CN104980933A - Binary sort tree-based local-bargaining spectrum allocation method - Google Patents

Abstract

The invention provides a binary sort tree-based local-bargaining spectrum allocation method. The method comprises the steps of dividing a scanned frequency spectrum into a plurality of signal channels at the initial moment, arranging the signal channels in an ascending sequence from small to large to construct a single-branch binary sort tree, allocating the signal channels to users based on the allocation algorithm to acquire an allocation matrix, updating users at each node of the binary sort tree according to the allocation matrix, updating the binary sort tree according to the variation of the scanned frequency spectrum during the next time period, and simultaneously reallocating the signal channels according to the variation of users that apply for signal channels. The above process is repeated until a command for stopping the frequency spectrum scanning operation is received. According to the technical scheme of the invention, the overall scanning result of frequency bands is in the form of a tree. In this way, the spectrum allocation is realized only through the inserting operation and the deleting operation on the binary sort tree. Therefore, the allocation time is saved. Under the circumstance of the lack of frequency spectrum resources, the spectrum utilization rate is increased and the satisfaction level of users is improved at the same time.

Description

The frequency spectrum distributing method negotiated a price in a kind of local based on binary sort tree

Technical field

The invention belongs to cognitive radio technology field, concrete, the present invention relates to the frequency spectrum distributing method negotiated a price in a kind of local based on binary sort tree.

Background technology

Cognitive radio technology mainly comprises frequency spectrum detection, spectrum allocation may, frequency spectrum switches and power controls these key technologies.

Due to the resource that the mandate frequency range distributing to primary user is in special time, frequency and geographical position.So idle frequency spectrum detected, for different cognitive user, it is the resource of time, frequency and spatially dynamic change.The object of frequency spectrum detection picks out when previous user can be used for the authorization channel of cognition radio communication.

The dynamic variation characteristic of idle frequency spectrum determines cognitive user that cognition wireless network must take dynamic frequency spectrum deployment mode could meet constantly change to the demand of frequency spectrum and cognitive user (also the claiming time user) frequency spectrum share with primary user.When carrying out dynamic frequency spectrum deployment for time user, cognition wireless network must ensure that time user can not cause interference to authorizing primary user's proper communication in frequency range.Under such prerequisite, the usable spectrum of secondary user takies the change of situation and dynamic change to authorizing frequency range along with primary user.Spectrum allocation may algorithm is just used to determine which time user can access network and how to distribute these user and could realize throughput of system and obtain maximized optimum allocation.Spectrum allocation may algorithm carries out under the prerequisite both known in frequency spectrum detection.Spectrum allocation may algorithm is based on three models, and one of them comparatively typical common model is exactly graph theory model.It becomes abstract for user a little, and by the line be expressed as between 2 interfering with each other between user.Have many algorithms based on graph theory model, wherein Color-sensitive Turing pattern formation (CSGC) is comparatively common algorithm.CSGC algorithm hypothesis user in one-period does not change.The spectrum information scanned is converted into available channel matrix by this algorithm, and the topological structure scanned is converted into interference matrix and allocation matrix, is used for carrying out spectrum allocation may, until available channel is assigned with complete.If in multiple computation of Period allocation result, need, according to single step, repeatedly to call this partition function.

The frequency spectrum distributing method negotiated a price in local is a kind of CSGC method of improvement, the model that the mobility due to user can cause topological structure to change by the method, organizes themselves into negotiation team and self adaptation formation optimal allocation model.With ignore both know assignment information traditional optimal topology model compared with, the amount of calculation of the method significantly reduces.The whole process of the frequency spectrum distributing method negotiated a price in local: the spectrum allocation may often completing one-period, each user searches for whole frequency range and the channel that takies of perception primary user again, if occur, primary user uses this channel, then need to allow the secondary user having distributed this channel exit channel occupancy in time.The channel perceived being changed into the matrix be made up of binary system array---available channel matrix, represents that can each user use the situation of this channel respectively.The position of the primary user perceived and time user and transmitting power are converted into binary system interference constraints matrix, and then utilize label criterion to carry out spectrum allocation may.Allocation result binary assignment matrix represents, it characterizes each user and obtains frequency spectrum whether situation.When conflict graph changes, this adaptive algorithm, compared with traditional CSGC algorithm, does not need to redistribute frequency spectrum, only need compensate the minor variations of the regional area that mobility causes.When the node perceiving change is primary user, upgrade allocation matrix, make to take this primary user and authorize secondary user's allocation matrix of frequency spectrum to become 0.And when time user adds in established topological diagram, originally desirable distribution structure is upset, adjacent node and insert node will adaptive formation conflict graph, from composition agreed-upon price group, in group, change the method for salary distribution, reach the object maximizing the availability of frequency spectrum.

The frequency spectrum distributing method negotiated a price in local can save the spectrum allocation may time greatly relative to traditional CSGC method; But the method does not consider this factor of user's request, after making to distribute all usable spectrums according to distribution rule, if there is new user to apply for just contributing redundant frequency spectrum when using this frequency spectrum.These were not the frequency spectrum of user's request originally, occupy certain distribution time in the assignment procedure, the method needs to apply for and organize themselves in for agreed-upon price group process taking portion of time again in agreed-upon price, can time loss be increased undoubtedly, even can make originally to be submerged to the advantage reducing time overhead in order to reduce number of users of participating in the distribution; And in the method, if the spectrum allocation may information perceived occurs in graph form, need to transform into binary system available channel matrix, the extra distribution time can be consumed so undoubtedly.

Summary of the invention

The object of the invention is to the distribution overlong time of the frequency spectrum distributing method existence overcoming existing local agreed-upon price, do not consider that user's request causes the defect of unreasonable distribution, propose the frequency spectrum distributing method negotiated a price in a kind of local based on binary sort tree, the method considers user's request, the frequency spectrum scanned is built binary sort tree, for carrying out the distribution of frequency spectrum and sharing, the method has saves the spectrum allocation may time, improves the feature of the availability of frequency spectrum.

To achieve these goals, the invention provides the frequency spectrum distributing method negotiated a price in a kind of local based on binary sort tree, described method comprises: at initial time, be several channels by the spectrum division that scans, and described channel is propped up binary sort tree according to ascending order sequential build list from small to large; Utilize allocation algorithm to carry out channel allocation to user, obtain allocation matrix, and upgrade the user on each node of binary sort tree according to described allocation matrix; In the next time cycle, the change according to scanning frequency spectrum upgrades binary sort tree, carries out redistributing of channel according to the change of request channel user simultaneously; Repeat said process, stop the order of scanning frequency spectrum until receive.

In technique scheme, described method specifically comprises:

Step 1) at initial time, the frequency spectrum scanned is carried out division and obtains multiple channel, and described channel is propped up binary sort tree according to ascending order sequential build list from small to large;

Step 2) according to the maximum target of throughput of system, utilize allocation algorithm to carry out channel allocation to user, obtain allocation matrix, and upgrade the user on each node of binary sort tree according to described allocation matrix;

Step 3) make k=1;

Step 4) frequency spectrum is scanned; According to the change of frequency spectrum after scanning, binary sort tree is upgraded, re-start redistributing of channel according to the change of time user simultaneously;

Step 5) make k=k+1, proceed to step 4); The order of scanning frequency spectrum is stopped until receiving.

In technique scheme, described step 2) specifically comprise:

Step 201) user's request is greater than the user's request of number of available channels object is set to 0, its label value is set to 0 simultaneously; Add up the total number of users that all user's request numbers are less than self usable spectrum sum;

Step 202) calculate the user obtaining maximum label value in each channel; Pick out this user, give this user by this channel allocation;

For each channel m, due to user list and interference constraints relation along with other Channel Processing constantly changes; Therefore the interference of channel also can constantly change; The interference value of user n use channel is:

$R_{n,m}=\underset{k=1,k\≠n}{\overset{N}{\Σ}}{c}_{n,k,m}$

Wherein, when user uses channel, if user shares this channel and does not produce interference, then c _{n, k, m=0}; If produce interference, then c _{n, k, m=1}; 1≤n≤N, 1≤m≤M, M is total number of channels;

Collaboration type maximum ratio equity criterion is selected to carry out label to user; Tag number size is determined jointly by target function and benefit weight, and the priority that tag number is assigned with more greatly is higher; In label criterion, the corresponding a kind of color of each tag number, characterizes the channel taken when user obtains maximum label value; The maximum label value of user is::

