Straight-line Grid Drawings of 3-Connected 1-Planar Graphs

Similar documents
Planar Upward Drawings

Paths. Connectivity. Euler and Hamilton Paths. Planar graphs.

Graphs. Graphs. Graphs: Basic Terminology. Directed Graphs. Dr Papalaskari 1

Graph Isomorphism. Graphs - II. Cayley s Formula. Planar Graphs. Outline. Is K 5 planar? The number of labeled trees on n nodes is n n-2

An undirected graph G = (V, E) V a set of vertices E a set of unordered edges (v,w) where v, w in V

Graphs. CSC 1300 Discrete Structures Villanova University. Villanova CSC Dr Papalaskari

12/3/12. Outline. Part 10. Graphs. Circuits. Euler paths/circuits. Euler s bridge problem (Bridges of Konigsberg Problem)

The University of Sydney MATH2969/2069. Graph Theory Tutorial 5 (Week 12) Solutions 2008

5/9/13. Part 10. Graphs. Outline. Circuits. Introduction Terminology Implementing Graphs

Cycles and Simple Cycles. Paths and Simple Paths. Trees. Problem: There is No Completely Standard Terminology!

1 Introduction to Modulo 7 Arithmetic

Algorithmic and NP-Completeness Aspects of a Total Lict Domination Number of a Graph

Math 61 : Discrete Structures Final Exam Instructor: Ciprian Manolescu. You have 180 minutes.

b. How many ternary words of length 23 with eight 0 s, nine 1 s and six 2 s?

COMPLEXITY OF COUNTING PLANAR TILINGS BY TWO BARS

(2) If we multiplied a row of B by λ, then the value is also multiplied by λ(here lambda could be 0). namely

Constructive Geometric Constraint Solving

Outline. 1 Introduction. 2 Min-Cost Spanning Trees. 4 Example

Exam 1 Solution. CS 542 Advanced Data Structures and Algorithms 2/14/2013

Why the Junction Tree Algorithm? The Junction Tree Algorithm. Clique Potential Representation. Overview. Chris Williams 1.

Outline. Circuits. Euler paths/circuits 4/25/12. Part 10. Graphs. Euler s bridge problem (Bridges of Konigsberg Problem)

V={A,B,C,D,E} E={ (A,D),(A,E),(B,D), (B,E),(C,D),(C,E)}

Module graph.py. 1 Introduction. 2 Graph basics. 3 Module graph.py. 3.1 Objects. CS 231 Naomi Nishimura

Garnir Polynomial and their Properties

Trees as operads. Lecture A formalism of trees

CSE 373. Graphs 1: Concepts, Depth/Breadth-First Search reading: Weiss Ch. 9. slides created by Marty Stepp

V={A,B,C,D,E} E={ (A,D),(A,E),(B,D), (B,E),(C,D),(C,E)}

0.1. Exercise 1: the distances between four points in a graph

MAT3707. Tutorial letter 201/1/2017 DISCRETE MATHEMATICS: COMBINATORICS. Semester 1. Department of Mathematical Sciences MAT3707/201/1/2017

On Local Transformations in Plane Geometric Graphs Embedded on Small Grids

Solutions to Homework 5

CSE 373: More on graphs; DFS and BFS. Michael Lee Wednesday, Feb 14, 2018

, each of which is a tree, and whose roots r 1. , respectively, are children of r. Data Structures & File Management

Uniform 2D-Monotone Minimum Spanning Graphs

Page 1. Question 19.1b Electric Charge II Question 19.2a Conductors I. ConcepTest Clicker Questions Chapter 19. Physics, 4 th Edition James S.

Solutions for HW11. Exercise 34. (a) Use the recurrence relation t(g) = t(g e) + t(g/e) to count the number of spanning trees of v 1

Outline. Computer Science 331. Computation of Min-Cost Spanning Trees. Costs of Spanning Trees in Weighted Graphs

Steinberg s Conjecture is false

CSC Design and Analysis of Algorithms. Example: Change-Making Problem

CSE 373: AVL trees. Warmup: Warmup. Interlude: Exploring the balance invariant. AVL Trees: Invariants. AVL tree invariants review

CS200: Graphs. Graphs. Directed Graphs. Graphs/Networks Around Us. What can this represent? Sometimes we want to represent directionality:

CS 241 Analysis of Algorithms

CS 461, Lecture 17. Today s Outline. Example Run

COMP108 Algorithmic Foundations

Research Article On the Genus of the Zero-Divisor Graph of Z n

Numbering Boundary Nodes

ECE COMBINATIONAL BUILDING BLOCKS - INVEST 13 DECODERS AND ENCODERS

# 1 ' 10 ' 100. Decimal point = 4 hundred. = 6 tens (or sixty) = 5 ones (or five) = 2 tenths. = 7 hundredths.

arxiv: v1 [math.co] 15 Dec 2015

Present state Next state Q + M N

CS61B Lecture #33. Administrivia: Autograder will run this evening. Today s Readings: Graph Structures: DSIJ, Chapter 12

A 4-state solution to the Firing Squad Synchronization Problem based on hybrid rule 60 and 102 cellular automata

On a Class of Planar Graphs with Straight-Line Grid Drawings on Linear Area

Designing A Concrete Arch Bridge

Section 10.4 Connectivity (up to paths and isomorphism, not including)

1. Determine whether or not the following binary relations are equivalence relations. Be sure to justify your answers.

DUET WITH DIAMONDS COLOR SHIFTING BRACELET By Leslie Rogalski

5/7/13. Part 10. Graphs. Theorem Theorem Graphs Describing Precedence. Outline. Theorem 10-1: The Handshaking Theorem

Generalized swap operation for tetrahedrizations

Outline. Binary Tree

Nefertiti. Echoes of. Regal components evoke visions of the past MULTIPLE STITCHES. designed by Helena Tang-Lim

Floating Point Number System -(1.3)

Floating Point Number System -(1.3)

Similarity Search. The Binary Branch Distance. Nikolaus Augsten.

QUESTIONS BEGIN HERE!

This chapter covers special properties of planar graphs.

Witness-Bar Visibility Graphs

QUESTIONS BEGIN HERE!

The University of Sydney MATH 2009

Weighted graphs -- reminder. Data Structures LECTURE 15. Shortest paths algorithms. Example: weighted graph. Two basic properties of shortest paths

Announcements. Not graphs. These are Graphs. Applications of Graphs. Graph Definitions. Graphs & Graph Algorithms. A6 released today: Risk

A 43k Kernel for Planar Dominating Set using Computer-Aided Reduction Rule Discovery

learning objectives learn what graphs are in mathematical terms learn how to represent graphs in computers learn about typical graph algorithms

EE1000 Project 4 Digital Volt Meter

Greedy Algorithms, Activity Selection, Minimum Spanning Trees Scribes: Logan Short (2015), Virginia Date: May 18, 2016

TOPIC 5: INTEGRATION

Jonathan Turner Exam 2-10/28/03

Using the Printable Sticker Function. Using the Edit Screen. Computer. Tablet. ScanNCutCanvas

(4, 2)-choosability of planar graphs with forbidden structures

Graph Theory. Vertices. Vertices are also known as nodes, points and (in social networks) as actors, agents or players.

Integration Continued. Integration by Parts Solving Definite Integrals: Area Under a Curve Improper Integrals

SOLVED EXAMPLES. be the foci of an ellipse with eccentricity e. For any point P on the ellipse, prove that. tan

arxiv: v1 [math.mg] 5 Oct 2015

CS September 2018

Straight-line Grid Drawings of 3-Connected 1-Planar Graphs

12. Traffic engineering

On the Characterization of Level Planar Trees by Minimal Patterns

Construction 11: Book I, Proposition 42

Announcements. These are Graphs. This is not a Graph. Graph Definitions. Applications of Graphs. Graphs & Graph Algorithms

S i m p l i f y i n g A l g e b r a SIMPLIFYING ALGEBRA.

Discovering Pairwise Compatibility Graphs

An Integer Linear Programming Based Routing Algorithm for Flip-Chip Design

Properties of Hexagonal Tile local and XYZ-local Series

IEEE TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS, VOL. TK, NO. TK, MONTHTK YEARTK 1. Hamiltonian Walks of Phylogenetic Treespaces

FSA. CmSc 365 Theory of Computation. Finite State Automata and Regular Expressions (Chapter 2, Section 2.3) ALPHABET operations: U, concatenation, *

O n t h e e x t e n s i o n o f a p a r t i a l m e t r i c t o a t r e e m e t r i c

Seven-Segment Display Driver

NP-Completeness. CS3230 (Algorithm) Traveling Salesperson Problem. What s the Big Deal? Given a Problem. What s the Big Deal? What s the Big Deal?

Register Allocation. Register Allocation. Principle Phases. Principle Phases. Example: Build. Spills 11/14/2012

Graph Contraction and Connectivity

Transcription:

Stright-lin Gri Drwings o 3-Connt 1-Plnr Grhs M. Jwhrul Alm 1, Frnz J. Brnnurg 2, n Sthn G. Koourov 1 1 Drtmnt o Comutr Sin, Univrsity o Arizon, USA {mjlm, koourov}@s.rizon.u 2 Univrsity o Pssu, 94030 Pssu, Grmny rnn@inormtik.uni-ssu. Astrt. A grh is 1-lnr i it n rwn in th ln suh tht h g is ross t most on. In gnrl, 1-lnr grhs o not mit strightlin rwings. W show tht vry 3-onnt 1-lnr grh hs stright-lin rwing on n intgr gri o urti siz, with th xtion o singl g on th outr tht hs on n. Th rwing n omut in linr tim rom ny givn 1-lnr ming o th grh. 1 Introution Sin Eulr s Königsrg rig rolm ting k to 1736, lnr grhs hv rovi intrsting rolms in thory n in rti. Using th lort thnius o nonil orring n Shnyr rlizrs, vry lnr grh n rwn on gri o urti siz, n suh rwings n omut in linr tim [15, 21]. Th r oun is symtotilly otiml, sin th nst tringl grhs r lnr grhs n ruir Ω(n 2 ) r [10]. Th rwing lgorithms wr rin to imrov th r ruirmnt or to mit onvx rrsnttions, i.., whr h innr is onvx [5, 8] or stritly onvx [1]. Howvr, most grhs r nonlnr n rntly, thr hv n mny ttmts to stuy lrgr lsss o grhs. O rtiulr intrst r 1-lnr grhs, whih in sns r on st yon lnr grhs. Thy wr introu y Ringl [20] in n ttmt to olor lnr grh n its ul. Although it is known tht 3-onnt lnr grh n its ul hv stright-lin 1-lnr rwing [24] n vn on urti gri [13], littl is known out gnrl 1-lnr grhs. It is NP-hr to rogniz 1- lnr grhs [16, 18] in gnrl, lthough thr is linr-tim tsting lgorithm [11] or mximl 1-lnr grhs (i.., whr no itionl g n without violting 1-lnrity) givn th th irulr orring o inint gs roun h vrtx. A 1-lnr grh with n vrtis hs t most 4n 8 gs [4, 14, 19] n this ur oun is tight. On th othr hn stright-lin rwings o 1-lnr grhs my hv t most 4n 9 gs n this oun is tight [9]. Hn not ll 1-lnr grhs mit stright-lin rwings. Unlik lnr grhs, mximl 1-lnr grhs n muh srsr with only 2.64n gs [6]. 1 Rsrh suort in rt y NSF grnts CCF-1115971 n DEB 1053573. 2 Rsrh suort y th Dutsh Forshungsgminsht, DFG, grnt Br 835/18-1.

i 11 10 j 9 g 8 7 6 h h 5 4 3 j g g 2 1 i h 0 1 0 1 2 3 4 5 6 7 8 9 10 () () () 8 7 6 5 4 3 2 1 0 h g 0 1 2 3 4 5 6 7 8 () Fig. 1. () () A 3-onnt 1-lnr grh n its stright-lin gri rwing (with on n in on g), () () nothr 3-onnt 1-lnr grh n its stright-lin gri rwing. Thomssn [23] rrs to 1-lnr grhs s grhs with ross inx 1 n rov tht n m 1-lnr grh n turn into stright-lin rwing i n only i it xlus B- n W -onigurtions; s Fig. 2. Ths orin onigurtions wr irst isovr y Egglton [12] n us y Hong t l. [17], who show tht th onigurtions n tt in linr tim i th ming is givn. Thy lso rov tht thr is linr tim lgorithm to onvrt 1-lnr ming without B- n W - onigurtions into stright-lin rwing, ut without ouns or th rwing r. In this r w sttl th stright-lin gri rwing rolm or 3-onnt 1- lnr grhs. First w omut norml orm or n m 1-lnr grh with no B-onigurtion n t most on W -onigurtion on th outr. Thn, tr ugmnting th grh with s mny lnr gs s ossil n thn lting th rossing gs, w in 3-onnt lnr grh, whih is rwn with stritly onvx s using n xtnsion o th lgorithm o Chrok n Knt [8]. Finlly th irs o rossing gs r rinsrt into th onvx s. This givs stright-lin rwing on gri o urti siz with th xtion o singl g on th outr, whih my n on n (n this xtion is unvoil); s Fig. 1. In ition, th rwing is otin in linr tim rom givn 1-lnr ming. 2 Prliminris A rwing o grh G is ming o G into th ln suh tht th vrtis r m to istint oints n h g is Jorn r twn its noints. A rwing is lnr i th Jorn rs o th gs o not ross n it is 1-lnr i h g is ross t most on. Not tht rossings twn gs inint to th sm vrtx r not llow. For xml, K 5 n K 6 r 1-lnr grhs. An ming o grh is lnr (rs. 1-lnr) i it mits lnr (rs. 1-lnr) rwing. An ming siis th s, whih r toologilly onnt rgions. Th unoun is th outr. A in lnr grh is sii y yli sun o gs on its ounry (or uivlntly y th yli sun o th noints o th gs). Aoringly, 1-lnr ming E(G) siis th s in 1-lnr rwing o G inluing th outr. A 1-lnr ming is witnss or 1-lnrity. In rtiulr, E(G) sris th irs o rossing gs n th s whr th gs

() () () Fig. 2. () An ugmnt X-onigurtion, () n ugmnt B-onigurtion, () n ugmnt W -onigurtion. Th grhs inu y th soli gs r ll n X-onigurtion (), B-onigurtion (), n W -onigurtion (). ross n hs linr siz. Eh ir o rossing gs (, ) n (, ) inus rossing oint. Cll th sgmnt o n g twn th vrtx n th rossing oint hlg. Eh hl-g is imrml, nlogous to th gs in lnr rwings, in th sns tht no g n ross suh hl-g without violting th 1-lnrity o th ming. Th non-ross gs r ll lnr. A lnriztion G is otin rom E(G) y using th rossing oints s rgulr vrtis n rling h rossing g y its two hl-gs. A 1-lnr ming E(G) n its lnriztion shr uivlnt mings, n h is givn y list o gs n hl-gs ining it, or uivlntly, y list o vrtis n rossing oints o th gs n hl gs. Egglton [12] ris th rolm o rognizing 1-lnr grhs with rtilinr rwings. H solv this rolm or outr-1-lnr grhs (1-lnr grhs with ll vrtis on th outr-yl) n roos thr orin onigurtions. Thomssn [23] solv Egglton s rolm n hrtriz th rtilinr 1-lnr mings y th xlusion o B- n W-onigurtions; s Fig. 2. Hong t l. [17], otin similr hrtriztion whr th B- n W -onigurtions r ll th Bulgri n Gui grhs. Thy lso show tht ll ourrns o ths onigurtions n omut in linr tim rom givn 1-lnr ming. Dinition 1. Consir 1-lnr ming E(G): A B-onigurtion onsists o n g (, ) n two gs (, ) n (, ) whih ross in som oint suh tht n li in th intrior o th tringl (,, ). Hr (, ) is ll th s o th onigurtion. An X-onigurtion onsists o ir (, ) n (, ) o rossing gs whih os not orm B-onigurtion. A W-onigurtion onsists o two irs o gs (, ), (, ) n (, ), (, ) whih ross in oints n, suh tht,,, li in th intrior o th urngl,,,. Hr gin th g (, ), i rsnt is th s. Osrv tht or ll ths onigurtions th s gs my ross y nothr g, whrs th rossing gs r imrml; s Fig 2. Thomssn [23] n Hong t l. [17] rov tht or 1-lnr ming to mit stright-lin rwing, B- n W -onigurtions must xlu:

Proosition 1. A 1-lnr ming E(G) mits stright-lin rwing with toologilly uivlnt ming i n only i it os not ontin B- or W -onigurtion. Augmnt givn 1-lnr ming E(G) y ing s mny gs to E(G) s ossil so tht G rmins siml grh n th nwly gs r lnr in E(G). W ll suh n ming lnr-mximl ming o G n th ortion lnr-mximl ugmnttion. (Not tht Hong t l. [17] olor th lnr gs o 1-lnr ming r n ll lnr-mximl ugmnttion r ugmnttion.) Th lnr sklton P(E(G)) onsists o th lnr gs o lnr-mximl ugmnttion. It is lnr m grh, sin ll irs o rossing gs r omitt. Not tht th lnr ugmnttion n th lnr sklton r in or n ming, not or grh. A grh my hv irnt mings whih giv ris to irnt onigurtions n ugmnttions. Th notion o lnr-mximl ming is irnt rom th notions o mximl 1-lnr mings n mximl 1-lnr grhs, whih r suh tht th ition o ny g violts 1-lnrity (or simliity) [6]. Th ollowing lim, rovn in mny rlir rs [6, 14, 17, 22, 23], shows tht rossing ir o gs inus K 4 in lnr-mximl ming, sin missing gs o K 4 n without inuing nw rossings. Lmm 1. Lt E(G) lnr-mximl 1-lnr ming o grh G n lt (, ) n (, ) two rossing gs. Thn th our vrtis {,,, } inu K 4. By Lmm 1, or lnr-mximl ming h X-, B-, n W -onigurtion is ugmnt y itionl gs. Hr w in ths ugmnt onigurtions. Dinition 2. Lt E(G) lnr-mximl 1-lnr ming o grh G. An ugmnt X-onigurtion onsists o K 4 with vrtis (,,, ) suh tht th gs (, ) n (, ) ross insi th urngl. An ugmnt B-onigurtion onsists o K 4 with vrtis (,,, ) suh tht th gs (, ) n (, ) ross yon th ounry o th urngl. An ugmnt W-onigurtion onsists o two K 4 s (,,, ) n (,,, ) on o whih is in n ugmnt X-onigurtion n th othr in n ugmnt B-onigurtion. For n ugmnt X- or ugmnt B-onigurtion, th gs not inuing rossing with othr gs in th onigurtion in yl, w ll it th sklton. In h onigurtion, th gs on th outr-ounry o th m onigurtion n not inuing rossing with othr gs in th onigurtion r th s gs. Using th rsults o Thomssn [23] n Hong t l. [17], w n now hrtriz whn lnr-mximl 1-lnr ming o grh mits stright-lin rwing: Lmm 2. Lt E(G) lnr-mximl 1-lnr ming o grh G. Thn thr is stright-lin 1-lnr rwing o G with toologilly uivlnt ming s E(G) i n only i E(G) os not ontin n ugmnt B-onigurtion. Proo. Assum E(G) ontins n ugmnt B-onigurtion. Thn it ontins B- onigurtion n hs no stright-lin 1-lnr rwing y Proosition 1. Convrsly, i E(G) hs no stright-lin 1-lnr rwing thn y Proosition 1 it ontins t lst on B- or W -onigurtion. Sin Γ is lnr-mximl ming, y Lmm 1 h rossing g ir in E(G) inus K 4. Thus th ott gs in Fig. 2() () must rsnt in ny B- or W - onigurtion, inuing n ugmnt B-onigurtion.

Th norml orm or n m 1-lnr grh E(G) is otin y irst ing th our lnr gs to orm K 4 or h ir o rossing gs whil routing thm losly to th rossing gs n thn rmoving ol ulit gs i nssry. Suh n ming o 1-lnr grh is norml ming o it. A norml lnrmximl ugmnttion or n m 1-lnr grh is otin y irst ining norml orm o th ming n thn y lnr-mximl ugmnttion. Lmm 3. Givn 1-lnr ming E(G), th norml lnr-mximl ugmnttion o E(G) n omut in linr tim. Proo. First ugmnt h rossing o two gs (, ) n (, ) to K 4, suh tht th gs (, ), (, ), (, ), (, ) r n in s o ulit th ormr g is rmov. Thn ll ugmnt X-onigurtions r mty n ontin no vrtis insi thir skltons. Nxt tringult ll s whih o not ontin hl-g, rossing g, or rossing oint. Eh st n on in linr tim. 3 Chrtriztion o 3-Connt 1-Plnr Grhs Hr w hrtriz 3-onnt 1-lnr grhs y norml ming, whr th rossings r ugmnt to K 4 s suh tht th rsulting ugmnt X-onigurtions hv vrtx-mty skltons n thr is no ugmnt B-onigurtion xt or t most on ugmnt W-onigurtion with ir o rossing gs in th outr. Lt E(G) 1-lnr ming o grh G. Eh ir o rossing gs inus rossing oint n th rossing gs n thir hl-gs r imrml s thy nnot ross y othr gs without violting 1-lnrity. An imrml th in E(G) is n intrnlly-isjoint sun P = v 1, 1, v 2, 2,..., v n, n, v n+1, whr v 1, v 2,..., v n+1 r (rgulr) vrtis o G, 1, 2,..., n r rossing oints in E(G) n (v i, i ), ( i, v i+1 ) or h i {1, 2,..., n} r hl gs. I v n+1 = v 0, thn P is n imrml yl. An imrml yl is srting whn it hs vrtis oth insi n outsi o it, sin lting its vrtis isonnts G. Lmm 4. Lt G = (V, E) 3-onnt 1-lnr grh with lnr-mximl 1-lnr ming E(G). Thn th ollowing onitions hol. A. (i) Two ugmnt B-onigurtions or two ugmnt X-onigurtions nnot on th sm si o ommon s g. (ii) Suos n ugmnt B-onigurtion B n n ugmnt X-onigurtion X r on th sm si o ommon s g (, ). Lt n th rossing oints or X n B, rstivly n lt R(X) n R(B) th rgions insi th skltons o X n B. Thn ll vrtis o V \ {, } r insi th imrml yl i R(X) R(B); othrwis ll vrtis o V \ {, } r outsi th imrml yl. B. (i) I two ugmnt B-onigurtions r on oosit sis o ommon s g (, ), with rossing oints n, rstivly, thn ll th vrtis o V \ {, } r insi th imrml yl.

() () () () () () (g) Fig. 3. Illustrtion or th roo o Lmm 4. (ii) I two ugmnt X-onigurtions r on oosit sis o ommon s g (, ), with rossing oints n, rstivly, thn ll th vrtis o V \ {, } r outsi th imrml yl. (iii) An ugmnt B-onigurtion n n ugmnt X-onigurtion nnot shr ommon s g rom oosit sis. Proo. Conition A.(i) n B.(iii) hol us h o ths onigurtions inus srting imrml yl in E(G) with only two (rgulr) vrtis rom G, ontrition with th 3-onntivity o G; s Fig. 3() () n (). Similrly, i ny o th Conitions A.(ii) n B.(i) (ii) is not stisi, thn th imrml yl oms srting n hn th ir {, } oms srtion ir o G, gin ontrition with th 3-onntivity o G; s Fig. 3() (), () n (g). Corollry 1. Lt G 3-onnt 1-lnr grh with lnr-mximl 1-lnr ming E(G). Thn no thr rossing g-irs in E(G) shr th sm s g. Proo. Eh rossing g ir inus ithr n ugmnt B- or n ugmnt X- onigurtion. This t long with Lmm 4[A.(i), B.(iii)] yils th orollry. Lmm 5. Lt G 3-onnt 1-lnr grh. Thn thr is lnr-mximl 1- lnr ming E(G ) o surgrh G o G so tht E(G ) ontins t most on ugmnt W-onigurtion in th outr n no othr ugmnt B-onigurtion, n h ugmnt X-onigurtion in E(G ) ontins no vrtx insi its sklton. Proo. Lt E(G) 1-lnr ming o G. W lim tht y norml lnrmximl ugmnttion o E(G) w gt th sir ming o surgrh o G. Not tht u to th g-rrouting this ortion onvrts ny B-onigurtion whos s is not shr with nothr onigurtion into n X-onigurtion; s Fig. 4(). I s g is shr y two B-onigurtions, thy r onvrt into on W - onigurtion n y Lmm 4 this W -onigurtion is on th outr ; s Fig. 4().

() () () Fig. 4. Illustrtion or th roo o Lmm 5. By Corollry 1, s g nnot shr y mor thn two B-onigurtions. Furthrmor this ortion os not rt ny nw B-onigurtion. It lso mks th sklton o ny ugmnt X-onigurtion vrtx-mty; y Lmm 4 s g n shr y t most two ugmnt X-onigurtions rom oosit sis n i it is shr y two ugmnt X-onigurtions, th intrior o th inu imrml yl is mty; s Fig. 4(). Lmm 5 togthr with Proosition 1 imlis th ollowing: Thorm 1. A 3-onnt 1-lnr grh mits stright-lin 1-lnr rwing xt or t most on g in th outr. 4 Gri Drwings In th rvious stion w show tht 3-onnt 1-lnr grh hs strightlin 1-lnr rwing, with th xtion o singl g in th outr. W now strngthn this rsult n show tht thr is stright-lin gri rwing with O(n 2 ) r, whih n onstrut in linr tim rom givn 1-lnr ming. Th lgorithm tks n ming E(G) n omuts norml lnr-mximl ugmnttion. Consir th lnr sklton P(E(G)) or th ming. I thr is n ugmnt W-onigurtion n rossing in th outr, on rossing g on th outr is kt n th othr rossing g is trt srtly. Thus th outr o P(E((G)) is tringl n th innr s r tringls or urngls. Eh urngl oms rom n ugmnt X-onigurtion. It must rwn stritly onvx, suh tht th rossing gs n r-insrt. This is hiv y n xtnsion o th onvx gri rwing lgorithm o Chrok n Knt [8], whih itsl is n xtnsion o th

