CN103987065A - Method for predicting behavior model and monitor strategy of relay node - Google Patents

Abstract

The invention discloses a method for predicting a behavior model and a monitor strategy of a relay node and relates to the field of physical layers of the wireless communication. The method includes the following steps of building a communication model, obtaining the secrecy capacity of the communication model, building a game model between the untrusted relay node and a monitor system according to the secrecy channel capacity, calculating Nash equilibrium and obtaining the probabilities Pa and Pb for the two parties to adopt a certain strategy. A prediction model and related algorithms are provided based on the non-zero-sum games. The behavior strategy adopted by the node is predicted, meanwhile, the optimal monitor strategy is provided, influence of selfish and hostile attack behaviors of the node in the cooperative communication is reduced, and the safety transmission of the physical layers is guaranteed.

Description

A kind of method of predicting via node behavior model and monitoring policy

Technical field

The present invention relates to radio communication physical layer field, relate in particular to a kind of prediction via node behavior model and optimum monitoring policy, to ensure the method for physical layer security performance optimum.

Background technology

The broadcast characteristic of wireless transmission makes nodes easily suffer attack and the eavesdropping of malicious node.And collaboration communication utilize the forwarding of cooperative node become can practical application physical layer safe practice.When node cooperation, if all via nodes are all observed cooperation criterion, can realize stable physical layer security performance.Owing to lacking effective supervision mechanism, there is malice and selfish behavior in system interior nodes, has a strong impact on physical layer security performance.In collaboration communication, the selfishness of node is different also different to system safety influence degree from malice degree.For ensureing the overall security of system, need related algorithm look-ahead facility strategy, and take optimum monitoring policy.

Current in collaboration communication physical layer safety problem, adopt single node model more and all only consider that node is safe node.And nodes ' behavior is changeable in radio communication, and adopt multinode model more.Have not yet to see and can realize by predicting that insincere nodes ' behavior is to ensure the algorithm of physical layer security performance.

Summary of the invention

The invention provides a kind of method of predicting via node behavior model and monitoring policy, to ensure the method for physical layer security performance optimum.In the time that under collaboration communication pattern, via node is insincere node, the present invention adopts nonzero sum game, proposes prediction via node behavior model and optimum monitoring policy, ensures physical layer security performance optimum, described below:

A method of predicting via node behavior model and monitoring policy, said method comprising the steps of:

Set up traffic model; The secrecy capacity of obtaining communication model;

Set up the betting model between insincere via node and monitoring system according to cryptochannel capacity;

Calculate Nash Equilibrium, obtain game both sides and take the probability P of a certain strategy _aand P _b.

Described traffic model is specially:

There is n via node, a source node, a destination node, an eavesdropping node; Source node is to insincere via node R _i, insincere via node R _ito destination node, and insincere via node R _ichannel to eavesdropping node is quasistatic flat fading channel;

Interchannel noise n _sD, n _sR, n _rDbe variances sigma ²additive white Gaussian noise CN (0, σ ²i _n), wherein, I _nfor unit diagonally opposing corner matrix; n _sDfor the interchannel noise between source node S and destination node D; n _sRfor source node S and insincere via node R _ibetween interchannel noise; n _rDfor insincere via node R _iand interchannel noise between destination node D.

Described betting model is specially:

Game effectiveness is the poor V of channel capacity ^t:

$V^t={\tilde{C}}_s-C_s\left(s_a^t\right)$

The utility schedule of via node is shown:

$u_a(s_a^t,s_b^k)=V^t{\mathrm{\η}}^i{D}^{\mathrm{tk}}$

The utility schedule of surveillance is shown:

$u_b(s_a^t,s_b^k)={\mathrm{\η}}^i{D}^{\mathrm{tk}}{C}_s\left(s_a^t\right)$

Wherein, η ⁱbe verification and measurement ratio, represent that monitoring system successfully detects the probability of the insincere behavior of i node; for secrecy capacity; for taking strategy when channel time secrecy capacity; D ^tkfor the income of the punishment to insincere node and monitoring system; represent k the strategy of monitoring system b, represent t the strategy of insincere node a.

The beneficial effect of technical scheme provided by the invention is:

The behavior that 1, may exist for node changes, and the present invention is based on nonzero sum game, proposes forecast model and related algorithm.The present invention has predicted the behavioral strategy that node is taked, and has proposed optimum monitoring policy simultaneously, has reduced the impact of the selfishness of collaboration communication interior nodes and malicious attack behavior, has ensured the safe transmission of physical layer.

2, the present invention characterizes the parameter beta that nodes ' behavior changes, span 0-1 by proposition.Parameter value more malicious attack or the selfish behavior of minor node is more serious.Parameter value is to show that nodes ' behavior is reliable at 1 o'clock, does not have malicious attack or selfish behavior, embodies cooperation behavior.Parameter value is the malicious act of 0 expression node, only to node transmitted noise.

Brief description of the drawings

Fig. 1 is the schematic diagram of many relay node cooperations traffic model;

Fig. 2 is a kind of flow chart of the method for predicting via node behavior model and monitoring policy.

Embodiment

For making the object, technical solutions and advantages of the present invention clearer, below embodiment of the present invention is described further in detail.

101: set up traffic model;

Referring to Fig. 1, in traffic model, there is n via node, a source node S,, there is an eavesdropping node E in a destination node D.Source node S is to insincere via node R _i(1≤i≤N), insincere via node R _ito destination node D, and insincere via node R _ichannel to eavesdropping node E is quasistatic flat fading channel, source node S and insincere via node R _ibetween channel gain index h _sRrepresent; Insincere via node R _iand channel gain index h between destination node D _rDrepresent; Insincere via node R _iand channel gain index h between eavesdropping node E _rErepresent.Interchannel noise n between source node S and destination node D _sDrepresent; Source node S and insincere via node R _ibetween interchannel noise n _sRrepresent; Insincere via node R _iand interchannel noise n between destination node D _rDrepresent, and interchannel noise n _sD, n _sR, n _rDbe variances sigma ²additive white Gaussian noise CN (0, σ ²i _n), wherein, I _nfor unit diagonally opposing corner matrix.

Wherein, the weighted of different insincere via nodes, represents that with w (N node is [w to weight ¹..., w ⁿ]).The transmitting power of source node S is P _s, system gross power is P, insincere via node R _ipower be P-P _s, use P _rrepresent.

102: the secrecy capacity of obtaining communication model;

Whole collaboration communication process is divided into two stages, first stage, and source node S sends information x to N insincere via node R, second stage, insincere relay node cooperation forwards the information of receiving to destination node D.First stage is while end, the information y that all insincere via nodes are received _rfor

$y_r=\sqrt{P_s}{h}_{\mathrm{SR}}^{*}x+n_{\mathrm{SR}}$

Wherein h _sRfor h _sr ^*for channel gain matrix h _sRconjugate matrices.() ^*, () ^tand the conjugation of representing matrix respectively, transposition and conjugate transpose, black matrix representing matrix. represent source node S and via node R ₁between channel gain, represent source node S and via node R ₂between channel gain, represent source node S and via node R _nbetween channel gain.X represents the information that source node S is sent.

Second stage, parameter beta ⁱembody the selfish of insincere via node or malice situation.β ⁱ∈ [0,1], β ⁱbe worth the harmful degree of less explanation higher.Work as β ⁱbe 1 o'clock, insincere via node for cooperation (be insincere via node can in cooperation, selfish and malice any one, by parameter beta ⁱdetermine that concrete behavior, how known for those skilled in the art, the embodiment of the present invention does not limit this) work as β ⁱbe 0 to 1 o'clock, insincere via node is selfish; Work as β ⁱbe 0 o'clock, insincere via node is malice.Parameter μ ⁱauxiliary variable during for mathematical computations, works as β ⁱbe 1 or 0 to 1 o'clock, μ ⁱbe 0; Work as β ⁱbe 0 o'clock, μ ⁱbe 1.The information y that destination node D receives _dfor:

Wherein ${\mathrm{\Ω}}^i=\frac{\sqrt{{\mathrm{\β}}^i{P}_R}}{\sqrt{P_s{\left|h_{\mathrm{SR}}^i\right|}^2}+{\mathrm{\σ}}^2},{\mathrm{\λ}}^i=\frac{\sqrt{{\mathrm{\μ}}^i{P}_R}}{\sqrt{P_s{\left|h_{\mathrm{SR}}^i\right|}^2+{\mathrm{\σ}}^2}}$

represent source node S and via node R _ibetween decline index, diag{.} represents diagonal matrix.

Calculate for convenient, establish:

$A=\sqrt{P_S}\mathrm{diag}\left\{{\mathrm{\Ω}}^{*}\right\}\mathrm{diag}\left\{h_{\mathrm{SR}}\right\}{h}_{\mathrm{RD}}---\left(3\right)$

$B=\mathrm{diag}\left\{h_{\mathrm{RD}}^{*}\right\}\mathrm{diag}\left\{\mathrm{\Ω}\right\}---\left(4\right)$

$O=\sqrt{P_S}\mathrm{diag}\left\{{\mathrm{\λ}}^{*}\right\}\mathrm{diag}\left\{h_{\mathrm{SR}}\right\}{h}_{\mathrm{RD}}---\left(5\right)$

So y _dbecome:

The information y that eavesdropping node E receives _efor

Calculate for convenient, establish:

$G=\sqrt{P_S}\mathrm{diag}\left\{{\mathrm{\Ω}}^{*}\right\}\mathrm{diag}\left\{h_{\mathrm{SR}}\right\}{h}_{\mathrm{RE}}---\left(8\right)$

$H=\mathrm{diag}\left\{h_{\mathrm{RE}}^{*}\right\}\mathrm{diag}\left\{\mathrm{\Ω}\right\}---\left(9\right)$

$1=\sqrt{P_S}\mathrm{diag}\left\{{\mathrm{\λ}}^{*}\right\}\mathrm{diag}\left\{h_{\mathrm{SR}}\right\}{h}_{\mathrm{RE}}---\left(10\right)$

Wherein

Cryptochannel capacity C _sfor C _d-C _e, C _dfor source node is to the capacity of the channel of destination node.C _efor source node is to the capacity of the channel of eavesdropping node.

103: set up the betting model between insincere via node and monitoring system according to cryptochannel capacity;

In the time of the whole cooperation of insincere via node, secrecy capacity is when channel is taked strategy time, secrecy capacity is

Game effectiveness is the poor V of channel capacity ^t:

$V^t={\tilde{C}}_S-C_S\left(s_a^t\right)---\left(12\right)$

The utility schedule of via node is shown:

$u_a(s_a^t,s_b^k)=V^t-{\mathrm{\η}}^i{D}^{\mathrm{tk}}---\left(13\right)$

Wherein η ⁱbe verification and measurement ratio, represent that monitoring system successfully detects the probability of the insincere behavior of i node.

The utility schedule of surveillance is shown:

$u_b(s_a^t,s_b^k)={\mathrm{\η}}^i{D}^{\mathrm{tk}}{C}_s\left(s_a^t\right)---\left(14\right)$

Wherein D ^tkfor the income of the punishment to insincere node and monitoring system, after the bad behavior of a certain node is monitored to, this node is taked cooperation behavior, secrecy capacity D now ^tkrepresent. represent k the strategy of monitoring system b, represent t the strategy of insincere node a.

104: calculate Nash Equilibrium, obtain game both sides and take the probability P of a certain strategy _aand P _b.

Calculate after game both sides effectiveness according to (12)-(14), utilize following formula to carry out the calculating of Nash Equilibrium

Maxf(S _a,S _b,U _a,U _b)＝p _a(U _a+U _b)p _b-l ₁-l ₂?(15)

s.t.U _ap _b≤l ₁i ₁；U _bp _a≤l ₂i ₂；

i ₁ ^Tp _a＝1；i ₂ ^Tp _b＝1；

$0<p_a^t<1;0<p_b^k<1;$

Wherein l ₁and l ₂for scale, represent to expect maximum return, P _aand P _bfor game both sides take a certain tactful probability.Thereby calculate the most possible insincere behavioral strategy occurring of node, and the monitoring policy of optimizing.

Wherein, u _a, u _b, i ₁, i ₂, P _aand P _bbe all vector set, , for the element in vector set.

Non-zero game model is five yuan of Vector Groups G (Q, S, Z, P, U), wherein:

Q={Q _a, Q _bgame both sides set.

Q _arepresent N insincere relaying, Q _brepresent monitoring system.

S={S _a, S _bit is the set of game strategies.

represent the strategy of insincere via node, represent the strategy of monitoring system.

it is the set of actions under specific policy.

$Z_a^t=\left\{{\left[z_a^i\right]}_{i=1,...,N}\right|s_a^t\},Z_b^k=\{{\left[z_b^i\right]}_{i=1,...,N}\left|s_b^k\right\}.$

P={P _a, P _bit is Making by Probability Sets.

$P_a={\left[p_a^t\right]}_{t=1,...,M},P_b={\left[p_b^k\right]}_{t=1,...,M}$ 。 represent that insincere node takes the probability of strategy, expression monitoring system is taked the probability of strategy.

U={U _a, U _bit is the utility matrix of insincere node and monitoring system.

It will be appreciated by those skilled in the art that accompanying drawing is the schematic diagram of a preferred embodiment, the invention described above embodiment sequence number, just to describing, does not represent the quality of embodiment.

The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any amendment of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (3)

1. a method of predicting via node behavior model and monitoring policy, is characterized in that, said method comprising the steps of:

Set up traffic model; The secrecy capacity of obtaining communication model;

Set up the betting model between insincere via node and monitoring system according to cryptochannel capacity;

Calculate Nash Equilibrium, obtain game both sides and take the probability P of a certain strategy _aand P _b.

2. a kind of method of predicting via node behavior model and monitoring policy according to claim 1, is characterized in that, described traffic model is specially:

There is n via node, a source node, a destination node, an eavesdropping node; Source node is to insincere via node R _i, insincere via node R _ito destination node, and insincere via node R _ichannel to eavesdropping node is quasistatic flat fading channel;

Interchannel noise n _sD, n _sR, n _rDbe variances sigma ²additive white Gaussian noise CN (0, σ ²i _n), wherein, I _nfor unit diagonally opposing corner matrix; n _sDfor the interchannel noise between source node S and destination node D; n _sRfor source node S and insincere via node R _ibetween interchannel noise; n _rDfor insincere via node R _iand interchannel noise between destination node D.

3. a kind of method of predicting via node behavior model and monitoring policy according to claim 1, is characterized in that, described betting model is specially:

Game effectiveness is the poor V of channel capacity ^t:

$V^t={\tilde{C}}_s-C_s\left(s_a^t\right)$

The utility schedule of via node is shown:

$u_a(s_a^t,s_b^k)=V^t{\mathrm{\&eta;}}^i{D}^{\mathrm{tk}}$

The utility schedule of surveillance is shown:

$u_b(s_a^t,s_b^k)={\mathrm{\&eta;}}^i{D}^{\mathrm{tk}}{C}_s\left(s_a^t\right)$

Wherein, η ⁱbe verification and measurement ratio, represent that monitoring system successfully detects the probability of the insincere behavior of i node; for secrecy capacity; for taking strategy when channel time secrecy capacity; D ^tkfor the income of the punishment to insincere node and monitoring system; represent k the strategy of monitoring system b, represent t the strategy of insincere node a.