1: \begin{abstract}

2: Fully distributed estimation and tracking solutions to large-scale multi-agent networks suffer slow convergence and are vulnerable to network failures.  

3: % the bottlenecks of scalability and fault resilience. 

4: In this paper, we aim to speed up the convergence and enhance the resilience of state estimation and tracking using a simple hierarchical system architecture wherein agents are clusters into smaller networks, and a parameter server exists to aid the information

5: exchanges among networks. The information exchange among networks is expensive and occurs only once in a while. %We consider the challenging communication failures -- packet-dropping links -- wherein a communication link may drop the transmitted messages unexpectedly and without notifying the sender. 

6: 

7: We propose two ``consensus + innovation'' algorithms for the state estimation and tracking problems, respectively. 

8: In both algorithms, we use a novel hierarchical push-sum consensus component. For the state estimation, we use dual averaging as the local innovation component. 

9: State tracking is much harder to tackle in the presence of dropping-link failures and the standard integration of the consensus and innovation approaches are no longer applicable. Moreover, dual averaging is no longer feasible. Our algorithm introduces a pair of additional variables per link and ensure the relevant local variables evolve according to the state dynamics, and use projected local gradient descent as the local innovation component. 

10: We also characterize the convergence rates of both of the algorithms under linear local observation model and minimal technical assumptions. 

11: We numerically validate our algorithm through simulation of both state estimation and tracking problems.

12: \end{abstract}

13: