Minimum Linear Arrangement

Minimum Linear Arrangement
Name	Minimum Linear Arrangement
Alt	MLA
Field	Graph theory
Problems	Combinatorial optimization
Complexity	NP-hard
Related	Seriation problem

Minimum Linear Arrangement

Minimum Linear Arrangement is a combinatorial optimization problem in graph theory that asks for an ordering of the vertices of a given graph to minimize the sum of edge lengths measured along a line embedding. It appears in contexts ranging from VLSI layout and compiler optimization to bioinformatics and social network analysis, and connects to classical problems studied by researchers associated with institutions such as Bell Labs, MIT, Stanford University, and IBM Research.

Definition and Problem Statement

Given an undirected graph G = (V, E) with |V| = n, the task is to find a bijection π: V → {1,...,n} that minimizes Σ_{(u,v)∈E} |π(u) − π(v)|. Instances are often specified with unit edge weights, though weighted variants assign nonnegative weights w(u,v) and minimize Σ w(u,v)|π(u)−π(v)|. The problem statement has been formalized in literature by authors affiliated with Princeton University, University of California, Berkeley, University of Toronto, and ETH Zurich and is related to ordering schemes studied in works circulated at conferences such as STOC, FOCS, and SODA.

Computational Complexity and Variants

The decision and optimization versions are NP-hard, with hardness proofs reducible from problems like Minimum Cut and 3-SAT via constructions used by researchers at Cornell University and Carnegie Mellon University. Variants include the weighted Minimum Linear Arrangement, the bandwidth problem (minimizing maximum |π(u)−π(v)|), the minimum logarithmic arrangement, and the bipartite arrangement specialized to bipartite graphs. Parameterized complexity analyses consider parameters such as treewidth and cutwidth; influential results originated from groups at Institut des Hautes Études Scientifiques and University of Warsaw. Approximation hardness results reference PCP-theorem developments credited to researchers at Rutgers University and University of California, San Diego.

Algorithms and Approximation Techniques

Exact algorithms use branch-and-bound, dynamic programming on tree decompositions, and integer linear programming formulations implemented by teams at Microsoft Research and Google Research. Approximation techniques include spectral methods based on the second eigenvector of the Laplacian inspired by work at Harvard University and Columbia University, greedy heuristics such as Cuthill–McKee originating in Daresbury Laboratory contexts, and multilevel coarsening frameworks developed in analog with graph partitioning toolchains from Lawrence Livermore National Laboratory and Sandia National Laboratories. Semidefinite programming relaxations and rounding schemes were advanced by researchers at Courant Institute and EPFL to derive approximation guarantees, while local improvement methods (pairwise swaps, sifting) were refined by practitioners at Siemens and Nokia.

Special Cases and Exact Solutions

Polynomial-time solvable cases include trees, caterpillars, and certain series-parallel graphs; classical exact solutions and characterizations were proved by mathematicians associated with University of Cambridge, University of Oxford, and Imperial College London. For trees, dynamic programming and centroid decompositions yield optimal arrangements; for outerplanar graphs and graphs of bounded pathwidth, exact algorithms exploit decompositions studied at Tokyo Institute of Technology and Seoul National University. Bipartite instances with constrained degree sequences admit specialized algorithms developed in joint work involving ETH Zurich and Max Planck Institute for Informatics.

Applications and Practical Use Cases

In VLSI physical design and chip routing problems tackled by engineers at Intel and TSMC, Minimum Linear Arrangement models channel routing and wirelength minimization. In compiler optimization and register allocation research at Bell Labs and IBM Research, linear arrangements inform code layout and memory locality. Bioinformatics applications include genome scaffolding and sequencing assembly pipelines used by teams at Broad Institute and Wellcome Sanger Institute. In computational sociology and information retrieval, ordering users or documents to reduce interaction cost has been explored by researchers at Yahoo Research and Facebook AI Research.

Related Optimization Problems

Related problems include the bandwidth problem, the minimum cutwidth problem, the seriation problem in archaeology and paleontology contexts, and the optimal linear arrangement’s connection to minimum linear extensions studied in combinatorics by scholars at University of Illinois Urbana-Champaign and Ohio State University. Connections to graph partitioning, ordering heuristics like Cuthill–McKee, and spectral ordering techniques link to work presented at IEEE symposia and workshops hosted by ACM SIGPLAN and ACM SIGGRAPH.

Category:Graph theory Category:Combinatorial optimization Category:NP-hard problems