[WUSTCTF2020]dp_leaking_1s_very_d@angerous

题目:

e = 65537
n = 156808343598578774957375696815188980682166740609302831099696492068246337198792510898818496239166339015207305102101431634283168544492984586566799996471150252382144148257236707247267506165670877506370253127695314163987084076462560095456635833650720606337852199362362120808707925913897956527780930423574343287847
c = 108542078809057774666748066235473292495343753790443966020636060807418393737258696352569345621488958094856305865603100885838672591764072157183336139243588435583104423268921439473113244493821692560960443688048994557463526099985303667243623711454841573922233051289561865599722004107134302070301237345400354257869
dp = 734763139918837027274765680404546851353356952885439663987181004382601658386317353877499122276686150509151221546249750373865024485652349719427182780275825

首先,观察题目,一道典型的rsa——dp泄露问题,我们都知道dp=d mod p-1

=>dp*e=d*e mod p-1=>dp*e=(p-1)(k1(q-1)+k2)+1

=>由于我们可知dp<p-1,所以(k1(q-1)+k2)取值范围为(1.e),暴力求解即可

for i in range(1,e):
    if (dp*e-1)%i==0:
        if n%((dp*e-1)//i+1)==0:
            p=(dp*e-1)//i+1
            q=n//p
            phi=(p-1)*(q-1)
            d=gmpy2.invert(e,phi)
            print(libnum.n2s(int(gmpy2.powmod(c,d,n))))
            break

写出dp泄露脚本

最终代入数据:

import gmpy2
import libnum
e = 65537
n = 156808343598578774957375696815188980682166740609302831099696492068246337198792510898818496239166339015207305102101431634283168544492984586566799996471150252382144148257236707247267506165670877506370253127695314163987084076462560095456635833650720606337852199362362120808707925913897956527780930423574343287847
c = 108542078809057774666748066235473292495343753790443966020636060807418393737258696352569345621488958094856305865603100885838672591764072157183336139243588435583104423268921439473113244493821692560960443688048994557463526099985303667243623711454841573922233051289561865599722004107134302070301237345400354257869
dp = 734763139918837027274765680404546851353356952885439663987181004382601658386317353877499122276686150509151221546249750373865024485652349719427182780275825


for i in range(1,e):
    if (dp*e-1)%i==0:
        if n%((dp*e-1)//i+1)==0:
            p=(dp*e-1)//i+1
            q=n//p
            phi=(p-1)*(q-1)
            d=gmpy2.invert(e,phi)
            print(libnum.n2s(int(gmpy2.powmod(c,d,n))))
            break


 得到:

b'wctf2020{dp_leaking_1s_very_d@angerous}'

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
以下是使用ADMM训练回声状态网络(ESN)的MATLAB代码,用于MackeyGlass_t17数据集的预测: ``` % Load data load('MackeyGlass_t17.mat'); data = MackeyGlass_t17'; % Define ESN parameters n = 1000; % number of neurons rho = 0.99; % spectral radius alpha = 0.5; % leaking rate win = randn(n,1); % input weights w = randn(n,n)*0.4; % reservoir weights % Train ESN using ADMM lambda = 10; % regularization parameter maxIter = 100; % maximum number of iterations rho_admm = 1.5; % ADMM penalty parameter x = zeros(n,1); % initial state z = zeros(n,maxIter); % ADMM auxiliary variable u = zeros(n,maxIter); % ADMM dual variable for i = 1:maxIter % Update reservoir state x = (1-alpha)*x + alpha*tanh(w*x + win*data(i)); % Update auxiliary variable z(:,i) = (w + rho_admm*eye(n)) \ (win*data(i) - w*x + u(:,i)); % Update dual variable u(:,i+1) = u(:,i) + rho_admm*(w*x + z(:,i) - win*data(i)); end % Compute output weights using ridge regression X = z(:,1:end-1)'; Y = data(1:end-1)'; Wout = (X'*X + lambda*eye(size(X,2))) \ (X'*Y); % Test ESN on new data numSamples = 500; % number of test samples y = zeros(numSamples,1); % ESN predictions x = zeros(n,1); % initial state for i = 1:numSamples x = (1-alpha)*x + alpha*tanh(w*x + win*y(i)); y(i) = Wout'*tanh(w*x + win*data(maxIter+i)); end % Plot results figure; plot(data(maxIter+1:maxIter+numSamples),'b'); hold on; plot(y,'r--'); legend('True','Predicted'); title('Mackey-Glass prediction using ADMM-trained ESN'); ``` 在这个代码中,我们首先加载MackeyGlass_t17数据集,定义ESN的参数,并使用ADMM算法训练ESN。在训练期间,我们更新ESN的状态,计算ADMM辅助变量和双重变量,并使用岭回归计算输出权重。最后,我们使用训练好的ESN对新数据进行测试,并绘制真实值和预测值。 请注意,这是一个简化的代码示例,可能需要根据具体情况进行调整。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值