
The star chromatic index of a graph G, denoted by χ′st(G), is the smallest number of colors required to properly color E(G) such that every connected bicolored subgraph is a path with no more than three edges. A graph is K2,t-free if it contains no K2,t as a subgraph. This paper proves that every K2,t-free planar graph G satisfies χ′st(G)≤1.5Δ+20t+20, which is sharp up to the constant term. In particular, our result provides a common generalization of previous results on star edge coloring of outerplanar graphs by Bezegová et al.(2016) and of C4-free planar graphs by Wang et al.(2018), as those graphs are subclasses of K2,3-free planar graphs.
Citation: Yunfeng Tang, Huixin Yin, Miaomiao Han. Star edge coloring of K2,t-free planar graphs[J]. AIMS Mathematics, 2023, 8(6): 13154-13161. doi: 10.3934/math.2023664
[1] | Yuehua Bu, Qi Jia, Hongguo Zhu, Junlei Zhu . Acyclic edge coloring of planar graphs. AIMS Mathematics, 2022, 7(6): 10828-10841. doi: 10.3934/math.2022605 |
[2] | Xiaoxue Hu, Jiangxu Kong . An improved upper bound for the dynamic list coloring of 1-planar graphs. AIMS Mathematics, 2022, 7(5): 7337-7348. doi: 10.3934/math.2022409 |
[3] | Yanyi Li, Lily Chen . Injective edge coloring of generalized Petersen graphs. AIMS Mathematics, 2021, 6(8): 7929-7943. doi: 10.3934/math.2021460 |
[4] | Fugang Chao, Donghan Zhang . Neighbor sum distinguishing total choice number of IC-planar graphs with restrictive conditions. AIMS Mathematics, 2023, 8(6): 13637-13646. doi: 10.3934/math.2023692 |
[5] | Jin Cai, Shuangliang Tian, Lizhen Peng . On star and acyclic coloring of generalized lexicographic product of graphs. AIMS Mathematics, 2022, 7(8): 14270-14281. doi: 10.3934/math.2022786 |
[6] | Xin Xu, Xu Zhang, Jiawei Shao . Planar Turán number of double star $ S_{3, 4} $. AIMS Mathematics, 2025, 10(1): 1628-1644. doi: 10.3934/math.2025075 |
[7] | Ningge Huang, Lily Chen . AVD edge-colorings of cubic Halin graphs. AIMS Mathematics, 2023, 8(11): 27820-27839. doi: 10.3934/math.20231423 |
[8] | Meiqin Jin, Ping Chen, Shuangliang Tian . Interval edge colorings of the generalized lexicographic product of some graphs. AIMS Mathematics, 2024, 9(11): 30597-30611. doi: 10.3934/math.20241477 |
[9] | Baolin Ma, Chao Yang . Distinguishing colorings of graphs and their subgraphs. AIMS Mathematics, 2023, 8(11): 26561-26573. doi: 10.3934/math.20231357 |
[10] | Jing Su, Qiyue Zhang, Bing Yao . The connection between the magical coloring of trees. AIMS Mathematics, 2024, 9(10): 27896-27907. doi: 10.3934/math.20241354 |
The star chromatic index of a graph G, denoted by χ′st(G), is the smallest number of colors required to properly color E(G) such that every connected bicolored subgraph is a path with no more than three edges. A graph is K2,t-free if it contains no K2,t as a subgraph. This paper proves that every K2,t-free planar graph G satisfies χ′st(G)≤1.5Δ+20t+20, which is sharp up to the constant term. In particular, our result provides a common generalization of previous results on star edge coloring of outerplanar graphs by Bezegová et al.(2016) and of C4-free planar graphs by Wang et al.(2018), as those graphs are subclasses of K2,3-free planar graphs.
The graphs discussed in this paper are simple. When possibly multiple edges occurs, we call it a multigraph. The edge set, vertex set, minimum degree and maximum degree of G are denoted by E(G),V(G),δ(G), and Δ(G), respectively. We sometimes use E,V,δ, and Δ for short if G is understood in the context. In a multigraph G, an edge coloring satisfying any two adjacent edges assigned distinct colors is called a proper edge coloring. The chromatic index χ′(G) denotes the minimum integer s such that G admits a proper edge coloring using s colors. In 1964, Vizing [1] obtained a celebrated theorem stating that Δ+1≥χ′(G)≥Δ for every simple graph G.
In 1983, Fouquet and Jolivet [2] first introduced and studied the notion of strong edge coloring. Here strong edge coloring means proper edge coloring such that any two edges with a distance at most 2 have different colors. Similarly, we use χ′s(H), called the strong chromatic index, to denote the minimum integer c such that H has a strong edge c-coloring. Introduced by Liu and Deng [3] in 2008, a star edge coloring of a graph H is a proper edge coloring such that the edges of all cycles and paths of length 4 are colored by at least three distinct colors, i.e., there is neither a bicolored path nor cycle with length 4. The star chromatic index χ′st(H) is the minimum integer s such that H admits a star edge coloring using color set {1,2,…,s}. According to the above definition, it is natural to obtain χ′(G)≤χ′st(G)≤χ′s(G).
From the computational complexity point of view, Lei, Shi, and Song [4] showed that determining whether χ′st(G)≤3 or not for a given graph is NP-complete. Given the difficulty of computing χ′st(G), it is natural to consider its valid upper bounds. Liu and Deng [3] first proved a general result that, when Δ≥7,
χ′st(G)≤⌈16(Δ−1)32⌉. |
For sufficiently large Δ, Dvořák et al. [5] further showed that
χ′st(G)≤Δ⋅2O(1)√logΔ. |
In addition, the star edge coloring problem remains hard for many graph classes. It is difficult even for complete graph Kt, where in [5] it is proved that
2t(1+o(1))≤χ′st(Kt)≤t22√2(1+o(1))√logt(logt)14. |
On the other hand, much better results are known for sparse graphs. In [5] it is obtained that every subcubic graph H satisfies χ′st(H)≤7. For trees, Deng et al. [6] and Bezegová et al. [7] showed a sharp bound below, independently.
Theorem 1.1 ([7], [6]). For any tree T with maximum degree Δ, we have
χ′st(T)≤⌊1.5Δ⌋, |
and the bound is sharp.
Planar graphs are graphs that can be drawn on a plane without any edge crossing. If a graph has a plane embedding such that all vertices of the graph belong to the unbounded face of the drawing, we call it an outerplanar graph. For the star edge coloring of outerplanar graphs, Bezegová et al. [7] showed the results blow.
Theorem 1.2. [7] Suppose G is an outerplanar graph.
(i) If Δ≥4, then χ′st(G)≤⌊1.5Δ⌋+12;
(ii) If Δ≤3, then χ′st(G)≤5.
Wang et al. [8] improved the above bounds of outerplanar graphs and studied more general graphs without 4-cycles by introducing some novel edge-partition techniques. Their results are as below.
Theorem 1.3. [8] (i) For a planar graph G, χ′st(G)≤2.75Δ+18.
(ii) For a planar graph G having no 4-cycles, χ′st(G)≤⌊1.5Δ⌋+18.
(iii) For any outerplanar graph H, χ′st(H)≤⌊1.5Δ⌋+5.
If a planar graph G has no 4-cycles or G is outerplanar, then G is certainly K2,3-free, i.e., G contains no complete bipartite graph K2,3 as a subgraph. By extending Theorem 1.3, we study the star edge coloring problem of more general planar graphs without K2,t for every t≥2 and prove the following main result.
Theorem 1.4. Let t≥2 be an integer. Then every K2,t-free planar graph satisfies
χ′st(G)≤1.5Δ+20t+20. |
Note that, as constructed in [9], there exists a family of planar graphs with maximum degree Δ containing K2,Δ as a subgraph such that their star chromatic indices are at least 138Δ−34. So the upper bound in Theorem 1.4 is essentially sharp up to the constant term.
The rest of this paper is divided into several steps for discussion. We first describe the properties of K2,t-free graphs in Lemma 2.1. Then, using those properties and some edge-partition methods, we show that K2,t-free planar graphs can be edge decomposed into several parts and colored with different strategies to obtain the desired star edge coloring, thus proving our main result Theorem 1.4. Finally, we end this paper with a few concluding remarks.
For vertex x∈V(G) in graph G, let dG(x)(or d(x) for short) be the degree of vertex x. If d(x)=k (d(x)≥k, or d(x)≤k, respectively), then x is called a k-vertex(k+-vertex, or k−-vertex, respectively). Let NG(x) denote the set of neighbors of x. A neighbor of vertex x in NG(x) with degree k (at most k, or at least k, respectively) is called a k-neighbor(k−-neighbor, or k+-neighbor, respectively).
For a planar graph G with a fixed plane embedded, let F(G) be its face set. For f∈F(G), dG(f) is used to denote the number of edges contained in face f. If dG(f)≥4, then f is called a 4+-face. A planar embedding of a planar graph G is said an elegant drawing if it satisfies the following condition: for each pair of vertices v,u∈V(G) with {w1,w2,⋯,wt} being their common 2-neighbors, in such an embedding the 2-vertices w1,w2,⋯,wt are all located successively in clockwise (or anticlockwise) order. That is, we require that each 4-cycle vwiuwi+1 for i∈{1,2,⋯,t−1} is not separating in such an embedding for any pair of vertices. For instance, in Figure 1 graph H2 is an elegant drawing, while graph H1 is not.
The key to our proof is the following lemma concerning the structure of planar graphs, which we prove below by discharging method.
Lemma 2.1. Given an integer t≥2, let G be a simple planar graph with δ(G)≥2. Then either there exists an edge xy∈E(G) with max{d(x),d(y)}≤5t+5 or there exist t 2-vertices with two common neighbors.
Proof. By contradiction, suppose that such counterexamples to Lemma 2.1 exist. Let P denote the set of graphs G satisfying the following two conditions:
(i) G is a counterexample to Lemma 2.1,
(ii) for every 2-vertex u with NG(u)={u1,u2}, u1 and u2 are adjacent in G, i.e., u1u2∈E(G).
We first claim that set P is nonempty.
Claim 1. We have P≠∅.
Proof Let G be an arbitrary counterexample. That is, we have max{d(x),d(y)}≥5t+6 for every edge xy∈E(G) and there do not exist t 2-vertices with two common neighbors in graph G.
If condition (ⅱ) is not satisfied, let u be a 2-vertex with two neighbors a,b which are non-adjacent. Since G is a counterexample, considering the edges ua,ub, we must have d(a)≥5t+6 and d(b)≥5t+6. Now we obtain a new graph from G by connecting b and a. Since the degrees of b and a are at least 5t+7 now, the obtained graph is still a counterexample. So by a finite number of edge addition operations, the condition (ii) can be satisfied. Note that the resulting graph is still a simple graph. Hence P≠∅.
The rest of the proof will involve multigraphs, in which multiple edges may exist. Note that a simple graph itself is also a multigraph. Let Q be the collection of multigraphs H satisfying the following three conditions:
(α-i) there do not exist t 2-vertices with two common neighbors in H, and for each edge yx∈E(H) we have max{dH(y),dH(x)}≥5t+6,
(α-ii) for every 2-vertex u with NH(u)={u1,u2}, u1 and u2 are adjacent in H;
(β) the multigraph G′=H−V2(H) is a planar triangulation, where G′ is obtained from H by eliminating all 2-vertices of H, denoted by V2(H).
We now prove that the set Q is nonempty.
Claim 2. We have Q≠∅.
Proof By Claim 1, P≠∅, and so there exists a graph or multigraph H∈P satisfying conditions (α-i) and (α-ⅱ). Choose a multigraph H satisfying conditions (α-i) and (α-ⅱ) such that the total length of all the 4+-faces in G′=H−V2(H) is minimized. We shall show below that G′=H−V2(H) is a triangulation.
By contradiction, fix arbitrary 4+-face f in G′, and denote the vertices in the boundary of f by v1,v2,v3,v4,…,vp in a clockwise order. If max{dH(v1),dH(v3)}≥5t+6, we add a new edge v1v3 in H to obtain a new multigraph H1. If max{dH(v1),dH(v3)}≤5t+5, then we must have dH(v2)≥5t+6 as v1v2∈E(H) and by (α-i). So we add a new edge v2v4 in H to obtain a new multigraph H2. Let J=H1 or H2. Then as
min{dJ(v1),dJ(v2),dJ(v3),dJ(v4)}≥min1≤i≤4{dH(vi)}≥3, |
the multigraph J still satisfies conditions (α-i) and (α-ⅱ). But by the construction of adding a new edge in a 4+-face, the total length of all the 4+-faces in J is strictly smaller than that of H. This is a contradiction to the minimality of H. Thus there exists a multigraph H∈Q satisfying conditions (α-i), (α-ⅱ), and (β), proving this lemma. Note that H may be a multigraph.
Now by Claim 2 we choose a multigraph G∈Q. Denote G′=G−V2(G) as a planar triangulation by Claim 2. Then we provide some restrictions on the degrees in G′ and G.
Claim 3. We have tdG′(v)≥dG(v)≥dG′(v) for any vertex v.
Proof Let v∈V(G′) be a vertex in G′. Denote NG′(v)={u1,u2,⋯,udG′(v)}. Note that G′=G−V2(G) is constructed from G by eliminating all 2-vertices. By condition (α-i), for any edge vui∈E(G′), where 1≤i≤dG′(v), there exists at most t−1 2-vertices in G with v,ui as common neighbors. In other words, the edge vui corresponds to at most t−1 2-vertices in an elegant drawing of G. Hence we obtain the desired inequality
tdG′(v)=dG′(v)+(t−1)dG′(v)≥dG(v)≥dG′(v). |
Claim 4. We have δ(G′)≥3.
Proof Assume δ(G′)≤2 and a minimum degree vertex of G′ is denoted by v. By the construction of G′ and by δ(G)≥2, we have dG(v)≥3, which implies that v is adjacent to no less than one 2-vertex u in G. By Claim 3, dG(v)≤2t and we obtain an edge uv∈E(G) satisfying max{dG(u),dG(v)}≤5t+5, which is a contradiction to condition (α-i) and Claim 2.
Claim 5. Each of the following holds.
(a) For each u∈V(G′) with dG′(u)≤5, dG(u)=dG′(u).
(b) Each 5−-vertex in G′ is neither adjacent to any 5−-vertex in G′ nor any 5−-vertex in G.
(c) Each 5−-vertex in G′ can not be adjacent to any 6-vertex in G′.
Proof If (a) is not true, then dG(v)>dG′(v). Hence the vertex v is adjacent to at least one 2-vertex w in G. By Claim 3, it holds dG(v)≤tdG′(v)≤5t. Now we get an edge vw∈E(G) with max{dG(v),dG(w)}≤5t+5, which is a contradiction to condition (α-i) and Claim 2.
To prove (b), suppose that a 5−-vertex is adjacent to a 5−-vertex in G′ or G. This implies that a (5t)−-vertex is adjacent to a (5t)−-vertex in G by Claim 4, which is a contradiction to condition (α-i).
To prove (c), suppose that a 5−-vertex v is adjacent to a 6-vertex u. By Claim 5 (a), we obtain dG(v)=dG′(v) which implies that the edge uv does not correspond to any 2-vertex of G in an elegant drawing by condition (α-i). For each other edge incident to u, by condition (α-i) it corresponds to at most t−1 2-vertices of G in an elegant drawing. Hence the vertex u has degree dG(u)≤6+5(t−1)<5t+5. Now vu∈E(G) satisfies max{dG(v),dG(u)}≤5t+5, which is a contradiction to condition (α-i) and Claim 2.
Claim 6. For any u∈V(G′), if u is adjacent to at most five 6+-vertices in G′, then any neighbor v of vertex u has dG′(v)≥6.
Proof By contradiction, suppose that a neighbor v of vertex u has dG′(v)≤5. By Claim 5(a), we have dG(v)=dG′(v)≤5. By Claim 5 (b) again, each 5−-neighbor of u have no 2-vertex of G as their common neighbor in G. For each 6+-neighbor of u, it has at most t−1 common 2-neighbors with u in G. Thus we have
dG(u)≤dG′(u)+5(t−1). |
Notice that G′ is a planar triangulation. By Claim 5 (b), considering the embedding of 6+-vertices neighbors around vertex u in an elegant drawing, there is at most one 5−-neighbor of u between two consecutive 6+-neighbors. Hence dG′(u)≤10 as u is adjacent to at most five 6+-neighbors in G′. Therefore, we have dG(u)≤dG′(u)+5(t−1)≤5t+5, and thus vu∈E(G) satisfies max{dG(v),dG(u)}≤5t+5, contrary to condition (α-i) and Claim 2.
Finally, we are able to complete the proof by a discharging method on G′. Assign initial charges by setting m(x)=dG′(x)−6 for each vertex x∈V(G′) and m(f)=2(dG′(f)−3) for each face f∈F(G′). Since G′ is a planar triangulation, we have m(f)=0 for every face f∈F(G′). By rewriting Euler formula, it follows that
∑x∈V(G′)∪F(G′)m(x)=∑u∈V(G′)(dG′(u)−6)+2∑f(dG′(f)−3)=−12. | (2.1) |
We shall use the following discharging rule.
Discharging Rule: Every 5−-vertex v takes charges 6−dG′(v)dG′(v) from each 7+-neighbor in G′.
We will derive the final contradiction from the next claim.
Claim 7. Each vertex in G′ has a non-negative charge after discharging.
Proof By Claim 5(c), every neighbor of each 5−-vertex in G′ has a degree of at least 7. Hence the final charge of each 5−-vertex is at least zero by the discharging rule.
Assume that u∈V(G′) has degree dG′(u)=k∈{7,8,9,10}. We claim that u is adjacent to at most k−6 5−-vertex in G′. Otherwise, u is adjacent to at most five 6+-vertices. Let a be any 5−-neighbor of u. Then we have dG(a)=dG′(a)≤5 by Claim 5(a). Since max{dG(u),dG(a)}>5t+5, we have dG(u)>5t+5>dG′(u). But as u has at most five 6+-neighbors, and for each 6+-neighbor at most t−1 2-vertices are added from G′ to G, and thus dG(u)≤5(t−1)+k≤5t+5, a contradiction. This verifies our claim that u is adjacent to at most k−6 5−-vertex in G′. By the discharging rule, the vertex u loses at most dG′(u)−6 charges and remains non-negative.
Assume instead that u∈V(G′) has degree dG′(u)≥11. If u is adjacent to at least six 6+-vertices, then u is adjacent to at most dG′(u)−6 5−-vertices in G′, and hence the final charge is non-negative by our discharging rule. If u is adjacent to at most five 6+-vertices, then it follows from Claim 6 that the final charge of u is dG′(u)−6>0. Therefore, we have proved that each vertex in G′ has a non-negative charge after discharging.
As G′ is a planar triangulation, the final charge of each face in G′ is zero. By Claim 7, each vertex has a non-negative final charge. Hence total final charge is non-negative, which is a contradiction to Equation (2.1). This contradiction proves Lemma 2.1.
Lemma 2.1 leads to the following edge-partition result on K2,t-free planar graphs.
Lemma 2.2. Let G be a K2,t-free planar graph, where t≥2 is an integer. Then we can edge-partition G into a forest T and a subgraph H⊆G with Δ(H)≤5t+5.
Proof. Let G be a minimal counterexample with |E(G)|+|V(G)| as small as possible. If δ(G)=1, then we denote a 1-vertex by u and its unique neighbor by v. By the minimal counterexample property, we can edge-partition the graph G′=G−{uv} into a forest T′ and a subgraph H′⊆G′ with Δ(H′)≤5t+5. Denote T=T′+uv and H=H′. By dG(u)=1, T has no cycle, and thus it is a forest with Δ(T)≤Δ(G), and we have further Δ(H)≤5t+5, which is a contradiction. Hence we obtain δ(G)≥2. It follows from Lemma 2.1 that we have an edge xy∈E(G) with max{dG(x),dG(y)}≤5t+5. Let G∗=G−{xy}. So E(G∗) can be decomposed into a forest T∗ and a subgraph H∗⊆G∗ with Δ(H∗)≤5t+5. Let H=H∗+{xy}. Since Δ(H)≤max{Δ(H∗),dG(x),dG(y)}, T=T∗ and H=H∗+{xy} are the desired partition of G, which is a contradiction.
We need one more result on edge coloring as part of the prerequisites.
Theorem 2.3 ([10], [11]). Let H be a planar graph with maximum degree Δ. If k=max{Δ,7}, then H is k-edge-colorable.
Now it is ready to demonstrate the main result of Theorem 1.4 restated below.
Theorem 2.4. Let G be a K2,t-free planar graph, where t≥2 is an integer. We have
χ′st(G)≤⌊1.5Δ⌋+4(5t+5). |
Proof. By Lemma 2.2, we can edge-partition G into a forest T with Δ(T)≤Δ(G) and a planar subgraph H⊆G with Δ(H)≤5t+5. By Theorem 1.1, χ′st(T)=k≤⌊1.5Δ⌋, thus we have a star k-edge-coloring ϕ:E(T)↦{f1,f2,…,fk} for T. By Theorem 2.3 and Δ(H)≤5t+5, H is (5t+5)-edge-colorable, and consequently H has an edge-partition E1,E2,…,E5t+5, where Ei denotes the collection of edges of color-i for each 1≤i≤5t+5. For fixed i∈{1,…,5t+5}, consider the graph Gi constructed from G by contracting edge set Ei. By the planarity of Gi, it follows from the Four Color Theorem that Gi is 4-vertex colorable with colors saying {ai,bi,ci,di}. Note that every edge e in Ei of G corresponds a contracted vertex in Gi, and we use its color to color the edge e. After performing this coloring process for all i∈{1,…,5t+5}, we obtain a 4(5t+5)-edge-coloring ψ:E(H)↦∪5t+5i=1{ai,bi,ci,di}. Furthermore, each pair of edges e1,e2∈E(H) with distance at most 2 receive different colors in the coloring ψ by its construction.
Then we show below that ϕ∪ψ is a star edge coloring of G using k+4(5t+5)≤⌊1.5Δ⌋+4(5t+5) colors. Clearly, ϕ∪ψ is a proper edge coloring of G, i.e., any adjacent edges get distinct colors. Let P=x1x2x3x4x5 denote a path or cycle with 4 edges. Note that each edge in E(T) receives a color in {f1,f2,…,fk}, and each edge in E(H) receives a color in ∪5t+5i=1{ai,bi,ci,di}. If x1x2 and x3x4 receive the same color, then by the construction of coloring ϕ∪ψ, we must have both x1x2∈E(T) and x3x4∈E(T), as any two edges with distance two in E(H) receive distinct colors. If x2x3 and x4x5 receive the same color, then we also have x2x3,x4x5∈E(T) for the same reason. But since ϕ is a star edge coloring of T, ϕ(x2x1)=ϕ(x4x3) and ϕ(x3x2)=ϕ(x5x4) can not occur simultaneously. Therefore, ϕ∪ψ is a star edge coloring of G by definition.
This paper has shown that every K2,t-free planar graph admits a star (⌊1.5Δ⌋+20t+20)-edge coloring, which is sharp up to the constant term. This can be evident by the planar graph constructed in [9], whose star chromatic index is at least 138Δ−34 and contains K2,Δ as a subgraph. On the other hand, it would be interesting to push further on the constant term to determine the optimal upper bound. In [7], a similar problem is proposed for outerplanar graphs by the authors, who conjectured that every outplanar graph is star (⌊1.5Δ⌋+1)-edge colorable. Another related research direction is to consider the list version of star edge coloring. The edge-partition techniques in this paper (and in the literature) could not be generalized to list star edge coloring. There are some other techniques, such as that of [9], which may be helpful in handling list star edge coloring. It should be of interest to search for a similar list version of Theorem 1.4, or its subclass of graphs like outerplanar graphs and C4-free planar graphs.
The authors would like to thank all the reviewers for their very helpful comments and constructive suggestions. This research is partially supported by the National Natural Science Foundation of China (No. 11901434).
The authors declare there is no conflict of interest.
[1] |
V. G. Vizing, On an estimate of the chromatic index of a p-graph, Diskret. Anal., 3 (1964), 25–30. http://dx.doi.org/10.1090/S0894-0347-1992-1124979-1 doi: 10.1090/S0894-0347-1992-1124979-1
![]() |
[2] |
J. L. Fouquet, J. L. Jolivet, Strong edge-colorings of graphs and applicantions to multi-k-gons, Ars Comb., 16 (A) (1983), 141–150. http://dx.doi.org/10.1090/S0894-0347-1992-1124979-1 doi: 10.1090/S0894-0347-1992-1124979-1
![]() |
[3] |
X. S. Liu, K. Deng, An upper bound on the star chromatic index of graphs with Δ≥7, J. Lanzhou Univ. (Nat. Sci.), 2 (2008), 98–99. http://dx.doi.org/10.1090/S0894-0347-1992-1124979-1 doi: 10.1090/S0894-0347-1992-1124979-1
![]() |
[4] |
H. Lei, Y. Shi, Z. Song, Star chromatic index of subcubic multigraph, J. Graph Theory, 88 (2018), 566–576. https://doi.org/10.1002/jgt.22230 doi: 10.1002/jgt.22230
![]() |
[5] |
Z. Dvořák, B. Mohar, R. Šámal, Star chromatic index, J. Graph Theory, 72 (2013), 313–326. https://doi.org/10.1002/jgt.21644 doi: 10.1002/jgt.21644
![]() |
[6] |
K. Deng, X. S. Liu, S. L. Tian, Star edge coloring of trees, (in Chinese) J. Shandong Univ. Nat. Sci., 46 (2011), 84–88. http://dx.doi.org/10.1090/S0894-0347-1992-1124979-1 doi: 10.1090/S0894-0347-1992-1124979-1
![]() |
[7] |
L'. Bezegová, B. Lužar, M. Mocko∨vciaková, R. Soták, R. Škrekovski, Star edge coloring of some classes of graphs, J. Graph Theory, 81 (2016), 73–82. https://doi.org/10.1002/jgt.21862 doi: 10.1002/jgt.21862
![]() |
[8] |
Y. Wang, W. Wang, Y. Wang, Edge-partition and star chromatic index, Appl. Math. Comput., 333 (2018), 480–489. https://doi.org/10.1016/j.amc.2018.03.079 doi: 10.1016/j.amc.2018.03.079
![]() |
[9] |
J. Li, K. Horacek, R. Luo, Z. Miao, Upper Bounds on List Star Chromatic Index of Sparse Graphs, Acta Math. Sin. (Engl. Ser.), 36 (2020), 1–12. https://doi.org/10.1007/s10114-019-8546-7 doi: 10.1007/s10114-019-8546-7
![]() |
[10] |
D. P. Sanders, Y. Zhao, Planar graphs of maximum degree seven are class 1, J. Combin. Theory Ser. B, 83 (2001), 202–212. https://doi.org/10.1006/jctb.2001.2047 doi: 10.1006/jctb.2001.2047
![]() |
[11] |
L. Zhang, Every planar graph with maximum degree 7 is of class 1, Graphs Combin., 16 (2000), 457–495. http://doi.org/10.1007/s003730070009 doi: 10.1007/s003730070009
![]() |