Maximum correntropy Kalman filter algorithm with two-step random delay and packet loss

Due to the limited width of the communication channel, data delay and packet loss occur at the receiver when receiving measurement data. The number of delay steps is random, but the probability of delay occurrence decreases as the number of steps increases. Therefore, in this paper, we use a binary Bernoulli distribution to model measurement data with known delay probability, and redesign the Kalman filter(KF) and Maximum correntropy Kalman filter (MCKF) algorithms to accommodate a system with two steps of random delay and packet loss. We compare the improved KF and MCKF algorithms under Gaussian noise and non-Gaussian noise systems, and the experimental results show that the improved KF and MCKF algorithms possess high estimation accuracy under Gaussian noise and non-Gaussian noise systems, respectively, when the systems undergoe two-step random delay and packet loss.