Deep reinforcement learning based computation offloading and resource allocation for MEC
The DOI matches its Crossref record; 17 of the 17 references it cites were found in a registry or an index.
Abstract
Mobile edge computing (MEC) has the potential to enable computation-intensive applications in 5G networks. MEC can extend the computational capacity at the edge of wireless networks by migrating the computation-intensive tasks to the MEC server. In this paper, we consider a multi-user MEC system, where multiple user equipments (UEs) can perform computation offloading via wireless channels to an MEC server. We formulate the sum cost of delay and energy consumptions for all UEs as our optimization objective. In order to minimize the sum cost of the considered MEC system, we jointly optimize the offloading decision and computational resource allocation. However, it is challenging to obtain an optimal policy in such a dynamic system. Besides immediate reward, Reinforcement Learning (RL) also takes a long-term goal into consideration, which is very important to a time-variant dynamic systems, such as our considered multi-user wireless MEC system. To this end, we propose RL-based optimization framework to tackle the resource allocation in wireless MEC. Specifically, the Q-learning based and Deep Reinforcement Learning (DRL) based schemes are proposed, respectively. Simulation results show that the proposed scheme achieves significant reduction on the sum cost compared to other baselines.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Mobile Edge Computing: A Survey on Architecture and Computation Offloading
Mach, P., & Becvar, Z. (2017) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2017.2682318 · Cited by 3,258 · cited 1×
“Therefore, the computation offloading and computational resource allocation themes have attracted great interests as a key point in the MEC system [3].”
BackgroundEfficient multi-user computation offloading for mobile-edge cloud computing
Chen, X., Jiao, L., Li, W., & Fu, X. (2016) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2015.2487344 · Cited by 2,675 · cited 1×
“But according to [14], [15], the download data rate is very high in general, and the data size of the result is much smaller than that of input data, so the delay and energy consumptions of this step are neglected in the rest of this paper.”
BackgroundCloud computing for mobile users: can offloading computation save energy?
Kumar, K., & Lu, Y.-H. (2010) · Computer · doi:10.1109/mc.2010.98 · Cited by 1,493 · cited 1×
“The work in [5] proposed general guidelines to make offloading decision for energy consumption minimization, where the communication links were easily assumed to have a fixed rate.”
BackgroundSelf-citationEnergy-Efficient Offloading for Mobile Edge Computing in 5G Heterogeneous Networks
Zhang, K., Mao, Y., Leng, S., Zhao, Q., Li, L., Peng, X., Pan, L., Maharjan, S., & Zhang, Y. (2016) · IEEE Access · doi:10.1109/access.2016.2597169 · Cited by 855 · cited 1×
“But according to [14], [15], the download data rate is very high in general, and the data size of the result is much smaller than that of input data, so the delay and energy consumptions of this step are neglected in the rest of this paper.”
BackgroundDelay-optimal computation task scheduling for mobile-edge computing systems
Liu, J., Mao, Y., Zhang, J., & Letaief, K. B. (2016) · 2016 IEEE International Symposium on Information Theory (ISIT) · doi:10.1109/isit.2016.7541539 · Cited by 759 · cited 1×
“The work in [8] designed the single-user delay-optimal task scheduling policies based on the theory of MDP, which controlled the states of the local processing and transmission units and the task buffer queue length based on the channel state.”
MethodJaszkiewicz, A. (2002) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2002.802873 · Cited by 474
Communicating While Computing: Distributed mobile cloud computing over 5G heterogeneous networks
Barbarossa, S., Sardellitti, S., & Di Lorenzo, P. (2014) · IEEE Signal Processing Magazine · doi:10.1109/msp.2014.2334709 · Cited by 468 · cited 2×
“Since the data rates for wireless communications are not constant, the optimal binary computation offloading decision using convex optimization was proposed in [6], where the the famous Shannon-Hartley formula gave the power-rate function.”
BackgroundMobile-edge computing architecture: the role of mec in the internet of things
Sabella, D., Vaillant, A., Kuure, P., Rauschenbach, U., & Giust, F. (2016) · IEEE Consumer Electronics Magazine · doi:10.1109/mce.2016.2590118 · Cited by 413 · cited 1×
“MEC is a new paradigm in C-RAN, which can enhance the computation capacity at the edge of mobile networks by deploying high-performance servers [2].”
BackgroundEnergy-optimal mobile application execution: Taming resource-poor mobile devices with cloud clones
Wen, Y., Zhang, W., & Luo, H. (2012) · 2012 Proceedings IEEE INFOCOM · doi:10.1109/infcom.2012.6195685 · Cited by 394 · cited 1×
“We set zn = 10−27 ( f l n )2 according to the practical measurement in [13].”
MethodPeng, M., Sun, Y., Li, X., Mao, Z., & Wang, C. (2016) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2016.2548658 · Cited by 386
Cited by
Items 101–200 of 560
Mustafa, E., Shuja, J., Bilal, K., Mustafa, S., Maqsood, T., Rehman, F. U., & Khan, A. U. R. (2022) · Cluster Computing · doi:10.1007/s10586-022-03700-5 · Cited by 55
“In [23], researchers considered a reinforcement approach to formulate the cost of delay and energy consumption where multiple wireless devices off-load tasks through wireless channels.”
BackgroundLi, H., Xiong, K., Lu, Y., Chen, W., Fan, P., & Letaief, K. B. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3496536 · Cited by 57
“The cell radius is set to 200 m [37], [38], and the UDs are randomly located in each cell.”
Self-citationShan, N., Cui, X., & Gao, Z. (2020) · Computer Communications · doi:10.1016/j.comcom.2020.05.037 · Cited by 56
Energy-aware task offloading with deadline constraint in mobile edge computing
Li, Z., Chang, V. I. C., Ge, J., Pan, L., Hu, H., & Huang, B. (2021) · EURASIP Journal on Wireless Communications and Networking · doi:10.1186/s13638-021-01941-3 · Cited by 59
OSTTD: Offloading of Splittable Tasks With Topological Dependence in Multi-Tier Computing Networks
Zhang, R., Chu, X., Ma, R., Zhang, M., Lin, L., Gao, H., & Guan, H. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3227023 · Cited by 55
“In [25], an RL-based method is used to jointly optimize the offloading decision and computational resource allocation with the objective of minimizing the sum of delay and energy consumption.”
MethodSelf-citationJoint offloading decision and resource allocation for mobile edge computing enabled networks
Liao, Y., Shou, L., Yu, Q. G., Ai, Q., & Liu, Q. (2020) · Computer Communications · doi:10.1016/j.comcom.2020.02.071 · Cited by 53
Nduwayezu, M., Pham, Q., & Hwang, W. (2020) · IEEE Access · doi:10.1109/access.2020.2997925 · Cited by 54
“in [21] dedicated their efforts to design a reinforcement learning (RL)-based MEC computation offloading system as a replacement of Markov decision process (MDP).”
BackgroundArtificial Intelligence: An Energy Efficiency Tool for Enhanced High performance computing
Kelechi, A. H., Alsharif, M. H., Bameyi, O. J., Ezra, P. J., Joseph, I. K., Atayero, A. A., Geem, Z. W., & Hong, J. (2020) · Symmetry · doi:10.3390/sym12061029 · Cited by 63
“In an unknown, potentially complex environment, the agent learns how to reach a goal [86,87].”
Model-Free Ultra Reliable Low Latency Communication (URLLC): A Deep Reinforcement Learning Framework
Kasgari, A. T. Z., & Saad, W. (2019) · International Conference on Communications · doi:10.1109/icc.2019.8761721 · Cited by 55
“Moreover, since these deepRL works [14], [16] limited their problem’s action space, used discretization to limit the size of the action space [5], or did not address the limitation of deep-RL when dealing with large action spaces, they are not suitable to solve the problem of URLLC resource allocation with rate constr…”
BackgroundMethodInfluentialJoint Offloading and Resource Allocation Using Deep Reinforcement Learning in Mobile Edge Computing
Zhang, X., Zhang, X., & Yang, W. (2022) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2022.3184642 · Cited by 48
“Inspired by recent works on DRL-based offloading for MEC systems [20]–[22], we propose an energy-efficient DRL framework (EEDRL) based on an actor-critic network architecture for partial off-loading, where the actor network learns the mapping between wireless channel gains to solve the offloading…”
MethodRegional Intelligent Resource Allocation in Mobile Edge Computing Based Vehicular Network
Wang, G., & Xu, F. (2020) · IEEE Access · doi:10.1109/access.2020.2964018 · Cited by 50
“Li et al. in [6] proposed a multi-user MEC system, where multiple user equipments realized the computational offloading via wireless channels to a MEC server.”
BackgroundDistributed Learning for Wireless Communications: Methods, Applications and Challenges
Qian, L., Yang, P., Xiao, M., Dobre, O. A., Renzo, M. D., Li, J., Han, Z., Qin, Y., & Zhao, J. (2022) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2022.3156756 · Cited by 60
Self-citationSamanta, R. K., Sadhukhan, B., Samaddar, H., Sarkar, S., Koner, C., & Ghosh, M. (2022) · CAAI Transactions on Intelligence Technology · doi:10.1049/cit2.12114 · Cited by 53
Intelligent Offloading in Blockchain-Based Mobile Crowdsensing Using Deep Reinforcement Learning
Chen, Z., & Yu, Z. (2023) · IEEE Communications Magazine · doi:10.1109/mcom.001.2200223 · Cited by 47
“We consider a mobile edge cloud consisting of one edge server and one AP [9], where the radius of the network coverage is 250 meters and the bandwidth is 20 Mbps.”
BackgroundMethodElgendy, I. A., Muthanna, A., Hammoudeh, M. A. A., Shaiba, H. A., Ünal, D., & Khayyat, M. M. (2021) · Big Data · doi:10.1089/big.2020.0284 · Cited by 52
“35 Therefore, DQN is considered to handle the Q-learning limitation through estimating the Q-value function instead of storing the Q-table, as shown in the next subsection.”
Deep Reinforcement Learning based Path Planning for UAV-assisted Edge Computing Networks
Peng, Y., Liu, Y. K., & Zhang, H. (2021) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc49053.2021.9417292 · Cited by 52
“Afterward, the DRL has been employed to the design of online resource allocation and scheduling in wireless networks [7]–[9].”
BackgroundMethodMahbub, M., & Shubair, R. M. (2023) · Journal of Network and Computer Applications · doi:10.1016/j.jnca.2023.103726 · Cited by 56
Leveraging the edge and cloud for V2X-based real-time object detection in autonomous driving
Hawlader, F., Robinet, F., & Frank, R. (2023) · Computer Communications · doi:10.1016/j.comcom.2023.11.025 · Cited by 54
Chen, Z., Zhang, J., Huang, Z., Wang, P., Yu, Z., & Miao, W. (2023) · Future Generation Computer Systems · doi:10.1016/j.future.2023.12.004 · Cited by 47
Li, Y., Huynh, D. V., Do‐Duy, T., Garcia‐Palacios, E., & Duong, T. Q. (2022) · IET Signal Processing · doi:10.1049/sil2.12128 · Cited by 49
“[33], the author proposes a Reinforcement Learning (RL)‐based optimisation framework to minimise the cost of delay and energy consumptions for user equipment in a time‐ variant dynamic MEC system.”
BackgroundArtificial Intelligence Empowered UAVs Data Offloading in Mobile Edge Computing
Fragkos, G., Kemp, N., Tsiropoulou, E. E., & Papavassiliou, S. (2020) · International Conference on Communications · doi:10.1109/icc40277.2020.9149115 · Cited by 45
“A reinforcement learning approach is introduced in [7] to jointly optimize the users’ offloading decision and the computational resource allocation towards minimizing the sum cost of users’ delay and energy consumption.”
MethodXue, F., Hai, Q., Dong, T., Cui, Z., & Gong, Y. (2022) · Information Sciences · doi:10.1016/j.ins.2022.06.078 · Cited by 50
Wang, J., Wang, L., Zhu, K., & Dai, P. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3382242 · Cited by 43
Multi-user Multi-channel Computation Offloading and Resource Allocation for Mobile Edge Computing
Nath, S., Li, Y., Wu, J., & Fan, P. (2020) · International Conference on Communications · doi:10.1109/icc40277.2020.9149124 · Cited by 46
“The works in [7], [8], [11] adopts the Deep Q-Network (DQN) method [12], while Deep Deterministic Policy Gradient (DDPG) algorithm [13] is utilized in [9].”
BackgroundMethodResultInfluentialImpatient Queuing for Intelligent Task Offloading in Multiaccess Edge Computing
Han, B., Sciancalepore, V., Xu, Y., Feng, D., & Schotten, H. D. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3191287 · Cited by 46
Machine learning-based computation offloading in edge and fog: a systematic review
Taheri-abed, S., Moghadam, A. M. E., & Rezvani, M. H. (2023) · Cluster Computing · doi:10.1007/s10586-023-04100-z · Cited by 45
“[140] proposed a combination of Q-learning and DRL to achieve MEC’s best offload policy and resource allocation.”
MethodXie, M., Ye, J. D., Zhang, G., & Ni, X. (2024) · Computer Networks · doi:10.1016/j.comnet.2024.110564 · Cited by 42
DQN-enabled content caching and quantum ant colony-based computation offloading in MEC
Li, C., Zhang, Y., & Luo, Y. (2022) · Applied Soft Computing · doi:10.1016/j.asoc.2022.109900 · Cited by 44
Distributed Intelligence in Wireless Networks
Liu, X., Yu, J., Liu, Y., Gao, Y., Mahmoodi, T., Lambotharan, S., & Tsang, D. H. K. (2023) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2023.3265425 · Cited by 48
“Besides, the joint computation offloading and resource allocation problem has also been formulated and solved with DRL algorithms in recent studies [53], [121], [122].”
BackgroundMethodInfluentialValentino, R., Jung, W.-S., & Ko, Y. (2018) · Sensors · doi:10.3390/s18113751 · Cited by 41
“[19] aim to optimize the offloading decision and computational resource allocation for multiple user equipments (UEs) that are trying to access the offloading service of a MEC server.”
BackgroundInfluentialXu, S., Liu, Q., Gong, B., Qi, F., Guo, S., Qiu, X., & Yang, C. (2020) · IEEE Internet of Things Journal · doi:10.1109/jiot.2020.3002427 · Cited by 46
Chen, J., & Wu, Z. (2021) · IEEE Communications Letters · doi:10.1109/lcomm.2021.3094842 · Cited by 41
Cai, J., Fu, H., & Liu, Y. (2022) · International Journal of Intelligent Systems · doi:10.1002/int.22841 · Cited by 41
“Li et al.(22) proposed a task offloading and resource allocation method based on reinforcement learning to reduce task completion latency and device energy consumption.”
MethodAdaptive service function chaining mappings in 5G using deep Q-learning
Li, G., Feng, B., Zhou, H., Zhang, Y., Sood, K., & Yu, S. (2020) · Computer Communications · doi:10.1016/j.comcom.2020.01.035 · Cited by 41
Collaborative Edge Computing and Caching With Deep Reinforcement Learning Decision Agents
Ren, J., Wang, H., Hou, T., Zheng, S., & Tang, C. (2020) · IEEE Access · doi:10.1109/access.2020.3007002 · Cited by 45
“[16] proposed an RL-based optimization framework to solve the resource allocation problem in wireless MEC.”
MethodYang, C., Xu, X., Zhou, X., & Qi, L. (2022) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3548687 · Cited by 43
“Cai et al. [40] employed a distributed algorithm based on multi-agent DRL to handle a hybrid problem of data sensing and task offloading in a mobile crowd-sensing system, which achieved outstanding performance in reducing time delay and energy consumption and improving task completion rate.”
BackgroundInfluentialTask Offloading Based on Lyapunov Optimization for MEC-Assisted Vehicular Platooning Networks
Cui, T., Hu, Y., Shen, B., & Chen, Q. (2019) · Sensors · doi:10.3390/s19224974 · Cited by 40
Q‐learning based flexible task scheduling in a global view for the Internet of Things
Ge, J., Liu, B., Wang, T., Yang, Q., Liu, A., & Li, A. (2020) · Transactions on Emerging Telecommunications Technologies · doi:10.1002/ett.4111 · Cited by 43
“19-21 Task offload is proposed for the sensing devices with weak computing power and limited energy in IoT.”
BackgroundMethodInfluentialJoint Computation Offloading and Radio Resource Allocations in Small-Cell Wireless Cellular Networks
Chen, H., Zhao, D., Chen, Q., & Chai, R. (2020) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2020.2976932 · Cited by 39
“A similar objective is considered in [19], where the sum cost of delay and energy consumptions for all MDs is minimized by using deep reinforcement learning that jointly considers offloading decisions and computational resource allocations.”
BackgroundIslam, S., Badsha, S., Khalil, I., Atiquzzaman, M., & Konstantinou, C. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3172936 · Cited by 43
“Moreover, in [24] and [20] DRL-based offloading scheme has been proposed for vehicular edge network and mobile edge devices, respectively.”
BackgroundInfluentialChen, X., Dai, W., Ni, W., Wang, X., Zhang, S., Xu, S., & Sun, Y. (2023) · IEEE Transactions on Communications · doi:10.1109/tcomm.2023.3251353 · Cited by 38
“The same technique was applied to minimize the total cost of latency and power consumptions in dynamic multiuser MEC networks in [23].”
BackgroundMethod6G-Empowered Offloading for Realtime Applications in Multi-Access Edge Computing
Huang, H., Ye, Q., & Zhou, Y.-T. (2022) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2022.3188921 · Cited by 39
“Recently, many reinforcement learning-based offloading methods, which can learn an offloading model through realtime interactions with the external environment, have been proposed [10] [25][26] [27][28] [29][30] [31].”
BackgroundIntelligentCrowd: Mobile Crowdsensing via Multi-Agent Reinforcement Learning
Chen, Y., & Wang, H. (2020) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2020.3042244 · Cited by 37
Zhao, T., Li, F., & He, L. (2023) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2023.3280314 · Cited by 37
“Luckily, DRL provides a feasible idea to solve this kind of NP-hardproblemandhasbeenprovedtobeefficientinap-proaching the optimal solution [45], [46].”
MethodFederated Learning Assisted Intelligent IoV Mobile Edge Computing
Quan, H., Zhang, Q., & Zhao, J. (2024) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2024.3421357 · Cited by 40
Xie, Y., Xu, Z., Zhong, Y., Xu, J., Gong, S., & Wang, Y. (2019) · IEEE/CIC International Conference on Communications in China - Workshops (CIC/ICCC) · doi:10.1109/iccchinaw.2019.8849964 · Cited by 39
“Considering limited capacity at the MEC server, DQN is also used in [8] to learn the optimal resource allocation and the offloading decision, based on the cost of all users and the available capacity of the MEC server.”
Edge Network Optimization Based on AI Techniques: A Survey
Pooyandeh, M., & Sohn, I. (2021) · Electronics · doi:10.3390/electronics10222830 · Cited by 40
“The application is authenticated and then decrypted and run in a private environment using virtualization, resulting in the AI model being safe.”
MethodDeep reinforcement learning-based low-latency task offloading for mobile-edge computing networks
Yang, W., Liu, Z., Liu, X., & Ma, Y. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.112164 · Cited by 40
Khan, I. A., Tao, X., Rahman, G., Rehman, W. U., & Salam, T. (2020) · IEEE Access · doi:10.1109/access.2020.2991057 · Cited by 39
“In [25], [26] a single edge server based edge-computing system is discussed.”
BackgroundReinforcement Learning for Security-Aware Computation Offloading in Satellite Networks
Sthapit, S., Lakshminarayana, S., He, L., Epiphaniou, G., & Maple, C. R. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3135632 · Cited by 36
“cost, energy consumption, and computation delay by offloading tasks to a mobile-edge computing server was considered in [52] and [53], and solved using Deep RL techniques.”
Toward Intelligent and Adaptive Task Scheduling for 6G: An Intent-Driven Framework
Qingqing, W., Zou, S., Sun, Y., Liwang, M., Wang, X., & Ni, W. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3391318 · Cited by 36
“DRL framework presents a viable and widely adopted solution for navigating the MPMT scheduling challenges [28].”
MethodDeep Reinforcement Learning Based Dynamic Resource Allocation in 5G Ultra-Dense Networks
Liu, Z., Chen, X., Chen, Y., & Li, Z. (2019) · IEEE International Conference on Smart Internet of Things (SmartIoT) · doi:10.1109/smartiot.2019.00034 · Cited by 40
“In [18], DQN was used to solve the problem of resource allocation in wireless MEC, and the total cost was significantly reduced.”
MethodUAV‐enabled computation migration for complex missions: A reinforcement learning approach
Zhu, S., Gui, L., Cheng, N., Zhang, Q., Sun, F., & Lang, X. (2020) · IET Communications · doi:10.1049/iet-com.2019.1188 · Cited by 34
Joint Optimization Across Timescales: Resource Placement and Task Dispatching in Edge Clouds
Wei, X., Rahman, A., Cheng, D., & Wang, Y. (2021) · IEEE Transactions on Cloud Computing · doi:10.1109/tcc.2021.3113605 · Cited by 36
“Li et al. [52] considered the joint offloading and resource allocation in a multi-user edge system, where multiple users can perform computation offloading via wireless channels to an edge server.”
BackgroundDeng, X., Yang, H., Zhang, J., Gui, J., Lin, S., Wang, X., & Min, G. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3531887 · Cited by 39
“Li et al. [28] applied DQN to address computation offloading and resource allocation in multi-user MEC environments.”
Cuttlefish: Neural Configuration Adaptation for Video Analysis in Live Augmented Reality
Chen, N., Quan, S., Zhang, S., Qian, Z., Jin, Y., Wu, J., Li, W., & Lu, S. (2020) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2020.3035044 · Cited by 35
“In [43], [44], [45], [46], the authors considered a multi-user MEC system, and proposed A3C based optimization framework to tackle resource allocation for MEC.”
BackgroundA Q-learning based Method for Energy-Efficient Computation Offloading in Mobile Edge Computing
Jiang, K. Y., Zhou, H., Li, D., Liu, X., & Xu, S. (2020) · International Conference on Computer Communications and Networks (ICCCN) · doi:10.1109/icccn49398.2020.9209738 · Cited by 37
“Without any prior knowledge of the environment information, Li et al. [17] achieved a RL based adaptive framework to tackle the computation offloading decision in MEC, that framework was also used in [18] to study the dependency-aware task offloading.”
MethodA Survey on Green Enablers: A Study on the Energy Efficiency of AI-Based 5G Networks
Ezzeddine, Z., Khalil, A., Zeddini, B., & Ouslimani, H. H. (2024) · Sensors · doi:10.3390/s24144609 · Cited by 32
Zhu, Y. L., Hu, Y., Yang, T., Yang, T., Vogt, J., & Schmeink, A. (2021) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2021.3073791 · Cited by 34
“, execution latency and energy consumption [22], [23], computation rate [24] and long-term utility [25].”
BackgroundQadeer, A., & Lee, M. J. (2023) · IEEE Access · doi:10.1109/access.2023.3249153 · Cited by 32
“• DQN [21], [55]: A deep Q-network based agent proposed to address resource allocation problems.”
BackgroundJiao, X., Ou, H., Chen, S., Guo, S., Qu, Y., Xiang, C., & Shang, J. (2023) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2023.3263169 · Cited by 41
“In this mode, the energy consumption of IWD d i is caused by the CPU computing, and can be expressed as [37]: where c denotes the computing energy efficiency coefficient, and equals to 10 − 27 according to the practical evaluation results in [38].”
MethodZhu, L., Gong, T., Wei, S., & Yu, F. R. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3382747 · Cited by 31
“Therefore, we first present the communication and computational model of TACS in this subsection [35].”
MethodDeep Reinforcement Learning-Based Computation Offloading in Vehicular Edge Computing
Zhan, W., Luo, C., Wang, J., Min, G., & Duan, H. (2019) · Globecom · doi:10.1109/globecom38437.2019.9013982 · Cited by 33
“Recently, DRL-based offloading methods have begun to emerge, such as [11], [12] and [13].”
MethodAn Actor-Critic-Based Transfer Learning Framework for Experience-Driven Networking
Xu, Z., Yang, D., Tang, J., Tang, Y., Yuan, T., Wang, Y., & Xue, G. (2020) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2020.3037231 · Cited by 40
Salh, A., Audah, L., Kim, K. S., Alsamhi, S. H., Alhartomi, M. A., Abdullah, Q., Almalki, F. A., & Algethami, H. (2022) · IEEE Access · doi:10.1109/access.2022.3170447 · Cited by 33
“Previous works, as shown in [18], [19], could not satisfy the requirement of B5G URLLC because the systems did not gain considerable experience through rigorous training to improve unusual traffic patterns, extreme events, and unforeseen network congestion.”
BackgroundResultMulti-objective Deep Reinforcement Learning for Mobile Edge Computing
Yang, N., Wen, J., Zhang, M., & Tang, M. (2023) · International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt) · doi:10.23919/wiopt58741.2023.10349870 · Cited by 39
“, [4], [6]–[10]), most of them have focused exclusively on single-objective methods.”
BackgroundInfluentialZhang, Y., Zhou, Z., Shi, Z., Meng, L., & Zhang, Z. (2020) · IEEE Access · doi:10.1109/access.2020.2987574 · Cited by 31
“Furthermore, an MDP problem with a large scale of state set or action set is not handled in a normal TD learning method.”
BackgroundInfluentialBurger, A., Cichiwskyj, C., Schmeißer, S., & Schiele, G. (2020) · Future Generation Computer Systems · doi:10.1016/j.future.2020.07.035 · Cited by 29
Liang, S., Wan, H., Qin, T., Li, J., & Chen, W. (2020) · IEEE International Conference on Communication Technology (ICCT) · doi:10.1109/icct50939.2020.9295872 · Cited by 31
“We compare the proposed algorithms (i.e., Nash-Q-learning, Nash-DQN) in this paper with the other four meth-ods, which are as follows: 1) All users compute locally (i.e., full-local computation); 2) All users offload the task to the MEC server for processing (i.e., full-offloading-computation); 3) Traditional Q-learni…”
MethodSelf-citationJiao, X., Wang, Y., Guo, S., Zhang, H., Dai, H., Li, M., & Zhou, P. (2023) · IEEE Transactions on Communications · doi:10.1109/tcomm.2023.3283792 · Cited by 29
“Moreover, according to [37], we can achieve the energy consumption of computing the task locally by”
BackgroundA Deep Reinforcement Learning-Based Dynamic Computational Offloading Method for Cloud Robotics
Penmetcha, M., & Min, B. (2021) · IEEE Access · doi:10.1109/access.2021.3073902 · Cited by 34
“The time of execution for an application is proportional to the size of the input data [59].”
BackgroundMustafa, E., Shuja, J., Rehman, F. U., Riaz, A., Maray, M., Bilal, M., & Khan, M. K. (2024) · Journal of Network and Computer Applications · doi:10.1016/j.jnca.2024.103886 · Cited by 34
Intelligent Offloading for Multi-Access Edge Computing: A New Actor-Critic Approach
Liu, K.-H., & Liao, W. (2020) · International Conference on Communications · doi:10.1109/icc40277.2020.9149387 · Cited by 28
Gan, D., Ge, X., & Li, Q. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3292368 · Cited by 33
“The Q -function decomposition technique based on MDP [19] was adopted to minimize the delay and energy consumption.”
