CN103747530B - Wireless Mesh network channel allocation system and wireless Mesh network channel allocation method under partially overlapped channels - Google Patents

Abstract

The invention relates to a wireless Mesh network channel allocation system and a wireless Mesh network channel allocation method under partially overlapped channels. The system comprises a neighbor-interface binding module, a channel allocation order determination module and a greedy channel allocation module; the neighbor-interface binding module determines neighbor-interface binding relations according to the adjacent relations between nodes of a physical topology, and links which share the same interface serve as a whole and share the same channel when channels are allocated; the Rank descending order is used by the channel allocation order determination module, and the channel allocation order of each link is determined according to the principle that the link with the greatest Rank value is given priority in channel allocation; in order to minimize total interference in a network, the greedy channel allocation module uses the greedy algorithm to allocate channels to all the links in the network. The invention can effectively decrease the probability of network jamming, reduce interference in the network, increase the parallel transmission count in the network and increase the network capacity.

Description

Wireless mesh network channel assigning system and the method partly overlapping under channel

Technical field

The invention belongs to wireless communication technology field, more particularly, to a kind of partly overlapped channel based on ieee802.11b/g Under multi-signal-channel and multi-interface wireless mesh network channel assigning system and method.

Background technology

Wireless mesh network (wireless mesh networks, wmn) have high bandwidth, rapid deployment, be easily installed, Safeguard simple, the low advantage of up-front investment cost, be capable of the coverage of extended wireless network, be therefore expected to become solution " One kilometer afterwards " access problem ideal solution.

Being declined by the network capacity that cochannel interference causes is the significant challenge that wmn faces.Due to having of orthogonal channel number Sex-limited, it is based especially in all available channels of wmn of ieee802.11b/g and only have 3 orthogonal channels so that network is difficult to keep away Exempt to distribute same channel for adjacent link, therefore be difficult to solve interference problem merely with orthogonal channel.Partly overlap the drawing of channel Enter even to eliminate for wmn interference mitigation and bring new thinking, the use of the channel that partly overlapped by careful planning, can be effective Reduce the interference in network, increase the parallel transmission number in network, be obviously improved network capacity.From research feelings both domestic and external at present From the point of view of condition, the planning of part overlapped channels in wmn not yet obtains effectively solving using problem.The overwhelming majority partly overlaps under channel Channel assignment scheme assume network in there is business statistics device, therefore network can know the load on each link in advance；Or Person assumes that network topology gives, the routed path of each stream gives (shortest path first being usually used determine).Above-mentioned hypothesis is actual On determine the link set needing allocated channel it is often more important that for be used business load as channel distribution order according to According to method for channel allocation for, be equivalent to determine link be allocated channel order.The task of channel distribution is exactly to give Under the conditions of fixed these, the path computing along stream draws a kind of channel distribution and link scheduling result to complete the biography loading Pass.But the load of real network link is difficult to be known a priori by, and the channel distribution when link traffic load change Scheme may be inapplicable, needs to re-start channel distribution.Therefore it is contemplated that the dynamic of business in network, that is, connect Request has dynamic source, destination and arrival time, is that in network, all links distribute channel, is ensureing the same of network connectivty When minimize network in total interference.

Content of the invention

The technical problem to be solved in the present invention is to provide a kind of possibility that can reduce congestion generation, reduces in network Total interference, the wireless mesh network channel assigning system being partly overlapped under channel based on ieee802.11b/g of lifting wmn capacity And method.

In order to solve above-mentioned technical problem, the wireless mesh network channel assigning system under the channel that partly overlaps of the present invention Including:

Neighbours-interface binding module: calculate the number of degrees of each node, section according to the syntople between each node of physical topology The number of degrees of point are defined as the neighbor node number of node；When carrying out neighbours-interface binding according to avoiding the high node of the number of degrees as far as possible Determine neighbours-interface binding relationship with the principle of other neighbours' link sharing interfaces, share same connecing when carrying out channel distribution The link of mouth is as an integral dispensing same channel；

Channel distribution order determining module: using rank value descending, preferentially distribute letter according to the big link of rank value The principle in road determines the order of each link channel distribution；

Greedy channel assignment module: be calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen link j And the network being allocated in network between the link of channel always disturbs

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links during cj, c_j=1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network； Interference range reduction gear ratio when itr (τ) represents that the channel number between the channel that link i and j is used is spaced apart τ；d(i,j) Represent the distance between link i and j, refer between any one end points of link i and any one end points of link j distance Little value；R ' represents the interference range under cochannel, and itr (τ) × r ' is that the channel number between i and j is spaced apart partly overlapping during τ Channel reduces interference range；F represents the frequency of channel；Psd (f) represents power spectral density function；α is constant；K is that double footpaths are propagated Path-loss factor in model (two-ray ground propagation model), value is 2～4；

From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j；According to Secondary for all links distribute channel.

The present invention using neighbours-interface binding relationship node neighbours' number be more than number of ports when make some neighbours share connect Mouthful, pass through because the high node of the number of degrees might have a plurality of stream, easily become the bottleneck node in network, therefore should make such Node obtains compared with multiple resource, it is to avoid with other neighbours' shared interfaces, to reduce it, possibility of congestion occurs；Using greedy channel Allocation algorithm quantifies doing between link using the distance between link, corresponding to the interference range under certain particular channel interval Disturb, and using the interference range under theoretical calculation method fetching portion overlapped channels, it is more accurate that the interference between link is portrayed And it is portable poor to overcome the shortcomings of that the existing measurement method being directed to particular network parameter configuration obtains interference range, makes This method can be suitably used for the mesh network of any configuration；With minimize network always disturb for target use greedy algorithm as network in All links distribute channel；Using available 11 channels that partly overlap all in ieee802.11b/g frequency spectrum, net can be reduced Interference in network, increases the parallel transmission number in network, lifts network capacity.

