CN104113855A - Channel-based routing algorithm of wireless self-organizing network - Google Patents

Abstract

The invention relates to a channel-based routing algorithm of a wireless self-organizing network and discloses an optimized link state routing protocol based on physical channel characteristics. The channel-based routing algorithm of the wireless self-organizing network is characterized in that a processing mechanism for quality change of physical links is added into a route setup and selection algorithm. Demonstrated by taking an OLSR (optimized link state routing) protocol as for instance, the channel-based routing algorithm of the wireless self-organizing network can greatly improve the packet arrival rate and the packet delay under a scene of rapid node movement and frequent topological changes.

Description

The routing algorithm of wireless self-organization network based on channel

Technical field

The present invention relates to the routing algorithm of wireless self-organization network based on channel, belong to wireless sensor network field.

Background technology

Wireless self-organization network is comprised of one group of mobile node with radio communication R-T unit, is the provisional without central site network of a multi-hop.Node in network is terminal, is again router, and the inter-node communication in coverage each other can not complete through the forwarding of intermediate node, can be at any time, any place rapid build plays a mobile communications network.This network can work alone, and also can be connected with Internet or cellular radio.Wireless self-organization network is the vertiginous network of a kind of topological structure of multi-hop, need the support of dynamic routing protocol, Routing Protocol must be able to be got caught up in the rapid variation of the network topology structure that node motion brings, and efficient, the dynamic Routing Protocol of design becomes a challenge of programming wireless self-organizing network.

The dynamic topology of wireless self-organization network, limited transmission bandwidth, the node energy is limited and have the features such as one way link.Brought a lot of problems to the design of Routing Protocol.

(1) convergence of topological Dynamic Effect routing algorithm:

In wireless self-organization network, the movement of node has very large randomness, adds that wireless launcher transmit power variation, wireless channel influence each other and the factor such as terrain environment, and network topology structure cannot be predicted.The existence of these factors has proposed challenge to the constringency performance of the routing algorithm of wireless self-organization network, requires Routing Protocol to have quick reaction capability to change in topology, thereby can Fast Convergent guarantees that all nodes grasp up-to-date network topological information.The algorithm that convergence rate is slow, may cause route break or bust this.

(2) complexity of the limited constraint-based routing algorithm of node energy and expense:

Because the node energy of wireless self-organization network is mainly derived from battery supplied, over-burden can cause energy to exhaust too early for node, the logout ahead of time of having to.Therefore, the designing requirement of routing algorithm is simple, effective, when improving network throughput, control the expense of Routing Protocol.

(3) route loop and one way link:

Loop free is to guarantee the correct basic demand of Routing Protocol, because loop can cause routing error, not only has a strong impact on the performance of Routing Protocol, and wastes a large amount of bandwidth resources.Due to factors such as landform or transmitting powers, in network, may have one way link, and one way link easily brings the impacts such as perception one-way, route one-way and meeting point be unreachable to Routing Protocol, so the support of one way link is also the problem that routing Design faces.

(4) safety issue: nodes is random mobile has brought very large difficulty, the distributed nature of wireless self-organization network Routing Protocol to require to realize in net aspect the demands for security such as confidentiality, integrality, availability of service and safety certification to key management and AUC definite needs special consideration.The Routing Protocol draft of MANET working group issue is at present all to add that nodes has friendly, there is not malicious node, and actual node needs commutative Topology information when setting up route, this information exchanging process often becomes the target of attack of malicious node.

Through constantly research, the Routing Protocol of many designs of the characteristic study for self-organizing network is suggested in succession, and these Routing Protocols have this different mode classification according to different standards.According to the normal form classification of Route establishment, can be divided into table drives route and drives as required route.Show to drive Routing Protocol conventionally to adopt periodicity to send route control packet and carry out exchanging routing information, common agreement has DSDV, DYMO, OLSR etc.; And drive as required Routing Protocol, be to carry out route discovery according to send route requests grouping according to demand, set up transmission path, thereby the information that realizes transmits, yet safeguard by route maintenance and renewal process the route of having set up, as agreements such as AODV, DSR.

Being no matter that table drives Routing Protocol, still driving as required Routing Protocol, also or hybrid routing protocol, is mainly all the thought based on hierarchical design, and the design of route layer and MAC and physical layer design are separated from each other.This design philosophy, beneficial for the portability of Routing Protocol, but but make Routing Protocol sensitive not for the reaction of physical link mass change.When physical link quality changes, the hardware device of physical layer can perceive the variation tendency of link-quality the very first time, as the enhancing of signal to noise ratio with weaken.If Routing Protocol can be known from physical layer the variation of physical link, thereby when routing, avoid the link of mass attenuation, select quality link good and stable or that quality is constantly improved, just can between source node and destination node, set up and safeguard the healthy and strong route with long life span, to reduce the impact of network topology frequent variations on network performance.

In view of this, the inventor, for the defect further investigation of prior art, has this case to produce then.

Summary of the invention

Technical problem to be solved by this invention is to provide the routing algorithm of a kind of wireless self-organization network based on channel, at packet arrival rate and packet delay, has larger performance to improve.

The present invention is achieved in that

The routing algorithm of wireless self-organization network based on channel, comprises the steps:

First step: according to wireless link signal to noise ratio, under the channel model of supposition, set up link-quality and Stability Model; Wherein related definition is as follows:

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the link-quality of this node and a hop neighbor node can be designated as LinkQty (e[i, j], t), wherein LinkQty (e[i, j], t)=1-P _f(e[i, j], t),

$P_f\left(E\right)=1-{(1-\left(\frac{M-1}{M}\right)\{1-\sqrt{\frac{{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}{1+{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}}\left[\frac{M}{(M-1)\mathrm{\π}}\right][\frac{\mathrm{\π}}{2}+{\tan }^{-1}(1-\sqrt{\frac{{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}{1+{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}})\cot \left(\frac{\mathrm{\π}}{M}\right)]\left\}\right)}^n;$

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the stability of the link-quality of this node and a hop neighbor node be designated as LinkSty (e[i, j], t _k+1), wherein:

$\mathrm{LinkSty}\left(e\right[i,j],t_{k+1})=\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{[\mathrm{LinkQty}\left(e\right[i,j],t_i)-\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkQty}\left(e\right[i,j],t_i)]}^2;$

Make the directed edge of x representative from this node to this hop neighbor node, this node is designated as P with the stability forecast of the link in short-term probability of a hop neighbor node _u(x, t), wherein P _u(x, t)=Probability{1-P _f(x, t) > α };

The MPR that supposes to take computing node S is example, and a hop neighbor node set of S is designated as N (S), and two-hop neighbor node set is designated as N ₂(S), suppose that S has a hop neighbor node W ∈ N (S), at the two-hop neighbor node set N of S ₂(S) exist k node to be connected with node W, remember that these nodes are G _i(k>=i>=1), note W and G _ilink-quality between (k>=i>=1) is LinkQty (e[W, G _i], t), link-quality stability is LinkSty (e[W, G _i], t), that link stability forecast probability is P in short-term _u(e[W, G _i], t), wherein i is from 1 to k value, a hop neighbor node W with by the two-hop neighbor node G about node S of its covering _ilink-quality LinkQty between (k>=i>=1) _2-hopby following formula, provided:

${\mathrm{LinkQty}}_{2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkQty}\left(e\right[W,G_i],t)}{k};$

Link-quality stability LinkSty _2-hopby following formula, provided:

${\mathrm{LinkSty}}_{2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkSty}\left(e\right[W,G_i],t)}{k};$

Link probability of stability P in short-term _u2-hop(S _w, t) be following formula:

$P_{U2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{P}_U\left(e\right[W,G_i],t)}{k};$

Link-quality between W and S is designated as: LinkQty (e[S, W], t), node W is following formula about relative neighbours' link-quality of S:

RNLQty _W,S＝β·LinkQty _2-hop(S _W,t)+(1-β)LinkQty(e[S,W],t)，

Wherein β is the relevant optional parameters of concrete application;

Link-quality stability between W and S is designated as: LinkSty (e[S, W], t), node W is provided by following formula about relative neighbours' link-quality stability of S:

RNLSty _W,S＝η·LinkSty _2-hop(S _W,t)+(1-η)LinkSty(e[S,W],t)，

Equally, η is also the relevant optional parameters of concrete application;

The stability forecast of link in short-term probability between W and S is designated as: P _u(e[S, W], t), node W about the relative neighbours of S in short-term link stability forecast probability by following formula, provided:

RNLPty _W,S＝P _U2-hop(S _W,t)·P _U(e[S,W],t)

Node W is provided by following formula about the comprehensive coverage quality of the combination coverage degree of S:

${\mathrm{CoverQty}}_{W,S}=\mathrm{\γ}\·\frac{{\mathrm{RNLQty}}_{W,S}}{\sqrt{{\mathrm{RNLSty}}_{W,S}}+{\mathrm{RNLQty}}_{W,S}}+\mathrm{\τ}\·{\mathrm{RNLPty}}_{W,S}+(1-\mathrm{\γ}-\mathrm{\τ})\·C_W,$

γ wherein, τ is according to the selected non-negative parameter of concrete application, need meet constraint: 0≤γ+τ≤1;

Second step: the data structure expansion link information field to Routing Protocol control packet;

Third step: be the corresponding information table of Memory Extension of link information;

The 4th step: junction link information, transformation MPR selection algorithm and routed path Algorithms of Selecting.

Described MPR selection algorithm, specifically comprises:

The MPR set of first initialization node S, is set to empty set by MPR (S); If a two-hop neighbor node Z of S can only arrive S by a hop neighbor node Y of S, node Y is added in MPR (S); In N (S), the node in MPR (S) has been added in deletion to, is designated as N'(S); N ₂(S) in, deletion, by the node of coverage in MPR (S), is designated as N ₂' (S); N ₂if ' (S) empty set, algorithm finishes, otherwise continues;

According to

${\mathrm{CoverQty}}_{W,S}=\mathrm{\γ}\·\frac{{\mathrm{PNLQty}}_{W,S}}{\sqrt{{\mathrm{PNLSty}}_{W,S}}+{\mathrm{PNLQty}}_{W,S}}+\mathrm{\τ}\·{\mathrm{PNLPty}}_{W,S}+(1-\mathrm{\γ}-\mathrm{\τ})\·C_W$

Calculate N'(S) in each node be present in N ₂' two-hop neighbor node (S) is with respect to the comprehensive coverage quality CoverQty of joint S _w,S, and by its descending; The node that selection has maximum comprehensive coverage quality adds in MPR (S); The two-hop neighbor node of S is all covered, and algorithm finishes; Otherwise forward deletion in N (S) to, added the node in MPR (S) to, be designated as N'(S) this step.

Described routed path Algorithms of Selecting, specifically comprises:

First empty the routing table RTable (S) of node S; According to node S mono-hop neighbor node table, add route table items, if a hop neighbor node is not added routing table list item, add a route table items, destination address and next hop address that new list item is set are this hop neighbor node address, i=1 is set, distance is 1, and route quality is set to node S to the path quality of this node; I=2 is set, adds the i jumping route table items that arrives S, until all nodes that S knows all add in routing table entry:

If 1. do not exist in routing table, arrive the route performance of this destination address, add the route table items of new this destination address of arrival.The destination address that the destination address of new route table items is this topology list item is set; Next hop address is set to the next hop address of route table items corresponding to final jump address in topological list item; Apart from jumping figure, be set to i+1; By above formula, calculate the route quality in this path;

If 2. existed in routing table, arrive the list item of this destination node, according to above formula, calculate the route quality of new route, if the route quality of new route is greater than the route quality in the path of current existence, according to 1. method, add new route table items.

The route quality computing formula in described path is:

Wherein, parameter ξ and be according to the selected non-negative parameter of concrete application, need to satisfy condition:

The invention has the advantages that: the present invention introduces link quality metric mechanism and the stability evaluating mechanism based on link-quality time series variance based on characteristics of radio channels, improve MPR selection algorithm and the route selection algorithm of legacy protocol, improved the performance under the complex topology scene of node rapid movement.Simultaneously, the present invention adopts the link stability forecasting mechanism based on birth and death process, in " historical recently " data of collecting, link is predicted at the stability probability in the follow-up moment, suppress to a certain extent the loss bringing to protocol capabilities due to routing iinformation hysteresis quality.And, in the present invention, if arrive the route of destination node, exist, this algorithm can obtain in finite time; If arrive the route of destination node, do not exist, algorithm also can stop in finite time.Therefore, in routing algorithm of the present invention, do not exist and produce loop and be absorbed in the problem of count-to-infinity.

Accompanying drawing explanation

The invention will be further described in conjunction with the embodiments with reference to the accompanying drawings.

Fig. 1 is the realization mechanism schematic diagram that the present invention is applied to CB-OLSR agreement.

Fig. 2 is MPR election algorithm flow chart in the present invention.

Fig. 3 is routing table calculation flow chart in the present invention.

Embodiment

In prior art, less Routing Protocol adopts the cross-layer design philosophy with the direct combination of physical layer, and associated research work is also fewer.The present invention is directed to the Routing Protocol of described wireless self-organization network faces change in topology, route break problem above, by the research to node motion and channel time-varying characteristics, analyzing on the basis of more existing Routing Protocol, propose and improve relevant algorithm, improve the adaptive capacity of Routing Protocol to quick variation network topology, to advance the application of wireless self-organization network in the communications field, invention achievement has certain theory significance and actual using value.OLSR route selection agreement take below as example explanation the present invention.

OLSR agreement stems from the improvement to pure link-state routing algorithm, relies on the routing iinformation of constantly safeguarding to keep consistent with full mesh topology structure.Multiple spot relaying MPR is the crucial optimisation technique of OLSR agreement.MPR provides selectivity flooding mechanism, by reducing the repeated broadcast number of times of identical control packet in the same area, suppresses the broadcast storm that flooded packets causes.Rely on MPR technology, OLSR agreement can obviously reduce the control overhead of Routing Protocol than ospf protocol.Therefore, OLSR agreement is specially adapted to that network dimensioning is large, the densely distributed network of node, but when the node in networking is less, agreement just deteriorates to pure link-state routing protocol.

The defect of OLSR agreement is, MPR select and route computational process in only consider the coverage of node as the selection principle of MPR, take simple path jumping figure as path selection principle, ignore under the complex topology scene of node rapid movement the factors such as link unsteadiness, thereby caused data packet transfer failure.Therefore the improvement of traditional OLSR agreement is had to its necessity.

Due to the topological frequent variations of self-organizing network, node maintenance and obtain that link-state information is very difficult accurately, and inaccurate link-state information often causes routing failure.Based on discussing before, current various optimization and improvement strategy for OLSR agreement, mainly pay close attention to factor more than MAC layer, as the information such as motion of node, and less the most direct index---the characteristic of channel of paying close attention to reflection link-quality.The present invention is exactly the characteristic of channel from physical layer, and object is to find as early as possible the variation of link-quality, according to link quality information, carries out cross-layer design.Fig. 1 has provided the realization mechanism of the CB-OLSR agreement adding after thought of the present invention.

First, by the average signal-to-noise ratio of physical layer monitoring wireless link, according to average signal-to-noise ratio, under the channel model of supposition, calculate the frame error rate P of link _f(E) and using it as link-quality, set up radio link quality model.Now link no longer only has connected sum to disconnect two states but with probability (1-P _f(E)) keep being communicated with, in Route Selection, select link between node to be communicated with the high and stable path of probability as the transmission path of packet.Frame error rate P _f(E) can be calculated by following formula:

$P_f\left(E\right)=1-{(1-\left(\frac{M-1}{M}\right)\{1-\sqrt{\frac{{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}{1+{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}}\left[\frac{M}{(M-1)\mathrm{\π}}\right][\frac{\mathrm{\π}}{2}+{\tan }^{-1}(1-\sqrt{\frac{{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}{1+{\sin }^2\left(\frac{\mathrm{\π}}{M}\right)\stackrel{\‾}{{\mathrm{\γ}}_s}}})\cot \left(\frac{\mathrm{\π}}{M}\right)]\left\}\right)}^n$

In formula, what M represented is the modulation of M system, if the modulation system of baseband signal adopts 4 yuan of phase shift keying modulations, that is:, when 4PSK modulates, M just equals 4.

represent signal to noise ratio, wherein P ₀(dB) represent the transmitted power of transmitting terminal, average path loss be sending and receiving end apart from the function of d, it and d and reference distance d ₀the n of ratio be directly proportional, write as dB form as shown in the formula:

$\stackrel{\‾}{L_p}\left(d\right)\left(\mathrm{dB}\right)={\left(\frac{4\mathrm{\πd}}{\mathrm{\λ}}\right)}^2\left(\mathrm{dB}\right)+10n{\log }_{10}\left(\frac{d}{d_0}\right)$

So, wireless link e[i, j] at t link-quality constantly, be designated as: LinkQty (e[i, j], t)=1-P _f(e[i, j], t), the present invention adopts t ₁, t ₂, t ₃... t _kk link e[i constantly, j] the variance of frame error rate as at t _k+1link stability metric constantly, is designated as: and LinkSty (e[i, j], t _k+1).Following formula is exactly the computing formula of link stability:

$\mathrm{LinkSty}\left(e\right[i,j],t_{k+1})=\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\left\{\right[\mathrm{LinkQty}\left(e\right[i,j],t_i)-\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkQty}\left(e\right[i,j],t_i){]}^2\}$

In wireless self-organization network system, due to the motion of node and the complex characteristics of wireless channel, there is obvious unsteadiness in the link of transfer of data.In this case, the link quality information that Routing Protocol is collected is out-of-date information often.Out-of-date information can not accurate response network link up-to-date variation, so the present invention designs a kind of forecast model based on birth and death process, the link information according to collecting, carries out suitable prediction to the network link probability of stability.And, the hypothesis of this link stability forecast probability based on following: in a very short time range, the probability density function of channel frame error rate remains unchanged, in this of short duration time range, the variation of network topology is negligible simultaneously.

First whole mobile ad-hoc network is modeled as to a complete directed graph G=(V, E).For directed edge definition two states: Up (U) and the Down (D) in complete directed graph.In E, a limit shows that in U state this oriented link is stable.

So, set E in limit is just divided into two subclass E _up={ e|state (e)=Up} and E _down={ e|state (e)=Down}.The practical topology of network is exactly the subgraph consisting of all nodes and the U state limit between them.Note directed edge is e[i, j], wherein i is starting point, j is terminal.Simultaneously the present invention adopts link e[i, j] link-quality LinkQty constantly of t (e[i, j], t) probability that is greater than given threshold alpha is as e[i, j] at t constantly in U shape probability of state, so limit e[i, j] at the probability in stable state of moment t, can be provided by following formula:

P _stable(e[i,j],t)＝P _U(e[i,j],t)＝Probability{LinkQty(e[i,j],t)>α}

Wherein, P _u(e[i, j], t) refer to directed edge e[i, j] probability in U state (stable state) constantly at t.A path P ath of definition (s, d) is that a series of directed edges from source node s to destination node d form.On path, n+1 node associative list is shown:

S _v={ s, p ₁, p ₂..., p _k... .., p _n-1, d}, on Path (s, d) path, the set expression on n bar limit is:

S _e={ e ₁, e ₂..., e _k... .., e _n, so path P ath (s, d) t stability probability constantly as shown in the formula:

$P_{\mathrm{stable}}(\mathrm{Path}(s,d),t)=\underset{e_i\∈S_e}{\mathrm{\Π}}{P}_U(e_i,t)$

Provide related definition and the formula that need to use below:

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the link-quality of this node and a hop neighbor node can be designated as LinkQty (e[i, j], t), wherein LinkQty (e[i, j], t)=1-P _f(e[i, j], t).P _f(e[i, j], specific definition t) is referring to above.

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the stability of the link-quality of this node and a hop neighbor node be designated as LinkSty (e[i, j], t _k+1), wherein:

$\mathrm{LinkSty}\left(e\right[i,j],t_{k+1})=\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{[\mathrm{LinkQty}\left(e\right[i,j],t_i)-\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkQty}\left(e\right[i,j\left]{,t}_i\right)]}^2.$

Make the directed edge of x representative from this node to this hop neighbor node, this node is designated as P with the stability forecast of the link in short-term probability of a hop neighbor node _u(x, t), wherein P _u(x, t)=Probability{1-P _f(x, t) > α }

The MPR that supposes to take computing node S is example, and a hop neighbor node set of S is designated as N (S), and two-hop neighbor node set is designated as N ₂(S).Suppose that S has a hop neighbor node W ∈ N (S), at the two-hop neighbor node set N of S ₂(S) exist k node to be connected with node W, remember that these nodes are G _i(k>=i>=1), note W and G _ilink-quality between (k>=i>=1) is LinkQty (e[W, G _i], t), link-quality stability is LinkSty (e[W, G _i], t), that link stability forecast probability is P in short-term _u(e[W, G _i], t), wherein i is from 1 to k value, a hop neighbor node W with by the two-hop neighbor node G about node S of its covering _ilink-quality LinkQty between (k>=i>=1) _2-hopby following formula, provided:

${\mathrm{LinkQty}}_{2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkQty}\left(e\right[W,G_i],t)}{k}$

Link-quality stability LinkSty _2-hopby following formula, provided:

${\mathrm{LinkSty}}_{2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}\mathrm{LinkSty}\left(e\right[W,G_i],t)}{k}$

Link probability of stability P in short-term _u2-hop(S _w, t) as shown in the formula:

$P_{U2-\mathrm{hop}}(S_W,t)=\frac{\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{P}_U\left(e\right[W,G_i],t)}{k}$

Link-quality between W and S is designated as: LinkQty (e[S, W], t), node W about relative neighbours' link-quality of S as shown in the formula:

RNLQty _W,S＝β·LinkQty _2-hop(S _W,t)+(1-β)LinkQty(e[S,W],t)

Wherein β is the relevant optional parameters of concrete application.

Link-quality stability between W and S is designated as: LinkSty (e[S, W], t), node W is provided by following formula about relative neighbours' link-quality stability of S:

RNLSty _W,S＝η·LinkSty _2-hop(S _W,t)+(1-η)LinkSty(e[S,W],t)

Equally, η is also the relevant optional parameters of concrete application.

The stability forecast of link in short-term probability between W and S is designated as: P _u(e[S, W], t), node W about the relative neighbours of S in short-term link stability forecast probability by following formula, provided:

RNLPty _W,S＝P _U2-hop(S _W,t)·P _U(e[S,W],t)

Node W is provided by following formula about the comprehensive coverage quality of the combination coverage degree of S:

${\mathrm{CoverQty}}_{W,S}=\mathrm{\γ}\·\frac{{\mathrm{RNLQty}}_{W,S}}{\sqrt{{\mathrm{RNLSty}}_{W,S}}+{\mathrm{RNLQty}}_{W,S}}+\mathrm{\τ}\·{\mathrm{RNLPty}}_{W,S}+(1-\mathrm{\γ}-\mathrm{\τ})\·C_W$

γ wherein, τ is according to the selected non-negative parameter of concrete application, need meet constraint: 0≤γ+τ≤1, the assessment that their different values have embodied covering quality stresses degree for the difference of link stability, link stability forecast probability and coverage degree, can determine suitable span by emulation experiment.Can observe, when γ=0 and τ=0, algorithm of the present invention deteriorates to original algorithm simultaneously, and the coverage of node of take has maximum as choice criteria.When γ+τ=1, in this algorithm, do not consider the coverage of node, just do not met multiple spot relaying thought.

Link-quality is taken into account, will be transformed exactly its MPR algorithm, a hop neighbor node set of S is designated as N (S), and two-hop neighbor node set is designated as N ₂(S), MPR set is MPR (S).Concrete MPR selection algorithm step is as follows:

(1) MPR of initialization node S set, is set to empty set by MPR (S).

(2) if a two-hop neighbor node Z of S can only arrive S by a hop neighbor node Y of S, node Y is added in MPR (S).

(3) in reduction operation: N (S), the node in MPR (S) has been added in deletion to, is designated as N'(S); N ₂(S) in, deletion, by the node of coverage in MPR (S), is designated as N ₂' (S).N ₂if ' (S) empty set, algorithm finishes, otherwise continues.

(4) according to the formula in (10), calculate N'(S) in each node be present in N ₂' two-hop neighbor node (S) is with respect to the comprehensive coverage quality of joint S, and by its descending.The node that selection has maximum comprehensive coverage quality adds in MPR (S).

(5) two-hop neighbor node of S is all covered, and algorithm finishes; Otherwise forwarding step (3) to carries out.

Fig. 2 has provided the MPR election algorithm flow process that the present invention is based on comprehensive coverage quality.