${\mathrm{lable}}_n=max\frac{{\mathrm{\γ}}_n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Color value is:

${\mathrm{color}}_n=\mathrm{argmax}\frac{{\mathrm{\γ}}_n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Wherein, γ _nrepresent that user's request number is less than the user of available channel sum, b _n,mrepresent beneficial matrix, determined by the communication quality of channel and modulation system, this matrix each time cycle can upgrade; l _n,mrepresent the available channel matrix of user, it is determined by the geographical position of secondary user;

Step 203) dispensed matrix a _n,m, make the throughput of system reach maximum;

$f=max\underset{n=1}{\overset{N}{\Σ}}\underset{m=1}{\overset{M}{\Σ}}{a}_{n,m}\·{\mathrm{\γ}}_n\·b_{n,m}$

Step 204) according to described step 203) allocation matrix a _n,mupgrade the information of each node on binary sort tree.

In technique scheme, described step 4) specifically comprise:

Step 401) ready-portioned frequency spectrum is scanned, judge whether channel changes, if changed, enter step 402); Otherwise, enter step 406);

Step 402) judge whether that new channel produces, if judged result is affirmative, proceed to step 403); Otherwise, proceed to step 405);

Step 403) on binary sort tree, find the position of new channel corresponding node, binary sort tree inserts new node; And judged whether time user's request channel, if judged result is affirmative, proceed to step 404); Otherwise, proceed to step 5);

Step 404) by new channel allocation to the secondary user of request channel, utilize step 2) allocation algorithm recalculate allocation matrix a _n,m, and upgrade the information of each node on binary sort tree thus; Proceed to step 5);

Step 405) on binary sort tree, find the position of the channel corresponding node of minimizing, delete this node from binary sort tree; And by described allocation matrix a _n,min element corresponding to this node be set to 0;

Step 406) judge whether time user's request channel, if judged result is affirmative, proceed to step 407); Otherwise, proceed to step 5);

Step 407) judge that can time user share this channel with the user on its acquisition channel and do not produce interference; If judged result is affirmative, then secondary user can use this channel, otherwise, proceed to step 5).

The invention has the advantages that:

1, the result of method of the present invention to whole band scan is the form of tree, only needs to insert binary sort tree and delete, does not need to rebuild topological diagram, just can realize the distribution of frequency spectrum; Save the distribution time.

2, when conflict graph changes, method of the present invention need not judge primary user or secondary user, directly enables frequency spectrum share; Improve user satisfaction;

When 3, again scanning whole frequency spectrum, method of the present invention only needs to operate the frequency spectrum of change, when frequency spectrum resource is in short supply, improves the availability of frequency spectrum at every turn;

4, the free message in spectrum pool stores by the storage organization of method of the present invention by consideration user's request and in conjunction with y-bend sequence number, realize the whole frequency range of each search just add the node of binary sort tree or delete, the time that both saved turn improves user satisfaction.

Accompanying drawing explanation

Fig. 1 is the flow chart of the frequency spectrum distributing method negotiated a price in the local based on binary sort tree of the present invention;

Fig. 2 is the analog channel schematic diagram in embodiments of the invention;

Fig. 3 is the frequency spectrum distributing method of local agreed-upon price and the comparison diagram of method of the present invention 100 time loss mean values under condition best-case;

Fig. 4 is the frequency spectrum distributing method of local agreed-upon price and the comparison diagram of method of the present invention 100 time loss mean values under condition worst case;

Fig. 5 is frequency spectrum distributing method and the comparison diagram of method of the present invention in user satisfaction of local agreed-upon price;

Fig. 6 is the frequency spectrum distributing method and the comparison diagram of method of the present invention on time loss that negotiate a price in local under a same subscriber said conditions.

Embodiment

In the frequency spectrum distributing method negotiated a price in traditional local, if the spectrum allocation may information perceived occurs in graph form, then need to transform into binary system available channel matrix, this mode can consume the extra distribution time undoubtedly, can more save time if can directly utilize the figure perceived to carry out distribution.

Method of the present invention carries out studying under the situation that represents with the form of binary sort tree of frequency range of search.In binary sort tree, each node represents a channel.What be associated with each node is by the set of multiple users that can use this channel.When completing primary distribution, allocation result will be reflected on binary sort tree.When again searching for whole frequency range, when acquisition primary user reclaims and takies application to this authorization channel, binary sort tree will by the Knots inserting representated by corresponding band.Equally, the node city of authorizing frequency range just corresponding on binary sort tree is discharged when searching primary user.When conflict graph will change, can only be the use that time user's application adds some frequency spectrum, frequency spectrum share algorithm realization can be utilized.

Spectrum pool is one and merges idle frequency spectrum to the leasing system a public pond from different frequency spectrum owners there.By the spectrum pool that this is public, user temporarily can rent frequency spectrum resource in the mandate frequency range of free time.In spectrum pool, the idle subband of the authoring system of authorized user is periodically detected and provides binary system allocation vector to be a very important link.After execution cycle property detects, multiple mobile terminal can converge to access point (AP) result.Access point needs to utilize inclusive-OR operation, gathers all binary system allocation vector detected separately, to reach object authorized user fail safe being maximized to protection.

After the process of AP access point, in frequency mode whole idle frequency spectrum split and sort, if the binary system allocation vector now received about idle frequency spectrum exists in the mode of y-bend sequence number, that just need not be translated into the topological diagram about user node, and directly utilizes this binary sort tree just can carry out spectrum allocation may.

There is no assignment information before supposing, the idle matrix expressed by binary sort tree, utilize a need-based CSGC algorithm (yet can be parallel algorithm) to carry out spectrum allocation may, obtain spectrum allocation may information.

When the whole frequency range of mobile terminal rescan, by scanning result compared with last scanning result, if there is new idle frequency spectrum to produce, then by the node city represented by this frequency range in binary sort tree, if when primary user occurs, reclaim mandate frequency range, the corresponding frequency spectrum node just in binary sort tree carries out channel deletion.The user being before assigned with channel can be the same with other users, participates in spectrum allocation may.Rebuild topological diagram thus utilize any particular algorithms to distribute frequency spectrum.When network configuration changes, new secondary user is inserted into topological structure, need not rebuild new tree.Because each node of binary sort tree is associated with distributed user's node.According to these related informations, find the node obtaining this channel fast, new node shares frequency spectrum at subrange and the node distributed, and can produce interference then wait for that topology upgrades next time if shared.

Below in conjunction with accompanying drawing and instantiation, the present invention is described in detail.

As shown in Figure 1, the frequency spectrum distributing method negotiated a price in a kind of local based on binary sort tree, described method comprises:

Step 1) at initial time, the frequency spectrum scanned is carried out division and obtain M channel, and described channel is propped up binary sort tree according to ascending order sequential build list from small to large;

Step 2) according to the maximum target of throughput of system, utilize allocation algorithm to carry out channel allocation to user, obtain allocation matrix, and upgrade the user on binary sort tree on each node according to described allocation matrix; Specifically comprise:

Step 201) user's request is greater than the user's request of number of available channels object is set to 0, its label value is set to 0 simultaneously; Add up the total number of users N that all user's request numbers are less than self usable spectrum sum;

This process can guarantee that user's request is greater than number of available channels object user and does not participate in spectrum allocation may process;

Step 202) calculate the user obtaining maximum label value in each channel; Pick out this user, give this user by this channel allocation;

For each channel m (1≤m≤M, M is total number of channels); Due to user list and interference constraints relation along with other Channel Processing constantly changes; Therefore the interference of channel also can constantly change; User n (1≤n≤N) uses the interference value of channel to be:

$R_{n,m}=\underset{k=1,k\≠n}{\overset{N}{\Σ}}{c}_{n,k,m}$

Wherein: when user uses channel, if user shares this channel and does not produce interference, then c _{n, k, m=0}; If produce interference, then c _{n, k, m=1};

Collaboration type maximum ratio equity criterion is selected to carry out label to user; Tag number size is determined jointly by target function and benefit weight, and the priority that tag number is assigned with more greatly is higher; In label criterion, the corresponding a kind of color of each tag number, characterizes the channel taken when user obtains maximum label value; The maximum label value of user is:

${\mathrm{lable}}_n=max\frac{{\mathrm{\γ}}_n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Color value is:

${\mathrm{color}}_n=\mathrm{argmax}\frac{{\mathrm{\γ}}_n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Wherein, γ _nrepresent that user's request number is less than the user of available channel sum, b _n,mrepresent beneficial matrix, determined by the communication quality of channel and modulation system, this matrix each time cycle can upgrade; l _n,mrepresent the available channel matrix of user, it is determined by the geographical position of secondary user.

Step 203) dispensed matrix, make the throughput of system reach maximum;

$f=max\underset{n=1}{\overset{N}{\Σ}}\underset{m=1}{\overset{M}{\Σ}}{a}_{n,m}\·{\mathrm{\γ}}_n\·b_{n,m}$

Wherein, a _n,mfor allocation matrix;

Step 204) according to described step 203) allocation matrix upgrade the information of each node on binary sort tree;

Step 3) make k=1;

Step 4) frequency spectrum is scanned; According to the change of frequency spectrum after scanning, binary sort tree is upgraded, re-start the distribution of channel simultaneously according to the change of time user; Specifically comprise:

Step 401) ready-portioned frequency spectrum is scanned, judge whether channel changes, if changed, enter step 402); Otherwise, enter step 406);

Step 402) judge whether that new channel produces, if judged result is affirmative, proceed to step 403); Otherwise, proceed to step 405);

Step 403) on binary sort tree, find the position of new channel corresponding node, binary sort tree inserts new node; And judged whether time user's request channel, if judged result is affirmative, proceed to step 404); Otherwise, proceed to step 5);

Step 404) by new channel allocation to the secondary user of request channel, utilize step 2) allocation algorithm recalculate allocation matrix a _n,m, and upgrade the information of each node in sorting tree thus; Proceed to step 5);

Step 405) on binary sort tree, find the position of the channel corresponding node of minimizing, delete this node from binary sort tree; And by described allocation matrix a _n,min element corresponding to this node be set to 0;

Step 406) judge whether time user's request channel, if judged result is affirmative, proceed to step 407); Otherwise, proceed to step 5);

Step 407) judge that can time user share this channel with the user on its acquisition channel and do not produce interference; If judged result is affirmative, then secondary user can use this channel, otherwise, proceed to step 5)

Produce interference if share, but do not have new frequency range to produce, this time user waits scanning next time, until there is new channel to produce;

Step 5) make k=k+1, proceed to step 4); The order of scanning frequency spectrum is stopped until receiving.

Below the performance of method of the present invention is analyzed.

The frequency spectrum distributing method of negotiating a price due to traditional local carries out under the condition of fixed-line subscriber and fixed channel number, and carry out under the channel number of the change at random that method of the present invention is taked and number of users condition, in order to make both, there is comparativity, the number of users of the frequency spectrum distributing method of being negotiated a price traditional local adds 1 at every turn, close with the condition of the inventive method.

The frequency spectrum distributing method negotiated a price in traditional local is before spectrum allocation may, and every run-down, carries out a minor sort.The Straight Insertion Sort being seen as the inventive method can be similar to.When it inserts certain element, the element that will sequence order with n-1 above compares, one by one until just insert after finding insertion position.

From the viewpoint of the time, if the result scanned is according to order arrangement from small to large, the new node inserted only needs to carry out comparing for n time, 0 switch; And if worst case occurs: when the result scanned arranges according to inverted order, the number of comparisons (CN) that carry out and exchange times (EN) are such as formula (6) and formula (7):

$CN={\mathrm{\Σ}}_{i=2}^{n}i=\frac{(n+2)(n-1)}{2}\≈\frac{n^2}{2}---\left(6\right)$

$EN={\mathrm{\Σ}}_{i=2}^n(i+1)=\frac{(n+4)(n-1)}{2}\≈\frac{n^2}{2}---\left(7\right)$

What the frequency spectrum distributing method negotiated a price in traditional local utilized when scanning frequency spectrum is Straight Insertion Sort mode, and its time complexity is O (n ²).And the sort method that our bright method uses is in fact a kind of quicksort recursion method; It with first element for benchmark, whole sequence is divided into two, left and right branch tree, as long as code value is less than benchmark element, is all inserted on the left of benchmark element in left branch tree mode, and code value is greater than benchmark element, all insert on the right side of benchmark element as the tree-like formula of right branch.Therefore the time complexity of method of the present invention is:

$T\left(n\right)=\left\{\begin{array}{cc}O\left(1\right)& n\≤1\\ \mathrm{\Γ}\left(n\right)& n\≥2\end{array}\right.---\left(8\right)$

Wherein n represents element number in sequence, for once comparing spent time fast to N number of element.Quick sort sorting time depends on the degree of depth of tree.Optimal situation is exactly that the left subtree of benchmark element has the identical degree of depth with right subtree, and the time that now algorithm consumes is:

T(n)≤O(nlog ₂n) (9)

Can prove that the average time complexity of the inventive method is O (nlog ₂n).With regard to average time complexity, the performance of method of the present invention is more excellent.

Simulating, verifying is carried out to method of the present invention below.

For the frequency range of 1.5-3GHz, during emulation, this frequency range is divided into 10 sections, becomes 10 channel A-J.Random generation channel take situation.First suppose that taking of channel provides with the form of single binary sort tree, its analog channel produced at random as shown in Figure 2, this Tu Mingbiao occupancy of authorization channel.

When the spectrum occupancy again scanned time, only the node code corresponding to the channel changed is reflected on this single binary sort tree.If namely there is primary user to occur reclaiming corresponding idle frequency spectrum, then the node of this single binary sort tree response is deleted; And if have new idle frequency range to add spectrum pool, then Knots inserting is carried out to this single binary sort tree.This avoid the time rebuilding binary sort tree, it also avoid simultaneously and transmit whole channel occupancy information and the unnecessary memory space that occupies.

As shown in Figure 3, time the idle frequency spectrum under best-case namely in spectrum pool does not change, during the increase of frequency spectrum distributing method of the present invention along with cycle-index, time loss does not increase; Reason is that method of the present invention is found in addition and can be supplied the sharable frequency spectrum of newly added node on the basis of original apportion model, and increase by one during emulation, it is close for can regarding the whole time as at every turn.But traditional method needs at every turn that newly adding a node will build a topological diagram, along with newly adding increasing of access customer, and channel number does not change so time loss will along with increase.

As shown in Figure 4, worst case refers to that each scanning frequency spectrum all can have authorization channel to use and change, and in each scan period, has new user to apply for using channel.Like this, to add or delete principle characterizing on the binary sort tree that authorization channel takies situation.And when negotiating a price in local, can new node will judge share this frequency spectrum with the user of allocated channel, if can not share, also needs again new idle frequency spectrum to be distributed to this user, so more directly distribute spectrum information will consume the longer time.But consider generally, the probability that the idle frequency spectrum of spectrum pool remains unchanged within a period of time wants large, local agreed-upon price method more traditional in this case with the inventive method more save time.

Because method of the present invention improves the frequency spectrum distributing method negotiated a price in traditional local after considering user's request, so need under identical storage organization condition, can compare time loss and the availability of frequency spectrum.As shown in Figure 5, when secondary number of users is 10, the method that the average user satisfaction of method of the present invention is more traditional is high.As shown in Figure 6, will stop carrying out continuation distribution to this user because user's request is met; Thus, when channel number is more sufficient, institute's consumes time is more; And when channel number is fewer, frequency spectrum resource is more in short supply time, method of the present invention and traditional method have the less distribution time.

Claims (4)

1. based on the frequency spectrum distributing method negotiated a price in the local of binary sort tree, described method comprises: at initial time, is several channels, and described channel is propped up binary sort tree according to ascending order sequential build list from small to large by the spectrum division scanned; Utilize allocation algorithm to carry out channel allocation to user, obtain allocation matrix, and upgrade each node on binary sort tree according to described allocation matrix; In the next time cycle, the change according to scanning frequency spectrum upgrades binary sort tree, carries out redistributing of channel according to the change of request channel user simultaneously; Repeat said process, stop the order of scanning frequency spectrum until receive.

2. the frequency spectrum distributing method negotiated a price in the local based on binary sort tree according to claims 1, it is characterized in that, described method specifically comprises:

Step 1) at initial time, the frequency spectrum scanned is carried out division and obtains multiple channel, and described channel is propped up binary sort tree according to ascending order sequential build list from small to large;

Step 2) according to the maximum target of throughput of system, utilize allocation algorithm to carry out channel allocation to user, obtain allocation matrix, and upgrade the user in sorting tree on each node according to described allocation matrix;

Step 3) make k=1;

Step 4) frequency spectrum is scanned; According to the change of frequency spectrum after scanning, sorting tree is upgraded, re-start redistributing of channel according to the change of time user simultaneously;

Step 5) make k=k+1, proceed to step 4); The order of scanning frequency spectrum is stopped until receiving.

3. the frequency spectrum distributing method negotiated a price in the local based on binary sort tree according to claims 2, is characterized in that, described step 2) specifically comprise:

Step 201) user's request is greater than the user's request of number of available channels object is set to 0, its label value is set to 0 simultaneously; Add up the total number of users N that all user's request numbers are less than self usable spectrum sum;

Step 202) calculate the user obtaining maximum label value in each channel; Pick out this user, give this user by this channel allocation;

For each channel m, due to user list and interference constraints relation along with other Channel Processing constantly changes; Therefore the interference of channel also can constantly change; The interference value of user n use channel is:

$R_{n,m}=\underset{k=1,k\≠n}{\overset{N}{\Σ}}{c}_{n,k,m}$

Wherein, when user uses channel, if user shares this channel and does not produce interference, then c _{n, k, m=0}; If produce interference, then c _{n, k, m=1}; 1≤n≤N, 1≤m≤M, M is total number of channels;

Collaboration type maximum ratio equity criterion is selected to carry out label to user; Tag number size is determined jointly by target function and benefit weight, and the priority that tag number is assigned with more greatly is higher; In label criterion, the corresponding a kind of color of each tag number, characterizes the channel taken when user obtains maximum label value; The maximum label value of user is:

${\mathrm{lable}}_n=max\frac{{\mathrm{\γ}}_n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Color value is:

${\mathrm{color}}_n=\mathrm{argmax}\frac{\mathrm{\γ}n\·\frac{b_{n,m}}{R_{n,m}+1}}{\underset{n=1}{\overset{N}{\Σ}}{1}_{n,m}\·b_{n,m}}$

Wherein, γ _nrepresent that user's request number is less than the user of available channel sum, b _n,mrepresent beneficial matrix, determined by the communication quality of channel and modulation system, this matrix each time cycle can upgrade; l _n,mrepresent the available channel matrix of user, it is determined by the geographical position of secondary user;

Step 203) dispensed matrix a _n,m, make the throughput of system reach maximum;

$f=max\underset{n=1}{\overset{N}{\Σ}}\underset{m=1}{\overset{M}{\Σ}}{a}_{n,m}\·{\mathrm{\γ}}_n\·b_{n,m}$

Step 204) according to described step 203) allocation matrix a _n,mupgrade the information of each node on binary sort tree.

4. the frequency spectrum distributing method negotiated a price in the local based on binary sort tree according to claims 3, is characterized in that, described step 4) specifically comprise:

Step 401) ready-portioned frequency spectrum is scanned, judge whether channel changes, if changed, enter step 402); Otherwise, enter step 406);

Step 402) judge whether that new channel produces, if judged result is affirmative, proceed to step 403); Otherwise, proceed to step 405);

Step 403) on binary sort tree, find the position of new channel corresponding node, binary sort tree inserts new node; And judged whether time user's request channel, if judged result is affirmative, proceed to step 404); Otherwise, proceed to step 5);

Step 404) by new channel allocation to the secondary user of request channel, utilize step 2) allocation algorithm recalculate allocation matrix a _n,m, and upgrade the information of each node on binary sort tree thus; Proceed to step 5);

Step 405) on binary sort tree, find the position of the channel corresponding node of minimizing, delete this node from binary sort tree; And by described allocation matrix a _n,min element corresponding to this node be set to 0;

Step 406) judge whether time user's request channel, if judged result is affirmative, proceed to step 407); Otherwise, proceed to step 5);

Step 407) judge that can time user share this channel with the user on its acquisition channel and do not produce interference; If judged result is affirmative, then secondary user can use this channel, otherwise, proceed to step 5).