Joint Communication, Computing, and Caching Resource Allocation in LEO Satellite MEC Networks

Abstract

Driven by the urgent requirement of ubiquitous and reliable coverage for global users, the low earth orbit (LEO) satellite network has attracted numerous attentions from the academic and industry circles. By deploying multi-access edge computing (MEC) servers in LEO satellites, computation offloading and content caching services can be provided for remote Internet-of-Things (IoT) devices without proximal servers. In this paper, the joint optimization of computation offloading, radio resource allocation and caching placement in LEO satellite MEC networks are investigated. The problem is formulated to minimize the total delay of all ground IoT devices while ensuring the energy, computing and caching constraints. To solve this mixed-integer and non-convex problem, a Lagrange dual decomposition (LDD)-based algorithm is proposed to obtain the closed-form optimal solution. Then, a heuristic algorithm is proposed to further reduce the computation complexity. Numerical results validate that both algorithms are effective compared to the optimal exhaustive search, the full local computing and the full MEC methods. Besides, the offloading ratio and the average delay of all IoT devices with different numbers and computing capacities of devices and satellites are also demonstrated.INDEX TERMS LEO satellites, multi-access edge computing, computation offloading, resource allocation, caching.

A comprehensive review on internet of things task offloading in multi-access edge computing

“…Over 40 % of the literature surveyed in this survey focused on task allocation as the primary problem-solving goal [ 73 , 74 , 77 , 79 , 81 , [84] , [85] , [86] , 88 , [138] , [139] , [140] , 143 , 145 , [90] , [91] , [147] , 94 , [96] , [97] , [98] , 100 , 104 , 106 , [108] , [109] , [110] , 125 , 130 , 131 , 137 ]. The goal of resource allocation is addressed in approximately 10 % of the reviewed literature [ 99 , 103 , 105 , 111 , 116 , 121 ].…”

“…An algorithm based on Lagrange Dual Decomposition (LDD) and a heuristic algorithm is proposed in Ref. [ 100 ], which solves the IoT task offloading optimization problem in MEC networks built based on low Earth orbit (LEO) satellite networks. The algorithm jointly optimizes task offloading decisions and allocation of limited available resources.…”

“… Ref. Problems addressed Technique Network Objective Evaluation tools Baseline Advantages Limitations [ 96 ] Task assignment BAT WAVE Minimize latency Python based implementation Classic greedy algorithm Multi-hop Lack of MEC available resources to consider [ 97 ] Task assignment Greedy winner selection Undefined Minimize latency, Maximize success rate EdgeCloudSim, ElasticSim Nearest, Selfish Support dependencies between tasks Lack of network feature support [ 98 ] Task assignment, resource allocation Greedy, gradient decent 5G, WiFi, and ZigBee Minimize switching latency, maximize QoS Mininet JONSSPE, SDENTO, JONSSPE Probabilistic guarantees of offloading Low resource utilization [ 99 ] Task assignment Greedy, game theory AAN Maximize total IoT data computed Matlab Exhaustive searching Support offloading to TD and MEC Channel utilization is not considered [ 100 ] Task assignment, resource allocation Greedy, lagrange dual decomposition Low-earth orbit satellite networks Minimize total delay Matlab Optimal exhaustive search …”

“… Ref. Problems addressed Technique Network Objective Evaluation tools Baseline Advantages Limitations [ 98 ] Task assignment SVM RAN Minimize service delay, computation time, and service lag OPNET IHRA, COM, PCOA Acceptable framework Difficult to implement on large-scale training samples [ 99 ] Resource allocation CNN, RNN, LSTM 5G Maximize accuracy and detection rate Python based implementation RP, DE Predict task offloading timeliness Lack of comprehensive consideration of network parameters [ 100 ] Task assignment, resource allocation DNN WPT Maximize efficiency of the edge server Python based implementation DROO, C D, OP, KNN Support task offloading between TDs If the MDs have a high mobility, the framework could be difficult to converge. [ 101 ] Task assignment, resource allocation DNN RAN Maximize system utility and bandwidth allocation for each MD Python based implementation Local-Only, Edge-Only, Central-Only, Local and Central Acceptable framework Lack of comprehensive consideration of network parameters [ 102 ] Task assignment, resource allocation DNN OFDMA, SDN Minimize delay and energy consumption Python based...…”

“…In [ 100 ], a parallel offloading framework is proposed that uses deep neural networks as parallel offloading executors to generate offloading actions. This framework decomposes the optimization problem into offloading decision sub-problems and resource allocation sub-problems.…”