1: \begin{abstract}

2: The communication matrix for two-way deterministic finite automata (2DFA) with $n$ states

3: is defined for an automaton over a full alphabet of all $(2n+1)^n$ possible symbols:

4: its rows and columns are indexed by strings,

5: and the entry $(u, v)$ is $1$ if $uv$ is accepted by the automaton,

6: and $0$ otherwise.

7: With duplicate rows and columns removed, this is a square matrix of order $n(n^n-(n-1)^n)+1$,

8: and its rank is known to be a lower bound

9: on the number of states necessary to transform an $n$-state 2DFA

10: to a one-way unambiguous finite automaton (UFA).

11: This paper determines this rank,

12: showing that it is exactly

13: $\boundfunc{n}=\sum_{k=1}^n \binom{n}{k-1} \binom{n}{k} \binom{2k-2}{k-1} =(1+o(1)) \frac{3\sqrt{3}}{8\pi n} 9^n$,

14: and this function becomes the new lower bound

15: on the state complexity of the 2DFA to UFA transformation,

16: thus improving a recent lower bound by S. Petrov and Okhotin

17: (\href{https://doi.org/10.1007/978-3-030-68195-1_7}

18: {``On the transformation of two-way deterministic finite automata to unambiguous finite automata''},

19: \emph{Inf.\ Comput.}, 2023).

20: The key element of the proof is determining the rank of a $k! \times k!$ submatrix,

21: with its rows and columns indexed by permutations,

22: where the entry $(\pi, \sigma)$ is $1$ if $\sigma \circ \pi$ is a cycle of length $k$,

23: and 0 otherwise;

24: using the methods of group representation theory

25: it is shown that its rank is exactly $\binom{2k-2}{k-1}$,

26: and this implies the above formula for $\boundfunc{n}$.

27: 

28: \textbf{Keywords: } Two-way finite automata,

29: unambiguous finite automata,

30: communication matrix,

31: matrix rank,

32: symmetric group,

33: group representation theory.

34: \end{abstract}