CN102438235A - Method for selecting optimal distributed type interference source in mobile collaborative network - Google Patents

Abstract

The invention relates to the technical field of mobile communications and provides a method for selecting an optimal distributed type interference source in a mobile collaborative network. In the method, the possible states and the state transition probability of candidate collaborative nodes are obtained through periodically transmitting a training sequence and feedback information; whether a channel source node can be in direct communication with a destination node is judged, if the direct communication is available, the selection of the optimal interference node is carried out, or else, the selection of an optimal relay-interference pair is carried out; and a corresponding priority symbol value of each possible state of the collaborative nodes is calculated, and a suitable node is selected to collaborate data transmission. With the adoption of the method provided by the invention, the optimal interference source can be selected according the specific states of a channel, so as to respond to the interception of illegal nodes in the mobile collaborative network, the power consumption and the transmission error rate of the collaborative nodes are sufficiently considered at the same time that the secure communication is guaranteed, and the quality of communication is ensured.

Description

Move Optimal Distribution formula interference source system of selection in the collaborative network

Technical field

The present invention relates to moving communicating field, Optimal Distribution formula interference source system of selection in particularly a kind of mobile collaborative network.

Background technology

In recent years, the wireless communication technology development is swift and violent, has been penetrated into the every aspect of people's daily life gradually, and the informational needs of magnanimity is through the wireless channel transmission.To the problem of wireless communication spectrum resource-constrained, industry has proposed the notion of cooperating relay network.In the cooperating relay network, a plurality of portable terminals with single antenna are shared antenna, form virtual antenna array to obtain the diversity gain effect of MIMO (Multiple-Input Multiple-Output) system, improve system spectrum utilance and reliability.In the cooperating relay network, from a plurality of candidate relay nodes, select one or more optimum via node through relay selection and bear communication task, thereby make system have preferable efficiency of transmission, improve systematic function.Existing cooperating relay strategy mainly contain and amplify to transmit (Amplify-and-Forward, AF), decoding transmits that (Decode-and-Forward is DF) with cooperation coding (Code Cooperation, CC) three kinds of patterns.These relaying strategies are many from the error rate, power consumption, relaying position, channel gain equal angles, select one or more relaying in the network to communicate by letter with information source node cooperation completion.

But mobile communication need rely on radio signal propagation information, because the broadcast characteristic of wireless signal, Content of Communication is easy to by illegal third party's eavesdropping.See from the angle of physical layer; When the channel between information source node and the destination node (main channel) situation during inferior to channel (tapping channel) state between the node of information source node and eavesdropping; Information source node just almost can't be carried out secure communication with destination node; Content of Communication is obtained by eavesdropping side easily, brings massive losses to the user.Though existing relay selection scheme from plurality of target, is selected optimum relaying in the various network model, almost do not consider the safety problem in the radio communication, has ignored the hidden danger of information leakage.Though some documents have been considered the problem of relaying fail safe; Estimate the fail safe of cooperative system such as utilizing safe capacity (Secrecy Capacity); But generally only noticed the fail safe of selected forward relay; And do not consider and select suitable interference source to cooperate selected forward relay, improve the safe capacity of system.In the system model that these documents are studied, channel is memoryless channel, and the rapidity of fading is very slow or suppose that channel status is constant, and this and actual channel are not inconsistent; It is when the selection problem of the safe relaying of research, and target is more single, such as only considering safe capacity, and has ignored other important indicators in the collaborative network, like network lifetime, the error rate etc.

Recently,, improve the fail safe and the reliability of network, have the scholar to propose cooperation and disturb (Cooperative Jamming, CJ) strategy in order to prevent that information from being eavesdropped.This strategy utilization cooperation interference source (being called interfering nodes again) sends artificial interfere information, disturbs the eavesdropping of illegal wiretapping node to legal inter-node communication content, the probability that reduction information is eavesdropped.But strategy is disturbed in existing cooperation; Design object is comparatively single; Utilize one or more cooperation interfering nodes to disturb listener-in's eavesdropping simply: simplicity of design, the power consumption of single interfering nodes are little, but consider the interference of the interference signal of transmission to destination node, influence communication quality; A plurality of interfering nodes adopt beam forming (Beamforming) mode, and interfering signal power is concentrated on eavesdropping one side, and are less to the destination node influence, but its system complexity is higher, and adopt a plurality of interfering nodes, and power consumption is bigger.

Summary of the invention

The technical problem that (one) will solve

Shortcoming to prior art; The present invention moves single, the form simple question of eavesdropping conflicting mode target in the collaborative network in order to solve in the prior art; Optimal Distribution formula interference source system of selection in a kind of mobile collaborative network has been proposed; Take all factors into consideration the fail safe and the reliability of system, and guaranteed each item important indicator in the system.

(2) technical scheme

To achieve these goals, the invention provides Optimal Distribution formula interference source system of selection in a kind of mobile collaborative network, said method comprises step:

S1, information source node S periodically send training sequence, and destination node D and each cooperative node C carry out channel estimating;

S2 divides the channel power gain-state of each channel, the dump energy state and the information source relaying SR channel signal to noise ratio state of candidate's cooperative node according to the feedback result of training sequence, and confirms corresponding state transition probability; Each possible state when confirming candidate's cooperative node thus and corresponding state transition probability and candidate's cooperative node each possible state and the corresponding state transition probability during as via node as interfering nodes;

S3, judge information source node S whether can with destination node D direct communication, if can direct communication, then execution in step S4 carries out optimum interfering nodes and selects; Otherwise execution in step S8 carries out optimum relaying-interference to selecting;

S4, each possibly state computation disturb the priority symbol value accordingly during as interfering nodes for candidate's cooperative node, sets up interfering nodes state-priority symbol value table of comparisons;

S5; The time slot that data frame transfer is taken is divided into two sub-slots; At first sub-slots; According to each candidate's cooperative node present located state, state of living in when predicting second each candidate's cooperative node of sub-slots as interfering nodes, choosing the minimum candidate's cooperative node of interference priority symbol value with reference to said interfering nodes state-priority symbol value table of comparisons is optimum interfering nodes;

S6, at second sub-slots of data frame transfer, Frame sends to destination node by information source node, and the optimum interfering nodes of being chosen by first sub-slots simultaneously sends interference signal;

Whether S7 judges all data end of transmission after each data frame transfer is accomplished, if then finish communication; Otherwise go back to the transmission that step S5 carries out next Frame;

S8; Respectively from the angle of interfering nodes and the angle of via node; Disturb priority symbol value and relaying priority symbol value accordingly for each possible interfering nodes state of candidate's cooperative node or via node state computation, set up relaying-interference state-priority symbol value table of comparisons;

S9; The time slot that data frame transfer is taken is divided into two sub-slots; At first sub-slots; According to each candidate's cooperative node present located state, predict second each candidate's cooperative node of sub-slots state of living in, with reference to said relaying-interference state-priority symbol value table of comparisons being chosen the minimum candidate's cooperative node of interference priority symbol value is optimum interfering nodes; With reference to said relaying-interference state-priority symbol value table of comparisons being chosen the minimum candidate's cooperative node of relaying priority symbol value is optimum via node, and it is right that optimum via node of selecting and optimum interfering nodes constitute the relaying-interference of second sub-slots cooperation;

S10, at second sub-slots of data frame transfer, Frame is sent by information source node, and the optimum via node of choosing via first sub-slots sends to destination node, and the optimum interfering nodes of being chosen by first sub-slots simultaneously sends interference signal;

Whether S11 judges all data end of transmission after each data frame transfer is accomplished, if then finish communication; Otherwise go back to the transmission that step S9 carries out next Frame.

Preferably, among the step S2, the step that said feedback result according to training sequence is divided the state of each channel is specially:

Feedback result according to training sequence; Confirm the probability density function of the channel power gain of each channel; Be the corresponding threshold value of channel power gain setting of each channel, thereby the state of each channel is divided into a plurality of states corresponding with the channel power gain threshold value of this channel.

Preferably; Among the step S4; Based on Restless Multi-armed Bandit model; According to by the dump energy that disturbs eavesdropping JE channel and the channel power gain of disturbing purpose JD channel and candidate's cooperative node with transmit the electric weight of Frame needs consumption and system's remuneration that definite interfering nodes provides, each the possible state computation for candidate's cooperative node during as interfering nodes is disturbed the priority symbol value accordingly.

Preferably; Among the step S8; Based on Restless Multi-armed Bandit model; According to by disturb eavesdropping JE channel, disturb purpose JD channel, the error rate of the electric weight of the dump energy of the channel power gain of relaying eavesdropping RE channeling and trunking purpose RD channel and candidate's cooperative node, the consumption of Frame needs of transmission and SR channel and the definite system remuneration of relaying-interference to providing, for each possible interfering nodes state of candidate's cooperative node or via node state computation are disturbed priority symbol value and relaying priority symbol value accordingly.

Preferably; Among step S5 or the step S9; The dump energy of the channel power gain of each channel of RTS/CTS message feedback through enhancement mode, signal to noise ratio, candidate's cooperative node, and it is referred in the corresponding state, thereby confirm each candidate's cooperative node present located state.

Preferably, among step S7 or the step S11, judge whether end of transmission of all data through upper-layer protocol.

Preferably, said calculating is specially: said remuneration is converted into carries out first-order linear after the linear programming and relax, find the solution through the priority symbol value heuristic that is evolved into by original binary heuristic.

Preferably; Among the step S4; With the safe capacity target value conversion is the channel power gain of JE channel and channel power gain two desired values of JD channel; For the candidate's cooperative node m as interfering nodes, system's remuneration that said interfering nodes provides is:

$R_{r_{J_m}\left(t\right)}^{{\mathrm{\β}}_m\left(t\right)}=E({\mathrm{\β}}_m\left(t\right)\·({\mathrm{\μ}}_1\·o_{{\mathrm{JE}}_m}+{\mathrm{\μ}}_2\·o_{{\mathrm{JD}}_m}+{\mathrm{\μ}}_3\·z_m+{\mathrm{\μ}}_4\·H));$

Wherein, the expectation computing carried out according to candidate's cooperative node m current state and state transition probability of operator E () expression; β _m(t) whether expression candidate cooperative node m is selected as interfering nodes at moment t, if be selected, and β then _m(t)=1, as be not selected, then β _m(t)=0; With

Represent the channel power gain of JE channel and JD channel respectively; z _mThe dump energy of expression candidate cooperative node m; When H representes to transmit a Frame, the electric weight that need consume as candidate's cooperative node m of interfering nodes; μ _iBe the weights of each desired value, need to satisfy

μ ₁>=0, μ ₂≤0, μ ₁=-μ ₂, μ ₃>=0, μ ₄≤0.

Preferably; Among the step S8; Be channel power gain

and channel power gain

four desired values of RD channel of channel power gain

RE channel of channel power gain

JD channel of JE channel with the safe capacity target value conversion, said relaying-interference comprises the overall system remuneration that provides representes relaying-system's remuneration that interference centering interfering nodes provides of selecting and system's remuneration that the relaying of selecting-interference centering via node provides;

For candidate's cooperative node m, remuneration

that it provides during as interfering nodes and the remuneration

that provides during as via node are respectively:

${\mathrm{Re}}_{J_m}=E({\mathrm{\β}}_m^J\left(t\right)\·({\mathrm{\μ}}_1\·o_{{\mathrm{JE}}_m}+{\mathrm{\μ}}_2\·o_{{\mathrm{JD}}_m}+{\mathrm{\μ}}_3\·z_m+{\mathrm{\μ}}_4\·H));$

${\mathrm{Re}}_{R_m}=E({\mathrm{\β}}_m^R\left(t\right)\·({\mathrm{\τ}}_1\·o_{{\mathrm{RE}}_m}+{\mathrm{\τ}}_2\·o_{{\mathrm{RD}}_m}+{\mathrm{\τ}}_3\·z_m+{\mathrm{\τ}}_4\·P_b));$

Wherein, the expectation computing carried out according to current state and state transition probability of operator E () expression; Represent respectively whether candidate's cooperative node m is selected as interfering nodes or via node at moment t, if be selected as interfering nodes, then

As be not selected as interfering nodes, then

If be selected as via node, then

As be not selected as via node, then

With

Represent the channel power gain of JE channel, JD channel, RE channel and RD channel respectively, z _mThe dump energy of expression candidate cooperative node m; When H representes to transmit a Frame, the electric weight that need consume as candidate's cooperative node m of interfering nodes; P _bThe error rate of expression SR channel; μ _i, τ _iBe the weights of each desired value, need to satisfy

μ ₁>=0, μ ₂≤0, μ ₁=-μ ₂, μ ₃>=0, μ ₄≤0;

τ ₃>=0, τ ₄≤0, τ ₁≤0, τ ₂>=0, τ ₁=-τ ₂

Preferably; Among step S5 or the step S9, said choosing disturbs the minimum candidate's cooperative node of priority symbol value to be specially: choose and disturb minimum candidate's cooperative node of priority symbol value or choose preceding several candidate's cooperative nodes that disturb the priority symbol value minimum;

Among the step S9, the said minimum candidate's cooperative node of relaying priority symbol value of choosing is specially: choose minimum candidate's cooperative node of relaying priority symbol value or choose minimum preceding several candidate's cooperative nodes of relaying priority symbol value.

(3) beneficial effect

In the scheme of the present invention; To the problem that moves secure communication in the collaborative network; Optimal Distribution formula interference source system of selection in a kind of mobile collaborative network is provided; This method according to the concrete state of channel select optimum disturb or optimum relaying-interferences to disturbing the illegally eavesdropping behavior of node in the network, thereby protect the proper communication of legal node, ensured communication security.In the method for the present invention, also, prolonged network lifetime through the electric quantity consumption of reasonable distribution candidate cooperative node; Also take all factors into consideration transmission error rates and system safety capacity simultaneously, ensured communication quality.

Description of drawings

Fig. 1 is for moving the network model sketch map of a typical scene of collaborative network among the present invention;

Fig. 2 is the flow chart of the method among the present invention;

Fig. 3 is the network application scene sketch map in the embodiments of the invention 1;

Fig. 4 is the network application scene sketch map in the embodiments of the invention 2;

Fig. 5 is the comparison diagram that the average safe capacity of three kinds of different schemes in the simulation example of the present invention changes with candidate's cooperative node number;

Fig. 6 is the comparison diagram that the average safe capacity of three kinds of different schemes in another simulation example of the present invention changes with candidate's cooperative node number;

Fig. 7 is the system's compensatory time accumulation and the time dependent comparison diagram of optimum interfering nodes selection scheme and channel quiet scheme in the simulation example of the present invention;

Fig. 8 is the comparison diagram of the network lifetime of optimum interfering nodes selection scheme and channel quiet scheme in the simulation example of the present invention with different changes of threshold;

Fig. 9 is the comparison diagram that the average safe capacity of three kinds of different schemes in another simulation example of the present invention changes with candidate's cooperative node number;

Figure 10 is the accumulation of overall system compensatory time and the time dependent comparison diagram of three kinds of different schemes in another simulation example of the present invention.

Embodiment

To combine the accompanying drawing in the embodiment of the invention below, the technical scheme in the embodiment of the invention is carried out clear, intactly description, obviously, described embodiment is a part of embodiment of the present invention, rather than whole embodiment.Based on the embodiment among the present invention, the every other embodiment that those of ordinary skills are obtained under the prerequisite of not making creative work belongs to the scope that the present invention protects.

At first, the concrete network model of a typical application scene of the present invention is as shown in Figure 1, in numerous nodes, has some information source node S to communicate by letter with a destination node D.Have several candidate's cooperative nodes C, C can be operated under jamming pattern J and the repeater mode R.In network, also there are some eavesdropping node E.S maybe be by some E eavesdroppings with communicating by letter of D, and the present invention is primarily aimed at the analysis of CJ mode expansion, and supposes the number of the number of the interfering nodes J that selects greater than eavesdropping node E, and a plurality of E is separate.Then the safe capacity SC between SD is expressed as:

$\mathrm{SC}=\max \underset{i}{\min }\{C_M-C_{E_i}\};$

This safe capacity SC is illustrated in E and almost can't obtains under the situation of any information, the information rate of maximum between S and the D.C wherein _MThe channel capacity of expression main channel (SD channel), Represent i tapping channel (SE _i) channel capacity.Particularly:

P wherein _SBe the transmitted power of information source node S, h _SDBe the channel gain of SD channel, σ ²Be destination node D and eavesdropping node E _iThe noise power of receiving,

Expression is by SE _iThe vector that the channel gain of channel constitutes, H _JEBe the channel gain formation matrix of a plurality of JE channels,

The i row of representing it, w represent that each interfering nodes J sends the weight vector of interference signal,

Expression conjugate transpose operator.It is pointed out that the D in this scene also can be a relaying R in the multi-hop collaborative network.

After single S, single D and a plurality of E confirm, as the transmitted power P of S _sAnd noise power σ ²When constant, make safe capacity maximize, the selection of promptly a plurality of interfering nodes J makes that safe capacity maximizes between the SD:

Especially, consider a kind of special situation, S and a plurality of via node R cooperation are sent information to D, and then the safe capacity between the RD can be expressed as:

Wherein, v representes that each R sends the weight vector of signal, h _RDThe vector that constitutes for channel gain by the RD channel.With preamble in like manner, after single D confirms, make that the safe capacity between RD is maximum, i.e. relaying-interferences to (R, selection J) makes safe capacity maximize:

In order to simplify analysis, the scene among Fig. 1 is reduced to embodiment 1 and embodiment 2, the method flow among the present invention is specifically as shown in Figure 2.To at real dimension deployment analysis, method of the present invention be described based on the scene among these two embodiment below.The person skilled of this area should be able to recognize that embodiments of the invention are property explanation as an example only, and practical application of the present invention also not only is confined to embodiment 1 and embodiment 2 described scenes.

Embodiment 1

The scene of the embodiment of the invention 1 is as shown in Figure 3, and each one of information source node S, destination node D and eavesdropping node E is arranged in network, candidate's cooperative node C several, each candidate's cooperative node only is considered as the situation (being expressed as C/J among the figure) of interfering nodes J.Each node all has been equipped with single antenna, and internodal channel is rayleigh fading channel, and bandwidth is equal, and separate, and the transmitted power of information source node is P _S, the transmitted power of each interfering nodes is P _JIt is said that a certain jumping or some jumping figures that the scene of embodiment 1 can be applicable in the multihop network are defeated; Especially; When main channel (SD channel) environment during inferior to tapping channel (SE channel); Traditional approach is can't realize S to the secure communication between the D, can only rely on to introduce via node and transmit data, and the power of via node is greater than interfering nodes usually.The present invention is directed to this problem and adopt the mode of introducing interfering nodes; But information source node S can with the situation of destination node D direct communication under; The present invention carries out optimum interfering nodes and selects; Information source node S and destination node D are communicated under the assistance of one or several interfering nodes, can effectively cut down the consumption of energy, improve safe capacity.

Destination node D also can be a via node in the scene of the embodiment of the invention 1, in the phase I of relaying cooperation, during the information source node broadcast singal, selects a suitable interfering nodes to send interference signal, and the information that can prevent is obtained by the side of eavesdropping E.

The flow process of optimum interfering nodes system of selection is shown in the flow chart of Fig. 2 left side:

Before data frame transfer began, information source node was periodically sent training sequence, and destination node and each candidate's cooperative node (being interfering nodes) carry out channel estimating.

According to the feedback result of training sequence, confirm the channel power gain o=|h| of each channel ²(h be respective channels gain) probability density function p (o), choose N+1 threshold value O ₀, O ₁, O ₂..., O _N, as channel power gain O _n＜o＜O _N+1The time, judge that channel is in state n, according to this method, channel is divided into N state, each channel is Markov model for a state of Markov chain with Channel Modeling.Analyze for simplifying, make the Stationary Distribution probability π of each state _nEquate, then have:

$\left\{\begin{array}{c}{\mathrm{\&pi;}}_n={\&Integral;}_{O_n}^{O_{n+1}}p\left(o\right)\mathrm{do}=\frac{1}{\mathrm{<figure-callout\; id="0"\; label="N\; O"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">N</figure-callout>}}& \\ O_{}{}_0=0,O_N=+\&infin;\end{array}\right.$

Can solve each threshold value O by above-mentioned equation group _nThe state dividing mode of it is pointed out that is not confined to above this method.

Feedback result and historical observation through training sequence can also statistical induction go out the state transition probability between each state.Analyze for simplifying, suppose that here state transitions only occurs between the adjacent states.

In the scene of embodiment 1, because candidate's cooperative node (being made as m) is only as interfering nodes, correlated channels is only considered JE channel and JD channel, is a with the JE channel distribution _JIndividual state, the corresponding states space State transition probability matrix is expressed as:

${\mathrm{\&Omega;}}_{J_m}\left(t\right)={\left[{\mathrm{\&omega;}}_{x_{J_m}{y}_{J_m}}\left(t\right)\right]}_{a_J\&times;a_J};$

Wherein ${\mathrm{\&omega;}}_{x_{J_m}{y}_{J_m}}\left(t\right)=\mathrm{Pr}(o_{{\mathrm{JE}}_m}(t+1)=y_{J_m}|o_{{\mathrm{JE}}_m}\left(t\right)=x_{J_m}),x_{J_m},y_{J_m}\&Element;A_J;$ The cooperate channel power gain of JE channel of interfering nodes m of expression candidate

At t constantly from state

Transfer to state

Probability.

Similarly, be b with the JD channel distribution _JIndividual state, the corresponding states space

State transition probability matrix is expressed as:

${\mathrm{\&Psi;}}_{J_m}\left(t\right)={\left[{\mathrm{\&psi;}}_{u_{J_m}{v}_{J_m}}\left(t\right)\right]}_{b_J\&times;b_J};$

Wherein ${\mathrm{\&psi;}}_{u_{J_m}{v}_{J_m}}\left(t\right)=\mathrm{Pr}(o_{{\mathrm{JD}}_m}(t+1)=v_{J_m}|o_{{\mathrm{JD}}_m}\left(t\right)=u_{J_m}),u_{J_m},v_{J_m}\&Element;B_J;$ The cooperate channel power gain of JD channel of interfering nodes m of expression candidate

At t constantly from state

Transfer to state

Probability.

Similar with the channel status classifying method, can the dump energy of candidate's cooperative node m be classified as some states, and the statistical induction transition probability that does well.The dump energy of candidate's interfering nodes is divided into c _JIndividual state, the corresponding states space

State transition probability matrix is expressed as:

${\mathrm{\&Theta;}}_{J_m}^{{\mathrm{\&beta;}}_m}\left(t\right)={\left[{\mathrm{\&theta;}}_{e_{J_m}{f}_{J_m}}^{{\mathrm{\&beta;}}_m}\left(t\right)\right]}_{c_J\&times;c_J};$

Wherein ${\mathrm{\&theta;}}_{e_{J_m}{f}_{J_m}}^{{\mathrm{\&beta;}}_m}\left(t\right)=\mathrm{Pr}(z_m(t+1)=f_{J_m}|z_m\left(t\right)=e_{J_m}),e_{J_m},f_{J_m}\&Element;C_J;$ The expression candidate cooperate the dump energy zm of interfering nodes m at t constantly from state Transfer to state

Probability.Whether the transition probability that it should be noted that dump energy is selected for use relevant (selected time spent β with this candidate's interfering nodes _m=1, not selected time spent β _m=0).If candidate's interfering nodes is selected for use, then its probability that jumps to low state of charge will increase.

Because the dump energy of candidate's interfering nodes and channel status are independent; And each channel is separate; The state of interfering nodes m can be determined by JE channel status associated therewith, JD channel status and dump energy state jointly so the candidate cooperates, and the candidate interfering nodes m that cooperates can be expressed as at the state of moment t:

$r_{J_m}\left(t\right)=[o_{{\mathrm{JE}}_m}\left(t\right),o_{{\mathrm{JD}}_m}\left(t\right),z_m\left(t\right)];$

Its state transition probability matrix can be expressed as:

$P_{J_m}\left(t\right)={[{\mathrm{\&omega;}}_{x_{J_m}{y}_{J_m}}\left(t\right),{\mathrm{\&psi;}}_{u_{J_m}{v}_{J_m}}\left(t\right),{\mathrm{\&theta;}}_{e_{J_m}{f}_{J_m}}^{{\mathrm{\&beta;}}_m}\left(t\right)]}_{q_J\&times;q_J},q_J=a_J\&times;b_J\&times;c_J;$

Matrix

element

indicates the candidate node collaboration interference from the state at time t m (JE channel state

JD channel state

and the remaining power status

) to the state

(JE channel state

JD channel state and the remaining power status

) probability.

After having divided state and having confirmed corresponding state transition probability, judge information source node S whether can with destination node D direct communication, if can direct communication, then carry out the selection of optimum interfering nodes; If can't direct communication, then scene becomes the scene that embodiment 2 considers, carries out the right selection of relaying-interferences, and this partial content will introduction in detail in embodiment 2.

The selection of optimum interfering nodes should be satisfied following two conditions:

1, guarantees that SD is with bigger safe capacity transmission data.Behind selected certain cooperation interfering nodes J, under real dimension, the safe capacity between SD can be expressed as:

When only having single E and single J, can derive by following formula:

$\mathrm{SC}=\max \{\frac{1}{2}{\log }_2(1+\frac{{\left|h_{\mathrm{SD}}\right|}^2{P}_S}{{\left|h_{\mathrm{JD}}\right|}^2{P}_J+{\mathrm{\&sigma;}}_1^2})-\frac{1}{2}{\log }_2(1+\frac{{\left|h_{\mathrm{SE}}\right|}^2{P}_S}{{\left|h_{\mathrm{JE}}\right|}^2{P}_J+{\mathrm{\&sigma;}}_1^2}),0\};$

Wherein

is the transmitted power of interfering nodes;

and

is respectively the additive white Gaussian noise power that D and E receive, and its value does not change in time.

Select optimum interference source, should make safe capacity maximum:

Correspondingly, be under the single situation, can derive by following formula at E and J:

$\arg \max \{\frac{P_S{\left|h_{\mathrm{SD}}\right|}^2}{P_J{\left|h_{\mathrm{JD}}\right|}^2}\&CenterDot;\frac{P_J{\left|h_{\mathrm{JE}}\right|}^2}{P_S{\left|h_{\mathrm{SE}}\right|}^2}\}$

$\&DoubleRightArrow;\arg \max (\frac{P_S{\left|h_{\mathrm{SD}}\right|}^2}{P_S{\left|h_{\mathrm{SE}}\right|}^2}\&CenterDot;\frac{P_J{\left|h_{\mathrm{JE}}\right|}^2}{P_J{\left|h_{\mathrm{JD}}\right|}^2})$

$\&DoubleRightArrow;\arg \max \left(\frac{{\left|h_{\mathrm{JE}}\right|}^2}{{\left|h_{\mathrm{JD}}\right|}^2}\right)$

J∈{J _m}

So aims of systems value safe capacity can be converted into | h _JE| ²=o _JEWith | h _JD| ²=o _JDThese two desired values.Bigger safe capacity can guarantee can be with bigger information rate transmission data by eavesdropping node decoder obtaining communication content between S and the D.

2, the balanced candidate electric quantity consumption of interfering nodes of cooperating.In moving collaborative network; It is very big that each node relies on battery powered possibility, and the electric weight of battery is limited, a good cooperation interfering nodes selection scheme; Should be able to balanced each candidate the cooperate electric quantity consumption of interfering nodes; Avoid the too fast approach exhaustion of some node electric weight, prolong the service time of candidate's cooperative node, and then prolong network lifetime.

Comprehensive above the analysis, in Restless Multi-armed Bandit model, system's remuneration that candidate's cooperative node m provides during as interfering nodes just can be defined as:

$R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m\left(t\right)}=E({\mathrm{\&beta;}}_m\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

The wherein expectation computing carried out according to candidate's cooperative node m current state and state transition probability of operator E () expression.β _m(t) represent whether this candidate's cooperative node m is selected as interfering nodes, if be selected, β then _m(t)=1, as be not selected, then β _m(t)=0.o _JEAnd o _JDRepresent respectively JE channel and JD channel channel power gain (

With

The channel gain of expression respective channels), in the scene of embodiment 1, like the target analysis that preceding text carried out, sign system safety capacity desired value is united in these two channel power gains; z _mThe dump energy of representing this candidate's cooperative node m; H representes to transmit a Frame, and the electric weight that need consume as candidate's cooperative node m of interfering nodes is with o _JEAnd o _JD, P _JAnd Frame length etc. is relevant.μ _i(i=1,2,3,4) are the weights of each desired value, can adjust accordingly according to the specific requirement of system, but need satisfy μ ₁>=0, μ ₂≤0, μ ₁=-μ ₂, μ ₃>=0, μ ₄≤0,

After the definition that it should be noted that reward function has embodied and has selected optimum cooperation interfering nodes, the remuneration that system can obtain, herein, and system's remuneration and safe capacity desired value indirect correlation, directly related with the dump energy desired value.

Optimum cooperation interfering nodes selection strategy u _J(u _J∈ U _J, U _JExpression Markov policy set) should be able to maximize system's remuneration of whole data transfer phase, that is:

$L_J^{*}=\underset{u_J\&Element;U_J}{\max }{E}_{u_J}\left[\underset{t=0}{\overset{T-1}{\mathrm{\&Sigma;}}}(R_{r_{J_1}\left(t\right)}^{{\mathrm{\&beta;}}_1\left(t\right)}+R_{r_{J_2}\left(t\right)}^{{\mathrm{\&beta;}}_2\left(t\right)}+\&CenterDot;\&CenterDot;\&CenterDot;+R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m\left(t\right)}){\mathrm{\&eta;}}_J^t\right];$

System's remuneration has time decline property, η _JExpress time decline parameter.

is t candidate's residing state of interfering nodes m of cooperating constantly.Following formula is converted into the linear programming expression formula:

$\left(\mathrm{LP}\right)L_J^{*}=\underset{g_J\&Element;G_J}{\max }\underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{r_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}\underset{{\mathrm{\&beta;}}_m\&Element;\left\{\mathrm{0,1}\right\}}{\mathrm{\&Sigma;}}{R}_{r_{J_m}}^{{\mathrm{\&beta;}}_m}{g}_{r_{J_m}}^{{\mathrm{\&beta;}}_m};$

In formula, M representes the set of candidate's interfering nodes, The state space of expression candidate interfering nodes m, $G_J=\{g_J={\left(g_{r_{J_m}}^{{\mathrm{\&beta;}}_m}\left(u_J\right)\right)}_{r_{J_m}\&Element;S_{J_m},{\mathrm{\&beta;}}_m\&Element;\left\{\mathrm{0,1}\right\},m\&Element;M}|u_J\&Element;U_J\}$ By performance vectors g _JAt all Markov policy u _J∈ U _JThe corresponding performance zones in following expansion back.Variable

The expression candidate cooperates interfering nodes m according to Markov policy u _J, when its state does

Shi Zhihang moves β _mThe desired value of total time.

If

The cooperate initial condition of interfering nodes m of expression candidate does

Probability,

Expression initial condition probability vector,

Expression restless bandit polyhedron exists The projection in space, K _JThe cooperation interfering nodes number of indicating to select, then above-mentioned linear programming expression formula can first-order linear relax for:

$\left({\mathrm{LP}}^1\right){\mathrm{<figure-callout\; id="1"\; label="L\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">L</figure-callout>}}_J^{}=\max \underset{m=M}{\mathrm{\&Sigma;}}\underset{r_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}\underset{{\mathrm{\&beta;}}_m\&Element;\left\{\mathrm{0,1}\right\}}{\mathrm{\&Sigma;}}{R}_{r_{J_m}}^{{\mathrm{\&beta;}}_m}{g}_{r_{J_m}}^{{\mathrm{\&beta;}}_m}$

subject?to

$g_{J_m}\&Element;{\mathrm{\&rho;}}_{J_m}^{-1},m\&Element;M$

$\underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{r_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{g}_{{\mathrm{<figure-callout\; id="1"\; label="r\; J\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^1=\frac{K_J}{1-{\mathrm{\&eta;}}_J}$

This first-order linear relaxes the optimal basic solution of problem finds the solution through original binary heuristic (Primal-dual Heuristic) method for

again, and the dual expression formula that single order relaxes linear programming is:

$\left(D^1\right){\mathrm{<figure-callout\; id="1"\; label="L\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">L</figure-callout>}}_J^{}=\min \underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{S_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{\mathrm{\&pi;}}_{S_{J_m}}{\mathrm{\&gamma;}}_{S_{J_m}}+\frac{K_J}{1-{\mathrm{\&eta;}}_J}{\mathrm{\&gamma;}}_J$

subject?to

${\mathrm{\&gamma;}}_{r_{J_m}}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{p}_{r_{J_m}{s}_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;R_{{\mathrm{<figure-callout\; id="0"\; label="r\; J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^0,r_{J_m}\&Element;S_{J_m},m\&Element;M$

${\mathrm{\&gamma;}}_{r_{J_m}}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{p}_{r_{J_m}{s}_{{\mathrm{<figure-callout\; id="1"\; label="J\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^1{\mathrm{\&gamma;}}_{s_{J_m}}+{\mathrm{\&gamma;}}_J\&GreaterEqual;R_{{\mathrm{<figure-callout\; id="1"\; label="r\; J\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^1,r_{J_m}\&Element;S_{J_m},m\&Element;M$

γ _J≥0

Wherein,

expression candidate's not selected time spent of interfering nodes m jumps to the probability of state

from state

,

expression candidate's selected time spent of interfering nodes m jump to the probability of state

from state

; Not system's remuneration of selected time spent of

expression candidate interfering nodes m, system's remuneration of

expression candidate's selected time spent of interfering nodes m. separates for satisfying condition of requiring.Need to prove;

is the state of current time;

is next state constantly;

and

can be identical, and promptly candidate's interfering nodes is retained in previous status.

To separate to

be example with first constraints if above-mentioned single order relaxes dual expression formula optimum dual of linear programming:

${\mathrm{\&gamma;}}_{r_{J_m}}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{p}_{r_{J_m}{s}_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;R_{{\mathrm{<figure-callout\; id="0"\; label="r\; J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^0,r_{J_m}\&Element;S_{J_m},m\&Element;M;$

Candidate's interfering nodes is residing possibly to have 1,2,3...q by state _JBe total to q _J(q _J=a _J* b _J* c _J) individual,

Represent this q _JIn the individual state one, homographic solution

Get successively

First constraints can be write as respectively:

${\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;\{\mathrm{1,2,3},\&CenterDot;\&CenterDot;\&CenterDot;,q_J\}}{\mathrm{\&Sigma;}}{p}_{1s_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;{\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_1^0$

$\&DoubleRightArrow;{\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J(p_{11}^0{\mathrm{\&gamma;}}_1+p_{12}^0{\mathrm{\&gamma;}}_2+p_{13}^0{\mathrm{\&gamma;}}_3+\&CenterDot;\&CenterDot;\&CenterDot;+p_{1{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J})\&GreaterEqual;{\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_1^0;$

$\&DoubleRightArrow;(1-{\mathrm{\&eta;}}_J{p}_{11}^0){\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J{p}_{12}^0{\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{p}_{13}^0{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{1{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_1^0$

${\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;\{\mathrm{1.2.3}.\&CenterDot;\&CenterDot;\&CenterDot;,q_J\}}{\mathrm{\&Sigma;}}{p}_{2s_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_2^0$

$\&DoubleRightArrow;{\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J(p_{21}^0{\mathrm{\&gamma;}}_1+p_{22}^0{\mathrm{\&gamma;}}_2+p_{23}^0{\mathrm{\&gamma;}}_3+\&CenterDot;\&CenterDot;\&CenterDot;+p_{2{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J})\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_2^0;$

$\&DoubleRightArrow;-{\mathrm{\&eta;}}_J{p}_{21}^0{\mathrm{\&gamma;}}_1+(1+{\mathrm{\&eta;}}_J{p}_{22}^0){\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{p}_{23}^0{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{2{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_2^0$

${\mathrm{\&gamma;}}_3-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;\{\mathrm{1,2,3},\&CenterDot;\&CenterDot;\&CenterDot;,q_J\}}{\mathrm{\&Sigma;}}{p}_{3s_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;{\mathrm{<figure-callout\; id="3"\; label="R"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">R</figure-callout>}}_3^0$

$\&DoubleRightArrow;{\mathrm{\&gamma;}}_3-{\mathrm{\&eta;}}_J(p_{31}^0{\mathrm{\&gamma;}}_1+p_{32}^0{\mathrm{\&gamma;}}_2+p_{33}^0{\mathrm{\&gamma;}}_3+\&CenterDot;\&CenterDot;\&CenterDot;+p_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J})\&GreaterEqual;{\mathrm{<figure-callout\; id="3"\; label="R"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">R</figure-callout>}}_3^0;$

$\&DoubleRightArrow;-{\mathrm{\&eta;}}_J{p}_{31}^0{\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J{p}_{32}^0{\mathrm{\&gamma;}}_2+(1-{\mathrm{\&eta;}}_J{p}_{33}^0){\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="3"\; label="R"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">R</figure-callout>}}_3^0$

......；

${\mathrm{\&gamma;}}_{q_J}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;\{\mathrm{1.2.3}.\&CenterDot;\&CenterDot;\&CenterDot;,q_J\}}{\mathrm{\&Sigma;}}{p}_{q_J{s}_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{J_m}}\&GreaterEqual;{\mathrm{<figure-callout\; id="0"\; label="R\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_{q_J}^{{}_{}}$

$\&DoubleRightArrow;{\mathrm{\&gamma;}}_{q_J}-{\mathrm{\&eta;}}_J({\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}{\mathrm{\&gamma;}}_1+{\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}{\mathrm{\&gamma;}}_2+{\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}{\mathrm{\&gamma;}}_3+\&CenterDot;\&CenterDot;\&CenterDot;+p_{q_{\mathrm{<figure-callout\; id="0"\; label="J\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}{q}_J{}_{}}^0{\mathrm{\&gamma;}}_{q_J})\&GreaterEqual;{\mathrm{<figure-callout\; id="0"\; label="R\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_{q_J}^{{}_{}}.$

$\&DoubleRightArrow;-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}{\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}{\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;+(1-{\mathrm{\&eta;}}_J{p}_{q_J{q}_J}^0){\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="0"\; label="R\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_{q_J}^{{}_{}}$

Through concluding, formula is arranged:

$(1-{\mathrm{\&eta;}}_J{p}_{11}^0){\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_1{p}_{12}^0{\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{p}_{13}^0{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{1{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_1^0$

$-{\mathrm{\&eta;}}_J{p}_{21}^0{\mathrm{\&gamma;}}_1+(1-{\mathrm{\&eta;}}_J{p}_{22}^0){\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{p}_{23}^0{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{2{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_2^0$

$-{\mathrm{\&eta;}}_J{p}_{31}^0{\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J{p}_{32}^0{\mathrm{\&gamma;}}_2+(1-{\mathrm{\&eta;}}_J{p}_{33}^0){\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;-{\mathrm{\&eta;}}_J{p}_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0{\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="3"\; label="R"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">R</figure-callout>}}_3^0;$

...

$-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}{\mathrm{\&gamma;}}_1-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}{\mathrm{\&gamma;}}_2-{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}{\mathrm{\&gamma;}}_3-\&CenterDot;\&CenterDot;\&CenterDot;+(1-{\mathrm{\&eta;}}_J{p}_{q_J{q}_J}^0){\mathrm{\&gamma;}}_{q_J}\&GreaterEqual;{\mathrm{<figure-callout\; id="0"\; label="R\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_{q_J}^{{}_{}}$

Can be write as matrix form, had:

$\left[\begin{array}{ccccc}1-{\mathrm{\&eta;}}_J{p}_{11}^0& -{\mathrm{\&eta;}}_J{p}_{12}^0& -{\mathrm{\&eta;}}_J{p}_{13}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{1q_J}^0\\ -{\mathrm{\&eta;}}_J{p}_{21}^0& 1-{\mathrm{\&eta;}}_J{p}_{22}^0& -{\mathrm{\&eta;}}_J{p}_{23}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{2q_J}^0\\ -{\mathrm{\&eta;}}_J{p}_{31}^0& -{\mathrm{\&eta;}}_J{p}_{32}^0& 1-{\mathrm{\&eta;}}_J{p}_{33}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0\\ \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;\\ -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}& -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}& -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}& \&CenterDot;\&CenterDot;\&CenterDot;& 1-{\mathrm{\&eta;}}_J{p}_{q_{\mathrm{<figure-callout\; id="0"\; label="J\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}{q}_J{}_{}}^0\end{array}\right]\left[\begin{array}{c}{\mathrm{\&gamma;}}_1\\ {\mathrm{\&gamma;}}_2\\ {\mathrm{\&gamma;}}_3\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ {\mathrm{\&gamma;}}_{q_J}\end{array}\right]\&GreaterEqual;\left[\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_1^0\\ {\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_2^0\\ {\mathrm{<figure-callout\; id="3"\; label="R"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">R</figure-callout>}}_3^0\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ {\mathrm{<figure-callout\; id="0"\; label="R\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_{q_J}^{{}_{}}\end{array}\right].$

The coefficient matrix of

can further be converted into:

$\left[\begin{array}{ccccc}1-{\mathrm{\&eta;}}_J{p}_{11}^0& -{\mathrm{\&eta;}}_J{p}_{12}^0& -{\mathrm{\&eta;}}_J{p}_{13}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{1q_J}^0\\ -{\mathrm{\&eta;}}_J{p}_{21}^0& 1-{\mathrm{\&eta;}}_J{p}_{22}^0& -{\mathrm{\&eta;}}_J{p}_{23}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{2q_J}^0\\ -{\mathrm{\&eta;}}_J{p}_{31}^0& -{\mathrm{\&eta;}}_J{p}_{32}^0& 1-{\mathrm{\&eta;}}_J{p}_{33}^0& \&CenterDot;\&CenterDot;\&CenterDot;& -{\mathrm{\&eta;}}_J{p}_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0\\ \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;\\ -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}& -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}& -{\mathrm{\&eta;}}_J{\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}& \&CenterDot;\&CenterDot;\&CenterDot;& 1-{\mathrm{\&eta;}}_J{p}_{q_{\mathrm{<figure-callout\; id="0"\; label="J\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}{q}_J{}_{}}^0\end{array}\right]$

$=\left[\begin{array}{ccccc}1& 0& 0& \&CenterDot;\&CenterDot;\&CenterDot;& 0\\ 0& 1& 0& \&CenterDot;\&CenterDot;\&CenterDot;& 0\\ 0& 0& 1& \&CenterDot;\&CenterDot;\&CenterDot;& 0\\ \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;\\ 0& 0& 0& \&CenterDot;\&CenterDot;\&CenterDot;& 1\end{array}\right]-{\mathrm{\&eta;}}_J\left[\begin{array}{ccccc}{p}_{11}^0& p_{12}^0& p_{13}^0& \&CenterDot;\&CenterDot;\&CenterDot;& p_{1q_J}^0\\ p_{21}^0& p_{22}^0& p_{23}^0& \&CenterDot;\&CenterDot;\&CenterDot;& p_{2q_J}^0\\ p_{31}^0& p_{32}^0& p_{33}^0& \&CenterDot;\&CenterDot;\&CenterDot;& p_{3{\mathrm{<figure-callout\; id="0"\; label="q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">q</figure-callout>}}_J}^0\\ \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;\\ {\mathrm{<figure-callout\; id="1"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}1}& {\mathrm{<figure-callout\; id="2"\; label="p\; q\; J"\; filenames="HDA0000085225250000024.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}2}& {\mathrm{<figure-callout\; id="3"\; label="p\; q\; J"\; filenames="HDA0000085225250000023.png,HDA0000085225250000024.png"\; state="\{\{state\}\}">p</figure-callout>}}_{q_J}^{{}_{}3}& \&CenterDot;\&CenterDot;\&CenterDot;& p_{q_{\mathrm{<figure-callout\; id="0"\; label="J\; q\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}{q}_J{}_{}}^0\end{array}\right]$

$=E_{q_J}-{\mathrm{\&eta;}}_{\mathrm{<figure-callout\; id="0"\; label="J\; P\; J"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}{P}_J^{}{}_0^{}$

Expression q _J* q _JUnit matrix,

State transition probability matrix when expression candidate interfering nodes is not selected, η _JExpress time decline parameter.

calculated by the definition of system's reward function,

be the optimal solution that meets constraints that need solve.

If

expression reduces overhead factor, the optimum overhead factor

that reduces is defined as:

${\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{J_m}}}}^0=\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{r_{J_m}}}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{p}_{r_{J_m}{s}_{{\mathrm{<figure-callout\; id="0"\; label="J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^0\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{s_{J_m}}}-R_{{\mathrm{<figure-callout\; id="0"\; label="r\; J\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^0$

${\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{J_m}}}}^1=\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{r_{J_m}}}-{\mathrm{\&eta;}}_J\underset{s_{J_m}\&Element;S_{J_m}}{\mathrm{\&Sigma;}}{p}_{r_{J_m}{s}_{{\mathrm{<figure-callout\; id="1"\; label="J\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">J</figure-callout>}}_m}}^1\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{s_{J_m}}}+\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_J}-R_{{\mathrm{<figure-callout\; id="1"\; label="r\; J\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">r</figure-callout>}}_{J_m}}^1$

Can the solution procedure of original binary heuristic be subdivided into two stages: primitive stage and binary stage.

In the primitive stage, if candidate's interfering nodes satisfies

then it will be regarded as candidate's optimum interfering nodes.If satisfy the both candidate nodes of this condition k is arranged _JIndividual, can be divided into following three kinds of situation and also discuss respectively:

If k _J=K _J, then all optimum interfering nodes are selected to finish.

If k _J＜K _J, all k _JIndividual node is chosen as optimum interfering nodes, remaining K _J-k _JIndividual node was selected in the binary stage.In the binary stage, the remaining K that select _J-k _JIndividual node its

Be 0, when being in

When candidate's interfering nodes m of state is selected, choose action β in execution _m=1 total time expectation Under the situation of increase equal increments (as increasing a unit), it reduces overhead factor

Big more, its single order relaxes the desired value of linear programming

To reduce speed just big more.So in second stage, K _J-k _JIndividual

The candidate's that value is less node is optimum interfering nodes.

If k _J＞K _J, then do not take any action in the primitive stage, all optimum interfering nodes were selected in the binary stage.The binary stage will be at k _JIndividual

Both candidate nodes in select optimum interfering nodes, when being in

When candidate's interfering nodes m of state is not selected, do not choose action β in execution _m=0 total time expectation

Under the situation of increase equal increments (as increasing a unit), it reduces overhead factor

Big more, its single order relaxes the desired value of linear programming

To reduce speed just big more, so in second stage, K _JIndividual have bigger

The both candidate nodes of value is optimum interfering nodes.

Original binary heuristic can also further be converted into priority symbol value heuristic (Priority-index Heuristic), and the priority symbol value (Priority-index) of the correspondence of the candidate's cooperative node that is in state

during as interfering nodes is defined as:

${\mathrm{\&chi;}}_{r_{J_m}}={\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{J_m}}}}^1-{\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{J_m}}}}^0.$

Adopt above method, be candidate's corresponding priority symbol value of residing each the possible state computation of interfering nodes of cooperating, and set up the table of comparisons.So far the preparatory stage accomplishes before the transfer of data.

At data transfer phase; Transmit each Frame and take a time slot; Each time slot is divided into two sub-slots again, at first sub-slots, estimates the channel power gain (supposing that here the channel power gain between J and the E is known) of each channel through sending enhancement mode RTS/CTS (Request To Send/Clear To Send) between J and the D; And it is referred in the corresponding state; The dump energy state that combines candidate's cooperative node is again confirmed the residing state of each candidate's cooperative node, and each candidate's cooperative node is according to own residing state; In interfering nodes state-priority symbol value table of comparisons, check in corresponding priority symbol value, and inform information source node.Information source node is selected priority symbol value minimum one or several less (K _J＜M) candidate's cooperative node is as the interfering nodes of next son time slot cooperation.It is to be noted; Enhancement mode RTS/CTS has also carried the channel condition information (mainly comprising channel power gain, signal to noise ratio etc.) of candidate's cooperative node J and destination node D feedback, the dump energy information of candidate's cooperative node etc. except bearing the task of shaking hands here.

At second sub-slot transmission Frame: Frame sends to destination node D by information source node S, and the optimum interfering nodes J that is chosen by first sub-slots simultaneously sends interference signal.After data frame transfer is accomplished, judge whether all end of transmissions (for example, in the Transmission Control Protocol, transmit leg receives that the ack of last message segment confirms) of all data, if the equal end of transmission then accomplish communication of all data through upper-layer protocol; If do not finish, first sub-slots that then gets into next Frame time slot repeats said process.

Embodiment 2

The scene of the embodiment of the invention 2 is as shown in Figure 4; Information source node S, destination node D are arranged in network and eavesdrop each one of node E; Candidate's cooperative node C several, candidate's cooperative node C possesses two kinds of collaboration modes simultaneously: repeater mode (promptly cooperating as via node R) and jamming pattern (promptly cooperating as interfering nodes J).Each node all has been equipped with single antenna, and internodal channel is rayleigh fading channel, and bandwidth is equal, and separate.The transmitted power of information source node is P _S, the transmitted power when each cooperative node is operated in jamming pattern is P _J, the transmitted power when being operated in repeater mode is P _R

Compared to embodiment 1; Situation about considering in the scene of embodiment 2 is: when between information source node S and destination node D; Owing to exist barrier or channel to be in the deep fade state, can't direct communication, then need realize the data relaying by cooperative node R; And choose suitable interfering nodes simultaneously, can obviously improve the safe capacity between the RD.Information source node and destination node perhaps communicate under the common assistance of several via nodes with one at one or several interfering nodes.Special, when having only candidate's cooperative node, this node had both been taken relay task, bore the interference task again.

Optimum relaying-interference to the selection scheme flow process shown in the flow chart of Fig. 2 right side.Adopt and embodiment 1 similar method, before data frame transfer began, channel training that carries out and state division etc. were handled basic identical, repeat no more at this.

Need to prove; In present embodiment 2; Candidate's cooperative node both can be used as interfering nodes and had been operated in jamming pattern; Can be operated in repeater mode by via node again, so before the mode of operation of confirming cooperative node, can unify when dividing channel status for dividing the state of cooperative node-destination node CD channel and cooperative node-two kinds of channels of eavesdropping node CE channel.But for convenience of description, it is example that hereinafter is operated in repeater mode (R) with candidate's cooperative node, and the dividing mode of the channel status of RD channel, RE channel, SR channel is described; But it is noted that when candidate's cooperative node was operated in jamming pattern, the RD channel promptly was the JD channel; The RE channel promptly is the JE channel; This moment, the channel status division was identical with embodiment 1 with state transition probability, directly adopted the mode of embodiment 1 to get final product, and repeated no more at this.

In the scene of embodiment 2, consider that candidate's cooperative node m is operated in repeater mode, be a with the RE channel distribution _RIndividual state, the corresponding states space State transition probability matrix is expressed as:

${\mathrm{\&Omega;}}_{R_m}\left(t\right)={\left[{\mathrm{\&omega;}}_{x_{R_m}{y}_{R_m}}\left(t\right)\right]}_{a_R\&times;a_R};$

Wherein ${\mathrm{\&omega;}}_{x_{R_m}{y}_{R_m}}\left(t\right)=\mathrm{Pr}(o_{{\mathrm{RE}}_m}(t+1)=y_{R_m}|o_{{\mathrm{RE}}_m}\left(t\right)=x_{R_m}),x_{R_m},y_{R_m}\&Element;A_R,$ The channel power gain of the RE channel of expression candidate relay node m

At t constantly from state

Transfer to state Probability.

Similarly, be b with the RD channel distribution _RIndividual state, the corresponding states space

State transition probability matrix is expressed as:

${\mathrm{\&Psi;}}_{R_m}\left(t\right)={\left[{\mathrm{\&psi;}}_{u_{R_m}{v}_{R_m}}\left(t\right)\right]}_{b_R\&times;b_R};$

Wherein ${\mathrm{\&psi;}}_{u_{R_m}{v}_{R_m}}\left(t\right)=\mathrm{Pr}(o_{{\mathrm{RD}}_m}(t+1)=v_{R_m}|o_{{\mathrm{RD}}_m}\left(t\right)=u_{R_m}),u_{R_m},v_{R_m}\&Element;B_R,$ The channel power gain of the RD channel of expression candidate relay node m

At t constantly from state

Transfer to state Probability.

Similar with the channel status classifying method, can the dump energy of candidate's cooperative node and the signal to noise ratio of SR channel be classified as some states, and the statistical induction transition probability that does well.The dump energy of candidate's cooperative node is divided into c state, and (no matter be interfering nodes or via node, it is the same that the state of its dump energy is divided threshold value.The state transition probability of dump energy with still be that interfering nodes is relevant as via node; It is big when each ratio that consumes is as interfering nodes during as via node; But the state of dump energy and cooperative node are that via node or interfering nodes are irrelevant), corresponding states space C={C ₀, C ₁..., C _C-1, state transition probability matrix is expressed as:

${\mathrm{\&Theta;}}_m^{{\mathrm{\&beta;}}_m}\left(t\right)={\left[{\mathrm{\&theta;}}_{e_m{f}_m}^{{\mathrm{\&beta;}}_m^{R/J}}\left(t\right)\right]}_{c\&times;c};$

Wherein ${\mathrm{\&theta;}}_{e_m{f}_m}^{{\mathrm{\&beta;}}_m^{R/J}}\left(t\right)=\mathrm{Pr}(z_m(t+1)=f_m|z_m\left(t\right)=e_m),e_m,f_m\&Element;C,$ The dump energy z of expression candidate cooperative node m _mAt t constantly from state e _mTransfer to state f _mProbability.Whether the transition probability that it should be noted that dump energy is selected for use relevant (selected time spent β with this candidate's cooperative node _m=1, not selected time spent β _m=0), the collaboration mode (repeater mode R or jamming pattern J) with cooperative node also has relation.When node was selected, it was increased by the probability that high state of charge jumps to low state of charge, and transmitted power is bigger when being operated in repeater mode, and the transmitted power of jamming pattern is less, so the electric quantity consumption under two kinds of patterns is also different.

Signal to noise ratio υ with the SR channel _SRBe divided into d _RIndividual state, the corresponding states space

State transition probability matrix is expressed as:

Wherein ${\mathrm{\&xi;}}_{k_{R_m}{l}_{R_m}}\left(t\right)=\mathrm{Pr}({\mathrm{\&upsi;}}_{{\mathrm{SR}}_m}(t+1)=l_{R_m}|{\mathrm{\&upsi;}}_{{\mathrm{SR}}_m}\left(t\right)=k_{R_m}),k_{R_m},l_{R_m}\&Element;D_R,$ The signal to noise ratio of the SR channel of expression candidate relay node m At t constantly from state

Transfer to state

Probability.Under the situation that modulation demodulation system is confirmed, the error rate P of SR channel _bOnly by signal to noise ratio υ _SRDecision.

Because the dump energy of candidate's cooperative node and channel status are independent; And each channel is separate; So the state of candidate's cooperative node m during as via node can determine by the signal to noise ratio state of RD channel status associated therewith, RE channel status, dump energy state and SR channel jointly, candidate's cooperative node m during as via node the state at moment t can be denoted as:

$r_{R_m}\left(t\right)=[o_{{\mathrm{RD}}_m}\left(t\right),o_{{\mathrm{RE}}_m}\left(t\right),z_m\left(t\right),{\mathrm{\&upsi;}}_{{\mathrm{SR}}_m}\left(t\right)].$

Its state transition probability matrix can be expressed as:

$P_{R_m}\left(t\right)={[{\mathrm{\&omega;}}_{x_{R_m}{y}_{R_m}}\left(t\right),{\mathrm{\&psi;}}_{u_{R_m}{v}_{R_m}}\left(t\right),{\mathrm{\&theta;}}_{e_m{f}_m}^{{\mathrm{\&beta;}}_m^{R/J}}\left(t\right),{\mathrm{\&xi;}}_{k_{R_m}{l}_{R_m}}\left(t\right)]}_{q_R\&times;q_R},$

q _R＝a _R×b _R×c×d _R；

Matrix

element

indicates the candidate cooperative nodes as a relay node m at time t from the state

to state

probability.

The state of candidate's cooperative node m during as interfering nodes is identical with embodiment 1 with corresponding state transition probability matrix, directly adopts the mode of embodiment 1 to obtain getting final product.

After having divided state and having confirmed corresponding state transition probability, judge information source node S whether can with destination node D direct communication, if can direct communication, then carry out the selection of the optimum interfering nodes described in the embodiment 1; If can't direct communication, then scene becomes the scene that embodiment 2 considers, carries out the right selection of relaying-interferences, and the analysis of hereinafter will be the example expansion to (promptly from candidate's cooperative node, selecting via node and interfering nodes) to select a relaying-interference.

Optimum relaying-right selection of interference should be satisfied following three conditions:

1, guarantees to communicate with bigger safe capacity between RD.In real dimension, the safe capacity between the RD can be expressed as:

When only having single E, single R, single J, can derive by following formula:

$\mathrm{SC}=\max \{\frac{1}{2}{\log }_2(1+\frac{{\left|h_{\mathrm{RD}}\right|}^2{P}_R}{{\left|h_{\mathrm{JD}}\right|}^2{P}_J+{\mathrm{\&sigma;}}_1^2})-\frac{1}{2}{\log }_2(1+\frac{{\left|h_{\mathrm{RE}}\right|}^2{P}_R}{{\left|h_{\mathrm{JE}}\right|}^2{P}_J+{\mathrm{\&sigma;}}_2^2}),0\};$

Wherein,

is the transmitted power of interfering nodes;

is the transmitted power of via node;

and

is respectively the additive white Gaussian noise power that D and E receive, and its value does not change in time.Know by following formula, at P _S, P _J,

Under the situation about confirming, SC is by | h _SD| ², | h _SE| ², | h _JD| ², | h _JE| ²Common decision, and under S and the definite situation of D, | h _SD| ², | h _SE| ², | h _JD| ², | h _JE| ²Four parameters are only relevant with J with selected relaying R, promptly SC by relaying-interference of selecting to common decision.

Select optimum relaying-interference to (R J), should make safe capacity maximum, that is:

Same, be under the single situation at E, R, J, can derive by following formula:

$\arg \max \{\frac{{\left|h_{\mathrm{RD}}\right|}^2}{{\left|h_{\mathrm{JD}}\right|}^2}\&CenterDot;\frac{{\left|h_{\mathrm{JE}}\right|}^2}{{\left|h_{\mathrm{RE}}\right|}^2}\}$

$\&DoubleRightArrow;\arg \max (\frac{{\left|h_{\mathrm{RD}}\right|}^2}{{\left|h_{\mathrm{RE}}\right|}^2}\&CenterDot;\frac{{\left|h_{\mathrm{JE}}\right|}^2}{{\left|h_{\mathrm{JD}}\right|}^2})$

Can the safe capacity desired value be decomposed into | h _RD| ², | h _RE| ², | h _JE| ², | h _JD| ²Four desired values.Because same cooperative node is as

when getting maximum;

must obtain minimum value and (notice that the RD channel of same node and JD channel are same channels; RE channel and JE channel also are same channels), must not same node so have via node optimum under the situation of a plurality of candidate's cooperative nodes and optimum interfering nodes.

2, the electrical source consumption of balanced each cooperative node prolongs network lifetime.Similar with embodiment 1; A good relaying-interference is to selection scheme; Select relaying-interferences to the time, reasonably select cooperative node (as selecting the more node of dump energy), electric quantity consumption that can each cooperative node of equilibrium; Avoid too fast the exhausting of electric weight of candidate's cooperative node, prolong network lifetime.

3, reduce the error rate of SR channel.The error rate of SR channel directly influences the decoding of relaying R and encodes, and then influence arrives the signal quality at destination node D place.Select relaying-interferences to the time, should select the node of SR channel status good (signal to noise ratio height) is via node.

Take all factors into consideration each factor, the overall remuneration R of whole system _EtotalJust can be by relaying-interference to R is provided jointly _EJSystem's remuneration that relaying-interference centering interfering nodes provides that expression is selected, R _ERSystem's remuneration that relaying-interference centering via node provides that expression is selected.The overall system remuneration refers to selected relaying-interference to doing the as a whole remuneration that provides for overall system.

The system's remuneration that provides when wherein, candidate's cooperative node m is as interfering nodes J is:

${\mathrm{Re}}_{J_m}=R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m^J\left(t\right)}=E({\mathrm{\&beta;}}_m^J\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

This remuneration is identical with system's remuneration reality of consideration among the embodiment 1, and the concrete implication of each parameter can repeat no more at this referring to embodiment 1 for the mode of priority symbol value with evolution.

System's remuneration that candidate's cooperative node m provides during as via node R is:

${\mathrm{Re}}_{R_m}=R_{r_{R_m}\left(t\right)}^{{\mathrm{\&beta;}}_m^R\left(t\right)}=E({\mathrm{\&beta;}}_m^R\left(t\right)\&CenterDot;({\mathrm{\&tau;}}_1\&CenterDot;o_{{\mathrm{RE}}_m}+{\mathrm{\&tau;}}_2\&CenterDot;o_{{\mathrm{RD}}_m}+{\mathrm{\&tau;}}_3\&CenterDot;z_m+{\mathrm{\&tau;}}_4\&CenterDot;P_b));$

Among them, the operator E (·) represents the candidate cooperative node m under current state and desired state transition probability calculation;

indicates that the candidate node m at time t collaboration is being selected as a relay node, if it is selected, the without being selected,

and

denote RE and RD channel channel channel power gain,

With

The channel gain of expression respective channels, P _bThe error rate of expression SR channel (under the situation that modulation demodulation system is confirmed, it only with the signal to noise ratio υ of SR channel _SRRelevant); τ _iBe the weights of each desired value, need to satisfy

τ ₃>=0, τ ₄≤0, τ ₁≤0, τ ₂>=0, τ ₁=-τ ₂At this, system's remuneration and safe capacity desired value indirect correlation, directly related with dump energy desired value, SR channel bit error rate desired value.

Separate because of each node again, and optimum interfering nodes and optimum via node must not be same nodes under the situation of a plurality of candidate's cooperative nodes, select so optimum interfering nodes can separate with optimum via node.The hereinafter emphasis is with the routine deployment analysis of being chosen as of optimum via node (embodiment 1 is seen in the selection of optimum interfering nodes): optimum trunk node selection strategy u _R(u _R∈ U _R, U _RThe set of expression Markov policy) target is the system's remuneration maximization that makes whole data transfer phase, that is:

$L_R^{*}=\underset{u_R\&Element;U_R}{\max }{E}_{u_R}\left[\underset{t=0}{\overset{T-1}{\mathrm{\&Sigma;}}}(R_{r_{R_1}\left(t\right)}^{{\mathrm{\&beta;}}_1^R\left(t\right)}+R_{r_{R_2}\left(t\right)}^{{\mathrm{\&beta;}}_2^R\left(t\right)}+\&CenterDot;\&CenterDot;\&CenterDot;+R_{r_{R_m}\left(t\right)}^{{\mathrm{\&beta;}}_m^R\left(t\right)}){\mathrm{\&eta;}}_R^t\right];$

System's remuneration has time decline property, η _RExpress time decline parameter.

is the t residing state of candidate relay node m constantly.Following formula is converted into the linear programming expression formula:

$\left(\mathrm{LP}\right)L_R^{*}=\underset{g_R\&Element;G_R}{\max }\underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{r_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}\underset{{\mathrm{\&beta;}}_m^R\&Element;\left\{\mathrm{0,1}\right\}}{\mathrm{\&Sigma;}}{R}_{r_{R_m}}^{{\mathrm{\&beta;}}_m^R}{g}_{r_{R_m}}^{{\mathrm{\&beta;}}_m^R};$

In formula, M representes the set of candidate relay node,

The state space of expression candidate relay node m,

By performance vectors g _RAt all Markov policy u _R∈ U _RThe corresponding performance zones in following expansion back.Variable

Expression candidate cooperative node m is according to Markov policy u _R, when its state does

The Shi Zhihang action

The desired value of total time.

If

Expression candidate relay node m initial condition does Probability,

Expression initial condition probability vector,

Expression restless bandit polyhedron exists

The projection in space, K _RThe relaying number of indicating to select, then above-mentioned linear programming expression formula can first-order linear relax for:

$\left({\mathrm{LP}}^1\right){\mathrm{<figure-callout\; id="1"\; label="L\; R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">L</figure-callout>}}_R^{}=\max \underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{r_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}\underset{{\mathrm{\&beta;}}_m^R\&Element;\left\{\mathrm{0,1}\right\}}{\mathrm{\&Sigma;}}{R}_{r_{R_m}}^{{\mathrm{\&beta;}}_m^R}{g}_{r_{R_m}}^{{\mathrm{\&beta;}}_m^R}$

subject?to

$g_{R_m}{\mathrm{\&rho;}}_{R_m}^{-1},m\&Element;M$

$\underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{r_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{g}_{{\mathrm{<figure-callout\; id="1"\; label="r\; R\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">r</figure-callout>}}_{R_m}}^1=\frac{K_R}{1-{\mathrm{\&eta;}}_R}$

This first-order linear relaxes the optimal basic solution of problem finds the solution through original binary heuristic (Primal-dual Heuristic) method for

again, and the dual expression formula that single order relaxes linear programming is:

$\left(D^1\right){\mathrm{<figure-callout\; id="1"\; label="L\; R"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">L</figure-callout>}}_R^{}=\min \underset{m\&Element;M}{\mathrm{\&Sigma;}}\underset{s_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{\mathrm{\&pi;}}_{S_{R_m}}{\mathrm{\&gamma;}}_{S_{R_m}}+\frac{K_R}{1-{\mathrm{\&eta;}}_R}{\mathrm{\&gamma;}}_R$

subject?to

${\mathrm{\&gamma;}}_{r_{R_m}}-{\mathrm{\&eta;}}_R\underset{s_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{p}_{r_{R_m}{s}_{{\mathrm{<figure-callout\; id="0"\; label="R\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_m}}^0{\mathrm{\&gamma;}}_{s_{R_m}}\&GreaterEqual;R_{r_{R_{\mathrm{<figure-callout\; id="0"\; label="m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">m</figure-callout>}}}}^0,r_{R_m}\&Element;S_{R_m},m\&Element;M$

${\mathrm{\&gamma;}}_{r_{R_m}}-{\mathrm{\&eta;}}_R\underset{s_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{p}_{r_{R_m}{s}_{{\mathrm{<figure-callout\; id="1"\; label="R\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_m}}^1{\mathrm{\&gamma;}}_{s_{R_m}}+{\mathrm{\&gamma;}}_R\&GreaterEqual;R_{r_{R_{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">m</figure-callout>}}}}^1,r_{R_m}\&Element;S_{R_m},m\&Element;M$

γ _R≥0

Wherein,

expression candidate relay not selected time spent of node m jumps to the probability of state

from state

.In like manner,

expression candidate relay selected time spent of node m jumps to the probability of state

from state

.

separates for satisfying condition of requiring.Need to prove;

is the state of current time;

is next state constantly;

and

can be identical, and promptly the candidate relay node is retained in previous status.

If above-mentioned single order relax dual expression formula optimum dual of linear programming separate to

with first constraints be the example process of deriving with embodiment 1, no longer repeat here.If

expression reduces overhead factor, the optimum overhead factor

that reduces is defined as:

${\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{R_m}}}}^0=\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{r_{R_m}}}-{\mathrm{\&eta;}}_R\underset{s_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{p}_{r_{R_m}{s}_{{\mathrm{<figure-callout\; id="0"\; label="R\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">R</figure-callout>}}_m}}^0\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{s_{R_m}}}-R_{{\mathrm{<figure-callout\; id="0"\; label="r\; R\; m"\; filenames="HDA0000085225250000031.png,HDA0000085225250000032.png"\; state="\{\{state\}\}">r</figure-callout>}}_{R_m}}^0$

${\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{R_m}}}}^1=\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{r_{R_m}}}-{\mathrm{\&eta;}}_R\underset{s_{R_m}\&Element;S_{R_m}}{\mathrm{\&Sigma;}}{p}_{r_{R_m}{s}_{{\mathrm{<figure-callout\; id="1"\; label="R\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">R</figure-callout>}}_m}}^1\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_{s_{R_m}}}+-\stackrel{\&OverBar;}{{\mathrm{\&gamma;}}_R}-R_{{\mathrm{<figure-callout\; id="1"\; label="r\; R\; m"\; filenames="HDA0000085225250000024.png,HDA0000085225250000033.png"\; state="\{\{state\}\}">r</figure-callout>}}_{R_m}}^1$

Can the solution procedure of original binary heuristic be subdivided into two stages: primitive stage and binary stage.

In the primitive stage, if the candidate relay node satisfies

then it will be regarded as candidate's optimum via node.Concrete discussion process is similar to embodiment 1, no longer repeats at this.

Original binary heuristic can also further be converted into priority symbol value heuristic (Priority-index Heuristic), and the priority symbol value (Priority-index) of correspondence that is in the candidate relay node of state

is defined as:

${\mathrm{\&chi;}}_{r_{R_m}}={\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{R_m}}}}^1-{\stackrel{\&OverBar;}{{\mathrm{\&chi;}}_{r_{R_m}}}}^0.$

Adopt above method, be the corresponding priority symbol value of residing each the possible state computation of candidate relay node, and set up via node state-priority symbol value table of comparisons.Same method is set up interfering nodes state-priority symbol value table of comparisons among employing and the embodiment 1.So far the preparatory stage accomplishes before the transfer of data.

At data transfer phase; The transmission of each Frame takies a time slot; Each time slot is divided into two sub-slots again, at first sub-slots, estimates the information such as channel power gain (supposing that here the channel power gain between R/J and the E is known), signal to noise ratio of each channel through sending enhancement mode RTS/CTS (Request To Send/Clear To Send) between node; And it is referred in the corresponding state; The dump energy state that combines candidate's cooperative node is again confirmed the residing state of each candidate's cooperative node, and each candidate's cooperative node is according to own residing state; In interfering nodes state-priority symbol value table of comparisons and via node state-priority symbol value table of comparisons, check in corresponding priority symbol value (disturbing priority symbol value and relaying priority symbol value), and the notice information source node.Information source node selects to disturb or several less (K of priority symbol value minimum _R＜M) candidate's cooperative node to form relaying-interferences as interfering nodes and relaying priority symbol value minimum or less several candidate's cooperative nodes as via node right, assist information source node with destination node completion communicate by letter.

Second sub-slots that is transmitted in of Frame is accomplished: Frame is sent by information source node, and the via node of choosing via first sub-slots sends to destination node, and the interfering nodes of being chosen by first sub-slots simultaneously sends interference signal.After a data frame transfer is accomplished; All end of transmission has been (for example to judge whether all data through upper-layer protocol; In the Transmission Control Protocol, transmit leg receives that the ack of last message segment confirms), if the equal end of transmission then accomplish communication of all data; If do not finish, first sub-slots that then gets into next Frame time slot repeats said process.

The present invention mainly disturbs (CJ) pattern based on cooperation; The optimally in distributed mode interference source that research is exposed in the eavesdropping environment mobile collaborative network is down selected problem, thus following main be that the network scenarios that example is directed against two kinds of reality carries out simulation and analysis with the safe capacity desired value.

Scene one:

JE channel and JD channel respectively are divided into 4 kinds of states; Every kind of corresponding channel power gain

and

of state is 0.001; 0.01; 0.1,1.Corresponding state transition probability matrix is:

${\mathrm{\&Omega;}}_{J_m}\left(t\right)={\mathrm{\&Psi;}}_{J_m}\left(t\right)=\left[\begin{array}{cccc}0.3& 0.7& 0& 0\\ 0.7& 0.2& 0.1& 0\\ 0& 0.1& 0.5& 0.4\\ 0& 0& 0.4& 0.6\end{array}\right]$

Emulation is example with the safe capacity desired value, so put aside candidate's cooperative node dump energy at this.The state of candidate's cooperative node m is by

and

common decision, and system's remuneration is defined as:

$R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m\left(t\right)}=E({\mathrm{\&beta;}}_m\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}));$

Set μ ₁=-μ ₂=0.5, time decline parameter η _R=0.6, in a plurality of candidate's cooperative nodes, select a cooperative node as interfering nodes.The transmitted power P of information source node _S=10W, the transmitted power of interfering nodes is P _J=2W, noise power is

If the average channel power of SD channel gain o _SDBe 0.7 with the average channel power of SE channel gain o _SEBe 0.3.Fig. 5 has contrasted the situation of change of the average safe capacity (at the mean value of setting a plurality of Frame safe capacities that transmit under the situation) of three kinds of different schemes with candidate's cooperative node number.Three kinds of schemes are respectively: 1 of embodiment puies forward optimum interfering nodes selection scheme, channel quiet scheme (suppose that channel status is almost constant, and the optimum interfering nodes during as the second sub-slots transfer of data with the optimum interfering nodes of first sub-slots) simply and traditional noiseless scheme.As can be seen from the figure, the average safe capacity under set environment of noiseless scheme is starkly lower than the scheme of interfering nodes, and average safe capacity approaches 0.5, and does not change with the increase of candidate's cooperative node number.Under different candidate's cooperative node numbers; The average safe capacity of optimum interfering nodes selection scheme is all greater than the channel quiet scheme; This is because the time-varying characteristics of channel; The optimum interfering nodes of first sub-slots may not be the optimum interfering nodes of second sub-slots, and the channel quiet scheme can't embody the variation of channel; And the optimum interfering nodes selection scheme that embodiment 1 is carried then can dynamically be selected the optimum interfering nodes of second sub-slots according to the current state and the state transition probability of candidate's cooperative node.At last, along with the increase of candidate's cooperative node number, the average safe capacity of optimum interfering nodes selection scheme and channel quiet scheme all rises to some extent, and in the end tends towards stability.This is because of the increase along with candidate's cooperative node number, and the possibility that is in the both candidate nodes existence of preferable states (remuneration is bigger) increases, and chooses the node cooperation communication that is in preferable states can improve average safe capacity.Thereby after candidate's cooperative node number is greater than some determined values, almost always have node at one's best, it is little to the systematic function influence to increase interstitial content again.

Suppose a kind of more abominable scene further, establish the average channel power gain o of SD channel _SDBe 0.3 with the average channel power of SE channel gain o _SEBe 0.7.The state of main channel (SD channel) is inferior to tapping channel (SE channel), and Fig. 6 has compared the average safe capacity of three kinds of schemes under this channel conditions.In this case, the average safe capacity of traditional noiseless scheme almost is 0, and promptly information source node and destination node can't be carried out safe communication; And after the introducing interfering nodes, even the main channel state inferior to tapping channel, information source node and destination node still can be carried out secure communication with certain speed.

The system compensatory time accumulation that Fig. 7 has showed optimum interfering nodes selection scheme and channel quiet scheme is situation over time with

.

Suppose 1 second consuming time of the transmission of each Frame simply, promptly 1 time slot is 1 second (if hereinafter do not specify, all suppose emulation by this).Significantly, the system compensatory time of optimum interfering nodes selection scheme accumulation and be greater than the channel quiet scheme, and along with the increase of time, the two gap is also widening.

Further consider the network lifetime index.Suppose that initial time has 10 candidates interfering nodes of cooperating, the cooperate dump energy of interfering nodes of candidate can be divided into one of four states, and the electric weight mean value that each state is corresponding is respectively 0.125; 0.315,0.625,0.875; Initial electric weight is 1, and state transition probability matrix is:

${\mathrm{\&Theta;}}_{J_m}^1\left(t\right)=\left[\begin{array}{cccc}1& 0& 0& 0\\ 0.1& 0.9& 0& 0\\ 0& 0.1& 0.9& 0\\ 0& 0& 0.1& 0.9\end{array}\right],$ ${\mathrm{\&Theta;}}_{J_m}^0\left(t\right)=\left[\begin{array}{cccc}1& 0& 0& 0\\ 0.03& 0.97& 0& 0\\ 0& 0.03& 0.97& 0\\ 0& 0& 0.03& 0.97\end{array}\right].$

The reward function that the selection of one suboptimum interfering nodes (promptly transmitting a Frame) system obtains can be defined as:

$R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m\left(t\right)}=E({\mathrm{\&beta;}}_m\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

Wherein, The electric weight H that Frame of hypothesis transmission consumes in this emulation is only relevant with ;

value is big more; Then interference effect is good more; Safe capacity is big more; Under the situation of identical information amount, it is just few more to send symbol, and the electric weight H that sends each frame data consumption is few more; Establishing

here need to prove; Managing of H value is not unique, and it is relevant with multiple factor, when practical application, can be provided with flexibly.If μ ₁=-μ ₂=0.1, μ ₃=0.7, μ ₄=-0.1.Network lifetime is defined as from the network individual candidate of th (threshold value) that brings into operation and cooperates the time that the interfering nodes electric weight exhausts.Fig. 8 has compared the network lifetime of selection scheme of two kinds of cooperation interfering nodes with the situation of change of different threshold values.The optimum network lifetime of cooperative node that disturbs will obviously be longer than the channel quiet scheme.

Scene two:

Similarly cooperative node is divided into 4 kinds of states to destination node and the internodal channel power gain of eavesdropping (i.e.

) with scene one; Every kind of corresponding channel power gain mean value of state is respectively 0.002; 0.03; 0.2,1.State transition probability matrix is identical with scene one.Time decline parameter η _R=η _J=0.6.It is right from a plurality of cooperative nodes, to select the relaying-interference of a via node and interfering nodes composition.With the example that is chosen as of optimum via node, the reward function of system is defined as:

${\mathrm{Re}}_{R_m}=R_{r_{R_m}\left(t\right)}^{{\mathrm{\&beta;}}_m^R\left(t\right)}=E({\mathrm{\&beta;}}_m^R\left(t\right)\&CenterDot;({\mathrm{\&tau;}}_1\&CenterDot;o_{{\mathrm{RE}}_m}+{\mathrm{\&tau;}}_2\&CenterDot;o_{{\mathrm{RD}}_m}+{\mathrm{\&tau;}}_3\&CenterDot;z_m+{\mathrm{\&tau;}}_4\&CenterDot;P_b));$

Because system's first concern is the safe capacity performance, to analyze for simplifying, each weights can be made as τ ₁=-τ ₂=-0.5, τ ₃=τ ₄=0.In like manner:

${\mathrm{Re}}_{J_m}=R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m^J\left(t\right)}=E({\mathrm{\&beta;}}_m^J\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

If each weights is: μ ₁=-μ ₂=0.5, μ ₃=μ ₄=0.

When supposing that cooperative node is operated in repeater mode, transmitted power P _R=10W, when being operated in jamming pattern, transmitted power P _J=2W.Fig. 9 has compared the situation of change of the average safe capacity of three kinds of selection schemes with the number of candidate's cooperative node.Three kinds of schemes are respectively: 2 of embodiment propose optimum relaying-interference to selection scheme, channel quiet scheme (supposing that channel status is constant, simply with the optimum interfering nodes of first sub-slots and via node interfering nodes and the via node as overall process) and selection scheme (in candidate's cooperative node, selecting two points randomly as interfering nodes and via node) at random.As shown in Figure 9, the safe capacity performance of selection scheme is the poorest at random, and because randomness, its average safe capacity does not change along with the increase of candidate's cooperative node number, and average safe capacity is stabilized near the 0.3bit/symbol.Optimum relaying-interference then obviously is superior to selection scheme at random to the average safe capacity of selection scheme and channel quiet scheme, and along with the increase of candidate's cooperative node number, increases owing to the possibility of preferable node occurs, and average safe capacity also increases thereupon.But because the time-varying characteristics of channel, the average safe capacity of channel quiet scheme still is less than optimum relaying-interference to selection scheme.The overall system compensatory time accumulation that Figure 10 has compared three kinds of different schemes and (i.e. ) be situation over time.

