ebook img

$L(1,1)-$ Labeling of Direct Product of Cycles PDF

0.25 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview $L(1,1)-$ Labeling of Direct Product of Cycles

L(1,1)− LABELING OF DIRECT PRODUCT OF CYCLES 5 1 0 TAYO CHARLES ADEFOKUN1 AND DEBORAH OLAYIDE AJAYI2 2 n Abstract. AnL(1,1)-labelingofagraphGisanassignmentoflabelsfrom{0,1··· ,k} a J totheverticesofGsuchthattwoverticesthatareadjacentorhaveacommonneigh- 9 bor receive distinct labels. The λ11− number, λ11(G) of G is the minimum value k 2 such that G admits an L(1,1) labeling. We establish the λ11− numbers for direct product of cycles Cm ×Cn for all positive m,n ≥ 3, where both m,n are even or ] O when one of them is even and the other odd. C . h t a m 1. Introduction [ The L(h,k)-labeling problem (or (h,k)-coloring problem) is that of vertex labeling 1 of an undirected graph G with non-negative integers such that for every u,v ∈ V(G), v uv ∈ E(G), |l(u)−l(v)| ≥ h and for all u,v ∈ V(G),d(u,v) = 2, |l(u)−l(v)| ≥ k. 5 3 The difference between the largest label and the smallest label assigned is called the 3 span. The aim of L(h,k)−labeling is to obtain the smallest non negative integer 7 λk(G), such that there exists an L(h,k)-labeling of G with no label on V(G) greater 0 h . than λk(G). 1 h 0 Motivated by Hales’ 1980 paper [8], which provided a new model for frequency 5 assignment problems as a graph coloring problem, Griggs and Yeh [7] formulated the 1 : L(2,1) problem to model the channel assignment problem. The general notion of v L(h,k)- labeling was first presented by Georges and Mauro [6] in 1995. The topic i X has since then been an object of extensive research for various graphs. Calamonerri’s r a survey paper [4] contains known results on L(h,k)-labeling of graphs. L(1,1)-labeling(or strong labeling condition) of a graphis a labeling ofG such that vertices with a common neighbor are assigned distinct labels. The usual labeling (or propervertex coloring)conditionisthatadjacentvertices havedifferent colors, butfor L(1,1), also all neighbors of any vertex are colored differently. This is equivalent to a proper vertex-coloring of the square of a graph G. Note that a proper k- coloring of a graph is a mapping α : V(G) → {1,··· ,k} such that for all uv ∈ E(G) α(u) 6= α(v) and the square G2 of G has vertex V(G) with an edge between two vertices which are adjacent in G or have a common neighbor in G. The chromatic number χ(G) of G is the smallest k for which G admits a k-coloring. Therefore, χ(G2) = λ1(G)+1 1 for a graph G. Key words and phrases. L(1,1)-labeling,D-2 Coloring,DirectProductofGraphs,CrossProduct of Graphs, cycle graphs 2010 Mathematics Subject Classification. Primary: 05C35. 1 2 T.C ADEFOKUNAND D.O. AJAYI Labelingofgraphpowersisoftenmotivatedbyapplicationsinfrequencyassignment and has attracted much attention [See for example, [1]]. L(1,1)-labeling has applica- tions in computing approximation to sparse Hessian matrices, design of collision-free multi-hop channel access protocols in radio networks segmentation problem for files in a network and drawings of graphs in the plane [3, 13, 15, 16] to mention a few. For graphs G and H, the direct product G × H have vertex set V(G) × V(H) where two vertices (x ,x ) and (y ,y ) are adjacent if and only if (x ,y ) ∈ E(G) 1 2 1 2 1 1 and (x ,y ) ∈ E(H). This product is one of the most important graph products with 2 2 potential applications in engineering, computer science and related disciplines. [11]. The L(h,k)- labeling of direct product of graphs was investigated in [2, 5, 9, 12, 14, 18, 19, 20]. In particular, Jha et al [12] gave upper bounds for λk-labeling of the multiple direct 1 product and cartesian product of cycles with some conditions on k and the length of the cycles. They also presented some cases where we have exact values. In addition by using backtracking algorithm, they computed λd(C × C ) for 2 ≤ d ≤ 4 and 1 m n 4 ≤ m,n ≤ 10. Since every L(2,1)-labeling is an L(1,1)-labeling, then λ1(G) ≤ 1 λ2(G). Therefore, their results for d = 2 provided upper bounds for L(1,1)-labeling 1 of C ×C for 4 ≤ m,n ≤ 10. The only result for λ1-labeling for direct product of m n 1 two cycles in the paper is that if m,n ≡ 0mod5 then λ1(C ×C ) = 4. 1 m n In this paper, we solve the L(1,1)-labeling problem for direct product of cycles C , m C , m,n ≥ 3, except form ∈ {16,18,22,26,32,36,46}, n ∈ {14,16,18,26,28,34}and n for these outstanding cases we conjecture that λ1(C ×C ) = 5. 1 m n The paper is organized as follows: We give some preliminaries in Section 2 and obtain the λ1 labeling numbers for C ×C for m ≥ 3 and n = 4 and 6 and some of 1 m n their multiples in Section 3. Section 4 deals with labeling of direct product of bigger cycles. 2. Preliminaries Let G be a finite simple undirected graph with at least two vertices. For subgraph V′ ⊆ V(G), we denote by L(V′) the set of L(1,1)−labeling on V′ and for a non- negative integer, say, k, we take [k(ǫ)] as the set of even integers and zero in [k] while [k(o)] is the set of odd integers in [k]. Suppose further that v ∈ V(G), we denote d v as the degree of v. The following results, remarks and definitions are needed in the work. Theorem 2.1. [10] Graph G×H is connected if and only if G and H are connected and at least one of G and H is non-bipartite. Remark 2.2. (i) Let G = C ×C , where m,n are even positive integers. Then, m n G= G ∪ G , where G and G are the connected components of C × C , 1 2 1 2 m n where {V(G ) = u v : i ∈ [(m−1)(ǫ)],j ∈ [(n−1)(ǫ)] or 1 i j i ∈ [(m−1)(o)];j ∈ [(n−1)(0)]} and {V(G ) = u v : i ∈ [(m−1)(ǫ)],j ∈ [(n−1)(o)] or 2 i j LABELING PRODUCTS OF CYCLES 3 i ∈ [(m−1)(o)];j ∈ [(n−1)(ǫ)]}. Note that G and G are isomorphic and it is demonstrated in the graph C ×C 1 2 4 6 below Fig. 1: The components of C4×C6 (ii) Suppose G = C ×C such that G = G′ ∪G′′, where G′,G′′ are components m n of G, then, λ1(G) = max{λ1(G′),λ1(G′′)}. 1 1 1 (iii) Let G = C ×C , where m is even and n odd positive integers. Then, G ≡ G , m n 1 where G is any of the two connected components of C ×C . 1 m 2n 0,0 0,1 0,2 0,3 0,4 0,0 3,0 3,0 2,0 2,0 1,0 1,0 0,0 0,1 0,2 0,3 0,4 0,0 Fig. 2: C4×C5 isisomorphic to a component of C4×10 Let P be a path of length m−1. The following results are from [2]: m Corollary 2.3. For m ≥ 3, λ1(P ×C ) = 5 1 m 6 Lemma 2.4. For m ≥ 5, n ≥ 9, n 6≡ 0 mod 5 λ1(P ×C ) ≥ 5 1 m n A useful lower bound on L(1,1)-labeling for any graph G is contained in the fol- lowing Lemma: Lemma 2.5. [6] If G is a graph with maximum degree △, and G includes a vertex with △ neighbors, each of which is of degree △, then λ1(G) ≥ △ 1 From the lemma, we have for m,n ≥ 3 λ1(C ×C ) = 4 1 m n 4 T.C ADEFOKUNAND D.O. AJAYI 3. Labeling of C ×C ,n = 4,6 m n In this section, we investigate the λ− numbers of graph product C × C and m 4 C ×C , where m ≥ 3. m 6 Let G′ be the connected component of the product graph under consideration. Lemma 3.1. For m ≥ 4, and even, λ1(C ×C ) ≥ 5. 1 m 4 Proof. Let G′ ⊂ C ×C , where m ≥ 4 and even. Suppose V ,V ,V ⊂ V(G′). Let m 4 i i+1 i+2 G′ bethesubgraphofG′ inducedbyV ,forallj ∈ [2].Then, V(G′)={u v ,u v ,u v , 1 i+j 1 i 0 i 2 i+1 1 u v ,u v ,u v }. Now it is clear that the diameter of G′ is 2. Thus for every i+1 2 i+2 0 i+2 2 1 pair v ,v ∈ V(G′), d(v ,v ) ≤ 2. Thus, l(v ) 6= l(v ) for all v ,v ∈ V(G′). Now, 1 2 1 1 2 1 2 1 2 1 |V(G′)| = 6. Therefore λ1(C ×C ) ≥ 5. (cid:3) 1 1 m 4 Remark 3.2. Note that if G′ ⊂ C × C , m ≥ 4 with m even and v ∈ V , for m 4 i i some i, V ⊂ V(G′), such that l(v ) = α ∈ [m], with λ1(G′) = m, then α ∈/ i i i 1 1 L{V V V V }. i−2 i−1 i+1 i+2 Theorem 3.3. λ1(C ×C ) = 7 1 4 4 Proof. Let G′ ⊂ C ×C and V ⊂ V(G′) for each i ∈ [3]. Clearly, V(G′) = ∪3 V . 4 4 i i=0 i Let G′ be a subgraph of G′ such that G′ is induced by V ,V ,V . By the proof of 1 1 0 1 2 Lemma 3.1, |V(G′)| = 6 and suppose α′,α′′,∈ L(V ), then by remark 3.2,α′,α′′,∈/ 1 k k 3 k k L(V(G′)). Thus there exists α′,α′′,∈/ [5] such that {α′,α′′} = L(V ), and α′ 6= α′′ 1 k k k k 3 k k since d(v′,v′′) = 2 for v′,v′′ ∈ V . Thus, |L(∪3 V )|=|L(V(G′))| = 6+2. Therefore, 3 3 3 3 3 i=0 i λ1(C ×C ) = λ1(G′) = 7. (cid:3) 1 4 4 1 Next we present the necessary and sufficient condition under which λ1(C ×C ) is 1 m 4 5. Theorem 3.4. For m ≥ 4,m even, λ1(C ×C ) = 5 if and only if m ≡ 0 mod 6. 1 m 4 Proof. Let m = 6n, n ∈ N. By Lemma 3.1, λ1(G) ≥ 5. Therefore, λ1(G′) ≥ 5, where 1 1 G′ ⊂ C × C . Let G′′ be the connected component of C ×C . By Corollary 3.2, m 4 6 4 L(V )∩L(V ) = ∅, L(V )∩L(V ) = ∅, L(V )∩L(V ) = ∅. Now, set L(V ) = L(V ), 0 1 1 2 2 0 0 3 L(V ) = L(V ),L(V ) = L(V ). But L(V ) ∩ L(V ) = ∅, L(V ) ∩ L(V ) = ∅. Thus, 1 4 2 5 5 0 5 1 λ1(G′′) ≤ 5 and λ1(C ×C ) = 5. Thus by re-occurrence along C and C , m = 0 1 1 6 4 n m mod 6 implies λ1(G) = 5. 1 Conversely, suppose λ1(G) = 5. Let G′ be a connected component of G = C × 1 m C , m ≥ 4,m even. Then, λ1(G′) = 5. Now, assume that m 6≡ 0 mod 6 , then 4 1 m = 6n′ +2 or m = 6n′ +4 where n′ ∈ N∪0. For n′ = 0, G = C ×C , for which 4 4 λ1(G′) = 7 by Theorem 3.3. 1 Case i: For m = 6n′ + 2, n′ ∈ N, let V ,V ,V be subsets of V(G′). By Corollary 0 1 2 3.2, L(V ) ∩ L(V ) = ∅, L(V ) ∩ L(V ) = ∅ and L(V ) ∩ L(V ) = ∅. Now let G′ be 0 1 1 2 2 0 1 the subgraph of G′ induced by V ,V ,V . Since L(V ) ∩ L(V ) = ∅ for V ⊂ V(G′) 0 1 2 3 2 3 and λ1(G′) = 5, then L(V ) = L(V ). Let V ⊂ V(G′). Then L(V )∩L(V ) = ∅ and 1 3 0 4 4 3 L(V )∩L(V ) = ∅. Thus L(V ) = L(V ). Let V ⊂ V(G′). Then L(V )∩L(V ) = ∅ 4 2 4 1 5 5 4 and L(V )∩L(V ) = ∅ and therefore L(V ) = L(V ). The scheme continues in such a 5 3 4 1 LABELING PRODUCTS OF CYCLES 5 way that L(V ) = L(V +3) for all i ∈ [m−1], that is L(V ) = L(V ) = L(V ) = ··· = i i 0 3 6 L(V6n′), L(V1) = L(V4) = L(V7) = ··· = L(V6n′+1), L(V2) = L(V5) = L(V8) = ··· = L(V0). Now, for all va ∈ V0 and vb ∈ V6n′, d(va,vb) = 2. For all vc ∈ V1, vd ∈ V6n′+1, d(v ,v ) = 2 and finally, L(V )∩L(V ) = ∅. Thus a contradiction. c d 0 2 Case ii: For m = 6n′ +4, n′ ∈ N, similar argument as in m = 6n+2 applies. Thus, λ1(C ×C ) = 5 if and only m ≡ 0 mod 6 (cid:3) 1 m 4 Corollary 3.5. Let m ≡ 0 mod 6 and n ≡ 0 mod 4. Then, λ1(C ×C ) = 5 1 m n Proof. The claim follows from Theorem 3.4 and the re-occurrence of the optimal labeling of C6n′ ×C4, n′ ∈ N. (cid:3) Corollary 3.6. For all m 6≡ 0 mod 6, λ1(C ×C ) ≥ 6. 1 m 4 Theorem 3.7. For C ×C , λ1(C ×C ) = 7 8 4 1 8 4 Proof. Suppose G′ is a connected component of C ×C and suppose that λ1(G′) = 6. 8 4 1 By Corollary 3.2, |L{V ,V ,V }| = 6. Likewise, for all α ∈ L(V ), α ∈/ L(V ) and 0 1 2 k 0 k 7 α ∈/ L(V ). Also, for all α ∈ L(V ), α ∈/ L(V ). Suppose L(V ) = L(V ), then by k 6 j 1 j 7 2 7 Corollary 3.2, if α ,α ∈ L(V ), then α ,α ∈/ L{V ,V ,V ,V }. Since λ1(G′) = 6, a b 2 a b 3 4 5 6 1 then there exists only five members of [6] that labels V ,V ,V ,V . However, this 3 4 5 6 contradicts Lemma 3.1. Thus, L(V ) 6= L(V ). Now suppose one of α ,α ∈ L(V ), 2 7 a b 2 say α , labels some vertex v ∈ L(V ), then there exists some α′ ∈ [6] such that a 1 7 a α′ ∈/ L{V ,V ,V } such that α′ = l(v ) ∈ V , with v 6= V . Now let α ,α ∈ a 0 1 2 a 2 7 1 2 c d L(V ). Suppose L(V ) = L(V ). Then by Corollary 3.2, α ,α ∈/ L(V ,V ,V ), which 0 3 0 a b 4 5 6 contradicts Lemma 3.1 since |L(V ,V ,V )| = 6 and [6]\2 = 5. Then, α′ ∈ L(V ) 4 5 6 a 3 and also also one of α ,α′ ∈ L(V ). Further, by Corollary 3.2, α ,α′ ∈/ L(V ,V ,V ). a b 3 a b 4 5 6 Thus, λ1(G′) ≥ 7. Conversely, λ1(G′) ≤ 7 follows directly from re-occurrence of the 1 1 labeling of C ×C . Thus, λ1(G′) = λ1(C ×C ) ≥ 7. (cid:3) 4 4 1 1 8 4 The next result focuses on the λ1−number of C × C , for m ≥ 9. Theorem 3.6 1 m 4 have already established the lower bound for λ1−number of C ×C to be 6 if m is 1 m 4 not a multiple of 6. So we only need to label C ×C with [6] such that it combines 10 4 perfectly with the labeling of C ×C with [5] to establish general bound for all cases 6 4 except when m = 14 which is dealt with separately. 0 1 2 0 0 1 5 1 2 0 5 3 4 5 6 2 6 4 3 4 5 3 3 4 0 3 5 3 2 0 1 2 3 4 0 1 0 1 2 0 0 1 5 1 2 0 Fig.3: 5−L(1,1)-labelingofC4×C6 Fig.4: 6−L(1,1)-labelingofC4×C10 Theorem 3.8. Let m′,m′′ ∈ N ∪ 0, with 10m′ + 6n′′ not a multiple of 6. Then λ11(C10m′+6m′′ ×C4) = 6. Proof. ByCorollary3.6, λ1(C ×C ) ≥ 6forallmnotmultipleof6. Theclaimfollows 1 m n requiredcombinationsofFigures3and4abovewhichshowsthatλ11(C10m′+6m′′×C4) ≤ 6 for 10m′ +6m′′ not a multiple of 6. (cid:3) 6 T.C ADEFOKUNAND D.O. AJAYI Clearly, every even number m ≥ 10, m 6= 14 can be obtained from 10m′ + 6m′′ defined above. Therefore, we can conclude that for all m ≥ 9, λ1(C ×C ) = 6 for 1 m n all m that is not a multiple of 6 if we can establish that the λ1−number of C ×C 1 14 4 is 6. We show this in the next result. Theorem 3.9. λ1(C ×C ) = 6 1 14 4 Proof. λ1(C ×C ) = 6 (cid:3) 1 14 4 We have now completely determined the λ1−numbers of (C ×C ) for all m ≥ 3. 1 m 4 In what follows, we investigate the values of λ1(C ×C ). 1 m 6 Proposition 3.10. Let G′ be a connected component of C ×C , m ∈ N,m ≥ 3. Let m 6 V ⊆ V(G′), i ∈ [m−1]. Then, i (i) λ1(C ×C ) ≥ 5. 1 m 6 (ii) Given v ,v ∈ V , d(v ,v ) ≤ 2. a b i a b (iii) For all V ⊆ V(G′), |L(V )| = 3 i i (iv) suppose α ∈ L(V ), then α ∈/ L(V ). k i k i+2 Proof. The proof of the claims above are as follows: (i) C ×C contains P ×C . Now from Corollary 2.3, λ1(P ×C ) = 5. Therefore m 6 m 6 1 m 6 λ1(C ×C ) ≥ 5. 1 m 6 (ii) Let V ⊆ V(G′). V = {u v ,u v ,u v }, where j ∈ {0,1}. Now, since C is a i i i j i j+2 i j+4 m cycle,then d(u v ,u v ) = 2. Clearly, d(u v ,u v ) = 2, d(u v ,u v ) = 2 and i j i j+4 i j i j+2 i j+2 i j+4 thus the claim. (iii) This is quite obvious. (iv) It is obvious that for all u v ∈ V and u v ∈ V , d(u v ,u v ) = 2. i j i i+2 k i+2 i j i+2 k Therefore, L(V )∩L(V ) = ∅. i i+2 (cid:3) The next result describes a property of L(1,1)-labeling of C ×C m 6 Lemma 3.11. Let α ∈ L(V ), i ∈ [m−1], V ⊆ V(G′), then α labels some vertex k i i k v ∈ V . In other words, L(V ) labels V . i+1 i+1 i i+1 Proof. Let{u v ,u v ,u v } = V and{u v ,u v ,u v } = V Clearly i j i j+2 i j+4 i i+1 j+1 i+1 j+3 i+1 j+5 i+1 d(u v ,u v ) = d(u v ,u v ) = d(u v ,u v ) = 3. Therefore, suppose i j i+1 j+3 i j+2 i+1 j+5 i j+4 i+1 j+1 α = l(v ), for some v ∈ V , then, there exists some unique v ∈ V such that k i i i i+k i+1 l(v ) = l(v ),with |(i−(i+k))| = 3. (The uniqueness of v results from Proposi- i i+k i+k (cid:3) tion 3.10(b).) Corollary 3.12. If L(V ) = L(V ), then, L(V )∩L(V ) = ∅ and L(V )∩L(V ) = i i+1 i+2 i i i+3 ∅. It is obvious from Proposition 3.10(d). Corollary 3.13. λ1(C ×C ) = 5 if and only if m ≡ 0 mod 4 1 m 6 LABELING PRODUCTS OF CYCLES 7 Proof. Let m ≡ 0 mod 4. For m = 4, clearly λ1(C ×C ) = 5, which is obtained from 1 4 6 Corollary 3.6. Now in the case of the general m ≡ 0 mod 4, by re-occurrence of the labeling of C ×C along C , it follows that λ1(C ×C ) = 5. Conversely, suppose 4 6 m 1 m 6 that λ1(C ×C ) = 5. We show that m ≡ 0 mod 4. Let L(V ) = {α ,α ,α }. By 1 m 6 i i j k Proposition 3.10 (c),α 6= α 6= α 6= α , that is, |L(V )| = 3. Suppose L(V ) = L(V ) i j k i i 0 1 by Lemma 3.11, then by Corollary 3.12, L(V ) ∩L(V ) = ∅ and L(V ) ∩L(V ) = ∅. 2 1 3 1 Since λ1(C × C ) = 5, Then L(V ) = L(V ) = [5]\L(V ). This scheme continues 1 m 6 2 3 0 such that L(V ) = L(V ) = L(V ); L(V ) = L(V ) = L(V )··· = L(V ) = L(V ) = 0 4 5 2 6 7 0 m−4 L(V ); and L(V ) = L(V ) = L(V ) = ··· = L(V ), n ∈ N, where 4n = (m−1)+ m−3 0 4 8 4(n) 1 = m since C is a cycle. Thus m ≡ 0 mod 4. (cid:3) m The implication of the last result is that the lower bound for the λ1−number of 1 graph product C ×C , m ≥ 3 is 6 except for when m ≡ 0 mod 4, in which case the m 6 optimal λ1−number reduces by 1. 1 Now we consider particular cases where the lower bound is strictly greater than 6. Theorem 3.14. λ1(C ×C ) = 8 1 6 6 Proof. Supposethatλ1(C ×C ) = 7. Let{V } ⊆ V(G′),foralli ∈ [5]. Byproposition 1 6 6 i 3.10 (d), L(V ) ∩ L(V ) = ∅; L(V ) ∩ L(V ) = ∅ and L(V ) ∩ L(V ) = ∅. Now, 0 2 0 4 4 2 L(V ) ⊆ [7]\L(V ) and L(V ) ⊆ [7]\L(V ). Note that |[7]\L(V )| = 5. Now set 2 0 4 0 0 [7]\L(V ) = [A′]. L(V ) ⊆ [A′]\L(V ) since L(V )∩L(V ) = ∅. Now, |[A′]\L(V )| = 2. 0 4 2 4 2 2 However, by Proposition 3.10 (c), |L(V )| = 3. Therefore a contradiction and hence 4 λ(C ×C ) ≥ 8. The labeling in Figure 6 confirms that λ1(C ×C ) ≤ 8, and thus, 6 6 1 6 6 λ1(C ×C ) = 8. 1 6 6 0 1 2 0 8 6 7 6 7 8 3 5 3 4 3 4 5 3 2 0 1 0 1 2 0 Fig.6: 8−L(1,1)-labelingofC6×C6 (cid:3) Theorem 3.15. λ1(C ×C ) = 7 1 10 6 Proof. Let G′ be a connected component of C ×C . and suppose that λ1(G′) = 6. 10 6 1 Let V ⊂ V(G′) such that L(V ) ⊂ [6]. By Proposition 3.10 (c), L(V ) ∩L(V ) = ∅. 0 0 0 2 Therefore L(V ) ⊂ [6]\L(V ), where |L(V )| = 3. Thus, |[6]\L(V )| = 4. Now, for 0 0 0 0 all v ∈ V and v ∈ V , V ,V ⊂ V(G′), d(v ,v ) = 2, since C is a cycle of 0 0 8 8 0 8 0 8 10 length 10. Therefore, L(V ) ⊂ [6]\L(V ). Now, suppose that, L(V ) = L(V ), by 8 0 8 2 Proposition 3.10, then there exists α ∈ [6] such that α ∈/ L(V ), andα ∈ L(V ). k k 0 k 2 Thus L(V ) ⊂ L(V ) ∪ α and L(V ) ⊂ L(V ) ∪ α . Now, |L(V )∪α | = 4. By 4 0 k 6 0 k 0 k 8 T.C ADEFOKUNAND D.O. AJAYI Proposition 3.10, (d), L(V ) ∩ L(V ) = ∅. Thus |L(V )∪L(V )| = 6, which is a 6 4 6 4 contradiction. Now, suppose L(V ) 6= L(V ) then it is not difficult to see that there exists 8 0 α ,α ∈ [6]\L(V ) such that L(V ) ∩ L(V ) = {α ,α }. Thus L(V ) = {α ,α ,α } a b 0 8 2 a b 8 a b c and L(V ) = {α ,α ,α } such that L(V ) ∪ L(V ) = [6]\L[V ]. Now by Proposition 2 a b d 2 8 2 3.10 (d) still, L(V ) ⊆ [6]\L(V ) = L(V ) ∪ α , such that α ∈/ L(V ), α ∈ L(V ). 4 2 0 k k 0 k 2 L(V ) ⊆ [6]\L(V ) = L(V ) ∪ α for α ∈/ L(V ), α ∈/ L(V ), α 6= α . Thus, 0 8 0 j j 2 k 0 j k |L(V )∪{α ∪α }| = 5. By |L(V ∪L(V ))| = 6, and for all v ∈ V , and v ∈ V , 0 j k 6 4 6 6 4 4 d(v ,v ) = 2. Thus a contradiction and hence λ(G′) ≥ 7. 4 6 Conversely, we consider the 7−L(1,1)-labeling of C ×C in Figure 5 below. 10 6 0 6 1 2 6 0 1 2 7 5 7 2 3 5 0 3 2 4 0 1 4 0 1 7 4 7 5 1 5 3 6 3 2 0 6 1 2 6 0 Fig.7: 7−L(1,1)-labelingofC10×C6 (cid:3) Theorem 3.16. λ1(C ×C ) = 6 1 14 6 Proof. Since 14 is not a multiple of 4 and by Corollary 3.13, λ1(C ×C ) ≥ 6. 1 14 6 0 6 5 2 4 6 3 0 1 3 0 3 0 1 4 2 4 1 6 3 2 5 2 0 6 5 2 4 6 3 1 3 0 3 0 1 4 1 2 4 1 6 3 2 5 0 6 5 2 4 6 3 0 Fig.8: 5−L(1,1)-labelingofC14×C6 (cid:3) We can conclude that for C ×C , if m is even, and m 6≡ 0 mod 4, then λ1(C × m 6 1 m C ) ≥ 6. For m ≥ 14, this class of direct product graphs can be obtained from 6 C14+4m′, where m′ is a non-negative integer. Theorem 3.17. Form = 14+4m′, where m′ is a non-negativeinteger, λ1(C ×C ) = 1 m 6 6 LABELING PRODUCTS OF CYCLES 9 Proof. Since for any non-negative integer m′, m = 14 + 4m′ 6≡ 0 mod 4, then by 3.13, λ1(C ×C ) ≥ 6. By combining the labeling in Figure 8 and m′−multiple of 1 m n the labeling in Figure 3, we have that λ1(C ×C ) ≤ 6 and the result follows. (cid:3) 1 m n Note that the following result was established in [2] Theorem 3.18. Let m′,n′ ≡ 0mod10 and A = {12,14,16,18}. Then, for all k ∈ Aandm,n, λ11(Cm′ ×Ck+n′) = 5. Also, let m,n ≡ 0mod5, then λ11(Cm ×Cn) = 4. 4. Labeling of C ×C , n ≥ 8 m n In this section, we obtainthe λ1−numbers of graphproduct C ×C , where n,m ≥ 1 m n 8. Now we establish the λ1−number for C ×C . Since labeling of product graphs is 1 m 8 commutative, we restrict our work in this section to m ≥ 8 since the cases for smaller graphs have been taken care of in the last sections. The following result are helpful to reveal some useful properties of L(1,1)- labeling of C ×C . m 8 Lemma 4.1. Let G′ be a connected component of C × C , m ≥ 4. Suppose there m 8 exist v ,v ∈ V such that α = l(v ) = l(v ) ∈ [p], p ∈ N. then α ∈/ L(V ∪V ). a b i k a b k i+1 i+2 Furthermore, α ∈/ L(V ∪V ). k i−1 i−2 Proof. Claim: Let α = l(v ) = l(v ), v ,v ∈ V , V ⊆ V(G′). Then, d(v ,v ) = 4. k a b a b i i a b Reason: Clearly, |V | = 4 for i ∈ [m − 1]. Let V = {u v ,u v ,u v ,u v }. So, i i i 0 i 2 i 4 i 6 d(u v ,u v )=d(u v ,u v )=d(u v ,u v )=2. also, d(u v ,u v ) = 2 since C is a cy- i 0 i 2 i 2 i 4 i 4 i 6 i 6 i 4 8 cle. However, d(u v ,u v )=d(u v ,u v ) = 4. Thus v = v v and v = v v or i 0 i 4 i 2 i 6 a i 0 b i 4 v = v v and v = v v . Now, suppose v = u v and v = u v . Let V = a i 2 b i 6 a i 0 b i 4 i+1 {u v ,u v ,u v ,u v }. Then that d(u v ,u v ) = 1 = d(u v ,u v ) fol- i+1 1 I+1 3 i+1 5 i+1 7 i 0 i+1 1 i 0 i+1 7 lows from the definition of C ×C . Likewise, d(u v ,u v ) = 1 = d(u v ,u v ). m n=8 i 4 i+1 3 i 4 i+1 5 Therefore, α ∈ L(V ). Also, let V = {u v ,u v ,u v ,u v }. Then k i+1 i+2 i+2 0 i+2 2 i+2 4 i+2 6 d(u v ,u v ) = 2=d(u v ,u v ) and d(u v ,u v )=2=d(u v ,u v ) since C i 0 i+2 0 i 0 i+2 2 i 0 i+2 6 i 0 i+2 2 m=8 is a cycle. Now d(u v ,u v ) = 2 and therefore α ∈/ L(V ). This argument is i 4 i+2 2(4,6) k i+2 valid for V and V . (cid:3) i−1 1−2 The consequence of Lemma 4.1 is that if a label is assigned to two vertices on V ⊂ V(G′), then the label could no longer be assigned to another vertex on the i vertex sets two step above or below it. The next result is similar. Proposition 4.2. Suppose v ∈ V and v ∈ V such that α = l(v ) = l(v ), i i i+2 i+1 k i i+1 then, d(v ,v ) = 3. i i+1 Proof. Suppose v = u v without loss of generality, then d(v ,u v ) = 2. Now. i i 0 i i+1 1(7) d(v ,u v ) = 3 and d(v ,u v ) = 3. (cid:3) i i+1 3 i i+1 5 Lemma 4.3. Suppose V ,V ⊂ V(G′) where G′ is a connected component of C × 1 i+1 m C . Let α ∈ L(V )∩L(V ) then α ∈/ L(V )∪L(V ). 8 k i i+1 k i−1 i+2 10 T.C ADEFOKUNAND D.O. AJAYI Proof. By Proposition 4.2, suppose that α = l(u v ) and that α ∈ L(V ) then, k i j k i+1 α = l(u v ) or α = l(u v ). Without loss of generality, suppose that in fact, k i+1 j+3 k i+1 j+5 α = l(u v +5). Thend(u v ,u v ) = 2. Meanwhile, d(u v ,u v ) = k i+1 j i j i+2 j(j+2,j+6) i+1 j+3(j+5) i+2 j+4 1. Thus, α ∈/ L(V ). Similar argument holds for α ∈/ L(V ). (cid:3) k i+2 k i−1 By to Lemma 4.3, it is quite clear that if α belongs L(V ) ∩ L(V + 1), then α k i i k does not belong to L(V ∪Vi+2). A similar result is as follows: i−1 Lemma 4.4. Suppose α ∈ L(V )∩L(V ) ⊂ V(G′), where G′ is a connected com- k i i+2 ponent of C ×C , m ≥ 4, then α ∈/ L(V ) m 8 k i+1 Proof. Let,v ∈ V beu v . Notethat,d(u v ,u v ) = 2sinceC isacycle. Then, i i i 0 i 0 i+2 0(2,6) 8 the remaining vertex v ∈ V such that l(v ) = α is u v andd(u v ,u v ) = i+2 i+2 i+2 k i+2 4 i 0 i+2 4 4. Now d(v ,u v ) = 1 and d(u v ,u v ) = 1. Thus, α ∈/ L(V ). (cid:3) i i+1 1(7) i+2 4 i+1 3(5) k i+1 The consequence of Lemma 4.4 is that if two vertices on V and V share the same i i+2 label, then that label can not be shared by another vertex on V given that V ,V i+1 i i+1 and V are all in V(G′). i+2 Next we establish the lower bound of λ1(C ×C ) where m ≥ 8 and m ≡ 2 mod 6. 1 m 8 We require the following definition. Let G′ be a connected component of G. Then, V is the class of all vertices on αk V(G′) labeled α . k Lemma 4.5. For m ≥ 8,m ≡ 2 mod 6, λ1(C ×C ) ≥ 6. 1 m 8 Proof. Case 1: Let α ∈ L(V(G′)) such that if α ∈ L(V ), V ⊂ V(G′), i ∈ [m−1], k k i i then there exist v′,v′′ ∈ V such that l(v′) = α = l(v′′). Let V¯ be a class of all V ∈ i i i i k i i V(G′)suchthatα ∈ L(V ). Nowsuppose, without lossofgenerality, thatV ∈ V¯. By k i 0 thisandLemma 4.1, andbyassuming thatα labelsV(G′) optimally, suppose V ∈ V¯, k i then i ≡ 0 mod 3,i 6= m−2. Since m ≡ 2 mod 6, then there exists n′ ∈ N, such that m = 6n′+2. Note that m−5 = (6n′+2)−5 = 3(2n′−1). Thus, V ∈ V¯. By m−5 Lemma 4.1 and since V ,V ∈ V¯, then α ∈/ L(V ∪V ∪V ∪V ). Thus 0 m−5 k m−4 m−3 m−2 m−1 V¯ = {V ,··· ,V }. Set V¯′ = V¯\{V }. Since V¯′ = 2n′ −1, then |V′| = 2n′. Now, 0 m−5 0 (cid:12) (cid:12) |V(G′)| = (6n′ +2)4. Clearly |V | = 2(2n′) = 4(cid:12)n′.(cid:12)Hence, |V(G′)|=6n′+2 > 6. αk |Vαk| n′ Case 2: Suppose that for all triple V ,V V ⊂ V(G), α ∈ L(V ∩V ) and by i i+1 i+2 k i i+1 Lemma 4.3 α ∈/ V . Without loss of generality, we select the initial triple to be k i+2 V ,V ,V , such that α ∈ L(V ∩V ), α ∈/ V ; (and α ∈ L(V ∩V ),α ∈/ L(V )···). 0 1 2 k 0 1 k 2 k 3 4 k 5 Therefore, α ∈/ V for all i ∈ [m − 1] such that i + 1 ≡ 0 mod 3. Now, m ≡ 2 k i mod 6 implies there exists n′ ∈ N such that m = 6n′+2. Thus, m−2 ≡ 0 mod 3 and hence α ∈/ L(V ). Now, since α ∈ L(V ) ∩ L(V ), then α ∈/ L(V ) by k m−3 k 0 1 k m−1 Lemma 4.3 and since C is a cycle. By Lemma 4.4, it is possible for α ∈ L(V ) m k m−2 since α ∈/ L(V ). Thus we, for maximality, assume that α ∈ L(V ). Now, k m−3 k m−2 Let V¯ ={V ,··· ,V } ⊂ V(G′). Then V¯ = (m−3)+1 = 2(6n′) = 4n′, where αk 0 m−3 (cid:12) αk(cid:12) h 3 i 3 n′ ∈ N. Thus, for all V ∈ V(G′), |V |(cid:12) = (cid:12)4n′ + 1 since α ∈ L(V ). Thus i αk k m−2 |V(G′)| = (6n′+2)4 > 6. |Vαk| 4n′+1

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.