Same, also link-quality can be considered in route selecting and gone, Routing Protocol before, based on topology table, calculates and arrives known destination node route according to the shortest principle as route selecting of jumping figure.Although this method is fairly simple, direct, ignored link and change in topology that node motion causes, depend merely on the real quality that hop count is difficult to reaction route.Therefore, the present invention introduces the mechanism of link-quality, the link stability based on link-quality time series variance and the stability forecast of the link in short-term probability based on birth and death process based on physical channel characteristics, thereby select link more stable, the packet transfer route that jumping figure is relatively little.

The above analysis, the present invention adopts the multi-parameter comprehensive route quality algorithm based on the characteristic of channel.The routing table of computing node S of take is example, and the step of this route computing method is as follows:

(1) empty the routing table RTable (S) of node S, that is: RTable (S)=φ.

(2) according to node S mono-hop neighbor node table, add route table items, if a hop neighbor node is not added routing table list item, add a route table items, destination address and next hop address that new list item is set are this hop neighbor node address, i=1 is set, distance is 1, and route quality is set to node S to the path quality of this node, and it calculates according to following formula:

Link-quality LinkQty (e wherein _i, t), link stability LinkSty (e _i, t), path stability forecast probability P _stable(Path (s, d), definition t) is all by above providing; Parameter ξ and be according to the selected non-negative parameter of concrete application, need to satisfy condition: effect is that route quality is averaged out between link-quality, path stability forecast probability and jumping figure.When ξ=0 and time, algorithm selects to have the shortest jumping figure path as optimal path, and now route computing method deteriorates to primary routing computational methods.When time, choose the best path of link-quality and stability as optimal path, deficiency is possible cause the jumping figure in path long.ξ and value can by emulation experiment, determine optimum value according to concrete application.

(3) i=2 is set, add the i jumping route table items that arrives S, until all nodes that S knows all add in routing table entry, detailed process is as follows: for each the topological list item in topology table, if the distance jumping figure that final jump address arrives between the destination address of this list item is i, check in routing table, whether there is this destination address:

If 1. do not exist in routing table, arrive the route performance of this destination address, add the route table items of new this destination address of arrival.The destination address that the destination address of new route table items is this topology list item is set; Next hop address is set to the next hop address of route table items corresponding to final jump address in topological list item; Apart from jumping figure, be set to i+1; By above formula, calculate the route quality in this path.

If 2. existed in routing table, arrive the list item of this destination node, according to above formula, calculate the route quality of new route, if the route quality of new route is greater than the route quality in the path of current existence, according to 1. method, add new route table items.

Routing Protocol of the present invention has considered quality, stability and the stable prediction probability of path link for the tolerance of route quality, evade to a certain extent the route that has unstable link, take into account the jumping figure of route simultaneously, although sometimes the jumping figure in path can be slightly large, but the stability of path uplink but can guarantee, thereby guarantee to divide into groups to transmit failed probability in reduced levels.Fig. 3 has provided routing table calculation process of the present invention.

In sum, the present invention is by link-quality and quality variation are analyzed, thereby the path of a better quality of selection carries out routed path, thereby has improved network quality.Above carried OLSR method is use as an example just, is in order to help reader understanding's principle of the present invention, should be understood to that the protection range of invention is not limited to such special statement and embodiment.Everyly according to foregoing description, make various possible being equal to and replace or change, be all considered to belong to the protection range of claim of the present invention.

Claims (4)

1. the routing algorithm of wireless self-organization network based on channel, is characterized in that: comprise the steps:

First step: according to wireless link signal to noise ratio, under the channel model of supposition, set up link-quality and Stability Model; Wherein related definition is as follows:

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the link-quality of this node and a hop neighbor node can be designated as LinkQty (e[i, j], t), wherein LinkQty (e[i, j], t)=1-P _f(e[i, j], t),

This node is designated as node i, and wherein certain hop neighbor node is designated as node j, the stability of the link-quality of this node and a hop neighbor node be designated as LinkSty (e[i, j], t _k+1), wherein:

Make the directed edge of x representative from this node to this hop neighbor node, this node is designated as P with the stability forecast of the link in short-term probability of a hop neighbor node _u(x, t), wherein P _u(x, t)=Probability{1-P _f(x, t) > α };

The MPR that supposes to take computing node S is example, and a hop neighbor node set of S is designated as N (S), and two-hop neighbor node set is designated as N ₂(S), suppose that S has a hop neighbor node W ∈ N (S), at the two-hop neighbor node set N of S ₂(S) exist k node to be connected with node W, remember that these nodes are G _i(k>=i>=1), note W and G _ilink-quality between (k>=i>=1) is LinkQty (e[W, G _i], t), link-quality stability is LinkSty (e[W, G _i], t), that link stability forecast probability is P in short-term _u(e[W, G _i], t), wherein i is from 1 to k value, a hop neighbor node W with by the two-hop neighbor node G about node S of its covering _ilink-quality LinkQty between (k>=i>=1) _2-hopby following formula, provided:

Link-quality stability LinkSty _2-hopby following formula, provided:

Link probability of stability P in short-term _u2-hop(S _w, t) be following formula:

Link-quality between W and S is designated as: LinkQty (e[S, W], t), node W is following formula about relative neighbours' link-quality of S:

RNLQty _W,S＝β·LinkQty _2-hop(S _W,t)+(1-β)LinkQty(e[S,W],t)，

Wherein β is the relevant optional parameters of concrete application;

Link-quality stability between W and S is designated as: LinkSty (e[S, W], t), node W is provided by following formula about relative neighbours' link-quality stability of S:

RNLSty _W,S＝η·LinkSty _2-hop(S _W,t)+(1-η)LinkSty(e[S,W],t)，

Equally, η is also the relevant optional parameters of concrete application;

The stability forecast of link in short-term probability between W and S is designated as: P _u(e[S, W], t), node W about the relative neighbours of S in short-term link stability forecast probability by following formula, provided:

RNLPty _W,S＝P _U2-hop(S _W,t)·P _U(e[S,W],t)

Node W is provided by following formula about the comprehensive coverage quality of the combination coverage degree of S:

γ wherein, τ is according to the selected non-negative parameter of concrete application, need meet constraint: 0≤γ+τ≤1;

Second step: the data structure expansion link information field to Routing Protocol control packet;

Third step: be the corresponding information table of Memory Extension of link information;

The 4th step: junction link information, transformation MPR selection algorithm and routed path Algorithms of Selecting.

2. the routing algorithm of wireless self-organization network as claimed in claim 1 based on channel, is characterized in that: described MPR selection algorithm, specifically comprises:

The MPR set of first initialization node S, is set to empty set by MPR (S); If a two-hop neighbor node Z of S can only arrive S by a hop neighbor node Y of S, node Y is added in MPR (S); In N (S), the node in MPR (S) has been added in deletion to, is designated as N'(S); N ₂(S) in, deletion, by the node of coverage in MPR (S), is designated as N ₂' (S); N ₂if ' (S) empty set, algorithm finishes, otherwise continues;

According to

Calculate N'(S) in each node be present in N ₂' two-hop neighbor node (S) is with respect to the comprehensive coverage quality CoverQty of joint S _w,S, and by its descending; The node that selection has maximum comprehensive coverage quality adds in MPR (S); The two-hop neighbor node of S is all covered, and algorithm finishes; Otherwise forward deletion in N (S) to, added the node in MPR (S) to, be designated as N'(S) this step.

3. the routing algorithm of wireless self-organization network as claimed in claim 1 based on channel, is characterized in that: described routed path Algorithms of Selecting, specifically comprises:

First empty the routing table RTable (S) of node S; According to node S mono-hop neighbor node table, add route table items, if a hop neighbor node is not added routing table list item, add a route table items, destination address and next hop address that new list item is set are this hop neighbor node address, i=1 is set, distance is 1, and route quality is set to node S to the path quality of this node; I=2 is set, adds the i jumping route table items that arrives S, until all nodes that S knows all add in routing table entry:

If 1. do not exist in routing table, arrive the route performance of this destination address, add the route table items of new this destination address of arrival.The destination address that the destination address of new route table items is this topology list item is set; Next hop address is set to the next hop address of route table items corresponding to final jump address in topological list item; Apart from jumping figure, be set to i+1; By above formula, calculate the route quality in this path;

If 2. existed in routing table, arrive the list item of this destination node, according to above formula, calculate the route quality of new route, if the route quality of new route is greater than the route quality in the path of current existence, according to 1. method, add new route table items.

4. the routing algorithm of wireless self-organization network as claimed in claim 3 based on channel, is characterized in that: the route quality computing formula in described path is:

Wherein, parameter ξ and be according to the selected non-negative parameter of concrete application, need to satisfy condition: .