On Hash-Based Work Distribution Methods for Parallel Best-First Search

TL;DR

This paper introduces Abstract Zobrist hashing and GRAZHDA* to improve load balancing and reduce communication overhead in parallel best-first search algorithms, especially for domain-independent planning.

Contribution

It proposes a new feature projection-based hashing method and an automatic graph-partitioning approach for better work distribution in parallel search.

Findings

Abstract Zobrist hashing reduces communication overhead compared to standard Zobrist hashing.

GRAZHDA* outperforms previous methods on domain-independent planning tasks.

The methods improve load balance and efficiency in parallel best-first search.

Abstract

Parallel best-first search algorithms such as Hash Distributed A* (HDA*) distribute work among the processes using a global hash function. We analyze the search and communication overheads of state-of-the-art hash-based parallel best-first search algorithms, and show that although Zobrist hashing, the standard hash function used by HDA*, achieves good load balance for many domains, it incurs significant communication overhead since almost all generated nodes are transferred to a different processor than their parents. We propose Abstract Zobrist hashing, a new work distribution method for parallel search which, instead of computing a hash value based on the raw features of a state, uses a feature projection function to generate a set of abstract features which results in a higher locality, resulting in reduced communications overhead. We show that Abstract Zobrist hashing outperforms previous methods on search domains using hand-coded, domain specific feature projection functions. We then propose GRAZHDA*, a graph-partitioning based approach to automatically generating feature projection functions. GRAZHDA* seeks to approximate the partitioning of the actual search space graph by partitioning the domain transition graph, an abstraction of the state space graph. We show that GRAZHDA* outperforms previous methods on domain-independent planning.