Described rank value adopts following formula to calculate:

Wherein rank_lRank value for link l；Neighbours' number of link l is defined as neighbours' collection of two end nodes of link l And the element number concentrated, link l apart from gateway minimum hop count be defined as link l two end-point distances gateways jete The mean value of number；When rank value is identical, the more link channel assignment order of neighbours' number are more forward.

The present invention, on the basis of neighbours-interface binding, is rank using the method that rank descending determines channel distribution order The big link of value preferentially distributes channel.In network, neighbours' number is more, apart from the fewer link of the minimum hop count of gateway, more easily becomes For the capacity bottleneck of network, the present invention has considered link neighbours' number and apart from two factors of gateway minimum hop count, according to chain Links all in network are carried out descending, are dropped using rank by the minimum hop count apart from gateway for the road and neighbours' number of link Sequence arrangement determines the order of each link channel distribution, is preferably the big link distribution channel of rank value；When rank value is identical, adjacent Occupy the more link channel assignment order of number more forward, the bottleneck link number in network can be reduced to greatest extent, lift network Overall performance.

Itr (τ) value of corresponding different channels interval τ is as shown in table 1；

Table 1

Yi Dan distributing after channel for a link, the channel that link uses will no longer change greedy algorithm, therefore greedy calculation Method restrains quickly.But due to just no longer changing Yi Dan being assigned with after channel, greedy algorithm is usually used and is extremely difficult to optimum letter Road distributes.

Present invention additionally comprises population channel allocation optimization module；This module is by the channel distribution of greedy channel assignment module Result, as the input of particle cluster algorithm, optimizes channel distribution using particle cluster algorithm.

Described population channel allocation optimization module includes following submodules:

Initial parameter setup module: a kind of channel assignment scheme of links all in network is mapped as a particle；Grain Subgroup is made up of multiple particles representing different channels allocative decision；The initial parameter of setting particle cluster algorithm, including population Scale, maximum allowable iterations m, total number of available channels n, to the study of contemporary optimum individual with to optimum of history Control coefrficient alpha2 and alpha1 of body study, inertia weight factor alpha3；Wherein inertia weight factor alpha3 is by formula (5) obtain；

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent maximum and the most respectively The flying speed of partcles of little permission；

Particle position and speed initialization module: the channel distribution result initialization being obtained using greedy channel assignment module The flying speed of particle and position；Each channel assignment scheme will be mapped as a particle, the initial bit of any of which particle i It is set to x_id(0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_d(0) History optimum individual with this particle；The initial position x of arbitrary particle i in remaining particle_id(0) optimum as the history of its own Individual p_id(0)；Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents and learns to contemporary optimum individual Control coefrficient；v_idAnd x (1)_id(0) value is rounded using the method for rounding up；

Particle position and speed iteration update module: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{\mathrm{id}}(t+1)=\mathrm{alpha}3\×v_{\mathrm{id}}\left(t\right)+\mathrm{alpha}1\×\mathrm{rand}\left(\right)\×(p_{\mathrm{id}}\left(t\right)-x_{\mathrm{id}}\left(t\right))\\ +\mathrm{alpha}2\×\mathrm{rand}\left(\right)\×(g_d\left(t\right)-x_{\mathrm{id}}\left(t\right))\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) flying speed of the particle i of the t time and the t+1 time iteration, x are represented respectively_id(t) And x_id(t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents The optimal location that particle i was searched to current by first generation, i.e. history optimum individual；g_dT () represents population current search The optimal location arriving, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；v_idAnd x (t+1)_id (t+1) value is rounded using the method for rounding up；

Iteration calculates the fitness function objf of each particle position after updating every time, selects to adapt in current particle group The minimum particle position of degree functional value is as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to The minimum position of fitness function, as the history optimum individual of this particle, updates contemporary optimum individual and optimum of history with this Body: select when reaching the maximum iteration time of permission contemporary optimum individual as last channel assignment scheme, partly overlap Ieee 802.11b/g multiplex roles Multichannel wireless mesh network channel distribution under channel terminates；

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jBetween interference ratio；rank_ijRepresent certain disturb the corresponding weighted value of ratio, value be both links rank value sort in relatively Big person.

The population channel allocation optimization module of the present invention using greedy channel allocation algorithm result as initial input, according to Definition fitness function guide population iteration update it is contemplated that in network each link significance level, be mesh network institute There is link to distribute suitable channel, optimize greedy channel allocation algorithm result, so that the total interference of network is reduced further, network performance Lifted further.

The wireless mesh network method for channel allocation of the present invention comprises the steps:

1) neighbours-interface binding: calculate the number of degrees of each node, node according to the syntople between each node of physical topology The number of degrees be defined as the neighbor node number of node；When carrying out neighbours-interface binding according to avoid as far as possible the high node of the number of degrees with The principle of other neighbours' link sharing interfaces determines neighbours-interface binding relationship, shares same interface when carrying out channel distribution Link as an integral dispensing same channel；

2) determination of channel distribution order: using rank value descending, preferentially distribute letter according to the big link of rank value The principle in road determines the order of each link channel distribution；

3) greedy channel distribution: be calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen link j with The network being allocated between the link of channel in network always disturbs

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links during cj, c_j=1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network； Interference range reduction gear ratio when itr (τ) represents that the channel number between the channel that link i and j is used is spaced apart τ；d(i,j) Represent the distance between link i and j, refer between any one end points of link i and any one end points of link j distance Little value；R ' represents the interference range under cochannel, and itr (τ) × r ' is that the channel number between i and j is spaced apart partly overlapping during τ Channel reduces interference range；F represents the frequency of channel；Psd (f) represents power spectral density function；α is constant；K is that double footpaths are propagated Path-loss factor in model (two-ray ground propagation model), value is 2～4；

From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j；According to Secondary for all links distribute channel.

