Matlab Code for "Robust one-shot estimation over shared networks in the presence of denial-of-service attacks"

Proactive Jammer

p2p channel

• OptimalJammingProbability.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer.
• OptimalJammingProbability_VS_d.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer as a function of d.
• OptimalJammingProbability_VS_var.m: Optimal jamming probabilities $\varphi^\star$ as a function of $\sigma^2$.
• OptimalJammingProbability_VS_cd.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer as a function of $c$ and $d$.

Large-scale network

• OptimalTransmission_Jamming_Policy.m: Optimal Optimal transmission policy and jamming probability $\varphi^\star$ for the proactive jammer over large-scale network.
• OptimalJammingProbability_VS_d.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer over large-scale network as a function of d.
• OptimalJammingProbability_VS_kappa.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer over large-scale network as a function of $\bar{\kappa}$.
• OptimalJammingProbability_VS_cd.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer over large-scale network as a function of c and d. Here, X ~ N(0,1).
• OptimalJammingProbability_VS_ckappa.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer over large-scale network as a function of c and $\bar{\kappa}$. Here, X ~ N(0,1).
• OptimalJammingProbability_VS_dkappa.m: Optimal jamming probability $\varphi^\star$ for the proactive jammer over large-scale network as a function of d and $\bar{\kappa}$. Here, X ~ N(0,1).

Reactive Jammer

1 Dimension Gaussian variable

functions

• MonteCarlo_AlgorithmComparison.m: Monte Carlo simulations for PGA-CCP and GDA in 1 dimension
• Optimalalphabeta_VS_var: Optimal jamming probabilities $\alpha^\star$ and $\beta^\star$ as a function of $\sigma^2$.
• PGA_CCP_Algorithm: Convergence of PGA-CCP
• GDA_Algorithm.m: Convergence of GDA

auxiliary functions

• FirstNashEqulibiumChecker.m: Check whether approximate First Nash Equilibrium is satisfied
• grad_PGA.m Gradients for PGA
• grad_CCP.m: Gradients for CCP
• grad_GD.m: Gradients for GD

Multidimensional Guassian vector

Description: the algorithm is accelerated by replacing iterations with matrix computation.

functions

• MonteCarlo_MultiDim_AlgorithmComparison.m: Monte Carlo simulations for PGA-CCP and GDA for multidimensional state, the expectation is estimated by using the average of random samples
• MultiDim_PGA_CCP_Algorithm Convergence of PGA-CCP for multidimensional state
• MultiDim_GDA_Algorithm.m: Convergence of GDA for multidimensional state

auxiliary functions

• batch_grad_PGA.m Estimation of Gradients for PGA
• batch_grad_CCP.m: Estimation of Gradients for CCP
• batch_grad_GD.m: Estimation of Gradients for GD
• batch_FirstNashEqulibiumChecker.m: Check whether the estimated approximate First Nash Equilibrium is satisfied