The Descriptive Complexity of Subgraph Isomorphism in the Absence of Order

Let $C$ be a class of graphs and $\pi$ be a graph parameter. Let $\Phi$ be a formula in the first-order language containing only the adjacency and the equality relations. We say that $\Phi$ \emph{defines $C$ on connected graphs with sufficiently large $\pi$} if there is a constant $k$ such that, for every connected graph $G$ with $\pi(G)\ge k$, $\Phi$ is true on $G$ exactly when $G$ belongs to $C$. For a fixed connected graph $F$, let $S(F)$ denote the class of all graphs containing $F$ as a subgraph. Let $D_\pi(F)$ denote the minimum quantifier depth of a formula $\Phi$ defining $S(F)$ on connected graphs with sufficiently large $\pi$. We have $D_v(F)\ge D_{tw}(F)\ge D_\kappa(F)$, where $v(G)$ denotes the number of vertices in a graph $G$, $tw(G)$ is the treewidth of $G$, and $\kappa(G)$ is the connectivity of~$G$. We obtain the following results. - There are graphs $F$ such that $D_v(F)$ is strictly smaller than the number $n$ of vertices in $F$. In particular, $D_v(P_n)=n-1$ for the path graphs on $n\ge4$ vertices. Moreover, there are some trees $F$ such that $D_v(F)\le n-3$. - On the other hand, $D_v(F)=D_{tw}(F)=n$ if $F$ has no vertex of degree 1. In general, $D_v(F)>n/2$ unless $F=P_2$ or $P_3$. - $D_{tw}(F)\ge tw(F)$ for every $F$. Over trees $F$ with $n$ vertices, the values of $D_{tw}(F)$ occupy the almost full spectrum $\{1,5,\ldots,n\}$. The minimum value $D_{tw}(F)=1$ is attained if $F$ is a subtree of a subdivided 3-star $K_{1,3}$. The maximum $D_{tw}(K_{1,n-1})=n$ is attained for the star graphs on $n\ge5$ vertices. - $D_\kappa(F)\ge\frac mn+2$ whenever the number $m$ of edges in $F$ is larger than the number $n$ of vertices. Over graphs $F$ with $n$ vertices, the values of $D_\kappa(F)$ occupy the almost full spectrum $\{1,3,\ldots,n\}$.

[1]  Stephan Olariu,et al.  Paw-Fee Graphs , 1988, Inf. Process. Lett..

[2]  Alexander A. Razborov,et al.  On the AC0 Complexity of Subgraph Isomorphism , 2014, 2014 IEEE 55th Annual Symposium on Foundations of Computer Science.

[3]  Noga Alon,et al.  The Probabilistic Method , 2015, Fundamentals of Ramsey Theory.

[4]  Andrzej Lingas,et al.  Induced subgraph isomorphism: Are some patterns substantially easier than others? , 2012, Theor. Comput. Sci..

[5]  Reinhard Diestel,et al.  Graph Theory , 1997 .

[6]  Alexander A. Razborov,et al.  On the AC0 Complexity of Subgraph Isomorphism , 2014, FOCS.

[7]  N. Wormald,et al.  Models of the , 2010 .

[8]  Anton Kotzig,et al.  Cross-cloning and antipodal graphs , 1988, Discret. Math..

[9]  Bruno Courcelle,et al.  The Monadic Second-Order Logic of Graphs. I. Recognizable Sets of Finite Graphs , 1990, Inf. Comput..

[10]  Noga Alon,et al.  Color-coding , 1995, JACM.

[11]  Saharon Shelah,et al.  Choiceless Polynomial Time , 1997, Ann. Pure Appl. Log..

[12]  Dániel Marx Can you beat treewidth? , 2007, FOCS.

[13]  Hans L. Bodlaender,et al.  A linear time algorithm for finding tree-decompositions of small treewidth , 1993, STOC.

[14]  C. R. Subramanian,et al.  Girth and treewidth , 2005, J. Comb. Theory, Ser. B.

[15]  Joel H. Spencer,et al.  Threshold functions for extension statements , 1990, J. Comb. Theory, Ser. A.

[16]  Alon Itai,et al.  Finding a minimum circuit in a graph , 1977, STOC '77.

[17]  Anuj Dawar A Restricted Second Order Logic for Finite Structures , 1998, Inf. Comput..

[18]  Maksim Zhukovskii Zero-one k-law , 2012, Discret. Math..

[19]  Oleg Verbitsky,et al.  Logical complexity of graphs: a survey , 2010, AMS-ASL Joint Special Session.

[20]  Ge Xia,et al.  Strong computational lower bounds via parameterized complexity , 2006, J. Comput. Syst. Sci..

[21]  Svante Janson,et al.  Random graphs , 2000, ZOR Methods Model. Oper. Res..

[22]  Neil Immerman,et al.  Descriptive Complexity , 1999, Graduate Texts in Computer Science.

[23]  Leonid Libkin,et al.  Elements Of Finite Model Theory (Texts in Theoretical Computer Science. An Eatcs Series) , 2004 .

[24]  Svatopluk Poljak,et al.  On the complexity of the subgraph problem , 1985 .

[25]  Chandra Chekuri,et al.  Polynomial bounds for the grid-minor theorem , 2013, J. ACM.

[26]  György Turán,et al.  On the definability of properties of finite graphs , 1984, Discret. Math..

[27]  Benjamin Rossman An improved homomorphism preservation theorem from lower bounds in circuit complexity , 2016, SIGL.

[28]  Martin Grohe,et al.  Is Polynomial Time Choiceless? , 2015, Fields of Logic and Computation II.

[29]  Svante Janson,et al.  Random graphs , 2000, Wiley-Interscience series in discrete mathematics and optimization.

[30]  François Le Gall,et al.  Powers of tensors and fast matrix multiplication , 2014, ISSAC.

[31]  J. Spencer The Strange Logic of Random Graphs , 2001 .

[32]  Benjamin Rossman,et al.  On the constant-depth complexity of k-clique , 2008, STOC.

[33]  Kazuyuki Amano k-Subgraph Isomorphism on AC0 Circuits , 2009, Computational Complexity Conference.

[34]  Noga Alon,et al.  Sparse Balanced Partitions and the Complexity of Subgraph Problems , 2011, SIAM J. Discret. Math..