Described step 2) in, rank value adopts following formula to calculate:

Wherein rank_lRank value for link l；Neighbours' number of link l is defined as neighbours' collection of two end nodes of link l And the element number concentrated, link l apart from gateway minimum hop count be defined as link l two end-point distances gateways jete The mean value of number；When rank value is identical, the more link channel assignment order of neighbours' number are more forward.

Described step 3) in, itr (τ) value of corresponding different channels interval τ is as shown in table 1；

Table 1

Present invention additionally comprises following step:

Step 4) population channel allocation optimization: using greedy channel distribution result as the input of particle cluster algorithm, use Particle cluster algorithm optimizes channel distribution.

Described population channel allocation optimization comprises the steps:

1) initial parameter setting: a kind of channel assignment scheme of links all in network is mapped as a particle；Particle Group is made up of multiple particles representing different channels allocative decision；The initial parameter of setting particle cluster algorithm, including population Scale, maximum allowable iterations m, total number of available channels n, to contemporary optimum individual study and to history optimum individual Control coefrficient alpha2 and alpha1 of study, inertia weight factor alpha3；Wherein inertia weight factor alpha3 is by formula (5) Obtain；

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent maximum and the most respectively The flying speed of partcles of little permission；

2) particle position and speed initialize: the channel distribution result initialization grain being obtained using greedy channel assignment module The flying speed of son and position；Each channel assignment scheme will be mapped as a particle, the initial position of any of which particle i For x_id(0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_d(0) and The history optimum individual of this particle；The initial position x of arbitrary particle i in remaining particle_id(0) optimum individual as the history of its own Body p_id(0)；Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents and learns to contemporary optimum individual Control coefrficient；v_idAnd x (1)_id(0) value is rounded using the method for rounding up；

3) particle position and speed iteration update: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{\mathrm{id}}(t+1)=\mathrm{alpha}3\×v_{\mathrm{id}}\left(t\right)+\mathrm{alpha}1\×\mathrm{rand}\left(\right)\×(p_{\mathrm{id}}\left(t\right)-x_{\mathrm{id}}\left(t\right))\\ +\mathrm{alpha}2\×\mathrm{rand}\left(\right)\×(g_d\left(t\right)-x_{\mathrm{id}}\left(t\right))\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) flying speed of the particle i of the t time and the t+1 time iteration, x are represented respectively_id(t) And x_id(t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents The optimal location that particle i was searched to current by first generation, i.e. history optimum individual；g_dT () represents population current search The optimal location arriving, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；v_idAnd x (t+1)_id (t+1) value is rounded using the method for rounding up；

Iteration calculates the fitness function objf of each particle position after updating every time, selects to adapt in current particle group The minimum particle position of degree functional value is as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to The minimum position of fitness function, as the history optimum individual of this particle, updates contemporary optimum individual and optimum of history with this Body: select when reaching the maximum iteration time of permission contemporary optimum individual as last channel assignment scheme, partly overlap Ieee 802.11b/g multiplex roles Multichannel wireless mesh network channel distribution under channel terminates；

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jBetween interference ratio；rank_ijRepresent certain disturb the corresponding weighted value of ratio, value be both links rank value sort in relatively Big person.

Brief description

With reference to the accompanying drawings and detailed description the present invention is described in further detail.

Fig. 1 is the wireless mesh network channel assigning system block diagram of the present invention.

Fig. 2 is the wireless mesh network method for channel allocation flow chart of the present invention.

Fig. 3 is neighbours-interface binding method explanatory diagram.

Fig. 4 is the flow chart of present invention greediness method for channel allocation.

Fig. 5 is the flow chart of particle group optimizing method for channel allocation of the present invention.

Specific embodiment

As shown in figure 1, the wireless mesh network channel assigning system of the present invention includes neighbours-interface binding module, channel Assignment order determining module, greedy channel assignment module and population channel allocation optimization module.

Described neighbours-interface binding module determines neighbours-interface binding relationship using following methods:

In view of the dynamic of business in wireless mesh network, need to distribute channel for links all in network.According to thing Manage the syntople between topological each node and determine neighbours-interface binding relationship, need when carrying out channel distribution same for sharing The link distribution same channel of one interface.Due to the restriction of hardware condition, actual wireless mesh nodes number of ports is limited, Generally neighbours' number is more than number of ports, and some neighbours need shared node interface.It is thus necessary to determine that adjacent before carrying out channel distribution The binding relationship of residence-interface.The present invention is come really using the number of degrees (i.e. neighbours' number of node) that node adjacency relation calculates each node Determine neighbours-interface binding relationship.As shown in figure 3, a～k is 11 nodes on wireless mesh network backbone.Node a has 3 to connect Mouthful, form 4 links ab, ac, ad and ae with 4 neighbours.The number of degrees of node b, c, d, e are respectively 4,3,2,2, from number of degrees highest Node start to determine neighbours-interface binding, be link ab distribution interface 1, link ac distribution interface 2, link ad distribution interface 3.According to the principle avoiding the neighbours shared interface big with the node number of degrees as far as possible, it is to comprise in link ae distribution and ab, ac, ad The minimum link ad identical interface of the node number of degrees, as link ae distribution interface 3.Thus complete the neighbours-interface of node a Binding.The neighbours of remaining node-interface binding procedure is similar to.

Described channel distribution order determining module uses rank value descending, preferentially distributes according to the big link of rank value The principle of channel determines the order of each link channel distribution, and concrete grammar is as follows:

On the basis of neighbours-interface binding, the minimum hop count according to link range gateway and neighbours' number of link, to network In all links carry out descending, determine the order of each link channel distribution using rank descending.Neighbours' number is more, Apart from the fewer link of the minimum hop count of gateway, more easily become the capacity bottleneck of network, described link distribution letter should be preferably Road, when rank value is identical, the more link channel assignment order of neighbours' number are more forward.Neighbours' number of described link is defined as chain The element number that is neighbours' collection and concentrating of two end nodes in road, the minimum hop count of described link range gateway is defined as link The mean value of the minimum hop count of two end-point distances gateways.

Still, the determination of channel distribution order is described taking 4 links ab, ac, ad and ae around node a as a example.Link ab Neighbours' number be 8, link ab apart from gateway minimum hop count be 2, the therefore rank value of link ab is 4；In the same manner, link ac, ad It is respectively 7,6,6 with neighbours' number of ae, link ac, ad and ae are respectively 2.5,2.5,1.5 apart from the minimum hop count of gateway, therefore The rank value of link ac, ad and ae is respectively 2.8,2.4,4.Rank value is bigger to represent that channel distribution order is more forward；Work as rank When being worth identical, the more node channel assignment order of neighbours' number are more forward, and the channel distribution order of therefore 4 links is followed successively by: Link ab → ae → ac → ad.

Determine during channel distribution order can also using link neighbours' number more more forward principle of multichannel assignment order or Apart from the more forward principle of gateway minimum hop count fewer channel distribution order.I.e. rank value can also adopt following two formula meter Calculate:

rank_lNeighbours' number of=link l

Or

rank_l=link l is apart from the minimum hop count of gateway

Described greediness channel assignment module with minimize network always disturb for target use greedy algorithm as network in all Link distributes channel, specific as follows:

It is calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen link j and network in be allocated Network between the link of channel always disturbs

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links during cj, c_j=1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network； Interference range reduction gear ratio when itr (τ) represents that the channel number between the channel that link i and j is used is spaced apart τ, only with letter Taoist monastic name interval is relevant, and unrelated with specific environment, therefore itr (τ) value can be using theoretical calculation side before carrying out channel distribution Method obtains；D (i, j) represents the distance between link i and j, refers to any one end points of link i and any one end of link j The minimum of a value of distance between point；R ' represents the interference range under cochannel, and therefore itr (τ) × r ' is the channel number between i and j It is spaced apart the channel reduction interference range that partly overlaps during τ；F represents the frequency of channel；Psd (f) represents power spectral density function； α is constant, for quantifying the annoyance level partly overlapping between channel on same node distinct interface, generally takes higher value, Such as 10, to avoid same node distinct interface using the channel that partly overlaps as far as possible；K is in two-ray ground propagation model Path-loss factor, value be 2～4.

The present invention uses the interference relationships between interference range reduction gear ratio itr (τ) quantized segment overlapped channels, and using reason Obtain itr (τ) value by computational methods.When the path-loss factor k value in hypothesis two-ray ground propagation model is 4, When transceiver is 1 raised cosine FIR filter using roll-off factor, itr (τ) value of corresponding different channels interval τ is such as Shown in table 1.

Raised cosine FIR filter that table 1 is 1 using roll-off factor and path-loss factor are itr (τ) value when 4

When transceiver uses dissimilar wave filter, network can be counted by theory before carrying out channel distribution Calculation method obtains itr (τ) value.As the interference range r ' under given cochannel, channel number is spaced apart the channel that partly overlaps during τ Reduction interference range is itr (τ) × r '.This method obtaining interference range has good transplantability it is adaptable to any join The network put.It is assumed that distributing channel c for link j when for link j distribution channel_j, calculate link j and net under this channel distribution Total interference between all links having distributed channel in network, the rest may be inferred, calculates link j and distributes always doing during other channels Disturb.From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j.Using above-mentioned All links in greedy algorithm traverses network, are the channel that each of the links selects to make network always disturb minimum.Greedy algorithm is once Distribute after channel for a link, the channel that link uses will no longer change, therefore greedy algorithm is restrained quickly.But due to one Denier just no longer changes after being assigned with channel, greedy algorithm is usually used and is extremely difficult to optimal channel assignment.

Described population channel allocation optimization module includes following submodules:

Initial parameter setup module: a kind of channel assignment scheme of links all in network is mapped as a particle (grain Submatrix is a vector, and each element representation in vector distributes to the channel of respective links).Population is represented not by multiple The particle composition of cochannel allocative decision.The distance between each node matrix is built by physical topology, same interface will be shared Link bundling, must distribute same channel in channel distribution.The initial parameter of setting particle cluster algorithm, including the rule of population Mould (i.e. often the number for particle), maximum allowable iterations m, total number of available channels n, learn to contemporary optimum individual With control coefrficient alpha2 and alpha1 learning to history optimum individual, inertia weight factor alpha3.Wherein control coefrficient Alpha1 and alpha2 is the constant that particle cluster algorithm itself gives, and inertia weight factor alpha3 is obtained by formula (5).

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent maximum and the most respectively The flying speed of partcles of little permission.

Particle position and speed initialization module: the channel distribution result initialization being obtained using greedy channel assignment module The flying speed of particle and position, each channel assignment scheme will be mapped as a particle, the initial bit of any of which particle i It is set to x_id(0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_d(0) History optimum individual with this particle；The initial position x of arbitrary particle i in remaining particle_id(0) optimum as the history of its own Individual p_id(0).Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents and learns to contemporary optimum individual Control coefrficient.Because channel number is all integer, to v_idAnd x (1)_id(0) value is rounded using the method for rounding up.

