1: \begin{abstract}

2: 

3:   Self-stabilizing algorithms are an important because

4:   of their robustness and guaranteed convergence.

5:   Starting from any

6:   arbitrary state, a self-stabilizing algorithm is guaranteed to converge to a legitimate

7:   state.%  A number of self-stabilizing

8:   % graph algorithms are used in process communication and

9:   % fault tolerance.

10:   % Using those algorithms for distributed graph processing, however,

11:   %is not straightforward.

12:   Those algorithms are not directly amenable to solving distributed graph

13:   processing problems when performance and scalability are important.

14: %

15:   %In this paper, we consider self-stabilizing Single Source Shortes Path(SSSP)

16:   %algorithms.

17:   In this paper, we show the ``Abstract Graph Machine'' (AGM) model that can be used

18:   to convert self-stabilizing algorithms into forms suitable for distributed

19:   graph processing.

20: %

21:   An AGM is a mathematical

22:   model of parallel computation on graphs that

23:   adds work dependency and ordering to self-stabilizing algorithms.

24: %

25:   Using the AGM model we show that some of the existing distributed

26:   Single Source Shortest Path (SSSP) algorithms

27:   are actually specializations of self-stabilizing SSSP.

28: %

29:   We extend the AGM model to apply more fine-grained orderings at different spatial

30:   levels to derive additional scalable variants of SSSP algorithms, essentially

31:   enabling the algorithm to be generated for a specific target architecture.

32: %

33:   Experimental results show that this approach can generate new algorithmic variants

34:   that out-perform standard distributed algorithms for SSSP.

35: 

36: \end{abstract}

37: