CN103428747B - A kind of aviation self-organizing network wireless link stability prediction method - Google Patents

Abstract

The invention discloses a kind of aviation self-organizing network wireless link stability prediction method.First the method constructs aviation self-organized network nodes mobility model, and mobility model is made up of seven states: accelerate take off, constant speed risings, smooth flight, turning, constant speed decline, slow down stopping and static; Secondly according to the motion feature of above-mentioned mobility model interior joint in each state, the joint movements speed probability density function in each state is determined, and the probability density function of relative speed between node; Then on this basis, in conjunction with the distance between mobile node, and node direction of relative movement information, determine the distribution function of wireless link duration; The last stability factor drawing multi-hop wireless link according to link continuous time and its distribution function.Aircraft node can using the important evidence of the link stability factor as Route Selection, thus set up the high wireless transmission link of stability for aviation self-organizing network.Simulation results show in the QualNet simulated environment validity of the method.

Description

A kind of aviation self-organizing network wireless link stability prediction method

Technical field

The invention belongs to field of wireless, particularly aviation self-organizing network wireless link stability prediction method.

Background technology

Wireless self-organization network (WirelessAdhocNetworks) is a kind of novel wireless communication network not relying on static infrastructure, and it is made up of one group of mobile node with wireless transmitter, and networking fast and flexible, reliability are high.Each node in network is terminal and router, can forward the packet of other node in automatic network.Can with the mode quickly networking of multi-hop ad hoc under the procotol of layering controls after node start, after part of nodes breaks down or is quit work by destroying, the operation of whole network can not be affected, and thus network has very strong survivability and self-healing ability.

In recent years, wireless self-organization network has become a kind of important wireless networking mode, is all widely used in civilian and military field.Aviation self-organizing network (AeronauticalAdhocNetworks, AANET) is the product that radio self organizing network technology is applied in the multi-platform networking of aviation.At civil area, aviation self-organizing network replaces the broadband connections satellite of high cost, long time delay, becomes the new selection of Future broadband multimedia air communications.Researcher proposes, between the passenger plane of flight, build self-organizing network, and the mode that can forward with multi-hop sets up the communication link between aircraft and ground base station, for passenger provides aerial broadband the Internet access service that is cheap, low time delay on the road.

Aircraft node high-speed mobile (cruising speed is generally between 700km/h to 1000km/h) in aviation self-organizing network, network topology Rapid Variable Design, thus the internodal wireless transmission link of aircraft frequently ruptures.When any hop link in multi-hop transmission path ruptures, aircraft node then must start route finding process, finds the transmission path that other are available.A large amount of route finding process consumes network bandwidth resources greatly, causes network congestion and data-message transmission time delay to increase, and finally causes the degradation of network in general performance.Therefore, in the design of aviation self-organizing network Routing Protocol, the duration of inter-node wireless links must be considered, thus find the high wireless transmission link of stability by route discovery, alleviate the impact of network topology change on network in general performance.

As the above analysis, the prediction of aircraft inter-node wireless links stability is of great significance the overall performance tool promoting aviation self-organizing network.Existing wireless link stability prediction method roughly can be divided three classes: based on received signal power link stability Forecasting Methodology, move the link stability Forecasting Methodology of (geometric properties) and the link stability Forecasting Methodology based on probability Estimation theory based on node location.General principle based on the link stability Forecasting Methodology of received signal power is: when node sends packet, radio signal power is constantly decayed with the increase of transmission range in transmitting procedure.After neighbor node receives this signal, according to received signal power and the distance between path loss model estimation and sending node, and upgrade the link-state information between sending node according to this distance.The method simple, intuitive, but in actual environment, the propagation of wireless signal may be subject to the impact of the many factors such as air, sexual intercourse, complex-terrain, inevitably there is deviation in the distance between the sending node obtained by said method estimation and receiving node, and then has influence on final routing strategy.

The general principle predicted based on the link stability of node location movement is: according to movable informations such as existing node location, speed, directions, derives and obtains the information of link-state change.Link stability prediction based on node location movement is mainly divided into two kinds again.The first is based on link Duration Prediction (Predictionbased, PBR), and the Forecasting Methodology of this agreement to link existent time is as follows: given two mobile node i and j, and the distance defined between them is | d _ij|, corresponding translational speed is respectively v _iand v _j, communication distance is R, and defines s and be used for distinguishing motion or counter motion in the same way, and two internodal vertical width are w, and deriving thus obtains the predicted value of link duration between two mobile node i and j.This Forecasting Methodology only it is considered that the identical or contrary situation of movement direction of nodes, cannot solve two movement direction of nodes not on same straight line time problem.And this algorithm only it is considered that the linear uniform motion of node, therefore cannot solve the problems such as the acceleration of aviation self-organizing network interior joint, deceleration, turning.The second is the prediction (MultipathDopplerRoutingAlgorithm, MUDOR) based on many Pood frequency displacement, and Doppler frequency shift is the key factor that MUDOR agreement is considered.Pass between relative velocity and Doppler frequency shift is v=c (flf ₀-1), wherein c is the light velocity, and f is expected frequency, f ₀it is observed frequency.When expected frequency is less than observing frequency, for reflecting that the doppler values (DopplerValue, DV) of link load is defined as-c (flf ₀, otherwise be defined as+2c (flf-1) ₀-1).Doppler values is less, and link stability is better.Whether MUDOR algorithm neither needs the support of GPS device also reliable without the need to considering the attenuation model of Received signal strength.But, owing to only considering the impact of relative speed, and have ignored link establishment moment internodal initial distance completely.Thus, it can only carry out qualitative forecasting to link property, and cannot obtain the predicted value of link duration accurately.

Link stability Forecasting Methodology based on probability Estimation theory is: for any two mobile nodes 1 and 2, the communication distance of node 1 is divided into n section, and every section represents a state S respectively _i, writ state S _n+1represent absorbing state, the region beyond representation node 1 communication range, p _ijrepresent that node 2 is by state S _istate S is moved to after the unit interval _jprobability, define a step transition probability matrix P=|p _ij|, i, j ∈ [1, n+1]. represent that communication initial time node 2 is in state S _iprobability, the initial distance distribution matrix of defined node 2 i ∈ [1, n+1], can draw the probability density function of linkage availability and link duration thus, analyze link stability accordingly.Because in the method, a step transition probability matrix is changeless, therefore, the link stability that the method analysis obtains only distributes relevant with node initial distance, and has nothing to do with the initial relative speed of node.

Summary of the invention

The object of the invention is to propose a kind of wireless link stability prediction method being applicable to aviation self-organizing network, for Design of Routing Protocol provides important evidence, alleviate the impact of network topology change on network in general performance.In order to realize this object, step of the present invention is:

Step 1: build aviation self-organized network nodes mobility model, mobility model is made up of seven states: accelerate to take off, constant speed risings, smooth flight, turning, constant speed decline, the stopping and static of slowing down;

Step 2: according to the motion feature of above-mentioned mobility model interior joint in each state, determines the joint movements speed probability density function in each state;

Step 3: according to the speed probability density function of node in each motion state, determine the probability density function of relative speed between node;

Step 4: according to the probability density function of relative speed between node, the distance between node, and node direction of relative movement, determine the distribution function of link duration;

Step 5: calculate multi-hop link stability factor according to link continuous time and its distribution function.

The wireless link stability prediction method of the aviation self-organizing network that the present invention proposes realizes in QualNet4.5 network simulation environment.Simulating area is 10000 × 10000 × 1000m ³, network node adds up to 100, in simulating area random distribution.Physical layer adopts DSSS model, and simulation time is 300s, and all the other simulation parameters are as shown in table 1.Relevant parameter value in node motion model is as shown in table 2

Table 1 simulation parameter

Relevant parameter value in table 2 node motion model

Fig. 4 shows the theoretical value of link continuous time and its distribution function and the contrast of simulation value that the present invention calculates.The consistency of simulation value and theoretical value describes the validity of determination inter-node link continuous time and its distribution functional based method of the present invention.Accompanying drawing 5 and accompanying drawing 6 give the contrast of the network performance simulation result before and after the link stability Forecasting Methodology introducing the present invention's proposition in AODV (adhocon-demanddistancevector) Routing Protocol.Simulation results show, the link stability Forecasting Methodology introducing the present invention's proposition in Routing Protocol effectively can provide the saturation throughput performance of aviation self-organizing network, and reduces Packet Delay simultaneously.

Accompanying drawing explanation

Fig. 1 is aviation self-organized network nodes mobility model state transition graph;

Fig. 2 is node relative velocity schematic diagram;

Fig. 3 is that node link continuous time and its distribution function calculates schematic diagram;

Fig. 4 is theoretical value and the simulation value comparison diagram of the link continuous time and its distribution that the present invention calculates;

Fig. 5 be introduce in AODV Routing Protocol the present invention propose link stability Forecasting Methodology before and after network throughput performance simulation comparison diagram;

Fig. 6 be introduce in AODV Routing Protocol the present invention propose link stability Forecasting Methodology before and after network delay performance simulation comparison diagram.

Embodiment

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

The aviation self-organizing network link stability Forecasting Methodology that the present invention proposes realizes in Wireless Network Simulation environment QualNet4.5.

Provide specific embodiment of the invention step below:

Step 1: build aviation self-organized network nodes mobility model.

As shown in Figure 1, this mobility model is divided into seven states: accelerate take off, constant speed risings, smooth flight, turning, constant speed decline, slow down stopping and static.Aircraft node, from accelerating takeoff phase, after speed accelerates to set-point, enters constant speed ascent stage.After rising to regulation flying height, aircraft will start cruising flight, i.e. smooth flight.Awing, if meet with special circumstances, as prominent chance thunder and lightning weather, aircraft needs to revise course line, and now, aircraft will enter the turning stage.After turning terminates, regular shipping lines got back to by aircraft, i.e. smooth flight state.When being about to arrive destination, aircraft enters the constant speed decline stage.When dropping to ground proximity, aircraft enters deceleration stop phase, slows down until finally stop gradually, enters quiescent phase.

The design parameter of each motion state is described as follows:

1, takeoff phase, i.e. α stage is accelerated

Targeted rate v _α∈ U [v _{α, min}, v _{α, max}], continue timeslot number α ∈ U [α _min, α _max], horizontal movement direction vertical motion direction θ _α=pi/2.Wherein U represents and is uniformly distributed.Therefore the acceleration a in this stage _αfor

$a_{\mathrm{\α}}=\frac{v_{\mathrm{\α}}-v_{t_0}}{t_{\mathrm{\α}}-t_0}=\frac{v_{\mathrm{\α}}}{\mathrm{\α}\mathrm{\Δ}t}---\left(1\right)$

Δ t represents time step, the whole period of motion with Δ t for time slot divides.

2, constant speed ascent stage, i.e. β stage

Accelerate after takeoff phase terminates, aircraft starts to climb at a certain angle, speed v _β=v _α, continue timeslot number β ∈ U [β _min, β _max], horizontal movement direction vertical motion direction θ _β∈ U [0, pi/2].

3, smooth flight, i.e. γ stage

Cruising phase in this step simulations airplane motion, aircraft keeps flying speed constant, flies at a constant speed along preferential direction level.V _γ=v _β, continue timeslot number γ ∈ U [γ _min, γ _max], θ _γ=pi/2.

4, turn, i.e. the ζ stage

In flight course, may need to change original course line because of certain reason, at this moment aircraft will enter the turning stage, and its turning can be approximately uniform circular motion.Before aircraft turn, according to random selected horizontal movement direction with vertical motion direction θ _ζ, make uniform circular motion, in this stage, the speed v of aircraft _ζ=v _γ.Turning phase lasts timeslot number ζ ∈ U [ζ _min, ζ _max].After the γ stage terminates, aircraft may enter the turning stage, also may enter the constant speed decline η stage because being about to arrive destination.Assuming that after each smooth flight terminates, the probability that aircraft enters the turning stage is P _γ-ζ, the probability entering the constant speed decline stage is P _γ-η, obvious P _γ-ζ+ P _γ-η=1.Suppose that aircraft this process before fall-retarding will perform K time, K=0,1,2 ...In this case, the γ stage performs K+1 time, and the ζ stage performs K time.The desired value of K is as follows

$E\[K\]=\underset{K=0}{\overset{\mathrm{\∞}}{\Σ}}K{\left(P_{\mathrm{\γ}-\mathrm{\ξ}}\right)}^K(1-P_{\mathrm{\γ}-\mathrm{\ξ}})=\frac{P_{\mathrm{\γ}-\mathrm{\ξ}}}{1-P_{\mathrm{\γ}-\mathrm{\ξ}}}---\left(2\right)$

5, constant speed declines, i.e. the η stage

When being about to arrive destination, enter this stage.In this stage, aircraft declines with fixed speed uniform rectilinear, and initial velocity magnitude and direction are respectively: v _η=v _γ, θ _η∈ U [pi/2, π].Continue timeslot number η ∈ U [η _min, η _max].

6, deceleration stops, i.e. the d stage

After arriving ground, aircraft will be decelerated to stopping, the motion in this stage can be regarded as uniformly retarded motion.Initial velocity magnitude and direction are respectively: v _d=v _η, θ _d=pi/2.Duration d ∈ U [d _min, d _max], can obtain acceleration is accordingly

$a_d=\frac{0-v_{\mathrm{\η}}}{d\mathrm{\Δ}t}=-\frac{v_{\mathrm{\η}}}{d\mathrm{\Δ}t}---\left(3\right)$

7, static, i.e. the p stage

The duration p ∈ U [p that aircraft remains static _min, p _max], then start flight next time.

Step 2: the probability density function determining each stage speed of node.

To decline this four-stage for constant speed rising, smooth flight, turning and constant speed, because joint movements speed all remains unchanged, therefore joint movements rate distribution should meet and is uniformly distributed, as shown in the formula

$f\left(v\right)=\left\{\begin{array}{cc}\frac{1}{v_{\max }-v_{\min }}& v\∈(v_{\min },v_{\max })\\ 0& \mathrm{else}\end{array}\right.---\left(4\right)$

For accelerating takeoff phase, node is with initial velocity v ₀=0, accelerate to given target velocity v _α, v _αevenly choose from [v _min, v _max].Meanwhile, due to v _α∈ [v _min, v _max], therefore no matter v _αvalue is how many, v ∈ [0, v _min] value acceleration takeoff phase will occur with probability 1, due to even acceleration, so their probability density is also identical, be set to k.And v ∈ [v _min, v _max] can occur in this stage will by v _αvalue determine.V _αvalue is in v _minwith v _minprobability between+Δ v is Δ v/ (v _max-v _min), then accelerate in take-off process, speed v comprises v _minthe probability of+Δ v is 1-Δ v/ (v _max-v _min).According to above analysis, can obtain accelerating takeoff phase, the probability density function of joint movements speed is as follows

$f_{\mathrm{\&alpha;}}\left(v\right)=\left\{\begin{array}{cc}k,& 0\&le;v\&le;v_{\min }\\ k(1-\frac{v-v_{\min }}{v_{\max }-v_{\min }})& v_{\min }<v\&le;v_{\max }\end{array}\right.---\left(5\right)$

Due to

${\&Integral;}_{-\&infin;}^{+\&infin;}{f}_{\mathrm{\&alpha;}}\left(v\right)\mathrm{dv}=1---\left(6\right)$

Can obtain

$k=\frac{2}{(v_{\max }+v_{\min })}---\left(7\right)$

Thus formula (5) can be expressed as

$f_{\mathrm{\&alpha;}}\left(v\right)=\left\{\begin{array}{cc}\frac{2}{v_{\max }+v_{\min }},& 0\&le;v\&le;v_{\min }\\ \frac{2}{v_{\max }+v_{\min }}(1-\frac{v-v_{\min }}{v_{\max }-v_{\min }}),& v_{\min }<v\&le;v_{\max }\end{array}\right.---\left(8\right)$

Equally, for deceleration stop phase, also the conclusion identical with formula (8) can be obtained according to the analysis be similar to above.

Step 3: the probability density function determining relative speed between aircraft node.

Suppose that the speed of optional two mobile nodes is respectively v ₁, v ₂, as shown in Figure 2, then the relative velocity v '=v of these two mobile nodes ₁-v ₂, angle ω ∈ [0, π], if v ₁, v ₂, v ' mould be respectively v ₁, v ₂, v ', then have

$\left\{\begin{array}{c}{v}^{\&prime;}=\sqrt{{v_1}^2+{v_2}^2-{2v}_1{v}_2\cos w}\\ w=\arccos \frac{{v_1}^2+{v_2}^2-v^{\&prime;2}}{{2v}_1{v}_2}\end{array}\right.---\left(9\right)$

Due to v ₁, v ₂, ω is separate, so, v ₁, v ₂, ω joint probability density function can be written as

$f_{w,v_1,v_2}(w,v_1,v_2)=\frac{1}{\mathrm{\&pi;}}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right)---\left(10\right)$

Converted by Jacobi, can v be obtained ₁, v ₂, v ' joint probability density function

$f_{v^{\&prime;},v_1,v_2}(v^{\&prime;},v_1,v_2)=f_{w,v_1,v_2}(w,v_1,v_2)\left|\frac{\&PartialD;w}{\&PartialD;v^{\&prime;}}\right|$

$=\frac{2v^{\&prime;}{f}_{w,v_1,v_2}(w,v_1,v_2)}{\sqrt{4{v_1}^2{v_2}^2-{({v_1}^2+{v_2}^2-v^{\&prime;2})}^2}}---\left(11\right)$

$=\frac{2v^{\&prime;}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right)}{\mathrm{\&pi;}\sqrt{4{v_1}^2{v_2}^2-{({v_1}^2+{v_2}^2-v^{\&prime;2})}^2}}$

Can show that the probability density function of relative speed v ' is by formula (11)

$f_{v^{\&prime;}}\left(v^{\&prime;}\right)={\&Integral;}_{v_{2,\min }}^{v_{2,\max }}{\&Integral;}_{v_{1,\min }}^{v_{1,\max }}\frac{2v^{\&prime;}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right)}{\mathrm{\&pi;}\sqrt{4{v_1}^2{v_2}^2-{({v_1}^2+{v_2}^2-v^{\&prime;2})}^2}}d{v}_{1}d{v}_2---\left(12\right)$

Definition status S set={ α, beta, gamma, ζ, η, d, p}.Then, according to the rate distribution of each state that formula (4) and (8) obtain, in conjunction with formula (12), the probability density function that can obtain the relative speed under any two states is

$f_{v^{\&prime;}}^{E,G}\left(v^{\&prime;}\right)={\&Integral;}_{v_{G\min }}^{v_{G\max }}{\&Integral;}_{v_{E\min }}^{v_{E\max }}\frac{2v^{\&prime;}f\left(v_E\right)f\left(v_G\right)}{\mathrm{\&pi;}\sqrt{4{v_E}^2{v_G}^2-{({v_E}^2+{v_G}^2-v^{\&prime;2})}^2}}d{v}_{E}d{v}_G---\left(13\right)$

Wherein, E, G all belong to S.

Definition p (E), p (G) is respectively the node moment and is in E, the probability in G stage, then the relative speed under aeronautical Ad hoc networks node motion model is distributed as

$f_{v^{\&prime;}}\left(v^{\&prime;}\right)=\underset{E\&Element;S}{\mathrm{\&Sigma;}}\underset{G\&Element;S}{\mathrm{\&Sigma;}}p\left(E\right)p\left(G\right)f_{v^{\&prime;}}^{E,G}\left(v^{\&prime;}\right)---\left(14\right)$

Step 4: the distribution function calculating the link duration.

As shown in Figure 3, consider the wireless link between node 1 and node 2, using node 2 as reference node, node 1 is d with the distance of node 2 ₀, the direction of two node speed of related movement v ' and d ₀between angle theta obey at interval [0, π] and be uniformly distributed, R is node-node transmission distance; Use geometric knowledge, obtain the distribution function of link duration

$F\left(t\right)=P(T\&le;t)=P(\frac{d_1}{v^{\&prime;}}\&le;t)=P(\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t)$

$\left(15\right)$

$=\underset{\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t}{\&Integral;\&Integral;\&Integral;}f(\mathrm{\&theta;},d_0,v^{\&prime;})\mathrm{d\&theta;}{\mathrm{dd}}_0{\mathrm{dv}}^{\&prime;}$

θ, d ₀, v ' tri-variablees are separate, therefore formula (15) can be expressed as

$F\left(t\right)=\underset{\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t}{\&Integral;\&Integral;\&Integral;}f\left(\mathrm{\&theta;}\right)f\left(d_0\right)f_{v^{\&prime;}}\left(v^{\&prime;}\right)\mathrm{d\&theta;}{\mathrm{dd}}_0{\mathrm{dv}}^{\&prime;}---\left(16\right)$

The probability density function of θ is

$f\left(\mathrm{\&theta;}\right)=\frac{1}{\mathrm{\&pi;}}---\left(17\right)$

By

$F\left(d_0\right)=P(d\&le;d_0)=\frac{\frac{4}{3}{{\mathrm{\&pi;d}}_0}^3}{\frac{4}{3}{\mathrm{\&pi;R}}^3}=\frac{{d_0}^3}{R^3}---\left(18\right)$

Can obtain

$f\left(d_0\right)=\frac{{{3d}_0}^2}{R^3}---\left(19\right)$

By qualifications

$\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t---\left(20\right)$

Can obtain

$\cos \mathrm{\&theta;}\&le;\frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{{2v}^{\&prime;}{\mathrm{td}}_0}---\left(21\right)$

Order

k=∫f(θ)dθ(22)

Then

$k=\left\{\begin{array}{cc}0& \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{{2v}^{\&prime;}{\mathrm{td}}_0}\&le;-1\\ \frac{\mathrm{\&pi;}-\arccos \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{{2v}^{\&prime;}{\mathrm{td}}_0}}{\mathrm{\&pi;}}& -1<\frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{{2v}^{\&prime;}{\mathrm{td}}_0}<1\\ 1& \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{{2v}^{\&prime;}{\mathrm{td}}_0}\&GreaterEqual;1\end{array}\right.---\left(23\right)$

Therefore, formula (16) can be expressed as

$F\left(t\right)={\&Integral;}_0^{{{2v}^{\&prime;}}_{\max }}{\&Integral;}_0^R\mathrm{kf}\left(d_0\right)f_{v^{\&prime;}}\left(v^{\&prime;}\right){\mathrm{dd}}_0{\mathrm{dv}}^{\&prime;}---\left(24\right)$

By formula (14), (17), (19), (23) substitute into formula (24), can try to achieve the expression formula of link continuous time and its distribution function F (t).

Step 5: calculate the link stability factor.

In order to predict the stability of inter-node wireless links, invention introduces the link stability factor and computational methods thereof.As shown in Figure 3, the present invention adjusts the distance according to the initial phase between node 1 and node 2 and initial velocity is derived and drawn two inter-node link duration T ₁₂

$T_{12}=\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}---\left(25\right)$

Then by T ₁₂compared with the link duration corresponding when being 0.9 with link continuous time and its distribution function F (t) value, obtain namely

${\hat{T}}_{12}=\frac{T_{12}}{T_{F\left(t\right)=0.9}}---\left(26\right)$

Wireless link stability factor between node 1 and node 2 is defined as by the present invention

$S_{12}=\min ({\hat{T}}_{12},1)---\left(27\right)$

For the multi-hop wireless link be made up of (N-1) bar one hop link, its wireless link stability factor is defined as the product of many one hop link stability factors, namely

S _1N=S ₁₂S ₂₃S ₃₄…S _(N-1)N(28)

Namely wireless link stability factor can be used for the stability state predicting current radio link.In actual applications, actual value according to motion state parameters each in aircraft mobility model determines F (t), then the stability factor of inter-node wireless links is calculated by formula (28), and it can be used as the important evidence of Route Selection, thus set up the high wireless transmission link of stability for aviation self-organized network nodes.

The content be not described in detail in the present patent application book belongs to the known prior art of professional and technical personnel in the field.

Claims (2)

1. an aviation self-organizing network wireless link stability prediction method, the step adopted is:

Step 1: build aviation self-organized network nodes mobility model, mobility model is made up of seven states: accelerate to take off, constant speed risings, smooth flight, turning, constant speed decline, the stopping and static of slowing down;

Step 2: according to the motion feature of above-mentioned mobility model interior joint in each state, determine the joint movements speed probability density function in each state, concrete grammar is:

To decline this four-stage for constant speed rising, smooth flight, turning and constant speed, because joint movements speed all remains unchanged, therefore joint movements rate distribution should meet and is uniformly distributed, as shown in the formula:

$f\left(v\right)=\left\{\begin{array}{cc}\frac{1}{v_{max}-v_{min}}& v\&Element;\left(v_{\min },v_{\max }\right)\\ 0& else\end{array}\right.---\left(1\right)$

Accelerating takeoff phase, node is with initial velocity v ₀=0, evenly accelerate to given target velocity v _α, v _αevenly choose from [v _min, v _max]; At deceleration stop phase, node is with initial velocity v _α, even retardation is to 0, v _αevenly choose from [v _min, v _max]; Therefore, at acceleration takeoff phase and deceleration stop phase, the probability density function of joint movements speed is:

$f_{\mathrm{\&alpha;}}\left(v\right)=\left\{\begin{array}{cc}\frac{2}{v_{\max }+v_{\min }},& 0\&le;v\&le;v_{min}\\ \frac{2}{v_{\max }+v_{\min }}\left(1-\frac{v-v_{\min }}{v_{\max }-v_{\min }}\right),& v_{\min }<v\&le;v_{\max }\end{array}\right.---\left(2\right)$

Each state interior joint movement rate probability density function can be obtained by said process;

Step 3: according to the speed probability density function of node in each motion state, determine the probability density function of relative speed between node, concrete grammar is:

Suppose that the speed of optional two mobile nodes is respectively v ₁, v ₂, then the relative velocity v '=v of these two mobile nodes ₁-v ₂, angle ω ∈ [0, π], if v ₁, v ₂, v ' mould be respectively v ₁, v ₂, v ', then have

$\left\{\begin{array}{c}{v}^{\&prime;}=\sqrt{{v_1}^2+{v_2}^2-2v_1{v}_2\cos w},\\ w=\arccos \frac{{v_1}^2+{v_2}^2-v^{\&prime;2}}{2v_1{v}_2}\&CenterDot;\end{array}\right.---\left(3\right)$

Due to v ₁, v ₂, ω is separate, so, v ₁, v ₂, ω joint probability density function can be written as

$f_{w,v_1,v_2}(w,v_1,v_2)=\frac{1}{\mathrm{\&pi;}}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right).---\left(4\right)$

Converted by Jacobi, can v be obtained ₁, v ₂, v ' joint probability density function

$\begin{array}{c}{f}_{v^{\&prime;},v_1,v_2}\left(v^{\&prime;},v_1,v_2\right)=f_{w,v_1,v_2}\left(w,v_1,v_2\right)\left|\frac{\&part;w}{\&part;v^{\&prime;}}\right|\\ =\frac{2v^{\&prime;}{f}_{w,v_1,v_2}\left(w,v_1,v_2\right)}{\sqrt{4{v_1}^2{v_2}^2-{\left({v_1}^2+{v_2}^2-v^{\&prime;2}\right)}^2}}\\ =\frac{2v^{\&prime;}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right)}{\mathrm{\&pi;}\sqrt{4{v_1}^2{v_2}^2-{\left({v_1}^2+{v_2}^2-v^{\&prime;2}\right)}^2}}.\end{array}---\left(5\right)$

Can show that the probability density function of relative speed v ' is by formula (5)

$f_{v^{\&prime;}}\left(v^{\&prime;}\right)={\&Integral;}_{v_{2,\min }}^{v_{2,\max }}{\&Integral;}_{v_{1,\min }}^{v_{1,\max }}\frac{2v^{\&prime;}{f}_{v_1}\left(v_1\right)f_{v_2}\left(v_2\right)}{\mathrm{\&pi;}\sqrt{4{v_1}^2{v_2}^2-{\left({v_1}^2+{v_2}^2-v^{\&prime;2}\right)}^2}}{\mathrm{dv}}_1{\mathrm{dv}}_2---\left(6\right)$

Definition status S set={ α, beta, gamma, ζ, η, d, p}; Then, according to the rate distribution of each state that formula (1) and (2) obtain, in conjunction with formula (6), the probability density function that can obtain the relative speed under any two states is

$f_{v^{\&prime;}}^{E,G}\left(v^{\&prime;}\right)={\&Integral;}_{v_{G\min }}^{v_{G\max }}{\&Integral;}_{v_{E\min }}^{v_{E\max }}\frac{2v^{\&prime;}f\left(v_E\right)f\left(v_G\right)}{\mathrm{\&pi;}\sqrt{4{v_E}^2{v_G}^2-{\left({v_E}^2+{v_G}^2-v^{\&prime;2}\right)}^2}}{\mathrm{dv}}_E{\mathrm{dv}}_G---\left(7\right)$

Wherein, E, G all belong to S;

Definition p (E), p (G) is respectively the node moment and is in E, the probability in G stage, then the distribution of node relative speed can be expressed as:

$f_{v^{\&prime;}}\left(v^{\&prime;}\right)=\underset{E\&Element;S}{\&Sigma;}\underset{G\&Element;S}{\&Sigma;}p\left(E\right)p\left(G\right)f_{v^{\&prime;}}^{E,G}\left(v^{\&prime;}\right)---\left(8\right)$

The probability density function of each state interior joint movement rate can be obtained by said process;

Step 4: according to the probability density function of relative speed between node, the distance between node, and node direction of relative movement, determine the distribution function of link duration, and concrete grammar is:

Consider the wireless link between node 1 and node 2, using node 2 as reference node, node 1 is d with the distance of node 2 ₀, the direction of two node speed of related movement v ' and d ₀between angle be θ, R be node-node transmission distance; Use geometric knowledge, obtain the distribution function of link duration

$\begin{array}{c}F\left(t\right)=P\left(T\&le;t\right)=P\left(\frac{d_1}{v^{\&prime;}}\&le;t\right)=P(\frac{d_{0}cos\mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t)\\ =\underset{\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t}{\&Integral;\&Integral;\&Integral;}f(\mathrm{\&theta;},d_0,v^{\&prime;}){\mathrm{d\&theta;dd}}_0{\mathrm{dv}}^{\&prime;}\end{array}---\left(9\right)$

θ, d ₀, v ' tri-variablees are separate, therefore formula (9) can be expressed as

$F\left(t\right)=\underset{\frac{d_0\cos \mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t}{\&Integral;\&Integral;\&Integral;}f\left(\mathrm{\&theta;}\right)f\left(d_0\right)f_{v^{\&prime;}}\left(v^{\&prime;}\right){\mathrm{d\&theta;dd}}_0{\mathrm{dv}}^{\&prime;}---\left(10\right)$

θ obeys at interval [0, π] and is uniformly distributed, and its probability density function is

$f\left(\mathrm{\&theta;}\right)=\frac{1}{\mathrm{\&pi;}}---\left(11\right)$

By

$F\left(d_0\right)=P(d\&le;d_0)=\frac{\frac{4}{3}{{\mathrm{\&pi;d}}_0}^3}{\frac{4}{3}{\mathrm{\&pi;R}}^3}=\frac{{d_0}^3}{R^3}---\left(12\right)$

Can obtain

$f\left(d_0\right)=\frac{3{d_0}^2}{R^3}---\left(13\right)$

By qualifications

$\frac{d_{0}cos\mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}\&le;t---\left(14\right)$

Can obtain

$cos\mathrm{\&theta;}\&le;\frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{2v^{\&prime;}{\mathrm{td}}_0}---\left(15\right)$

Order

k＝∫f(θ)dθ(16)

Then

$k=\left\{\begin{array}{cc}0& \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{2v^{\&prime;}{\mathrm{td}}_0}\&le;-1\\ \frac{\mathrm{\&pi;}-\arccos \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{2v^{\&prime;}{\mathrm{td}}_0}}{\mathrm{\&pi;}}& -1<\frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{2v^{\&prime;}{\mathrm{td}}_0}<1\\ 1& \frac{v^{\&prime;2}{t}^2+{d_0}^2-R^2}{2v^{\&prime;}{\mathrm{td}}_0}\&GreaterEqual;1\end{array}\right.---\left(17\right)$

Therefore, formula (10) can be expressed as

$F\left(t\right)={\&Integral;}_0^{2{v^{\&prime;}}_{\max }}{\&Integral;}_0^{R}kf\left(d_0\right)f_{v^{\&prime;}}\left(v^{\&prime;}\right){\mathrm{dd}}_0{\mathrm{dv}}^{\&prime;}---\left(18\right)$

By formula (8), (11), (13), (17) substitute into formula (18), can try to achieve the expression formula of link continuous time and its distribution function F (t);

Step 5: calculate multi-hop link stability factor according to link continuous time and its distribution function, concrete grammar is:

Consider the wireless link between node 1 and node 2, using node 2 as reference node, node 1 is d with the distance of node 2 ₀, the direction of two node speed of related movement v ' and d ₀between angle be θ, R be node-node transmission distance; The duration T of wireless link between node 1 and node 2 ₁₂for

$T_{12}=\frac{d_{0}cos\mathrm{\&theta;}+\sqrt{R^2-{d_0}^2{\sin }^2\mathrm{\&theta;}}}{v^{\&prime;}}---\left(19\right)$

By T ₁₂compared with the link duration corresponding when being 0.9 with link continuous time and its distribution function F (t) value, obtain namely

${\hat{T}}_{12}=\frac{T_2}{T_{F\left(t\right)=0.9}}---\left(20\right)$

Wireless link stability factor between node 1 and node 2 is defined as by the present invention

$S_{12}=min({\hat{T}}_{12},1)---\left(21\right)$

For the multi-hop wireless link be made up of (N-1) bar one hop link, its wireless link stability factor is defined as the product of many one hop link stability factors, namely

S _1N＝S ₁₂S ₂₃S ₃₄…S _(N-1)N(22)

Namely wireless link stability factor can be used for the stability state predicting current radio link, in actual applications, actual value according to motion state parameters each in aircraft mobility model determines F (t), then the stability factor of inter-node wireless links is calculated by formula (22), and it can be used as the important evidence of Route Selection, thus set up the high wireless transmission link of stability for aviation self-organized network nodes.

2. a kind of aviation self-organizing network wireless link stability prediction method according to claim 1, is characterized in that the concrete construction method of aviation self-organized network nodes mobility model is:

Aircraft node is from accelerating takeoff phase, and after speed accelerates to set-point, enter constant speed ascent stage, after rising to regulation flying height, aircraft will start cruising flight, i.e. smooth flight; Awing, if meet with special circumstances, as prominent chance thunder and lightning weather, aircraft needs to revise course line, and now, aircraft will enter the turning stage; After turning terminates, regular shipping lines got back to by aircraft, i.e. smooth flight state; When being about to arrive destination, aircraft enters the constant speed decline stage, and when dropping to ground proximity, aircraft enters deceleration stop phase, slows down until finally stop gradually, enters quiescent phase;

The design parameter of each motion state is described as follows:

(1) takeoff phase, i.e. α stage is accelerated

Targeted rate v _α∈ U [v _{α, min}, v _{α, max}], continue timeslot number α ∈ U [α _min, α _max], horizontal movement direction vertical motion direction θ _α=pi/2; Wherein U represents and is uniformly distributed; Therefore the acceleration a in this stage _αfor:

$a_{\mathrm{\&alpha;}}=\frac{v_{\mathrm{\&alpha;}}-v_{t_0}}{t_{\mathrm{\&alpha;}}-t_0}=\frac{v_{\mathrm{\&alpha;}}}{\mathrm{\&alpha;}\mathrm{\&Delta;}t}---\left(23\right)$

Δ t represents time step, the whole period of motion with Δ t for time slot divides;

(2) constant speed ascent stage, i.e. β stage

Accelerate after takeoff phase terminates, aircraft starts to climb at a certain angle, speed v _β=v _α, continue timeslot number β ∈ U [β _min, β _max], horizontal movement direction vertical motion direction θ _β∈ U [0, pi/2];

(3) smooth flight, i.e. γ stage

Cruising phase in this step simulations airplane motion, aircraft keeps flying speed constant, flies at a constant speed along preferential direction level; v _γ=ν _β, continue timeslot number γ ∈ U [γ _min, γ _max], θ _γ=pi/2;

(4) turn, i.e. the ζ stage

In flight course, may need to change original course line because of certain reason, at this moment aircraft will enter the turning stage, and its turning can be approximately uniform circular motion; Before aircraft turn, according to random selected horizontal movement direction with vertical motion direction θ _ζ, make uniform circular motion, in this stage, the speed v of aircraft _ζ=v _γ, turning phase lasts timeslot number ζ ∈ U [ζ _min, ζ _max]; After the γ stage terminates, aircraft may enter the turning stage, also may enter the constant speed decline η stage because being about to arrive destination; Assuming that after each smooth flight terminates, the probability that aircraft enters the turning stage is P _γ-ζ, the probability entering the constant speed decline stage is P _γ-η, obvious P _γ-ζ+ P _γ-η=1; Suppose that aircraft this process before fall-retarding will perform K time, K=0,1,2 ..., in this case, the γ stage performs K+1 time, and the ζ stage performs K time; The desired value of K is as follows:

$E\&lsqb;K\&rsqb;=\underset{K=0}{\overset{\mathrm{\&infin;}}{\&Sigma;}}K{\left(P_{\mathrm{\&gamma;}-\mathrm{\&xi;}}\right)}^K(1-P_{\mathrm{\&gamma;}-\mathrm{\&xi;}})=\frac{P_{\mathrm{\&gamma;}-\mathrm{\&xi;}}}{1-P_{\mathrm{\&gamma;}-\mathrm{\&xi;}}}---\left(24\right)$

(5) constant speed declines, i.e. the η stage

When being about to arrive destination, enter this stage; In this stage, aircraft declines with fixed speed uniform rectilinear, and initial velocity magnitude and direction are respectively: v _η=v _γ, θ _η∈ U [pi/2, π]; Continue timeslot number η ∈ U [η _min, η _max];

(6) deceleration stops, i.e. the d stage

After arriving ground, aircraft will be decelerated to stopping, the motion in this stage can be regarded as uniformly retarded motion; Initial velocity magnitude and direction are respectively: v _d=v _η, θ _d=pi/2; Duration d ∈ U [d _min, d _max], can obtain acceleration is accordingly

$a_d=\frac{0-v_{\mathrm{\&eta;}}}{d\mathrm{\&Delta;}t}=-\frac{v_{\mathrm{\&eta;}}}{d\mathrm{\&Delta;}t}---\left(25\right)$

(7) static, i.e. the p stage

The duration p ∈ U [p that aircraft remains static _min, p _max], then start flight next time.