Loading [MathJax]/jax/output/SVG/jax.js
Research article Special Issues

On preemptive scheduling on unrelated machines using linear programming

  • We consider a basic preemptive scheduling problem where n non-simultaneously released jobs are to be processed by m unrelated parallel machines so as to minimize maximum job completion time. An optimal LP-solution has been used to construct an optimal preemptive schedule for simultaneously released jobs in time O(n3). We propose fast O(m) time algorithm that finds an optimal schedule in case the above LP-solution possesses "small enough" number of non-zero elements. We propose another linear program for non-simultaneously released jobs and show how an optimal schedule can be constructed also in time O(m) from the optimal solution to that linear program. Based on another stronger linear program formulation, we extend the earlier known schedule construction procedure for non-simultaneously released jobs. The procedure is important, in particular, because there may exist no optimal schedule that agrees with an optimal LP-solution. An optimal LP-solution imposes a number of preemptions, and additional preemptions may occur during the schedule construction process, a job might be forced to be split on the same machine. We show that if no job split is allowed, even a restricted version of the problem on three unrelated machines is NP-hard. As a result, we obtain that, given an optimal LP-solution, it is NP-hard to find an optimal schedule that agrees with that LP-solution. As another side result, we obtain that it is NP-hard to find an optimal schedule with at most m1 preemptions.

    Citation: Nodari Vakhania. On preemptive scheduling on unrelated machines using linear programming[J]. AIMS Mathematics, 2023, 8(3): 7061-7082. doi: 10.3934/math.2023356

    Related Papers:

    [1] Ali N. A. Koam, Adnan Khalil, Ali Ahmad, Muhammad Azeem . Cardinality bounds on subsets in the partition resolving set for complex convex polytope-like graph. AIMS Mathematics, 2024, 9(4): 10078-10094. doi: 10.3934/math.2024493
    [2] Ali N. A. Koam . Metric based resolvability of cycle related graphs. AIMS Mathematics, 2024, 9(4): 9911-9925. doi: 10.3934/math.2024485
    [3] Naila Mehreen, Rashid Farooq, Shehnaz Akhter . On partition dimension of fullerene graphs. AIMS Mathematics, 2018, 3(3): 343-352. doi: 10.3934/Math.2018.3.343
    [4] Suliman Khan, Sakander Hayat, Asad Khan, Muhammad Yasir Hayat Malik, Jinde Cao . Hamilton-connectedness and Hamilton-laceability of planar geometric graphs with applications. AIMS Mathematics, 2021, 6(4): 3947-3973. doi: 10.3934/math.2021235
    [5] Moussa Benoumhani . Restricted partitions and convex topologies. AIMS Mathematics, 2025, 10(4): 10187-10203. doi: 10.3934/math.2025464
    [6] Li Liu, Long Zhang, Huaxiang Zhang, Shuang Gao, Dongmei Liu, Tianshi Wang . A data partition strategy for dimension reduction. AIMS Mathematics, 2020, 5(5): 4702-4721. doi: 10.3934/math.2020301
    [7] Sakander Hayat, Bagus Imanda, Asad Khan, Mohammed J. F. Alenazi . Three infinite families of Hamilton-connected convex polytopes and their detour index. AIMS Mathematics, 2025, 10(5): 12343-12387. doi: 10.3934/math.2025559
    [8] Dalal Awadh Alrowaili, Uzma Ahmad, Saira Hameeed, Muhammad Javaid . Graphs with mixed metric dimension three and related algorithms. AIMS Mathematics, 2023, 8(7): 16708-16723. doi: 10.3934/math.2023854
    [9] Ahmed Alamer, Hassan Zafar, Muhammad Javaid . Study of modified prism networks via fractional metric dimension. AIMS Mathematics, 2023, 8(5): 10864-10886. doi: 10.3934/math.2023551
    [10] Jesús Gómez-Gardeñes, Ernesto Estrada . Network bipartitioning in the anti-communicability Euclidean space. AIMS Mathematics, 2021, 6(2): 1153-1174. doi: 10.3934/math.2021070
  • We consider a basic preemptive scheduling problem where n non-simultaneously released jobs are to be processed by m unrelated parallel machines so as to minimize maximum job completion time. An optimal LP-solution has been used to construct an optimal preemptive schedule for simultaneously released jobs in time O(n3). We propose fast O(m) time algorithm that finds an optimal schedule in case the above LP-solution possesses "small enough" number of non-zero elements. We propose another linear program for non-simultaneously released jobs and show how an optimal schedule can be constructed also in time O(m) from the optimal solution to that linear program. Based on another stronger linear program formulation, we extend the earlier known schedule construction procedure for non-simultaneously released jobs. The procedure is important, in particular, because there may exist no optimal schedule that agrees with an optimal LP-solution. An optimal LP-solution imposes a number of preemptions, and additional preemptions may occur during the schedule construction process, a job might be forced to be split on the same machine. We show that if no job split is allowed, even a restricted version of the problem on three unrelated machines is NP-hard. As a result, we obtain that, given an optimal LP-solution, it is NP-hard to find an optimal schedule that agrees with that LP-solution. As another side result, we obtain that it is NP-hard to find an optimal schedule with at most m1 preemptions.



