The deployment of sensors in multi-sensor network systems needs to measure the actual state of the target. There may be significant errors in transmitting information to each sensor node through the sensor network system, which cannot accurately reflect the actual target state, resulting in the sensor network generating wrong measurements and instructions. The accuracy of target state estimation is affected by two factors: random noise in the process of sensor measurement and random delay of sensor data information in the network transmission process. Because of the above problems, this paper designs a probabilistic time window multi-sensor arithmetic average fusion multi-target tracking algorithm that can resist these interference factors, compensate for the random delay, and calculate the correct probability cumulative value of different sensor nodes in the time window, and then adjusts the probability hypothesis density filter update value as the multi-sensor fusion weight to reduce the negative impact of random interference on the sensor measurement. The simulation results show that under the condition of accurate probability information of random noise and random delay, compared with the traditional arithmetic average fusion and GCI fusion algorithms, the proposed probabilistic time window multi-sensor network arithmetic average fusion algorithm has the most minor tracking error and is more suitable for complex tracking scenarios with clutter and detection probability changes.

Aiming at the problem of asynchronous multi-target tracking, this paper studies the AA fusion optimization problem of multi-sensor networks. Firstly, each sensor node runs a PHD filter, and the measurement information obtained from different sensor nodes in the fusion interval is flood communicated into composite measurement information. The Gaussian component representing the same target is associated with a subset by distance correlation. Then, the Bayesian Cramér–Rao Lower Bound of the asynchronous multi-target-tracking error, including radar node selection, is derived by combining the composite measurement information representing the same target. On this basis, a multi-sensor-network-optimization model for asynchronous multi-target tracking is established. That is, to minimize the asynchronous multi-target-tracking error as the optimization objective, the adaptive optimization design of the selection method of the sensor nodes in the sensor network is carried out, and the sequential quadratic programming (SQP) algorithm is used to select the most suitable sensor nodes for the AA fusion of the Gaussian components representing the same target. The simulation results show that compared with the existing algorithms, the proposed algorithm can effectively improve the asynchronous multi-target-tracking accuracy of multi-sensor networks.

Multi‐sensor networks often encounter challenges such as inconsistent sampling times among local sensors and data loss during transmission. To address these issues, this paper employs a data loss compensation strategy to reconstruct missing data information. It designs the state estimation of local sensors utilising iterative state equations, leveraging multistep prediction techniques to estimate sensor states at unsampled points, thereby transforming the asynchronous sensor network system into a synchronous one. Furthermore, the projection theorem is applied to determine the fusion weights of local sensors, grounded on the principle of square‐averaging significance. Ultimately, data information pertaining to the same target is fused through arithmetic averaging, guided by distance correlation. Simulation outcomes demonstrate that the proposed algorithm balances estimation accuracy with communication overhead, achieved by designing an optimal number of communication iterations.

In response to the multitarget tracking problem of distributed sensors with a limited detection range, a distributed sensor measurement complementary Gaussian component correlation GCI fusion tracking method is proposed on the basis of the probabilistic hypothesis density filtering tracking theory. First, the sensor sensing range is extended by complementing the measurements. In this case, the multitarget density product is used to classify whether the measurements belong to the intersection region of the detection range. The local intersection region is complemented only once to reduce the computational cost. Secondly, each sensor runs a probabilistic hypothesis density filter separately and floods the filtering posterior with the neighboring sensors so that each sensor obtains the posterior information of the neighboring sensors. Subsequently, Gaussian components are correlated by distance division, and Gaussian components corresponding to the same target are correlated into the same subset. GCI fusion is performed on each correlated subset to complete the fusion state estimation. Simulation experiments show that the proposed method can effectively perform multitarget tracking in a distributed sensor network with a limited sensing range.