Socially trusted collaborative edge computing in ultra dense networks

Abstract

Small cell base stations (SBSs) endowed with cloud-like computing capabilities are considered as a key enabler of edge computing (EC), which provides ultra-low latency and location-awareness for a variety of emerging mobile applications and the Internet of Things. However, due to the limited computation resources of an individual SBS, providing computation services of high quality to its users faces significant challenges when it is overloaded with an excessive amount of computation workload.In this paper, we propose collaborative edge computing among SBSs by forming SBS coalitions to share computation resources with each other, thereby accommodating more computation workload in the edge system and reducing reliance on the remote cloud. A novel SBS coalition formation algorithm is developed based on the coalitional game theory to cope with various new challenges in small-cellbased edge systems, including the co-provisioning of radio access and computing services, cooperation incentives, and potential security risks. To address these challenges, the proposed method (1) allows collaboration at both the user-SBS association stage and the SBS peer offloading stage by exploiting the ultra dense deployment of SBSs, (2) develops a payment-based incentive mechanism that implements proportionally fair utility division to form stable SBS coalitions, and (3) builds a social trust network for managing security risks among SBSs due to collaboration. Systematic simulations in practical scenarios are carried out to evaluate the efficacy and performance of the proposed method, which shows that tremendous edge computing performance improvement can be achieved.

All one needs to know about fog computing and related edge computing paradigms: A complete survey

“…We believe scalability is critical in designing fog systems; fog systems should be scalable so that they could be implemented in IoT networks. For instance, a scalable algorithm for fog offloading is an online offloading scheme that does not need information of individual IoT nodes for decision making (e.g., [88,312,313]). We encourage researchers in the fog computing area to verify the scalability of their proposed algorithms and schemes (e.g., by an actual implementation).…”

Budget-Constrained Edge Service Provisioning With Demand Estimation via Bandit Learning

“…In this section, we discuss in detail the solutions for optimization problems in (8), (14), and (15). Since these optimization problems have the same form, we take the Oracle subproblem (8) as an example. Note that the problem is solved for each time slot t, the time index is dropped in this section for ease of notation.…”

“…In addition, we see that COERR achieves a higher cumulative utility when the size of rental decision set |F | is larger. A larger rental decision set F loosens the constraint in per-slot problem (8), e.g., the constraint in (8c) becomes looser if we change the rental decision set F n = [0, 2,4] to F n = [0, 2,4,6]. Therefore, we may have a higher utility in each exploitation phase with F n = [0, 2,4,6].…”

Secure Mobile Edge Computing in IoT via Collaborative Online Learning

“…Furthermore, let ρ j ∈ [0, 1] denote a device-specific weight on security; e.g., a larger ρ j can be used for safetysensitive tasks involved in autonomous vehicle and healthcare, while a smaller ρ j can be adopted by sensors in smart homes and smart grids where privacy is the first priority. Per slot t, the security risk for device j choosing server k is modeled by [26] r j t (k) = ρ j c j t γ 1,t (k) + (1 − ρ j )s j t γ 2,t (k).…”