1: \begin{abstract} 

2: Distributed control systems require high reliability and availability guarantees, despite

3: often being deployed at the edge of network infrastructure.  Edge computing resources

4: are less secure and less reliable than centralized resources in

5: data centers.  Replication and consensus protocols improve robustness to

6: network faults and crashed or corrupted nodes, but these volatile 

7: environments can cause non-faulty nodes to temporarily diverge, increasing

8: the time needed for replicas to converge on a consensus value, and 

9: give Byzantine attackers too much influence over the convergence process.

10: 

11: This paper proposes \emph{proximal Byzantine consensus}, a new

12: \emph{approximate} consensus protocol where clients use statistical models of

13: streaming computations to decide a consensus value. In addition, it provides an

14: interval around the decision value and the probability that the true

15: (non-faulty, noise-free) value falls within this interval.

16: Proximal consensus (PC)

17: tolerates unreliable network conditions, Byzantine behavior, and other sources of 

18: noise that cause honest replica states to diverge.  

19: We evaluate our approach for scalar values, and compare 

20: PC simulations against a vector consensus (VC) protocol simulation.

21: Our simulations demonstrate that consensus values selected by PC

22: have lower error and are more robust against Byzantine attacks. 

23: We formally characterize the security guarantees against Byzantine attacks 

24: and demonstrate attacker influence is bound with high probability.

25: Additionally, an informal complexity analysis suggests PC

26: scales better to higher dimensions than convex hull-based protocols such as VC.

27: \end{abstract}

28: