Axiomatizing Regular Words

Stephen L. Bloom
Stevens Institute of Technology

Wednesday, February 19, 2:00PM
Lieb 3rd floor Conference Room
 

Abstract


Fixed point equations in words have been used to express the behavior of hardware and software. For example, the equation
x = a b x (1)
describes the behavior of a device capable of performing the 'action' a, followed by the action b, forever.

In 1978, Courcelle showed that (initial) solutions certain systems of finite fixed point equations in words could be expressed by regular expressions formed from letters in a fixed alphabet using word operations of concatenation (or product), omega and omega-op power.

For example, the initial solution to the equation (1) is written (a.b)^omega.

Courcelle asked for a complete equational axiomatization of words with product, omega and omega-op power.

In 1981, Heilbrunner showed that all systems of finite fixed point equations in words could be expressed using the operations above together with infinitely many 'shuffle operations'. He asked for a complete equational axiomatization of words with all of these operations.

A 'regular expression' on the alphabet A is a term formed from letters in A using the operations of composition, omega, omega-op power and the shuffle operations.

In 1986, Thomas found an algorithm to decide when $|s|=|t|$, for regular expressions $s,t,$, where $|s|$ is the word named by the expression $s$. Thomas' algorithm requires $O(2^{2^n})$-steps, where $n$ is the total number of symbols in $s,t$.

In 2001, Bloom and Choffrut gave a complete equational axiomatization for product and omega power.

In 2002, Bloom and \'Esik found a complete equational axiomatization for product, omega and omega-op power.

The talk will describe the complete solution to Heilbrunner's question and a polynomial time algorithm to decide if $|s|=|t|$.