Edge computing deploys edges close to end-users to provide highly accessible resources and latency-sensitive services. It is invaluable for urban crowd/hazard management services, e.g., real-time dynamic route planning and hazard monitoring/analysis, etc. However, in such scenarios, various types of urban hazards jeopardize the usability of edge servers. Worsely, these hazards could be integrated, like gas fires caused by urban earthquakes. In this regard, the formulation of usability risks that servers face is intractable due to the complexity, incomplete real-time data and insufficient expert knowledge of these integrated hazards. Therefore, we innovatively define the usability risks asServer Hazard Riskmodel from the view of the spatial data field by utilizingInformation Diffusion techniquewhich can overcome the adverse conditions above. Then we involve it to formulate theServer Hazard Risk User Allocation(SR-UA) problem, and analyze three typical solutions from the perspective of optimality and efficiency, which are the Lexicographic Goal Programming approach (SR-UA-LGP), the Approximation approach (SR-UA-A) and the Particle Swarm Optimization-based approach (SR-UA-PSO). The extensive experiments based on two real-world datasets illustrate the superior performance of our model and solutions.

Edge computing is a promising solution to enabling highly accessible resources and latency-sensitive services for nearby users. In public safety, it can provide critical support for urban crowd/hazard management services, such as real-time path planning, hazard warning, etc. In a crowd/hazard scenario, crowds can be allocated to nearby edge servers for obtaining real-time support, e.g., evacuation instructions for those who want to evacuate and crowd flow updates for those who want to rescue, etc. In such scenarios, the behaviors of different roles (like rescuers and evacuees) and the positive/negative interactions among them must be considered in user allocation for reducing injuries and fatalities. In this paper, these issues are defined as a novelRole-Based Criticality(RBC) model to describe the fatal risks of different roles in the crowd/hazard scenarios. Based on the model, theRole-Based User Allocation(RUA) problem is formulated. To tackle this problem, we devise an optimal solution named RUA-ILP based on Integer Linear Programming. To accommodate large-scale scenarios, we propose two representative approximate approach named RUA-A and RUA-GA to ensure efficient and effectiveness user allocation respectively. They can maximize the overall role-based criticality which can reduce injuries and fatalities in crowd/hazard scenarios by theoretical proofing and extensive experiments conducted on a real-world dataset.

Edge computing provides a novel computing paradigm by deploying services on edge servers to serve nearby end-users with low latency. In this regard, a suitable allocation strategy is crucial that maximizes the number of users served at the minimum overall cost, which is referred to as the Edge User Allocation (EUA) problem. However, when edge computing meets public safety, some critical issues have not been fully considered by existing EUA approaches. Among these issues, the levels of danger to individuals quantitatively indicate whether individuals are in danger in an emergency. Hence, the inclusion of these levels impacts the priority for allocating resources in the EUA problem. In this paper, these levels are defined as individual criticalities formally. Then, we take them into account to formulate the novel CRiticality-EUA (CR-EUA) problem, and prove its NP-hardness. To solve this problem, an optimal approach, named CR-EUA-O, is proposed by utilizing the Integer Programming technique. Furthermore, we propose an approach with a proven approximation ratio, named CR-EUA-H, as an effective and efficient solution. Experiments are conducted on a real-world dataset to evaluate our approaches against four representative approaches. The results show the superior performance of our approaches in the overall criticality and execution time.