MethodAchieving Fast Environment Adaptation of DRL-Based Computation Offloading in Mobile Edge Computing
Hu, Z., Niu, J., Ren, T., & Guizani, M. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3320253 · Cited by 29
“…locally computing I tn can be calculated as where ξ is the energy consumption coefficient of MD’s computing unit and can be calculated according to [31] as, It is worth noting that the task delay of I tn in the local computing mode includes two parts: One is the time required to execute I tn…”
MethodDas, S. K., Mudi, R., Rahman, M. S., Rabie, K. M., & Li, X. (2024) · IEEE Open Journal of Vehicular Technology · doi:10.1109/ojvt.2024.3466858 · Cited by 27
“Researchers have formulated this joint problem as a combinatorial optimization challenge with non-linear constraints and have proposed computation offloading algorithms based on convex optimization [103], [104], Lyapunov optimization [105], [106], and game theory [107], [108].”
Enhancing Sustainable Edge Computing Offloading via Renewable Prediction for Energy Harvesting
Alhartomi, M. A., Salh, A., Audah, L., Alzahrani, S., & Alzahmi, A. (2024) · IEEE Access · doi:10.1109/access.2024.3404222 · Cited by 29
“Two benchmark schemes were evaluated via simulations, namely the DRL scheme from [42], and the Q-learning-based offloading sys-tem described in [43].”
MethodOnline Service Provisioning in NFV-Enabled Networks Using Deep Reinforcement Learning
Nouruzi, A., Zakeri, A., Javan, M. R., Mokari, N., Hussain, R., & Kazmi, S. M. A. (2022) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2022.3159670 · Cited by 28
“is not possible [16], [28], a DNN is deployed for estimating the Q-function values.”
MethodToward Network-Slicing-Enabled Edge Computing: A Cloud-Native Approach for Slice Mobility
Shah, S. D. A., Gregory, M. A., & Li, S. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3292520 · Cited by 29
“The computational model can be defined in three steps following the models introduced in other related works [29], [30], [31].”
BackgroundMethodResource Management using Machine Learning in Mobile Edge Computing: A Survey
Zamzam, M., Elshabrawy, T., & Ashour, M. (2019) · International Conference on Intelligent Computing and Information Systems (ICICIS) · doi:10.1109/icicis46948.2019.9014733 · Cited by 28
“The multi objective problem of minimizing both latency and energy consumption is studied in [3] using combination of both deep neural network and deep reinforcement learning, [5] using neural network, [8] using combination of deep reinforcement learning and Q learning, [10] using deep reinforcement leaning and [12] us…”
BackgroundMethodWang, Z., Wei, T., Sun, G., Liu, X., Yu, H., & Niyato, D. T. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3586874 · Cited by 26
Resource Allocation in Wireless Semantic Communications: A Comprehensive Survey
Zhang, C., Huang, L., & Ning, Q. (2025) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2025.3595168 · Cited by 34
“For example, neural networks are used to approximate the function in which closed-form expressions cannot be obtained, and deep reinforcement learning (DRL) has also become a powerful tool for solving complex resource allocation problems in recent years [79], [80], [81].”
Q-Learning Based Task Offloading and Resource Allocation Scheme for Internet of Vehicles
Jiang, F., Liu, W., Wang, J., & Liu, X. (2020) · IEEE/CIC International Conference on Communications in China (ICCC) · doi:10.1109/iccc49849.2020.9238925 · Cited by 26
“P e,c k is the power consumption of VUE k while waiting for the MEC server to complete calculation task [3].”
BackgroundMethodInfluentialA Novel Hybrid-ARPPO Algorithm for Dynamic Computation Offloading in Edge Computing
Yang, X., Luo, H., Sun, Y., & Guizani, M. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3188928 · Cited by 26
“The weights are, respectively, denoted by α and β, and α + β = 1 [9], [43].”
Wang, Y., Ge, H., Feng, A., Li, W., Liu, L., & Jiang, H. (2020) · International Conference on Cloud Computing and Big Data Analytics (ICCCBDA) · doi:10.1109/icccbda49378.2020.9095689 · Cited by 26
“Figure 5 shows the change of total system cost when the ultimate computing power of MEC server is between [1, 10].”
BackgroundMethodA Multi-Agent RL Algorithm for Dynamic Task Offloading in D2D-MEC Network with Energy Harvesting
Mi, X., He, H., & Shen, H. (2024) · Sensors · doi:10.3390/s24092779 · Cited by 25
“Li et al. [16] utilized the deep Q-network (DQN) to jointly manage computation offloading and resource allocation in multiuser MEC setups to minimize the total cost of delay and energy consumption.”
BackgroundLi, T. K., Xiao, Y., & Song, L. (2021) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2021.3073979 · Cited by 25
A Memetic Algorithm for Cooperative Complex Task Offloading in Heterogeneous Vehicular Networks
Ren, H., Liu, K., Yan, G., Li, Y., Zhan, C., & Guo, S. (2022) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2022.3206228 · Cited by 26
Deep reinforcement learning based offloading decision algorithm for vehicular edge computing
Hu, X., & Huang, Y. (2022) · PeerJ Computer Science · doi:10.7717/peerj-cs.1126 · Cited by 27
“Referring to the works (Li et al., 2018; Tran & Pompili, 2018; Wang et al., 2019; Wang, Wang & Liu, 2021), the similar values are set in our simulations.”
ResultAn intelligent resource management method in SDN based fog computing using reinforcement learning
Anoushee, M., Fartash, M., & Torkestani, J. A. (2023) · Computing · doi:10.1007/s00607-022-01141-x · Cited by 30
Ouyang, H., Liang, L., Li, S., & Gao, L. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124526 · Cited by 25
Deep Reinforcement Learning for Resource Allocation in 5G Communications
Tham, M., Iqbal, A., & Chang, Y. C. (2019) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipaasc47483.2019.9023112 · Cited by 25
“Furthermore, similar to [19-22], the training process is not in a self-supervised learning mode.”
BackgroundMethodFair Computation Efficiency Scheduling in NOMA-Aided Mobile Edge Computing
Huang, X., Zeng, S., Li, D., Zhang, P., Yan, S., & Wang, X. (2020) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2020.3001994 · Cited by 24
“focused on minimizing the system cost of delay and energy consumption under the tolerable delay guarantee [4].”
BackgroundFang, C., Zhang, T., Huang, J., Xu, H., Hu, Z., Yang, Y., Wang, Z., Zhou, Z., & Luo, X. (2022) · Symmetry · doi:10.3390/sym14102120 · Cited by 28
“[17] presented a heterogeneous MEC network based on reinforcement learning (RL) to optimize resource allocation in the wireless system.”
BackgroundSun, Y., & He, Q. (2023) · Electronics · doi:10.3390/electronics12061304 · Cited by 24
Hybrid Online–Offline Learning for Task Offloading in Mobile Edge Computing Systems
Sohaib, M., Jeon, S.-W., & Yu, W. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3335362 · Cited by 26
“In recent studies, deep reinforcement learning (DRL) has been widely adopted to develop optimal task offloading policies for various design objectives [13], [14], [15].”
BackgroundDeep reinforcement learning in edge networks: Challenges and future directions
Hazra, A., Tummala, V. M. R., Mazumdar, N., Sah, D. K., & Adhikari, M. (2024) · Physical Communication · doi:10.1016/j.phycom.2024.102460 · Cited by 35
Buffer-aware Wireless Scheduling based on Deep Reinforcement Learning
Chen, X., Wang, J., Yu, T., Kong, C., Huangfu, Y., Li, R., Ge, Y., & Wang, J. (2020) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc45663.2020.9120729 · Cited by 28
“A DRL-based solution for multi-user computation offloading and resource allocation with mobile edge computation (MEC) is presented in [5], where time and energy cost is jointly minimized.”
MethodOnline Learning for Orchestration of Inference in Multi-user End-edge-cloud Networks
Shahhosseini, S., Seo, D., Kanduri, A., Hu, T., Lim, S.-S., Donyanavard, B., Rahmani, A. M., & Dutt, N. D. (2022) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3520129 · Cited by 29
“[6, 8, 27, 33, 45] ✗ ✗ ✗ CO [19] ✓ ✗ ✗ CO,HW [1, 7, 17, 20, 24, 36, 43] ✗ ✓ ✗ CO Ours ✓ ✓ ✓ CO,APP”
Lin, K., Gao, J., Han, G., Wang, H., & Li, C. (2022) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2022.3169457 · Cited by 26
“Some resource optimization schemes are designed based on deep reinforcement learning to minimize network latency and energy consumption by optimizing offloading decisions and computing resource allocation [4].”
Background
Metadata
- Authors
- Li, J., Gao, H., Lv, T., & Lu, Y.
- Affiliations
- Beijing University of Posts and Telecommunications (CN)
- Venue
- 2018 IEEE Wireless Communications and Networking Conference (WCNC)
- Publisher
- IEEE
- Published
- 2018-04
- Pages
- 1-6
- ISSN
- 1525-3511 (ISSN-L)
- OpenAlex
- W2808381205
- Cited by
- 548
- References
- 17
- Citations by year
- 2026: 16 · 2025: 45 · 2024: 73 · 2023: 83 · 2022: 92 · 2021: 96 · 2020: 94 · 2019: 45 · 2018: 4
- FWCI
- 49.38
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE Wireless Communications and Networking Conference (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Wireless Network Optimization
- Physical Sciences › Engineering › Aerospace Engineering › UAV Applications and Optimization
- Keywords
- mobile edge computing, computation offloading, resource allocation, deep reinforcement learning, Q-learning, multi-user systems, energy consumption, delay cost, 5G
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.