    Let ψ be a simple, connected graph with vertex set V(ψ) and edge set E(ψ). The distance d(ρ1,ρ2), ρ1,ρ2V(ψ) is the length of shortest path between ρ1 and ρ2. Let Q={v1,v2,,vj} be an ordered set of vertices of ψ. Let ρ1V(ψ), the representations denoted by r(ρ1|Q) is the j-tuple distances as (d(ρ1|v1),d(ρ1|v2),,d(ρ1|vj)). If distinct vertices of ψ have distinct representation w.r.t. Q then Q is called the resolving set. The minimum number of j in the resolving set is known as the metric dimension of ψ and written as dim(ψ). Motivated by the problem of determining an intruder's location in a network in a unique way, Slater introduced the definition of metric dimension in [27] and later independently by Harary and Melter in [11]. The concept of resolving set, metric basis and metric dimension appeared in the literature [4,6,8,9,10,12,15,19,28,30,31].

    A partition of a set is collection of its subsets, no pair of which overlap, such that the union of all the subsets is the whole set and partition dimension is related to the partitioning of the vertex set V(Ω) and resolvability. The partition dimension is a generalized variant of matric dimension. Another type of dimension of a graph, is called partition dimension. Let Γ={Γ1,Γ2,Γj} and r(ρ1|Γ)={d(ρ1,Γ1),d(ρ1,Γ2),,d(ρ1,Γj)} are named as j-ordered partition of vertices and j-tuple representations respectively. If the representations of every ρ1 in V(ψ) w.r.t. Γ is different, then Γ is the resolving partition of the vertex set and the minimum count of the resolving partition set of V(ψ) is named as the partition dimension of ψ and it is represented by pd(ψ) [7]. The problem of determining the resolving set of a graph is NP-hard [20]. As, the problem of finding the partition dimension is a generalize version of metric dimension, therefore partition dimension is also a NP-complete problem. It is natural to think that there is a relation between metric and partition dimension, [7] proved for any non-trivial connected graph ψ,

    pd(ψ)dim(ψ)+1. (1.1)

    In [22], fullerene graph of chemical structure is discussed and proved that the graph has constant and bounded partition dimension. For more and interesting results on constant partition dimension can see [16,21,24]. To find the exact value of partition dimension of a graph is not easy therefore, various results on the bounds of the partition dimension are discussed in literature, such as the partition dimension of Cartesian product operation on different graphs are studies and provided extensive bounds on partition dimension [29]. In [1] different bounds of partition dimension of subdivision of different graphs are discussed. In [25,26] provide bounds of partition dimension of tree and uni-cyclic graphs in the form of subgraphs.

    The applications of partition resolving sets can be found in different fields such as robot navigation [19], Djokovic-Winkler relation [9], strategies for the mastermind game [10], network discovery and verification [5], in chemistry for representing chemical compounds [17,18] and in problems of pattern recognition and image processing, some of which involve the use of hierarchical data structures [23] for more applications see [6,11]. Following theorems are very helpful in finding the partition dimension of a graph.

    Theorem 1.1. [7] Let Γ be a resolving partition of V(ψ) and ρ1,ρ2V(ψ). If d(ρ1,z)=d(ρ2,z) for all vertices zV(ψ)(ρ1,ρ2), then ρ1,ρ2 belong to different classes of Γ.

    Theorem 1.2. [7] Let ψ be a simple and connected graph, then

    pd(ψ) is 2 iff ψ is a path graph

    pd(ψ) is n iff ψ is a complete graph,

    Let R be a family of connected graphs Gn:R=(Gn)n1, where |V(ψ)|=λ(n) and limnλ(n)=. If there exists a constant α1 such that pd(ψ)α,n1, then R has bounded partition dimension otherwise unbounded. Imran et al. [14] studied the metric dimension of Rpn, Dpn, and Qpn, convex polytopes which motivates us to find the partition dimension of same families of convex polytopes. In this paper, the partition dimension of same families of convex polytopes are studied. We determine the partition dimension of Rpn, in second section. In the third section, the partition dimension of the graph Dpn of a convex polytope with pendent edges is presented. The fourth section remains for the partition dimension of the graph Qpn.

    The convex polytope Rpn (p for pendant edges) is a planar graph and obtained from the convex polytope Rn defined in [13]. If we attach a pendant edge at each vertex of outer layer of Rn then we obtained a new planer graph Rpn as shown in Figure 1. The vertex set of Rpn, V(RPn)={V(Rn)}{xα:1αn} and edge set of Rpn, E(RPn)={E(Rn)}{wαxα:1αn}.

    Figure 1.  Convex polytope Rpn.

    For calculation, {uα:1αn} represents the inner cycle, the cycle induced by {vα:1αn} is interior cycle, exterior cycle containing {wα:1αn} set of vertices and pendant vertices named {xα:1αn}.

    Theorem 2.1. Let Rpn be a polytopes with n6. Then pd(Rpn)4.

    Proof. We splits the proof into following two cases.

    Case 1: When n=2β,β3,βN. We partition the vertices of Rpn into four partition resolving sets Θ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Rpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertex of Rpn have different representation w.r.t. resolving set Γ, then pd(Rpn)4. We give the representations of all vertices w.r.t. resolving partition set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If β+2α2β, then r(uβ|Γ)=(2βα+1,2βα+2,αβ1,0). There are no two vertices have same representation in inner cycle of Rpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(vβ|Γ)=(1,1,β,0). If 2αβ, then r(vβ|Γ)=(α,α1,βα+1,0). If α=β+1, then r(vβ|Γ)=(β,β,1,0). If β+2α2β, then r(vβ|Γ)=(2βα+1,2βα+2,αβ,0). There are also no two vertices have same representation in interior cycle of Rpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(wβ|Γ)=(2,2,β+1,0). If 2αβ+1, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+2, then r(wβ|Γ)=(β+1,β+1,2,0). If β+3α2β, then r(wβ|Γ)=(2βα+2,2βα+3,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Rpn. The representations of pendant vertices w.r.t. Γ are shown in Table 1. Again we can see that there are no two vertices have same representation of pendant vertices of Rpn.

    Table 1.  Representations of the pendant vertices w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    xαα=1 3 3 β+2 0
    xα2αβ α+2 α+1 βα+3 0
    xαα=β+1 β+2 β+2 3 0
    xαβ+2α2β 2βα+3 2βα+4 αβ+2 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Rpn have unique representation w.r.t. resolving partition Γ. Its means we can resolve the vertices of Rpn into four partition resolving sets, when n is even.

    Case 2: When n=2β+1,β3,βN. Again we resolve the vertices of Rpn into four partition resolving sets Γ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Rpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertices of Rpn have different representation w.r.t. resolving set Γ, then pd(Rpn)4. {We give the representations of all vertices Γ4 w.r.t. resolving set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If α=β+2, then r(uβ|Γ)=(β,β,1,0). If β+3α2β+1, then r(uβ|Γ)=(2βα+2,2βα+3,αβ1,0). There are no two vertices have same representation in inner cycle of Rpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(vβ|Γ)=(1,1,β,0). If 2αβ, then r(vβ|Γ)=(α,α1,βα+1,0). If α=β+1, then r(vβ|Γ)=(β+1,β,1,0). If β+2α2β+1, then r(vβ|Γ)=(2βα+2,2βα+3,αβ,0). There are also no two vertices have same representation in interior cycle of Rpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are: If α=1, then r(wβ|Γ)=(2,2,β+1,0). If 2αβ, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+1, then r(wβ|Γ)=(β+2,β+1,2,0). If β+2α2β+1, then r(wβ|Γ)=(2βα+3,2βα+4,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Rpn.

