Matrix analysis
Distributed Algorithms
A scheme for robust distributed sensor fusion based on average consensus
IPSN '05 Proceedings of the 4th international symposium on Information processing in sensor networks
Approximate distributed Kalman filtering in sensor networks with quantifiable performance
IPSN '05 Proceedings of the 4th international symposium on Information processing in sensor networks
| Hi-index | 0.00 |
This paper introduces new methods for sensor fusion in homogenous sensor networks in the presence of the communication time delay. It allows the sensor nodes to solve consensus problems despite the network induced delay. It is shown that such approaches guarantee tracking of each consensus filter output to the original value of the measured signal in both regular and non-regular network graph topology. In the non-regular case, there is a sufficient condition to prove the convergence. In the networks with communication time delay, consensus problem is solved by considering time delay as a constant as well as using the idea of the variable sampling period in order for time delay compensation. In the second case, each sensor sampling period is variable and it is taken as maximum predicted time delay of the all channels in the network at each sampling period. In this way, an MLP neural network is used for time delay prediction and a majority consensus filter is introduced for approximating maximum predicted time delay in the network. Simulation results are provided that demonstrate the effectiveness of these methods for distributed sensor fusion and solving consensus problem in the presence of the time delay.