Privacy-Aware Offloading in Mobile-Edge Computing

He, Xiaofan; Liu, Juan; Jin, Richeng; Dai, Huaiyu

doi:10.1109/glocom.2017.8253985

Cited by 81 publications

(63 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We represent a computation task by (µ (t) , ϑ) with µ (t) and ϑ being, respectively, the input data size (in bits) and the number of CPU cycles required to accomplish one input bit of the computation task. This work assumes that the task arrival sequence {A k n,(t) : k ∈ N + } follows a Markov process [48]. Two options are available for each computation task 4 : 1) being processed locally at the MU; and 2) being offloaded to the logical MEC gateway in the computation slice.…”

Section: B Computation and Communication Modelsmentioning

confidence: 99%

“…3 It's straightforward that given the mobility model, the average overhead of a MU incurred during inter-BS handovers is fixed. 4 The kind of computation tasks that can only be processed at the mobile devices [48] does not affect the optimization goal and hence is neglected. 5 For simplicity, we assume that the CPU power at a mobile device matches the maximum computation task arrivals and a MU can hence process A where ς is the effective switched capacitance that depends on chip architecture of the mobile device [50] and ̺ is the CPUcycle frequency at a mobile device.…”

Section: B Computation and Communication Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning Approach

Chen

Zhao

et al. 2019

IEEE J. Select. Areas Commun.

125

View full text Add to dashboard Cite

With the cellular networks becoming increasingly agile, a major challenge lies in how to support diverse services for mobile users (MUs) over a common physical network infrastructure. Network slicing is a promising solution to tailor the network to match such service requests. This paper considers a system with radio access network (RAN)-only slicing, where the physical infrastructure is split into slices providing computation and communication functionalities. A limited number of channels are auctioned across scheduling slots to MUs of multiple service providers (SPs) (i.e., the tenants). Each SP behaves selfishly to maximize the expected long-term payoff from the competition with other SPs for the orchestration of channels, which provides its MUs with the opportunities to access the computation and communication slices. This problem is modelled as a stochastic game, in which the decision makings of a SP depend on the global network dynamics as well as the joint control policy of all SPs. To approximate the Nash equilibrium solutions, we first construct an abstract stochastic game with the local conjectures of channel auction among the SPs. We then linearly decompose the per-SP Markov decision process to simplify the decision makings at a SP and derive an online scheme based on deep reinforcement learning to approach the optimal abstract control policies. Numerical experiments show significant performance gains from our scheme.

show abstract

Section: B Computation and Communication Modelsmentioning

confidence: 99%

Section: B Computation and Communication Modelsmentioning

confidence: 99%

Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning Approach

Chen

Zhao

et al. 2019

IEEE J. Select. Areas Commun.

125

View full text Add to dashboard Cite

show abstract

“…A number of studies on D2D offloading have been conducted in the literature [4], [11]- [16]. Based on their objectives, these works can be classified into the following categories: 1) reduction of task completion delay [11]- [13]; 2) provision of reliable task processing results [4], [14]; and 3) solution to the privacy leakage problem [15], [16].…”

Section: Related Workmentioning

confidence: 99%

“…Ni et al [15] proposed a random matrix-based location matching approach for mobile crowdsensing to allocate tasks without disclosing the location of the mobile devices. He et al [16] presented a task offloading scheduling algorithm based on a constrained Markov decision process problem to achieve a low task completion delay and low energy consumption while maintaining a sufficient level of privacy on the location and usage pattern. However, these works did not consider the reliability of the offloading results.…”

Section: Related Workmentioning

confidence: 99%

“…For example, if mobile devices arrive frequently in the vicinity of the task owner in D2D environments, the task owner cannot easily maintain historical information due to mobility. Furthermore, when the tasks are frequently offloaded to a specific mobile device with a high reputation, the privacy of the task owner cannot be preserved (i.e., the usage patterns of the task owner may be exposed to external mobile devices) [15], [16].…”

mentioning

confidence: 99%

See 1 more Smart Citation

Privacy-Preserving and Trustworthy Device-to-Device (D2D) Offloading Scheme

Baek

Pack

2020

IEEE Access

View full text Add to dashboard Cite

In device-to-device (D2D) offloading, frequent offloading to a specific mobile device (i.e., an offloadee) can violate the privacy of a task owner and proper offloading results cannot always be guaranteed due to untruthful mobile devices. To address these problems, we propose a privacy-preserving and trustworthy D2D offloading scheme (PPTS) with two steps: 1) a privacy-preserving offloading step and 2) a blockchain-based verification step. In the first step, a task owner selects several offloadees and offloads the tasks redundantly to them to obtain reliable offloading results with a minimum task completion delay while preserving its own privacy at a sufficient level. In addition, the task owner chooses another mobile device as a verifier. In the second step, the task owner and verifier exploit blockchain networks to check whether the offloading results are appropriately processed. We formulate a static game to decide an appropriate redundancy ratio of offloading tasks and incentives for the offloadees.The evaluation results demonstrate that PPTS can provide reliable offloading results with a reduced task completion delay and guarantee a sufficient level of privacy of the task owner.

show abstract

Machine learning in/for blockchain: Future and challenges

et al. 2021

View full text Add to dashboard Cite

Machine learning and blockchain are two of the most notable technologies of recent years. The first is the foundation of artificial intelligence and big data analysis, and the second has significantly disrupted the financial industry. Both technologies are data‐driven, and thus there are rapidly growing interests in integrating both for more secure and efficient data sharing and analysis. In this article, we review existing research on combining machine learning and blockchain technologies and demonstrate that they can collaborate efficiently and effectively. In the end, we point out some future directions and expect more research on deeper integration of these two promising technologies.

show abstract

Privacy-Aware Offloading in Mobile-Edge Computing

Cited by 81 publications

References 19 publications

Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning Approach

Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning Approach

Privacy-Preserving and Trustworthy Device-to-Device (D2D) Offloading Scheme

Machine learning in/for blockchain: Future and challenges

Contact Info

Product

Resources

About