    The pendant vertices having the representations w.r.t. Γ shown in Table 2. Again we can see that there are no two vertices have same representation of pendant vertices of Rpn.

    Table 2.  Representations of the pendant vertices w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    xαα=1 3 3 β+2 0
    xα2αβ α+2 α+1 βα+3 0
    xαα=β+1 β+3 β+2 3 0
    xαβ+2α2β+1 2βα+4 2βα+5 αβ+2 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Rpn have unique representation w.r.t. resolving partition Γ. Its means we can also resolve the vertices of Rpn into four partition resolving sets, when n is odd.

    We note that from Case 1 and 2, there are no two vertices having the same representations implying that pd(Rpn)4.

    The convex polytope DPn is a planar graph and if we attach a pendant edge at each vertex of outer cycle of Dn [2] then we obtained a new plane graph DPn as shown in Figure 2. The vertex and edge set V(DPn)={V(Dn)}{yα:1αn}, E(DPn)={E(Dn)}{xαyα:1αn} are respectively. For calculation, {uα:1αn} represents the inner cycle, the cycle induced by {vα:1αn} is interior cycle, exterior cycle containing {wα:1αn} set of vertices, {xα:1αn} labeled as outer cycle and pendant vertices named for {yα:1αn}.

    Figure 2.  Convex polytope graph DPn.

    Theorem 3.1. Let DPn be a polytopes with n6. Then pd(DPn)4.

    Proof. We split the proof of above theorem into following two cases.

    Case 1: When n=2β,β3,βN. We partition the vertices of Dpn into four partition sets Γ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Dpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertices of Dpn have different representation w.r.t. resolving set Γ, then pd(Dpn)4. We give the representations of all vertices Γ4 w.r.t. resolving set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If β+2α2β, then r(uβ|Γ)=(2βα+1,2βα+2,αβ1,0). There are no two vertices have same representation in inner cycle of Dpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(vβ|Γ)=(1,2,β+1,0). If 2αβ, then r(vβ|Γ)=(α,α1,βα+2,0). If α=β+1, then r(vβ|Γ)=(β,β,1,0). If β+2α2β, then r(vβ|Γ)=(2βα+2,2βα+3,αβ,0). There are also no two vertices have same representation in interior cycle of Dpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(wβ|Γ)=(2,2,β+1,0). If 2αβ, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+1, then r(wβ|Γ)=(β+1,β+1,2,0). If β+2α2β, then r(wβ|Γ)=(2βα+2,2βα+3,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Dpn.

    The vertices on outer cycle and pendant vertices having the representations w.r.t. Γ as shown in Tables 3 and 4. Again we can see that there are no two vertices have same representation in outer cycle and pendant vertices of Dpn.

    Table 3.  Representations of the vertices on outer cycle w.r.t. resolving set Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    xαα=1 3 3 β+2 0
    xα2αβ α+2 α+1 βα+3 0
    xαα=β+1 β+2 β+2 3 0
    xαβ+2α2β 2βα+3 2βα+4 αβ+2 0

     | Show Table
    DownLoad: CSV
    Table 4.  Representations of the pendant vertices w.r.t. resolving set Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    yαα=1 4 4 β+3 0
    yα2αβ α+3 α+2 βα+4 0
    yαα=β+1 β+3 β+3 4 0
    yαβ+2α2β1 2βα+4 2βα+5 αβ+3 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Dpn have unique representation w.r.t. resolving partition Γ. Its means we can resolve the vertices of Dpn into four partition resolving sets, when n is even.

    Case 2: When n=2β+1,β3,βN. Again we resolve the vertices of Dpn into four partition resolving sets Γ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Dpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertices of Dpn have different representation w.r.t. resolving set Γ, then pd(Dpn)4. We give the representations of all vertices Γ4 w.r.t. resolving set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If α=β+2, then r(uβ|Γ)=(β,β,1,0). If β+3α2β+1, then r(uβ|Γ)=(2βα+1,2βα+2,αβ1,0). There are no two vertices have same representation in inner cycle of Dpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(vβ|Γ)=(1,2,β+1,0). If 2αβ+1, then r(vβ|Γ)=(α,α1,βα+2,0). If α=β+2, then r(vβ|Γ)=(β+1,β+1,2,0). If β+3α2β+1, then r(vβ|Γ)=(2βα+2,2βα+3,αβ,0). There are also no two vertices have same representation in interior cycle of Dpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If α=1, then r(wβ|Γ)=(2,2,β+1,0). If 2αβ, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+1, then r(wβ|Γ)=(β+2,β+1,2,0). If β+2α2β+1, then r(wβ|Γ)=(2βα+3,2βα+4,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Dpn.

    The vertices on outer cycle and pendant vertices having the representations w.r.t. Γ as shown in Tables 5 and 6. Again we can see that there are no two vertices have same representation in outer cycle and pendant vertices of Dpn.

    Table 5.  Representations of the vertices on exterior cycle w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    xαα=1 3 3 β+2 0
    xα2αβ α+2 α+1 βα+3 0
    xαα=β+1 β+2 β+2 3 0
    xαβ+2α2β+1 2βα+4 2βα+5 αβ+2 0

     | Show Table
    DownLoad: CSV
    Table 6.  Representations of the pendant vertices w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    xαα=1 4 4 β+3 0
    xα2αβ α+3 α+2 βα+4 0
    xαα=β+1 β+3 β+3 4 0
    xαβ+2α2β+1 2βα+5 2βα+6 αβ+3 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Dpn have unique representation w.r.t. resolving partition Γ. Its means we can also resolve the vertices of Dpn into four partition resolving sets, when n is odd.

    We note that from Case 1 and 2, there are no two vertices having the same representations implying that pd(Tpn)4.

    The convex polytope QPn is a planar graph and If we attach a pendant edge at each vertex of outer cycle of Qn [3] then we obtained a new plane graph QPn as shown in Figure 3. The vertex and edge set V(QPn)={V(αn)}{yα:1αn}, E(QPn)={E(Qn)}{xαyα:1αn} are respectively.

    Figure 3.  Convex polytope graph Qpn.

    For convenience, {uα:1αn} represents the inner cycle, the cycle induced by {vα:1αn} is interior cycle, exterior cycle containing {wα:1αn} set of vertices, {xα:1αn} are exterior vertices, and pendant vertices named for {yα:1αn}.

    Theorem 4.1. Let QPn be a polytopes with n6. Then pd(QPn)4.

    Proof. Case 1: When n=2β,β3,βN. We partition the vertices of Qpn into four partition resolving sets Γ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Qpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertices of Qpn have different representation w.r.t. resolving set Γ, then pd(Qpn)4. We give the representations of all vertices Γ4 w.r.t. resolving set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If β+2α2β, then r(uβ|Γ)=(2βα+1,2βα+2,αβ1,0). There are no two vertices have same representation in inner cycle of Qpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(1,2,α+1,0). If 2αβ, then r(vβ|Γ)=(α,α1,βα+2,0). If β+2α2β, then r(vβ|Γ)=(2βα+2,2βα+3,αβ,0). There are also no two vertices have same representation in interior cycle of Qpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(2,2,α+1,0). If 2αβ, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+1, then r(vβ|Γ)=(α+1,α+1,2,0). If β+2α2β, then r(wβ|Γ)=(2βα+2,2βα+3,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Qpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(3,3,α+2,0). If 2αβ, then r(wβ|Γ)=(α+2,α+1,βα+3,0). If α=β+1, then r(vβ|Γ)=(α+2,α+2,3,0). If β+2α2β, then r(wβ|Γ)=(2βα+3,2βα+4,αβ+2,0). Again there are no two vertices have same representation also in exterior cycle of Qpn.

    The pendant vertices having the representations w.r.t. Γ as shown in Table 7. Again we can see that there are no two vertices have same representation in pendant vertices of Qpn.

    Table 7.  Representations of pendant vertices w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    yαα=1 4 4 β+3 0
    yα2αβ α+3 α+2 βα+4 0
    yαα=β+1 β+3 β+3 4 0
    yαβ+2α2β 2βα+4 2βα+5 αβ+3 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Qpn have unique representation w.r.t. resolving partition Γ. Its means we can resolve the vertices of Qpn into four partition resolving sets, when n is even.

    Case 2: When n=2β+1,β3,βN. Again we resolve the vertices of Qpn into four partition resolving sets Γ={Γ1,Γ2,Γ3,Γ4} where Γ1={u1}, Γ2={u2}, Γ3={uβ+1} and Γ4={V(Qpn)|{Γ1,Γ2,Γ3}}. It suffice to show that if every vertices of Qpn have different representation w.r.t. resolving set Γ, then pd(Qpn)4. We give the representations of all vertices Γ4 w.r.t. resolving set Γ are following.

    The vertices on inner cycle having the representations w.r.t. Γ which are:

    If 3αβ, then r(uβ|Γ)=(α1,α2,βα+1,0). If α=β+2, then r(uβ|Γ)=(β,β,1,0). If β+3α2β+1, then r(uβ|Γ)=(2βα+1,2βα+2,αβ1,0). There are no two vertices have same representation in inner cycle of Qpn.

    The vertices on interior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(1,2,α+1,0). If 2αβ, then r(vβ|Γ)=(α,α1,βα+2,0). If α=β+2, then r(vβ|Γ)=(β+1,β+1,2,0). If β+3α2β+1, then r(vβ|Γ)=(2βα+2,2βα+3,αβ,0). There are also no two vertices have same representation in interior cycle of Qpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(2,2,α+1,0). If 2αβ, then r(wβ|Γ)=(α+1,α,βα+2,0). If α=β+1, then r(wβ|Γ)=(β+2,β+1,2,0). If β+2α2β+1, then r(wβ|Γ)=(2βα+3,2βα+4,αβ+1,0). Again there are no two vertices have same representation also in exterior cycle of Qpn.

    The vertices on exterior cycle having the representations w.r.t. Γ which are:

    If β=1, then r(vβ|Γ)=(3,3,α+2,0). If 2αβ, then r(wβ|Γ)=(α+2,α+1,βα+3,0). If α=β+1, then r(wβ|Γ)=(β+2,β+2,3,0). If β+2α2β+1, then r(wβ|Γ)=(2βα+4,2βα+5,αβ+2,0). Again there are no two vertices have same representation also in exterior cycle of Qpn.

    The pendant vertices having the representations w.r.t. Γ as shown in Table 8. Again we can see that there are no two vertices have same representation in pendant vertices of Qpn.

    Table 8.  Representations of the pendant vertices w.r.t. Γ.
    Representation Γ1 Γ2 Γ3 Γ4
    yαα=1 4 4 β+3 0
    yα2αβ α+3 α+2 βα+4 0
    yαα=β+1 β+4 β+3 4 0
    yαβ+2α2β+1 2βα+5 2βα+6 αβ+3 0

     | Show Table
    DownLoad: CSV

    It is easy to verify that all the vertices of Qpn have unique representation w.r.t. resolving partition Γ. Its means we can also resolve the vertices of Qpn into four partition resolving sets, when n is odd.

    We note that from Case 1 and 2, there are no two vertices having the same representations implying that pd(Upn)4.

    The core of the problem of the partition dimension is deciding the resolving partition set for a graph. In this paper, we have studies the partition dimension of some families of convex polytopes graph such as Rpn, Dpn and Qpn, which are obtained from the convex polytopes by adding a pendant edge at each vertex of outer cycle. In this research work, we have proved that partition dimension of these convex polytopes are bounded. Consequently, we propose the following open problems.

    Conjecture 5.1. The following equalities hold:

    pd(Rpn)=pd(Dpn)=pd(Qpn)=4

    The authors declare there is no conflict of interest.



    [1] T. Gonzalez, S. Sahni, Preemptive scheduling of uniform processor systems, J. Assoc. Comput. Mach., 25 (1978), 92–101. https://doi.org/10.1145/322047.322055 doi: 10.1145/322047.322055
    [2] J. Labetoulle, E. L. Lawler, J. K. Lenstra, A. H. G. Rinnooy Kan, Preemptive scheduling of uniform machines subject to release dates, In: Progress in combinatorial optimization, New York: Academic Press, 245–261, 1984. https://doi.org/10.1016/B978-0-12-566780-7.50020-9
    [3] E. L. Lawler, J. Labetoulle, On preemptive scheduling of unrelated parallel processors by linear programming, J. Assoc. Comput. Mach., 25 (1978), 612–619. https://doi.org/10.1145/322092.322101 doi: 10.1145/322092.322101
    [4] C. N. Potts, Analysis of a linear programming heuristic for scheduling unrelated parallel machines, Discrete Appl. Math., 10 (1985), 155–164. https://doi.org/10.1016/0166-218X(85)90009-5 doi: 10.1016/0166-218X(85)90009-5
    [5] J. K. Lenstra, D. B. Shmoys, E. Tardos, Approximation algorithms for scheduling unrelated parallel machines, Math. Program., 46 (1990), 259–271. https://doi.org/10.1007/BF01585745 doi: 10.1007/BF01585745
    [6] E. Shchepin, N. Vakhania, An optimal rounding gives a better approximation for scheduling unrelated machines, Oper. Res. Lett., 33 (2005), 127–133. https://doi.org/10.1016/j.orl.2004.05.004 doi: 10.1016/j.orl.2004.05.004
    [7] T. Gonzalez, S. Sahni, Open shop scheduling to minimize finish time, J. Assoc. Comput. Mach., 23 (1976), 665–679. https://doi.org/10.1145/321978.321985 doi: 10.1145/321978.321985
    [8] T. Gonzalez, E. L. Lawler, S. Sahni, Optimal preemptive scheduling of two unrelated processors, ORSA J. Comput., 2 (1990), 209–301. https://doi.org/10.1287/ijoc.2.3.219 doi: 10.1287/ijoc.2.3.219
    [9] E. Shchepin, N. Vakhania, On the geometry, preemptions and complexity of multiprocessor and open shop scheduling, Ann. Oper. Res., 159 (2008), 183–213. https://doi.org/10.1007/s10479-007-0266-1 doi: 10.1007/s10479-007-0266-1
    [10] M. Foumani, K. Smith-Miles, The impact of various carbon reduction policies on green flowshop scheduling, Appl. Energ., 249 (2019), 300–315. https://doi.org/10.1016/j.apenergy.2019.04.155 doi: 10.1016/j.apenergy.2019.04.155
    [11] G. L. Gong, Q. W. Deng, X. R. Gong, D. Huang, A non-dominated ensemble fitness ranking algorithm for multi-objective flexible job-shop scheduling problem considering worker flexibility and green factors, Knowl. Based Syst., 231 (2021), 107430. https://doi.org/10.1016/j.knosys.2021.107430 doi: 10.1016/j.knosys.2021.107430
  • This article has been cited by:

    1. Xiujun Zhang, Muhammad Salman, Anam Rani, Rashna Tanveer, Usman Ali, Zehui Shao, Metric Identification of Vertices in Polygonal Cacti, 2023, 136, 1526-1506, 883, 10.32604/cmes.2023.025162
    2. Kamran Azhar, Sohail Zafar, Agha Kashif, Amer Aljaedi, Umar Albalawi, The Application of Fault-Tolerant Partition Resolvability in Cycle-Related Graphs, 2022, 12, 2076-3417, 9558, 10.3390/app12199558
    3. Wajdi Alghamdi, Muhammad Ahsan Asim, Akbar Ali, On the Bounded Partition Dimension of Some Generalised Graph Structures, 2022, 2022, 2314-4785, 1, 10.1155/2022/9531182
    4. Ali Al Khabyah, Ali N. A. Koam, Ali Ahmad, Niansheng Tang, Partition Resolvability of Nanosheet and Nanotube Derived from Octagonal Grid, 2024, 2024, 2314-4785, 1, 10.1155/2024/6222086
    5. Syed Waqas Shah, Muhammad Yasin Khan, Gohar Ali, Irfan Nurhidayat, Soubhagya Kumar Sahoo, Homan Emadifar, Ram Jiwari, On Partition Dimension of Generalized Convex Polytopes, 2023, 2023, 2314-4785, 1, 10.1155/2023/4412591
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1888) PDF downloads(78) Cited by(5)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog