}}_n={\mathrm{redu}}_{n-1}\&CenterDot;\frac{t_{n-1}}{t_{n-1}+t_n}+\frac{{\mathrm{\&Delta;l}}_n}{{\mathrm{\&Delta;t}}_n}\&CenterDot;\frac{t_n}{t_{n-1}+t_n}\end{array}\right.$ Upgrade reduction speed; Wherein Δ l ₁represent the copy number of the i time reduction, Δ t _irepresent the time of experience between the i-1 time reduction and the i time reduction, t _irepresent the timestamp of the i time reduction.

The 4th step, forwards.Use $\mathrm{MD}=\left\{\begin{array}{cc}\min \left\{\mathrm{dis}(v_d,v_i^m)\right\}& v_i^m\&Element;M.\mathrm{vertexs},\mathrm{type}\left(M\right)\&NotEqual;\mathrm{UM}\\ \mathrm{undefined}& \mathrm{type}\left(M\right)=\mathrm{UM}\end{array}\right.$ Calculate destination v _dand the distance between mobility model.Wherein function is used for calculating destination node v _dand intermediate node between distance, MD represents v _dwith vertexs _mthe distance of nearest node in summit.

3. the credible routing method of vehicle-mounted time delay tolerant network according to claim 2, is characterized in that defining three class models in vehicle-mounted time delay tolerant network and describes the movement of different vehicle, comprising:

Fixed model fixed model FM: vehicle moves according to fixing route;

Rule model regular model RM: certain statistical law is followed in the movement of vehicle; The place that vehicle often haunts and random mobile probability; The crossing that the former uses that vehicle often haunts is that fixed intersection represents, and the latter has reacted vehicle not at fixed intersection or the probability not moving to fixed intersection;

Ambiguous model uncertainty model UM: vehicle moves at random, does not possess stability factor.