This folder contains Matlab codes for the IEEE Internet of Things Journal paper titled "Multiuser Scheduling in Centralized Cognitive Radio Networks: A Multi-Armed Bandit Approach". The reader can find the required codes related to the algorithm provided in the paper. All the main files and corresponding functions include necessary in-line comments to ease following the algorithms through the codes.

The main parameters of the three CPS system need first go through the KalmanFilter.m to compute the steady state error covariance and \beta_c. Then, the channel and power level parameters can be set in the J-CAP algorithm folder and then with the input from the output of Kalman filter the algorithms runs over T time horizon.

We study security of remote state estimation in wireless cyber-physical systems (CPS) where a sensor sends its measurements to the remote state estimator over a multi-channel wireless link in presence of a jamming attacker. Most of the existing works study the sensor's defense scheme by adopting optimization-based methods and rely on the prior knowledge of the attacker’s attack policy.To relax this constraint, we propose a novel online learning-based policy called J-CAP (Joint Channel And Power selection) for the sensor to dynamically choose transmission channel and power. The proposed method assumes \emph{no prior} knowledge of the attacker's attack policy, nor of the channel state information. J-CAP jointly optimizes sensor's channel selection and power consumption, and guarantees the estimator’s asymptotic stability. We theoretically prove that J-CAP achieves a sublinear learning regret bound. We also show J-CAP's optimality by deriving and matching its regret lower and upper bound orders. Compared with the solution that directly applies the baseline solution, J-CAP improves the regret upper 0bound by a factor of $\sqrt{K+L}$, where $K$ and $L$ denote the number of channels and number of power levels, respectively. Numerical evaluations validate the analytical results under various CPS parameters, and compare the J-CAP's performance with the state-of-the-art solutions.

We propose a novel online learning-based algorithm called J-CAP, for joint channel and power level selection by the sensor. J-CAP is presented in Algorithm 1 in the paper. Based on J-CAP, at each time $t$, the sensor chooses channel $i_t\in[K]$, and power level $l_t\in[L]$ according to the probabilities $p_i(t)$ and $q_l(t)$ distributed over $K$ and $L$, respectively (see step 2 to 5 in J-CAP algorithm). These distributions are a mixture of the uniform distribution (i.e., the terms $\frac{\gamma}{K}$ and $\frac{\gamma}{L}$) and a distribution which depends exponentially on the past observations for that channel and power level (i.e., the first term in the definition of $p_i(t)$ and $q_l(t)$). Mixing the uniform distribution on both sets of $K$ and $L$ actions, enables the algorithm to explore all the actions in these sets to find the best channel and power level pair.

@ARTICLE{9380209, author={Alipour-Fanid, Amir and Dabaghchian, Monireh and Wang, Ning and Jiao, Long and Zeng, Kai}, journal={IEEE Internet of Things Journal}, title={Online Learning-Based Defense Against Jamming Attacks in Multi-Channel Wireless CPS}, year={2021}, volume={}, number={}, pages={1-1}, doi={10.1109/JIOT.2021.3066476}}