shiting mtho o Fryssix, Ph n Pollk [15]. Sin th s r t most urngls, w n voi thr ollinr vrtis n th gnrtion to tringl y n xtr unit shit. Not tht our rwing lgorithm hivs n r o (2n 2) (2n 3), whil th gnrl lgorithms or stritly onvx gri rwings [1, 7] ruir lrgr r, sin stritly onvx rwings o n-gons n Ω(n 3 ) r [2]. Brny n Rot giv stritly onvx gri rwing o lnr grh on 14n 14n gri i th s r t most 4-gons, n on 2n 2n gri i, in ition, th outr is tringl. Howvr, thir roh is uit omlx n os not immitly yil ths ouns. It is lso not lr how to us this roh or lnr grhs in our 1-lnr grh stting, in rtiulr whn w hv n unvoil W-onigurtion in th outr. Th lgorithm o Chrok n Knt n in rtiulr th omuttion o nonil omosition rsums 3-onnt lnr grh. Thus th lnr sklton o 3- onnt 1-lnr grh must 3-onnt, whih hols xt or th K 4, whn it is m s n ugmnt X-onigurtion. This rsults rllls th t tht th lnriztion o 3-onnt 1-lnr grh is 3-onnt [14]. Lmm 6. Lt G grh with lnr-mximl 1-lnr ming E(G) suh tht it hs no ugmnt B-onigurtion n h ugmnt X-onigurtion in E(G) hs no vrtx insi its sklton. Thn th lnr sklton P(E(G)) is 3-onnt. W will rov Lmm 6 y showing tht thr is no srtion ir in P(E(G)). First w otin lnr grh H rom G s ollows. Lt (, ) n (, ) ir o rossing gs tht orm n ugmnt X-onigurtion X in Γ. W thn lt th two gs (, ), (, ); vrtx u n th gs (, u), (, u), (, u), (, u) to tringult th. Cll v ross-vrtx n ll this ortion ross-vrtx insrtion on X. W thn otin H rom G y ross-vrtx insrtion on h ugmnt X-onigurtion. Cll H lnriztion o G n not th st o ll th ross-vrtis y U. Thn P(E(G)) = H \U. Bor roving Lmm 6 w onsir svrl rortis o H, th lnriztion o th 1-lnr grh. Lmm 7. Lt G = (V, E) grh with lnr-mximl 1-lnr ming E(G) suh tht E(G) ontins no ugmnt B-onigurtion n h ugmnt X- onigurtion in E(G) ontins no vrtx insi its sklton. Lt H lnriztion o G, whr U is th st o ross-vrtis. Thn th ollowing onitions hol. () H is mximl lnr grh (xt i H is th K 4 in n X-onigurtion) () Eh vrtx o U hs gr 4. () U is n innnt st o H. () Thr is no srting tringl o H ontining ny vrtx rom U. () Thr is no srting 4-yl o H ontining two vrtis rom U. Proo. For onvnin, w ll h vrtx in V U rgulr vrtx. () Sin H is lnr grh, w only show tht h o H is tringl. Eh rossing g ir in Γ inus n ugmnt X-onigurtion whos sklton hs no vrtx in its intrior. Hn h o H ontining rossing vrtx is tringl. Agin, Hong t l. [17] show tht in lnr-mximl 1-lnr ming with no rossing vrtis is tringl. Thus H is mximl lnr grh.

() () Fig. 5. Illustrtion or th roo o Lmm 6. () () Ths two onitions ollow rom th t tht th nighorhoo o h rossing vrtx onsists o xtly our rgulr vrtis tht orm th sklton o th orrsoning ugmnt X-onigurtion. () For ontrition suos vrtx u U rtiits in srting tringl T o H. Sin th nighorhoo o u orms th sklton o th orrsoning ugmnt X-onigurtion X, th othr two vrtis, sy n, in T r rgulr vrtis. Th g (, ) nnot orm s g or X, sin i it i, thn th intrior o th srting tringl T woul ontin in th intrior o th sklton or X n hn woul mty. Assum thror tht n r not onsutiv on th sklton o X. In this s th g (, ) is rossing g in G n hn hs n lt whn onstruting H; s Fig. 5(). () Suos two vrtis u, v U rtiit in srting 4-yl T o H. Du to Conition (), ssum tht T = uv, whr, r rgulr vrtis. I th two vrtis, r jnt in H, ssum without loss o gnrlity tht th g (, ) is rwn insi th intrior o T. Thn th intrior o t lst on o th two tringls u n v is non-mty, inuing srting tringl in H, ontrition with Conition (). W thus ssum tht th two vrtis n r not jnt in H. Thn or oth th ugmnt X-onigurtions X n Y, orrsoning to th two rossing vrtis u n v, th two vrtis u n v r not onsutiv on thir sklton. This imlis tht th rossing g (, ) rtiits in two irnt ugmnt X-onigurtions in Γ, gin ontrition; s Fig. 5(). W r now ry to rov Lmm 6. Proo (Lmm 6). Assum or ontrition tht P(E(G)) is not 3-onnt. Thn thr xists som srtion ir {, } in P(E(G)). Lt H th lnriztion o G, whr U is th st o ross-vrtis. Thn S = U {, } is srting st or H. Tk miniml srting st S S suh tht no ror sust o S is srting st. Sin H is mximl lnr grh (rom Lmm 7()), S orms srting yl [3]. Th 3-onntivity o th mximl lnr grh H imlis S 3. Agin sin S ontins t most two rgulr vrtis, n no two ross-vrtis n jnt in H (Lmm 7()), S < 5. Hn S is srting tringl or srting 4-yl with t most two rgulr vrtis; w gt ontrition with Lmm 7() ().

Finlly, w sri our lgorithm or stright-lin gri rwings. This rwing lgorithm is s on n xtnsion o th lgorithm o Chrok n Knt [8] or omuting onvx rwing o lnr 3-onnt grh. For onvnin w rr to this lgorithm s th CK-lgorithm n w gin with ri ovrviw. Lt G = (V, E) n m 3-onnt grh n lt (u, v) n g on th outr-yl o G. Th CK-lgorithm strts y omuting nonil omosition o G, whih is n orr rtition V 1, V 2,..., V t o V suh tht th ollowing onitions hol: (i) For h k {1, 2,..., t}, th grh G k inu y th vrtis V 1... V k is 2-onnt n its outr-yl C k ontins th g (u, v). (ii) G 1 is yl, V t is singlton {z}, whr z / {u, v} is on th outr-yl o G. (iii) For h k {2,..., t 1} th ollowing onitions hol: I V k is singlton {z}, thn z is on th outr o G k 1 n hs t lst on nighor in G G k. I V k is hin {z 1,..., z l }, h z i hs t lst on nighor in G G k, z 1, z l hv on nighor h on C k 1 n no othr z i hs nighors on G k 1. For h k {1, 2,..., t}, w sy tht th vrtis tht long to V k hv rnk k. W ll vrtx o G k sturt i it hs no nighor in G G k. Th CK-lgorithm strts y rwing th g (u, v) with horizontl lin-sgmnt o unit lngth. Thn or k = 1, 2,..., t, it inrmntlly omlts th rwing o G k. Lt C k 1 = {(u = w 1,..., w,..., w,..., w r = v)} with 1 < r whr w n w r th ltmost n th rightmost nighor o vrtis in V k. Thn th vrtis o V k r l ov th vrtis w,..., w. Assum tht V k = {z 1,..., z l }. Thn z 1 is l on th vrtil lin ontining w i w is sturt in G k ; othrwis it is l on th vrtil lin on unit to th right o w. On th othr hn, z l is l on th ngtiv igonl lin (i.., with 45 slo) ontining w. I v k is singlton thn z = z 1 = z l is l t th intrstion o ths two lins. Othrwis (tr nssry shiting o w n othr vrtis), th vrtis z 1,... z l r l on onsutiv vrtil lins on unit rt rom h othr. In orr to mk sur tht this shiting ortion os not istur lnrity or onvxity, h vrtx v is ssoit with n unr-st U(v) n whnvr v is shit, ll vrtis in U(v) r lso shit long with v. Thus th gs twn vrtis o ny U(v) r in sns rigi. Thorm 2. Givn 1-lnr ming E(G) o 3-onnt grh G, strightlin rwing on th (2n 2) (2n 3) gri n omut in linr tim. Only on g on th outr my ruir on n. Proo. Assum tht E(G) is norml lnr-mximl ming; othrwis w omut on y norml lnr-mximl ugmnttion in linr tim y Lmm 3. Consir th lnr sklton P(E(G)). I thr is no unvoil W-onigurtion on th outr o th mximl lnr ugmnttion, thn th outr-yl o P(E(G)) is tringl. Othrwis w on o th rossing gs in th outr to P(E(G)) to mk th outr-yl tringl. Th othr rossing g is trt srtly. By Lmm 6, P(E(G)) is 3-onnt, its outr is tringl (,, ) n th innr s r tringls or urngls, whr th lttr rsult rom ugmnt X-onigurtions n r in on-to-on orrsonn to irs o rossing gs.

W wish to otin lnr stright-lin gri rwing o P(E(G)) suh tht ll urngls r stritly onvx. Although th CK-lgorithm rws ny 3-onnt lnr grh o n vrtis on gri o siz (n 1) (n 1) with onvx s, th s r not nssrily stritly onvx [8]. Hn w must moiy th lgorithm so tht ll urngls r stritly onvx. Not tht y th ssignmnt o th unr-sts, th CK-lgorithm gurnts tht on is rwn stritly onvx, it woul rmin stritly onvx tr ny susunt shiting o vrtis. For P(E(G)) h V k is ithr singl vrtx or ir with n g, sin th s r t most urngls. I V k is n g (z 1, z 2 ) thn, y th inition o th nonil omosition, xtly on urngl w z 1 z 2 w is orm n y onstrution this is rwn onvx. W thus ssum tht V k ontins singl vrtx, sy v. Lt C k 1 = {(u = w 1,..., w,..., w,..., w r = v)} with 1 < r whr w n w r th ltmost n th rightmost nighors o vrtis in V k. Thn th nw s rt y th insrtion o v r ll rwn stritly onvx unlss thr is som urngl vw 1w w +1 whr < < n w 1, w, w +1 r ollinr in th rwing o G k 1. In this s w must sturt in G k 1 n this ours in th CK-lgorithm only whn th lin ontining w 1, w, w +1 is vrtil lin or ngtiv igonl (with 45 slo). In th ormr s, w 1 shoul hv lso n sturt in G k 1, whih is not ossil sin v is its nighor. It is thus suiint to nsur tht no sturt vrtx o G k is in th ngtiv igonl o oth its lt n right nighors on C k. W o this y th ollowing xtnsion o th CK-lgorithm. Suos v is l ov w with slo 45, w ws l ov its rightmost lowr nighor w with slo 45, n thr is urngl (v, w, w, u) or som vrtx u with highr rnk to l ltr. Thn shit w y on xtr unit to th right whn v or u is l. This imlis n t w n stritly onvx ngl ov w. Th CK-lgorithm strts y ling th irst two vrtis on unit wy n it ruirs unit shit to th right or h ollowing vrtx. On th othr hn, 1-lnr grh hs t most n 2 irs o rossing gs. Hn, thr r g n 3 ugmnt X-onigurtions, h o whih inus urngl in th lnr sklton. Thus th with n hight r n 1 + g, whih is oun y 2n 4. Th vrtis,, o th outr tringl r l t th gri oints (0, 0), (0, n 1 + g), (n 1 + g, 0). I th grh h n unvoil W -onigurtion in th outr, w n ostrossing hs to rw th xtr g (, ), whih inus rossing with th g (, ). Sin is th ltmost lowr nighor o whn is l n is not sturt, is l t (1, j) or som j < n 2 + g. Shit on unit to th right, insrt n t ( 1, n + g), on igonl unit lt ov n rout (, ) vi th n oint. 5 Conlusion n Futur Work W show tht 3-onnt 1-lnr grhs n m on O(n) O(n) intgr gri, so tht gs r rwn s stright-lin sgmnts (xt or t most on g on th outr tht ruirs n). Morovr, th lgorithm is siml n runs in linr tim givn 1-lnr ming. Not tht vn th my ruir xonntil r or givn ix 1-lnr ming,.g., [17]. Rognition o 1-lnr grhs is NP-hr [18]. How hr is th rognition o lnr-mximl 1-lnr grhs?

Rrns 1. Bárány, I., Rot, G.: Stritly onvx rwings o lnr grhs. Doumnt Mthmti 11, 369 391 (2006) 2. Bárány, I., Tokushig, N.: Th minimum r o onvx ltti n-gons. Comintori 24(2), 171 185 (2004) 3. Byrs, I.: On k-th hmiltonin mximl lnr grhs. Disrt Mthmtis 40(1), 119 121 (1982) 4. Bonik, R., Shumhr, H., Wgnr, K.: Bmrkungn zu inm Shsrnrolm von G. Ringl. Ah. us m Mth. Sminr r Univ. Hmurg 53, 41 52 (1983) 5. Bonihon, N., Flsnr, S., Mosh, M.: Convx rwings o 3-onnt ln grhs. Algorithmi 47(4), 399 420 (2007) 6. Brnnurg, F.J., Estin, D., Glißnr, A., Goorih, M.T., Hnur, K., Rislhur, J.: On th nsity o mximl 1-lnr grhs. In: Diimo, W., Ptrignni, M. (s.) Grh Drwing. LNCS, vol. 7704,. 327 338. Sringr (2013) 7. Chrok, M., Goorih, M.T., Tmssi, R.: Convx rwings o grhs in two n thr imnsions. In: Symosium on Comuttionl Gomtry.. 319 328 (1996) 8. Chrok, M., Knt, G.: Convx gri rwings o 3-onnt lnr grhs. Intrntionl Journl o Comuttionl Gomtry n Alitions 7(3), 211 223 (1997) 9. Diimo, W.: Dnsity o stright-lin 1-lnr grh rwings. Inormtion Prossing Lttr 113(7), 236 240 (2013) 10. Dolv, D., Lighton, T., Triky, H.: Plnr ming o lnr grhs. In: Avns in Comuting Rsrh.. 147 161 (1984) 11. Es, P., Hong, S.H., Ktoh, N., Liott, G., Shwitzr, P., Suzuki, Y.: Tsting mximl 1-lnrity o grhs with rottion systm in linr tim. In: Diimo, W., Ptrignni, M. (s.) Grh Drwing. LNCS, vol. 7704,. 339 345. Sringr (2013) 12. Egglton, R.B.: Rtilinr rwings o grhs. Utilits Mth. 29, 149 172 (1986) 13. Ertn, C., Koourov, S.G.: Simultnous ming o lnr grh n its ul on th gri. Thory o Comuting Systms 38(3), 313 327 (2005) 14. Frii, I., Mrs, T.: Th strutur o 1-lnr grhs. Disrt Mthmtis 307(7 8), 854 865 (2007) 15. Fryssix, H., Ph, J., Pollk, R.: How to rw lnr grh on gri. Comintori 10(1), 41 51 (1990) 16. Grigoriv, A., Bolnr, H.L.: Algorithms or grhs ml with w rossings r g. Algorithmi 49(1), 1 11 (2007) 17. Hong, S.H., Es, P., Liott, G., Poon, S.H.: Fáry s thorm or 1-lnr grhs. In: Gumunsson, J., Mstr, J., Vigls, T. (s.) Intrntionl Conrn on Comuting n Comintoris. LNCS, vol. 7434,. 335 346. Sringr (2012) 18. Korzhik, V.P., Mohr, B.: Miniml ostrutions or 1-immrsions n hrnss o 1-lnrity tsting. Journl o Grh Thory 72(1), 30 71 (2013) 19. Ph, J., Tóth, G.: Grhs rwn with w rossings r g. Comintori 17, 427 439 (1997) 20. Ringl, G.: Ein Shsrnrolm u r Kugl. Ah. us m Mth. Sminr r Univ. Hmurg 29, 107 117 (1965) 21. Shnyr, W.: Eming lnr grhs on th gri. In: Johnson, D.S. (.) Symosium on Disrt Algorithms.. 138 148 (1990) 22. Suzuki, Y.: R-mings o mximum 1-lnr grhs. SIAM Journl o Disrt Mthmtis 24(4), 1527 1540 (2010) 23. Thomssn, C.: Rtilinr rwings o grhs. Journl o Grh Thory 12(3), 335 341 (1988) 24. Tutt, W.T.: How to rw grh. Pro. Lonon Mth. Soity 13(52), 743 768 (1963)