Optimum relaying-interference is to system's compensatory time accumulation of selection scheme and channel quiet scheme with along with the time increases; And selection scheme is quite different at random; This is because some moment; Relaying-interferences of selecting at random be to possibly making that system's remuneration is a negative value, and relaying-interference that preceding two kinds of schemes are selected ninety-nine times out of a hundred to can both make system's remuneration be on the occasion of.System's compensatory time accumulation of optimum interfering nodes selection scheme and be greater than the channel quiet scheme, and along with the increase of time, the two gap is also widening.

Can find out that by above-mentioned simulation analysis compared with prior art, Optimal Distribution formula interference source system of selection in the mobile collaborative network that the present invention carried has following advantage:

1, compares with traditional communication, in network model, introduce interference source, can obviously improve the safe capacity of system, guarantee the safety of communication.Especially under some adverse circumstances (like the main channel state inferior to tapping channel); Traditional mode almost can't realize secure communication; And after the introducing interference source, information source node and destination node still can realize with certain safe speed communicating by letter and eavesdrop node and can't obtain any information.

2, the channel status when between information source node and destination node allows direct communication, but the main channel state is during inferior to tapping channel, and source node and destination node can't realize secure communication; Existing mode is between information source node and destination node, to introduce relaying; But the transmitted power of relaying is bigger, and this programme is introduced interfering nodes, and the transmitted power of interfering nodes is significantly less than relaying; Can when saving the energy, improve the system safety capacity effectively.

3, the channel when information source node and destination node does not allow direct communication (be in the deep fade state like channel and perhaps have barrier between the two); Though the mode of traditional introducing relaying can solve the communication issue of information source and destination node; But Information Security can not get ensureing, and generally cooperative node only has single antenna, can't be in relay data; Send interference eavesdropping node; And that this programme has been introduced relaying-interference is right, selects two kinds of dissimilar cooperative nodes simultaneously, the fail safe of common guarantee system and reliability.

4, scheme of the present invention has been accomplished the preparation that a part of optimum cooperative node is selected before transfer of data, has reduced data and has passed time delay and the complexity that the stage cooperative node is selected.

5, scheme of the present invention is united factors such as having considered the error rate, network lifetime, system safety capacity, and is had very strong extensibility and flexibility when the selection of cooperative node.

Above execution mode only is used to explain the present invention; And be not limitation of the present invention; The those of ordinary skill in relevant technologies field under the situation that does not break away from the spirit and scope of the present invention, can also be made various variations and modification; Therefore all technical schemes that are equal to also belong to category of the present invention, and invention protection range of the present invention should be defined by the claims.

Claims (10)

1. Optimal Distribution formula interference source system of selection in the mobile collaborative network is characterized in that said method comprises step:

S1, information source node S periodically send training sequence, and destination node D and each cooperative node C carry out channel estimating;

S2 divides the channel power gain-state of each channel, the dump energy state and the information source relaying SR channel signal to noise ratio state of candidate's cooperative node according to the feedback result of training sequence, and confirms corresponding state transition probability; Each possible state when confirming candidate's cooperative node thus and corresponding state transition probability and candidate's cooperative node each possible state and the corresponding state transition probability during as via node as interfering nodes;

S3, judge information source node S whether can with destination node D direct communication, if can direct communication, then execution in step S4 carries out optimum interfering nodes and selects; Otherwise execution in step S8 carries out optimum relaying-interference to selecting;

S4, each possibly state computation disturb the priority symbol value accordingly during as interfering nodes for candidate's cooperative node, sets up interfering nodes state-priority symbol value table of comparisons;

S5; The time slot that data frame transfer is taken is divided into two sub-slots; At first sub-slots; According to each candidate's cooperative node present located state, state of living in when predicting second each candidate's cooperative node of sub-slots as interfering nodes, choosing the minimum candidate's cooperative node of interference priority symbol value with reference to said interfering nodes state-priority symbol value table of comparisons is optimum interfering nodes;

S6, at second sub-slots of data frame transfer, Frame sends to destination node by information source node, and the optimum interfering nodes of being chosen by first sub-slots simultaneously sends interference signal;

Whether S7 judges all data end of transmission after each data frame transfer is accomplished, if then finish communication; Otherwise go back to the transmission that step S5 carries out next Frame;

S8; Respectively from the angle of interfering nodes and the angle of via node; Disturb priority symbol value and relaying priority symbol value accordingly for each possible interfering nodes state of candidate's cooperative node or via node state computation, set up relaying-interference state-priority symbol value table of comparisons;

S9; The time slot that data frame transfer is taken is divided into two sub-slots; At first sub-slots; According to each candidate's cooperative node present located state, predict second each candidate's cooperative node of sub-slots state of living in, with reference to said relaying-interference state-priority symbol value table of comparisons being chosen the minimum candidate's cooperative node of interference priority symbol value is optimum interfering nodes; With reference to said relaying-interference state-priority symbol value table of comparisons being chosen the minimum candidate's cooperative node of relaying priority symbol value is optimum via node, and it is right that optimum via node of selecting and optimum interfering nodes constitute the relaying-interference of second sub-slots cooperation;

S10, at second sub-slots of data frame transfer, Frame is sent by information source node, and the optimum via node of choosing via first sub-slots sends to destination node, and the optimum interfering nodes of being chosen by first sub-slots simultaneously sends interference signal;

Whether S11 judges all data end of transmission after each data frame transfer is accomplished, if then finish communication; Otherwise go back to the transmission that step S9 carries out next Frame.

2. method according to claim 1 is characterized in that, among the step S2, the step that said feedback result according to training sequence is divided the state of each channel is specially:

Feedback result according to training sequence; Confirm the probability density function of the channel power gain of each channel; Be the corresponding threshold value of channel power gain setting of each channel, thereby the state of each channel is divided into a plurality of states corresponding with the channel power gain threshold value of this channel.

3. method according to claim 1; It is characterized in that; Among the step S4; Based on Restless Multi-armed Bandit model, according to by the dump energy that disturbs eavesdropping JE channel and the channel power gain of disturbing purpose JD channel and candidate's cooperative node with transmit the electric weight of Frame needs consumption and system's remuneration that definite interfering nodes provides, each the possible state computation for candidate's cooperative node during as interfering nodes is disturbed the priority symbol value accordingly.

4. method according to claim 1; It is characterized in that; Among the step S8; Based on Restless Multi-armed Bandit model; According to by disturb eavesdropping JE channel, disturb purpose JD channel, the error rate of the electric weight of the dump energy of the channel power gain of relaying eavesdropping RE channeling and trunking purpose RD channel and candidate's cooperative node, the consumption of Frame needs of transmission and SR channel and the definite overall system remuneration of relaying-interference to providing, for each possible interfering nodes state of candidate's cooperative node or via node state computation are disturbed priority symbol value and relaying priority symbol value accordingly.

5. method according to claim 1; It is characterized in that; Among step S5 or the step S9; The dump energy of the channel power gain of each channel of RTS/CTS message feedback through enhancement mode, signal to noise ratio, candidate's cooperative node, and it is referred in the corresponding state, thereby confirm each candidate's cooperative node present located state.

6. whether method according to claim 1 is characterized in that, among step S7 or the step S11, judge all data end of transmission through upper-layer protocol.

7. according to claim 3 or 4 described methods, it is characterized in that said calculating is specially: said remuneration is converted into carries out first-order linear after the linear programming and relax, find the solution through the priority symbol value heuristic that is evolved into by original binary heuristic again.

8. method according to claim 3; It is characterized in that; Among the step S4; With the safe capacity target value conversion is the channel power gain

of JE channel and channel power gain

two desired values of JD channel; For the candidate's cooperative node m as interfering nodes, system's remuneration that said interfering nodes provides is:

$R_{r_{J_m}\left(t\right)}^{{\mathrm{\&beta;}}_m\left(t\right)}=E({\mathrm{\&beta;}}_m\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

Wherein, the expectation computing carried out according to candidate's cooperative node m current state and state transition probability of operator E () expression; β _m(t) whether expression candidate cooperative node m is selected as interfering nodes at moment t, if be selected, and β then _m(t)=1, as be not selected, then β _m(t)=0;

With Represent the channel power gain of JE channel and JD channel respectively; z _mThe dump energy of expression candidate cooperative node m; When H representes to transmit a Frame, the electric weight that need consume as candidate's cooperative node m of interfering nodes; μ _iBe the weights of each desired value, need to satisfy

μ ₁>=0, μ ₂≤0, μ ₁=-μ ₂, μ ₃>=0, μ ₄≤0.

9. method according to claim 4; It is characterized in that; Among the step S8; Be the channel power of channel power gain

and RD channel of channel power gain

RE channel of channel power gain

JD channel of JE channel

four desired values that gain with the safe capacity target value conversion, said relaying-interference comprises system's remuneration that relaying-system's remuneration that interference centering interfering nodes provides of selecting and relaying-interference of selecting provide via node to the overall system remuneration that provides;

For candidate's cooperative node m, remuneration

that it provides during as interfering nodes and the remuneration that provides during as via node are respectively:

${\mathrm{Re}}_{J_m}=E({\mathrm{\&beta;}}_m^J\left(t\right)\&CenterDot;({\mathrm{\&mu;}}_1\&CenterDot;o_{{\mathrm{JE}}_m}+{\mathrm{\&mu;}}_2\&CenterDot;o_{{\mathrm{JD}}_m}+{\mathrm{\&mu;}}_3\&CenterDot;z_m+{\mathrm{\&mu;}}_4\&CenterDot;H));$

${\mathrm{Re}}_{R_m}=E({\mathrm{\&beta;}}_m^R\left(t\right)\&CenterDot;({\mathrm{\&tau;}}_1\&CenterDot;o_{{\mathrm{RE}}_m}+{\mathrm{\&tau;}}_2\&CenterDot;o_{{\mathrm{RD}}_m}+{\mathrm{\&tau;}}_3\&CenterDot;z_m+{\mathrm{\&tau;}}_4\&CenterDot;P_b));$

Wherein, the expectation computing carried out according to current state and state transition probability of operator E () expression;

Represent respectively whether candidate's cooperative node m is selected as interfering nodes or via node at moment t, if be selected as interfering nodes, then

As be not selected as interfering nodes, then

If be selected as via node, then

As be not selected as via node, then

With

Represent the channel power gain of JE channel, JD channel, RE channel and RD channel respectively, z _mThe dump energy of expression candidate cooperative node m; When H representes to transmit a Frame, the electric weight that need consume as candidate's cooperative node m of interfering nodes; P _bThe error rate of expression SR channel; μ _i, τ _iBe the weights of each desired value, need to satisfy

μ ₁>=0, μ ₂≤0, μ ₁=-μ ₂, μ ₃>=0, μ ₄≤0;

τ ₃>=0, τ ₄≤0, τ ₁≤0, τ ₂>=0, τ ₁=-τ ₂

10. method according to claim 1; It is characterized in that; Among step S5 or the step S9, said choosing disturbs the minimum candidate's cooperative node of priority symbol value to be specially: choose and disturb minimum candidate's cooperative node of priority symbol value or choose preceding several candidate's cooperative nodes that disturb the priority symbol value minimum;

Among the step S9, the said minimum candidate's cooperative node of relaying priority symbol value of choosing is specially: choose minimum candidate's cooperative node of relaying priority symbol value or choose minimum preceding several candidate's cooperative nodes of relaying priority symbol value.

Images