CN104954116A - Image password construction method based on neuron chaotic response coding - Google Patents

Abstract

The invention discloses an image password construction method based on neuron chaotic response coding. According to the image password construction method, firstly, a digital password is converted into a position initial value of a Lorenz system through logarithmic transformation, so as to obtain a chaotic time sequence; linear transformation is performed on the chaotic time sequence combined with a user number, so as to obtain a low-level time sequence, thereby constructing a chaotic square wave exciting signal; the exciting signal is used as input current of a Hodgkin-Huxley neuron, so as to perform the neuron chaotic response coding and record the time for discharging neuron pulses; finally, a time interval sequence of discharging the neuron pulses is obtained, and a one-dimensional time interval sequence is converted into a two-dimensional image password through the gray mapping according to row and line parameters of the image password. According to the image password construction method, the chaotic initial value sensitivity of the Lorenz system and the pulse discharging chaotic parameter sensitivity of a neuron model are fully used, the irreversibility of the image password is guaranteed, and the safety of an identity authentication system is effectively increased.

Description

A kind of image password building method based on neuron chaos motion coding

Technical field

The invention belongs to a kind of in identity authorization system, relate to a kind of image password building method based on neuron chaos motion coding, in conjunction with digital cipher application in identity authorization system, the problem of conventional password cipher safety difference can be solved.

Background technology

Authentication, for differentiating user identity, is the major way that user ensures self privacy and property.Enter today of large data age in the Internet, ID authentication mechanism becomes the important component part of network security technology.Password mechanism is the most basic, most widely used ID authentication mechanism.System is that each legal user establishes user name/password password, and when logging in system by user, user inputs oneself the user name and password password, and itself and the user name and password stored compare by system, judge that whether the identity of user is effective.There is following problem in this ID authentication mechanism: 1. the password code of user stores in systems in which with form expressly, easily leaks; What 2. user inputted to system is the password of plaintext version, and assailant can intercept the password of user from communication line; 3. user cannot the identity of checking system, may have been gained correct password or other valuable information by cheating by the system of falseness.Therefore make to carry out authentication in this way fairly simple, convenient, but fail safe extreme difference.

Summary of the invention

The present invention is directed to the deficiencies in the prior art, propose a kind of image password building method based on neuron chaos motion coding.

After user registration success, user name, numerical ciphers and image password are preserved by user, do not carry out the storage of user cipher in system, only preserve the Customs Assigned Number corresponding to user name.The present invention is using the initial value of numerical ciphers as long-range navigation thatch system, and using Customs Assigned Number as the neuronic input signal parameter of Hodgkin-Huxley, carry out neuron chaos motion coding, utilize chaos initial value sensitivity and the sensitivity to parameter of two systems, final generation decodes the larger image password of difficulty.When logging in system by user, user inputs user name, numerical ciphers and image password, and system, according to user name and numerical ciphers, adopts image password building method of the present invention to produce image password.When the image password of user's input is consistent with the image that system produces, user logs in successfully, completes the authentication to user.

Image password building method of the present invention mainly comprises the following steps:

Step (1) user first input through system authentication registration user name and propose arbitrarily fixed 4-digit number password, wherein user name has uniqueness, and user name automatically can be converted to corresponding Customs Assigned Number by system;

The numerical ciphers that user inputs by step (2) system carries out logarithmic transformation, using transformation results as the position initial value on long-range navigation thatch chaos system three dimensions x direction, makes it to produce the chaos time sequence on x direction;

Step (3), in Self defined time window, is converted to the time series after normalization to the chaos time sequence that step (2) produces; Then using Customs Assigned Number as transformation parameter, linear transformation is carried out to the time series after normalization, rise time sequence;

Described Self defined time window will meet the requirement producing image password sequence length;

Step (4) structure chaos square wave excitation signal, wherein square-wave signal amplitude and high level time are encryption system inner parameter; And square wave low level time is a dynamic value, corresponding to the new time series generated in step (3);

Chaos square wave excitation signal as stimulating current input Hodgkin-Huxley neuron models, is produced neuron pulse and provides sequence by step (5);

The moment is provided in the pulse of step (6) record neuron, obtains the time interval sequence that the moment is provided in pulse;

The time interval sequence that step (7) intercepts particular sequence length carries out grey scale mapping, and presses the ranks parameter of image password, and One-dimension Time Series is converted to 2-D gray image Image, Image is image password; Described particular sequence length is the total number of pixel of image password.

The method that the numerical ciphers that user inputs by the system in described step (2) carries out logarithmic transformation is

$x_1=\left(ln\frac{M_1}{1000}\right)\×4.343---\left(1\right);$

Wherein x ₁as the position initial value on long-range navigation thatch system three dimensions x direction, M ₁for the numerical ciphers of user's input.

In step (3) using Customs Assigned Number as transformation parameter, carry out linear transformation to the time series after normalization, the method for rise time sequence D elay is:

Delay _i＝A+G _i×(ID_N/10 ^m) (2)；

Wherein Delay _irepresent that i-th element, the A of time series Delay are that system fixes constant, G _irepresent the time value after the chaos time sequence normalization that step (2) is produced, ID_N is Customs Assigned Number, 10 ^mrepresent system user total number of persons, wherein i, m are positive integer.

Beneficial effect of the present invention is:

1. the numerical ciphers of user and image password do not need to store in systems in which, are not easy to leak;

2. utilize numerical ciphers to realize the chaos initial value sensitivity of long-range navigation thatch system, strengthen the unpredictability of image password;

3. the chaotic parameter sensitiveness utilizing user name to realize the pulse of Hodgkin-Huxley neuron to provide, and neuron models structure has higher complexity, reduces the possibility that the predicted or construction process of image password is cracked;

4. what user inputted to system is image password, reduces password at communication line by the risk intercepted.

Accompanying drawing explanation

Fig. 1 is flow chart of the present invention;

Fig. 2 is neuron chaos square wave excitation signal configuration figure;

Fig. 3 a is numerical ciphers is M ₁=1627, Customs Assigned Number is the image password schematic diagram of ID_N=50;

Fig. 3 b is numerical ciphers is M ₁=1626, Customs Assigned Number is the image password schematic diagram of ID_N=50;

Fig. 4 a is numerical ciphers is M ₁=1627, Customs Assigned Number is the image password schematic diagram of ID_N=50;

Fig. 4 b is numerical ciphers is M ₁=1627, Customs Assigned Number is the image password schematic diagram of ID_N=51; Embodiment:

By reference to the accompanying drawings 1, the concrete implementation step of the present invention is:

First step (1) user inputs through the user name of system authentication registration, and user name can be converted to corresponding Customs Assigned Number ID_N by system automatically, and system user total number of persons is m the progression of 10, and user can propose surely arbitrary 4-digit number password M ₁; M is positive integer;

The numerical ciphers M that user inputs by step (2) system ₁carry out logarithmic transformation, shown in (1).

$x_1=\left(ln\frac{M_1}{1000}\right)\×4.343---\left(1\right)$

By transformation results x ₁as the position initial value on long-range navigation thatch system three dimensions x direction, and other parameters of long-range navigation thatch system and the initial value in y and z direction are generated automatically by identity authorization system, and long-range navigation thatch system is such as formula shown in (2).

$\begin{array}{c}\frac{dx}{dt}=a(y-x)\\ \frac{dy}{dt}=x(b-z)-y\\ \frac{dz}{dt}=xy-cz\end{array}---\left(2\right)$

In this exemplifying embodiment, the initial value that the initial value of a=10, b=28, c=8/3, y equals 2, z equals 10.Thus the chaos time sequence obtained on x direction, be designated as X, shown in (3).

X＝{x ₁,x ₂,x ₃,......} (3)

Wherein x _ibe illustrated respectively in the positional value on long-range navigation thatch system x direction that i-th sampling instant obtain.

Step (3) is normalized the time chaos sequence X (being such as taken as 10s) in special time window and producing, and obtains the time series G={G after normalization ₁, G ₂, G ₃... ..}, G _ishown in (4).

$G_i=\frac{(x_i-X_{min})}{(X_{max}-X_{min})}---\left(4\right)$

Wherein, X _minfor the minimum value in chaos time sequence X, X _maxfor the maximum in chaos time sequence X.Then in conjunction with Customs Assigned Number ID_N, linear transformation rise time sequence D elay={Delay is utilized ₁, Delay ₂, Delay ₃..., Delay _ishown in (5), wherein A is that system fixes constant, such as desirable A=10.The implication of m as described in step (1), such as desirable m=2.

Delay _i＝A+G _i×(ID_N/10 ^m) (5)

Step (4) structure chaos square wave excitation signal, wherein square-wave signal amplitude and high level time are encryption system inner parameter, and high level time is T_top, and low and high level difference in magnitude is F, and low level amplitude is F ₀.And square wave low level time is a dynamic value, corresponding with the time series Delay produced in step (3); Chaos square wave excitation signal is concrete as shown in Figure 2.

Step (5) carries out neuron response coding to chaos square wave excitation signal, using chaos square wave excitation signal as stimulating current input Hodgkin-Huxley neuron models, shown in (6) ~ (9).It is the time instant that film potential V exceedes threshold value (such as can be set as 10mv) that the moment is provided in the pulse of definition neuron, thus obtains pulse granting moment sequence Spike={S ₁, S ₂..., wherein S _jrepresent the moment that a jth pulse is provided.

$C_m\frac{dV}{dt}=I_{ext}-g_{Na}{m}^{3}h\left(V-V_{Na}\right)-g_K{n}^4\left(V-V_K\right)-g_L\left(V-V_L\right)---\left(6\right)$

K _t＝3 ^(T-6.3)/10(7)

$\begin{array}{c}\stackrel{\·}{m}=K_t\[{\mathrm{\α}}_m\left(V\right)(1-m)-{\mathrm{\β}}_m\left(V\right)m\]\\ \stackrel{\·}{h}=K_t\[{\mathrm{\α}}_h\left(V\right)(1-h)-{\mathrm{\β}}_h\left(V\right)h\]\\ \stackrel{\·}{n}=K_t\[{\mathrm{\α}}_n\left(V\right)(1-n)-{\mathrm{\β}}_n\left(V\right)n\]\end{array}---\left(8\right)$

$\begin{array}{c}{\mathrm{\α}}_m=0.1\×\frac{25-V}{e^{\frac{25-V}{10}}-1},\\ {\mathrm{\β}}_m=4\×e^{-\frac{V}{18}},\\ {\mathrm{\α}}_h=0.07\×e^{-\frac{V}{20}},\\ {\mathrm{\β}}_h=\frac{1}{e^{\frac{30-V}{10}}+1},\\ {\mathrm{\α}}_n=0.01\×\frac{10-V}{e^{\frac{10-V}{10}}-1},\\ {\mathrm{\β}}_m=0.125\×e^{-\frac{V}{80}}.\end{array}---\left(9\right)$

Wherein neuron parameter value C _m=1uF/cm ², T=6.3 DEG C, g _na=120mS/cm ², g _k=36mS/cm ², g _l=0.3mS/cm ², V _na=54.97mV, V _k=-71.96mV, V _l=-49mV.I _extfor the input current of neuron models, be taken as the chaos square wave excitation signal in step (4).

Step (6) provides moment sequence Spike={S to neuron pulse ₁, S ₂... ask difference, obtain the time interval sequence ISI={ISI that the moment is provided in pulse ₁, ISI ₂, ISI ₃..}, wherein ISI _i=S _i+1-S _i.

Step (7) supposes that the ranks parameter of image password is N (such as N=20), and intercepting sequence length is that the ISI sequence of N × N carries out grey scale mapping, obtains One-dimension Time Series M={M ₁, M ₂, M ₃... ..}, M _ishown in (10).

$M_i=\frac{({\mathrm{ISI}}_i-{\mathrm{ISI}}_{\min })}{({\mathrm{ISI}}_{max}-{\mathrm{ISI}}_{min})}\×256---\left(10\right)$

Wherein, ISI _minfor the minimum value in time interval sequence ISI, ISI _maxfor the maximum in time interval sequence ISI.

One-dimension Time Series M is converted by rows to the 2-D gray image Image of N × N, Image is image password, and schematic diagram is as shown in accompanying drawing 3a, Fig. 3 b, Fig. 4 a, Fig. 4 b.

Claims (3)

1., based on an image password building method for neuron chaos motion coding, it is characterized in that, the method specifically comprises the following steps:

Step (1) user first input through system authentication registration user name and propose arbitrarily fixed 4-digit number password, wherein user name has uniqueness, and user name automatically can be converted to corresponding Customs Assigned Number by system;

The numerical ciphers that user inputs by step (2) system carries out logarithmic transformation, using transformation results as the position initial value on long-range navigation thatch chaos system three dimensions x direction, makes it to produce the chaos time sequence on x direction;

Step (3), in Self defined time window, is converted to the time series after normalization to the chaos time sequence that step (2) produces; Then using Customs Assigned Number as transformation parameter, linear transformation is carried out to the time series after normalization, rise time sequence;

Described Self defined time window will meet the requirement producing image password sequence length;

Step (4) structure chaos square wave excitation signal, wherein square-wave signal amplitude and high level time are encryption system inner parameter; And square wave low level time is a dynamic value, corresponding to the new time series generated in step (3);

Chaos square wave excitation signal as stimulating current input Hodgkin-Huxley neuron models, is produced neuron pulse and provides sequence by step (5);

The moment is provided in the pulse of step (6) record neuron, obtains the time interval sequence that the moment is provided in pulse;

The time interval sequence that step (7) intercepts particular sequence length carries out grey scale mapping, and presses the ranks parameter of image password, and One-dimension Time Series is converted to 2-D gray image Image, Image is image password; Described particular sequence length is the total number of pixel of image password.

2. a kind of image password building method based on neuron chaos motion coding according to claim 1, is characterized in that: the method that the numerical ciphers that user inputs by the system in described step (2) carries out logarithmic transformation is

$x_1=\left(ln\frac{M_1}{1000}\right)\×4.343---\left(1\right);$

Wherein x ₁as the position initial value on long-range navigation thatch system three dimensions x direction, M ₁for the numerical ciphers of user's input.

3. a kind of image password building method based on neuron chaos motion coding according to claim 1, it is characterized in that: in step (3) using Customs Assigned Number as transformation parameter, carry out linear transformation to the time series after normalization, the method for rise time sequence D elay is:

Delay _i＝A+G _i×(ID_N/10 ^m) (2)；

Wherein Delay _irepresent that i-th element, the A of time series Delay are that system fixes constant, G _irepresent the time value after the chaos time sequence normalization that step (2) is produced, ID_N is Customs Assigned Number, 10 ^mrepresent system user total number of persons, wherein i, m are positive integer.