Particle position and speed iteration update module: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{\mathrm{id}}(t+1)=\mathrm{alpha}3\×v_{\mathrm{id}}\left(t\right)+\mathrm{alpha}1\×\mathrm{rand}\left(\right)\×(p_{\mathrm{id}}\left(t\right)-x_{\mathrm{id}}\left(t\right))\\ +\mathrm{alpha}2\×\mathrm{rand}\left(\right)\×(g_d\left(t\right)-x_{\mathrm{id}}\left(t\right))\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) represent the flying speed of the particle i of the t time and the t+1 time iteration respectively；x_id(t) And x_id(t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents The optimal location that particle i was searched to current by first generation, i.e. history optimum individual；g_dT () represents population current search The optimal location arriving, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；Due to channel number all It is integer, to v_idAnd x (t+1)_id(t+1) value is rounded using the method for rounding up.

Iteration calculates the fitness function objf of each particle position, the less expression of fitness function value after updating every time Particle position is more excellent.Update contemporary optimum individual and history optimum individual: select fitness function value in current particle group minimum Particle position as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to fitness function Little position is as the history optimum individual of this particle.Select contemporary optimum individual to make when reaching the maximum iteration time of permission For last channel assignment scheme, the ieee 802.11b/g multiplex roles multi-channel wireless mesh network letter under the channel that partly overlaps Road distribution terminates.

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jBetween interference ratio；rank_ijRepresent that certain disturbs the corresponding weighted value of ratio, value is larger in the rank sequence of both links Person.

As shown in Fig. 2 the wireless mesh network method for channel allocation of the present invention comprises the following steps:

1) neighbours-interface binding: in view of the dynamic of business in wireless mesh network, need for links all in network Distribution channel.Neighbours-interface binding relationship is determined according to the syntople between each node of physical topology, is carrying out channel distribution When need for share same interface link distribution same channel.Due to the restriction of hardware condition, in actual wireless mesh network Node interface number is limited, and usual neighbours' number is more than number of ports, and some neighbours need shared node interface.Carry out channel distribution it Front it is thus necessary to determine that the binding relationship of neighbours-interface.The present invention uses node adjacency relation to calculate the number of degrees (the i.e. node of each node Neighbours' number) determining neighbours-interface binding relationship.As shown in figure 3, a～k is 11 sections on wireless mesh network backbone Point.Node a has 3 interfaces, forms 4 links ab, ac, ad and ae with 4 neighbours.The number of degrees of node b, c, d, e respectively 4, 3rd, 2,2, start to determine neighbours-interface binding from number of degrees highest node, be link ab distribution interface 1, link ac distribution interface 2, link ad distribution interface 3.According to the principle avoiding the neighbours shared interface big with the node number of degrees as far as possible, be link ae distribution with The link ad identical interface of the node number of degrees minimum comprising in ab, ac, ad, as link ae distribution interface 3.Thus complete The neighbours of node a-interface binding.The neighbours of remaining node-interface binding procedure is similar to.

2) determination of channel distribution order: using rank value descending, preferentially distribute letter according to the big link of rank value The principle in road determines the order of each link channel distribution, and concrete grammar is as follows:

On the basis of neighbours-interface binding, the minimum hop count according to link range gateway and neighbours' number of link, to network In all links carry out descending, determine the order of each link channel distribution using rank descending.Neighbours' number is more, Apart from the fewer link of the minimum hop count of gateway, more easily become the capacity bottleneck of network, described link distribution letter should be preferably Road, when rank value is identical, the more link channel assignment order of neighbours' number are more forward.Neighbours' number of described link is defined as chain The element number that is neighbours' collection and concentrating of two end nodes in road, the minimum hop count of described link range gateway is defined as link Two end nodes are apart from the mean value of the minimum hop count of gateway.

Still, the determination of channel distribution order is described taking 4 links ab, ac, ad and ae around node a as a example.Link ab Neighbours' number be 8, link ab apart from gateway minimum hop count be 2, the therefore rank value of link ab is 4；In the same manner, link ac, ad It is respectively 7,6,6 with neighbours' number of ae, link ac, ad and ae are respectively 2.5,2.5,1.5 apart from the minimum hop count of gateway, therefore The rank value of link ac, ad and ae is respectively 2.8,2.4,4.Rank value is bigger to represent that channel distribution order is more forward；Work as rank When being worth identical, the more link channel assignment order of neighbours' number are more forward, and the channel distribution order of therefore 4 links is followed successively by: Link ab → ae → ac → ad.

Determine during channel distribution order can also using link neighbours' number more more forward principle of multichannel assignment order or Apart from the more forward principle of gateway minimum hop count fewer channel distribution order.I.e. rank value can also adopt following two formula meter Calculate:

rank_lNeighbours' number of=link l

Or

rank_l=link l is apart from the minimum hop count of gateway

3) greedy channel distribution: with minimize network always disturb for target using greedy algorithm as network in all links divide Allocating channel.As shown in figure 4, concrete grammar is as follows:

It is assumed to be link j distribution channel c_j, it is calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen chain The network being allocated between the link of channel in road j and network always disturbs

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links during cj, c_j=1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network； Interference range reduction gear ratio when itr (τ) represents that the channel number between the channel that link i and j is used is spaced apart τ, only with letter Taoist monastic name interval is relevant, and unrelated with specific environment, therefore itr (τ) value can be using theoretical calculation side before carrying out channel distribution Method obtains；D (i, j) represents the distance between link i and j, refers to any one end points of link i and any one end of link j The minimum of a value of distance between point；R ' represents the interference range under cochannel, and therefore itr (τ) × r ' is the channel number between i and j It is spaced apart the channel reduction interference range that partly overlaps during τ；F represents the frequency of channel；Psd (f) represents power spectral density function； α is constant, for quantifying the annoyance level partly overlapping between channel on same node distinct interface, generally takes higher value, Such as 10, to avoid same node distinct interface using the channel that partly overlaps as far as possible；K is in two-ray ground propagation model Path-loss factor, value be 2～4.

The present invention uses the interference relationships between interference range reduction gear ratio itr (τ) quantized segment overlapped channels, and using reason Obtain itr (τ) value by computational methods.When the path-loss factor k value in hypothesis two-ray ground propagation model is 4, When transceiver is 1 raised cosine FIR filter using roll-off factor, itr (τ) value of corresponding different channels interval τ is such as Shown in table 1.

Raised cosine FIR filter that table 1 is 1 using roll-off factor and path-loss factor are itr (τ) value when 4

When transceiver uses dissimilar wave filter, network can be counted by theory before carrying out channel distribution Calculation method obtains itr (τ) value.As the interference range r ' under given cochannel, channel number is spaced apart the channel that partly overlaps during τ Reduction interference range is itr (τ) × r '.This method obtaining interference range has good transplantability it is adaptable to any join The network put.It is assumed that distributing channel c for link j when for link j distribution channel_j, calculate link j and net under this channel distribution Total interference between all links having distributed channel in network, the rest may be inferred, calculates link j and distributes always doing during other channels Disturb.

From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j.

Using all links in above-mentioned greedy algorithm traverses network, it is the letter that each of the links selects to make network always disturb minimum Road.

Yi Dan distributing after channel for a link, the channel that link uses will no longer change greedy algorithm, therefore greedy calculation Method restrains quickly.But due to just no longer changing Yi Dan being assigned with after channel, greedy algorithm is usually used and is extremely difficult to optimum letter Road distributes.

4) population channel allocation optimization: optimize channel allocation using particle cluster algorithm, greedy algorithm channel is divided Join result as the input of particle cluster algorithm.As shown in figure 5, specifically including following step:

(1) the initial parameter setting of particle cluster algorithm: a kind of channel assignment scheme of links all in network is mapped as One particle.Population is made up of multiple particles representing different channels allocative decision.Built between each node by physical topology Distance matrix, will share same interface link bundling, same channel must be distributed in channel distribution.Setting population is calculated The initial parameter of method, scale including population (i.e. often the number for particle), maximum allowable iterations m, total available Channel number n, control coefrficient alpha2 and alpha1 to the study of contemporary optimum individual with to the study of history optimum individual, inertia Weight factor alpha3.Wherein control coefrficient alpha1 and alpha2 is the constant that particle cluster algorithm itself gives, inertia weight Factor alpha3 is obtained by formula (5).

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent maximum and the most respectively The flying speed of partcles of little permission.

(2) particle position and speed initialize: the channel distribution result initialization being obtained using greedy channel assignment module The flying speed of particle and position, each channel assignment scheme will be mapped as a particle, the initial bit of any of which particle i It is set to x_id(0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_d(0) History optimum individual with this particle；The initial position x of arbitrary particle i in remaining particle_id(0) optimum as the history of its own Individual p_id(0).Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents and learns to contemporary optimum individual Control coefrficient.Because channel number is all integer, to v_idAnd x (1)_id(0) value is rounded using the method for rounding up.

(3) particle position and speed iteration update: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{\mathrm{id}}(t+1)=\mathrm{alpha}3\×v_{\mathrm{id}}\left(t\right)+\mathrm{alpha}1\×\mathrm{rand}\left(\right)\×(p_{\mathrm{id}}\left(t\right)-x_{\mathrm{id}}\left(t\right))\\ +\mathrm{alpha}2\×\mathrm{rand}\left(\right)\×(g_d\left(t\right)-x_{\mathrm{id}}\left(t\right))\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) represent the flying speed of the particle i of the t time and the t+1 time iteration respectively；x_id(t) And x_id(t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents The optimal location that particle i was searched to current by first generation, i.e. history optimum individual；g_dT () represents population current search The optimal location arriving, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；Due to channel number all It is integer, to v_idAnd x (t+1)_id(t+1) value is rounded using the method for rounding up.

Iteration calculates the fitness function objf of each particle position, the less expression of fitness function value after updating every time Particle position is more excellent.Update contemporary optimum individual and history optimum individual: select fitness function value in current particle group minimum Particle position as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to fitness function Little position is as the history optimum individual of this particle.Select contemporary optimum individual to make when reaching the maximum iteration time of permission For last channel assignment scheme, the ieee 802.11b/g multiplex roles multi-channel wireless mesh network letter under the channel that partly overlaps Road distribution terminates.

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jBetween interference ratio；rank_ijRepresent certain disturb the corresponding weighted value of ratio, value be both links rank value sort in relatively Big person.

Claims (8)

1. the wireless mesh network channel assigning system under a kind of channel that partly overlaps is it is characterised in that include:

Neighbours-interface binding module: the number of degrees of each node are calculated according to the syntople between each node of physical topology, node The number of degrees are defined as the neighbor node number of node；When carrying out neighbours-interface binding according to avoid as far as possible the high node of the number of degrees and its The principle of his neighbours' link sharing interface determines neighbours-interface binding relationship, shares same interface when carrying out channel distribution Link is as an integral dispensing same channel；

Channel distribution order determining module: using rank value descending, preferentially distribute channel according to the big link of rank value Principle determines the order of each link channel distribution；Described rank value adopts following formula to calculate:

Wherein rank_lRank value for link l；Neighbours' number of link l is defined as the union of neighbours' collection of two end nodes of link l In element number, link l apart from gateway minimum hop count be defined as link l two end-point distances gateways minimum hop count Mean value；When rank value is identical, the more link channel assignment order of neighbours' number are more forward；

Greedy channel assignment module: be calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen link j and network In be allocated between the link of channel network always disturb

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links, c during cj_j= 1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network；itr (τ) interference range reduction gear ratio when representing that the channel number between the channel that link i and j is used is spaced apart τ；D (i, j) represents The distance between link i and j, refers to the minimum of distance between any one end points of link i and any one end points of link j Value；R ' represents the interference range under cochannel, and itr (τ) × r ' is that the channel number between i and j is spaced apart the letter that partly overlaps during τ Road reduces interference range；F represents the frequency of channel；Psd (f) represents power spectral density function；α is constant；K is double footpaths propagating mode Path-loss factor in type, value is 2～4；

From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j；It is followed successively by institute There is link distribution channel.

2. the wireless mesh network channel assigning system under the channel that partly overlaps according to claim 1 it is characterised in that Itr (τ) value of corresponding different channels interval τ is as shown in table 1；

Table 1

3. the wireless mesh network channel assigning system under the channel that partly overlaps according to claim 1 and 2, its feature exists In also including population channel allocation optimization module；This module is using the channel distribution result of greedy channel assignment module as particle The input of group's algorithm, optimizes channel distribution using particle cluster algorithm.

4. the wireless mesh network channel assigning system under the channel that partly overlaps according to claim 3 it is characterised in that Described population channel allocation optimization module includes following submodules:

Initial parameter setup module: a kind of channel assignment scheme of links all in network is mapped as a particle；Population It is made up of multiple particles representing different channels allocative decision；The initial parameter of setting particle cluster algorithm, including the rule of population Mould, maximum allowable iterations m, total number of available channels n, to contemporary optimum individual study and to history optimum individual Control coefrficient alpha2 and alpha1 of habit, inertia weight factor alpha3；Wherein inertia weight factor alpha3 is obtained by formula (5) ?；

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent minimum and maximum respectively to permit The flying speed of partcles permitted；

Particle position and speed initialization module: the channel distribution result initialization particle being obtained using greedy channel assignment module Flying speed and position；Each channel assignment scheme will be mapped as a particle, the initial position of any of which particle i is x_id(0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_d(0) and should The history optimum individual of particle；The initial position x of arbitrary particle i in remaining particle_id(0) as the history optimum individual of its own p_id(0)；Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents the control to the study of contemporary optimum individual Coefficient processed；v_idAnd x (1)_id(0) value is rounded using the method for rounding up；

Particle position and speed iteration update module: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{id}\left(t+1\right)=alpha3\×v_{id}\left(t\right)+alpha1\×rand\left(\right)\×\left(p_{id}\left(t\right)-x_{id}\left(t\right)\right)\\ +alpha2\×rand\left(\right)\×\left(g_d\left(t\right)-x_{id}\left(t\right)\right)\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) flying speed of the particle i of the t time and the t+1 time iteration, x are represented respectively_id(t) and x_id (t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents particle i The optimal location being searched to current by first generation, i.e. history optimum individual；g_dT () represents that population current search arrives Excellent position, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；v_idAnd x (t+1)_id(t+1) value Rounded using the method for rounding up；

Iteration calculates the fitness function objf of each particle position after updating every time, selects fitness letter in current particle group The minimum particle position of numerical value is as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to adaptation The minimum position of degree function as the history optimum individual of this particle, updates contemporary optimum individual and history optimum individual with this: Select contemporary optimum individual as last channel assignment scheme when reaching the maximum iteration time of permission, partly overlap channel Under ieee 802.11b/g multiplex roles Multichannel wireless mesh network channel distribution terminate；

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jIt Between interference ratio；rank_ijRepresent that certain disturbs the corresponding weighted value of ratio, value is larger in the rank value sequence of both links Person.

5. the wireless mesh network method for channel allocation under a kind of channel that partly overlaps is it is characterised in that comprise the steps:

1) neighbours-interface binding: calculate the number of degrees of each node, the degree of node according to the syntople between each node of physical topology Number is defined as the neighbor node number of node；When carrying out neighbours-interface binding according to avoid as far as possible the high node of the number of degrees and other The principle of neighbours' link sharing interface determines neighbours-interface binding relationship, shares the chain of same interface when carrying out channel distribution Road is as an integral dispensing same channel；

2) determination of channel distribution order: using rank value descending, preferentially distribute channel according to the big link of rank value Principle determines the order of each link channel distribution；Rank value adopts following formula to calculate:

Wherein rank_lRank value for link l；Neighbours' number of link l is defined as the union of neighbours' collection of two end nodes of link l In element number, link l apart from gateway minimum hop count be defined as link l two end-point distances gateways minimum hop count Mean value；When rank value is identical, the more link channel assignment order of neighbours' number are more forward；

3) greedy channel distribution: be calculated as link j distribution channel c using formula (1), (2), (3), (4)_jWhen link j and network in The network being allocated between the link of channel always disturbs

$i_{c_j}=\underset{i\&element;i}{\σ}ir(i,j,c_i,c_j)---\left(1\right)$

$itr\left(\mathrm{\τ}\right)=\sqrt[k]{\frac{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd\left(f\right)\×psd(f-5\·\mathrm{\τ})df}{{\&integral;}_{-\mathrm{\∞}}^{+\mathrm{\∞}}psd{\left(f\right)}^{2}df}}---\left(3\right)$

τ=| c_i-c_j| (4)

Wherein: ir (i, j, c_i,c_j) representing to be link i, j is respectively allocated channel ci, the interference ratio of both links, c during cj_j= 1,2 ... n, n can use the quantity of channel for link j；I represents the link set having been allocated that channel in network；itr (τ) interference range reduction gear ratio when representing that the channel number between the channel that link i and j is used is spaced apart τ；D (i, j) represents The distance between link i and j, refers to the minimum of distance between any one end points of link i and any one end points of link j Value；R ' represents the interference range under cochannel, and itr (τ) × r ' is that the channel number between i and j is spaced apart the letter that partly overlaps during τ Road reduces interference range；F represents the frequency of channel；Psd (f) represents power spectral density function；α is constant；K is double footpaths propagating mode Path-loss factor in type, value is 2～4；

From channel 1,2 can be used ... select the channel c making network always disturb minimum in n_j* distribute to link j；It is followed successively by institute There is link distribution channel.

6. the wireless mesh network method for channel allocation under the channel that partly overlaps according to claim 5 it is characterised in that Described step 3) in, itr (τ) value of corresponding different channels interval τ is as shown in table 1；Table 1

7. the wireless mesh network method for channel allocation under the channel that partly overlaps according to claim 5 or 6, its feature exists In also comprising the steps:

Step 4) population channel allocation optimization: using greedy channel distribution result as particle cluster algorithm input, using particle Group's algorithm optimization channel distribution.

8. the wireless mesh network method for channel allocation under the channel that partly overlaps according to claim 7 it is characterised in that Described population channel allocation optimization comprises the steps:

1) initial parameter setting: a kind of channel assignment scheme of links all in network is mapped as a particle；Population by Multiple particle compositions representing different channels allocative decision；Setting particle cluster algorithm initial parameter, including the scale of population, Maximum allowable iterations m, total number of available channels n, learn to the study of contemporary optimum individual with to history optimum individual Control coefrficient alpha2 and alpha1, inertia weight factor alpha3；Wherein inertia weight factor alpha3 is obtained by formula (5) ?；

$alpha3=w_{max}-\frac{w_{\max }-w_{\min }}{m}\×m---\left(5\right)$

Wherein m is the number of times of current iteration；M is maximum allowable iterations；w_maxAnd w_minRepresent minimum and maximum respectively to permit The flying speed of partcles permitted；

2) particle position and speed initialize: the channel distribution result being obtained using greedy channel assignment module initializes particle Flying speed and position；Each channel assignment scheme will be mapped as a particle, the initial position of any of which particle i is x_id (0)；Using the particle position of optimal channel assignment scheme mapping in channel assignment scheme as contemporary optimum individual g_dAnd this grain (0) The history optimum individual of son；The initial position x of arbitrary particle i in remaining particle_id(0) as the history optimum individual of its own p_id(0)；Calculate first time iteration flying speed v of each particle using formula (6)_id(1)；

v_id(1)=alpha2 × rand () × (g_d(0)-x_id(0)) (6)

Wherein: rand () represents the stochastic variable of change in the range of (0,1)；Alpha2 represents the control to the study of contemporary optimum individual Coefficient processed；v_idAnd x (1)_id(0) value is rounded using the method for rounding up；

3) particle position and speed iteration update: using formula (7), the speed of (8) more new particle and position；

$\begin{array}{c}{v}_{id}\left(t+1\right)=alpha3\×v_{id}\left(t\right)+alpha1\×rand\left(\right)\×\left(p_{id}\left(t\right)-x_{id}\left(t\right)\right)\\ +alpha2\×rand\left(\right)\×\left(g_d\left(t\right)-x_{id}\left(t\right)\right)\end{array}---\left(7\right)$

x_id(t+1)=x_id(t)+v_id(t+1) (8)

Wherein: v_id(t) and v_id(t+1) flying speed of the particle i of the t time and the t+1 time iteration, x are represented respectively_id(t) and x_id (t+1) represent the position of the particle i of the t time and the t+1 time iteration respectively, t=1,2..., m ... m；p_idT () represents particle i The optimal location being searched to current by first generation, i.e. history optimum individual；g_dT () represents that population current search arrives Excellent position, i.e. contemporary optimum individual；Rand () represents the stochastic variable of change in the range of (0,1)；v_idAnd x (t+1)_id(t+1) value Rounded using the method for rounding up；

Iteration calculates the fitness function objf of each particle position after updating every time, selects fitness letter in current particle group The minimum particle position of numerical value is as contemporary optimum individual；Selecting to be arrived currently by first generation certain particle search to adaptation The minimum position of degree function as the history optimum individual of this particle, updates contemporary optimum individual and history optimum individual with this: Select contemporary optimum individual as last channel assignment scheme when reaching the maximum iteration time of permission, partly overlap channel Under ieee 802.11b/g multiplex roles Multichannel wireless mesh network channel distribution terminate；

The fitness function objf of particle group optimizing be defined to correspond to total interference in network during certain channel distribution it may be assumed that

$objf=\underset{i,j\&element;l}{\σ}{\mathrm{rank}}_{ij}\×ir(i,j,c_i,c_j)---\left(9\right)$

rank_ij=max (rank_i,rank_j) (10)

Wherein: l represents all link set in network；ir(i,j,c_i,c_j) represent the channel c that link i and j is used_iWith c_jIt Between interference ratio；rank_ijRepresent that certain disturbs the corresponding weighted value of ratio, value is larger in the rank value sequence of both links Person.