Optimizing Task Allocation for Edge Micro-Clusters in Smart Cities

Alhaizaey, Yousef; Singer, Jeremy; Michala, Anna Lito

doi:10.1109/wowmom51794.2021.00062

Cited by 4 publications

(8 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The final work categorized here was proposed by Alhaizaey et al [ 53 ] and is based on Mixed-Integer Programming, which is an extension of Linear Programming including binary constraints. In the proposed approach, the binary variables are used to assign tasks to the nodes.…”

Section: State Of the Art On The Task Allocation Problem For Iotmentioning

confidence: 99%

Survey on Multi-Objective Task Allocation Algorithms for IoT Networks

Weikert

Steup

Mostaghim

2022

Sensors

View full text Add to dashboard Cite

The Internet of Things (IoT) has been an area of growing research interest for decades. Task allocation is an important problem for the optimized operation of Internet-of-Things networks. This paper provides an overview of recent research in the field of Internet-of-Things task allocation optimization. First, the task allocation problem for the IoT itself is analyzed and divided into distinct sub-problem categories, such as deployment optimization, static or dynamic optimization as well as single- or multi-objective optimization. Following that, the commonly used optimization objectives are explained. Various recent works in the field of task allocation optimization are then summarized and catalogued according to the problem categories. Finally, the paper concludes with a qualitative analysis of the categorized approaches and a description of open problems and highlights promising directions for future research.

show abstract

Section: State Of the Art On The Task Allocation Problem For Iotmentioning

confidence: 99%

Survey on Multi-Objective Task Allocation Algorithms for IoT Networks

Weikert

Steup

Mostaghim

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…T: denotes the Makespan Time m n : is a constant value representing the gradient of node n numT ask n : is the total number of tasks allocated to node n c n : is a constant value representing y-intercept of node n Similar to our earlier work [6], two constraints are introduced to model the micro-cluster heterogeneity and nodes capacities. First, we defined node capacity constraint (Equation 2), in which each node n has a defined capacity and should not accept more than its defined capacity.…”

Section: B System Objective Functionmentioning

confidence: 99%

“…Table V presents illustration examples of different particles and the allocation representation. For example, in solution 1, the particle is represented as array [7,7,7,4,7,5,7,3,7,6]. In our representation, tasks t1, t2, t3, t5, t7, t9 are allocated to node n7; task t4 is allocated to node n4, task t6 is allocated to node n5, and etc.…”

Section: Task Allocation Using Particle Swarm Optimisationmentioning

confidence: 99%

“…This paper expands on our previous work on optimising task allocation for edge micro-clusters [6], through proposing a linear-model objective function and evaluating metaheuristic Particle Swarm Optimisation (PSO) [7]. PSO is a wellknown optimisation technique used in many scientific areas, especially in cloud data centres to solve various resource management problems [8].…”

Section: Introductionmentioning

confidence: 97%

“…Incrementing to the micro-cluster setup configured in our earlier work [6], this micro-cluster testbed features several heterogeneous nodes to model the system. The current version of our micro-cluster comprises eight Raspberry Pi devices drawn from different generations to model the system's heterogeneity.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Optimizing Heterogeneous Task Allocation for Edge Compute Micro Clusters Using PSO Metaheuristic

Alhaizaey

Singer

Michala

2022

2022 Seventh International Conference on Fog and Mobile Edge Computing (FMEC)

Self Cite

View full text Add to dashboard Cite

Optimised task allocation is essential for efficient and effective edge computing; however, task allocation differs in edge systems compared to the powerful centralised cloud data centres, given the limited resource capacities in edge and the strict QoS requirements of many innovative Internet of Things (IoT) applications. This paper aims to optimise heterogeneous task allocation specifically for edge micro-cluster platforms. We extend our previous work on optimising task allocation for micro-clusters by presenting a linear-based model and propose a metaheuristic Particle Swarm Optimisation (PSO) technique to minimise the makespan time and the allocation overhead time of heterogeneous workloads in batch execution. We present a comparative performance evaluation of metaheuristic PSO, mixed-integer programming (MIP) and randomised allocation based on the computation overhead time and the quality of the solutions. Our results show a crossover implying that mixedinteger programming is efficient for small-scale clusters, whereas PSO scales better and provides near-optimal solutions for largerscale micro-clusters.

show abstract