Resource Management in Fog/Edge Computing: A Survey (1810.00305v1)

Abstract: Contrary to using distant and centralized cloud data center resources, employing decentralized resources at the edge of a network for processing data closer to user devices, such as smartphones and tablets, is an upcoming computing paradigm, referred to as fog/edge computing. Fog/edge resources are typically resource-constrained, heterogeneous, and dynamic compared to the cloud, thereby making resource management an important challenge that needs to be addressed. This article reviews publications as early as 1991, with 85% of the publications between 2013-2018, to identify and classify the architectures, infrastructure, and underlying algorithms for managing resources in fog/edge computing.

• The paper presents a comprehensive review of resource management schemes across architectures, infrastructure, and algorithms in fog/edge computing.
• It analyzes diverse approaches—data flow, control, and tenancy architectures—to enhance workload distribution and reduce latency.
• The study outlines future directions including lightweight processing, GPU integration, and advanced orchestration for dynamic edge environments.

Resource Management in Fog/Edge Computing: A Survey

The paper "Resource Management in Fog/Edge Computing: A Survey," authored by Cheol-Ho Hong and Blesson Varghese, provides an extensive review of the methodologies and technologies in resource management pertinent to fog and edge computing. It explores the nuances of computing paradigms that deviate from the centralized cloud model by leveraging decentralized computing resources near the network edge to cater to data processing requirements closer to end-users.

Overview of the Study

The authors categorize the survey into three primary domains: architectures, infrastructure, and algorithms aimed at optimizing resource management in fog/edge computing environments. This comprehensive review ranges from seminal works in the early '90s, with a significant body of the research evaluated spanning from 2013 to 2018, demonstrating a temporal focus on recent advancements.

Architectures in Fog/Edge Computing

The authors dissect architectural schemas into three distinct categories:

1. Data Flow Architectures: These include aggregation, sharing, and offloading methods. Aggregation focuses on minimizing communication overhead through data processing at the edge, while sharing involves distributing workloads among devices in a peer network. Offloading strategies are pivotal for reducing latencies by delegating computing tasks from user devices to nearby fog/edge resources.
2. Control Architectures: Control mechanisms are divided into centralized and distributed frameworks. Centralized control utilizes single controllers for task allocation, presenting a challenge in scale and robustness. Distributed control leverages sophisticated techniques such as blockchain, game theory, and genetic algorithms to achieve decentralized resource coordination.
3. Tenancy Architectures: The paper considers single and multi-tenancy systems to underscore the importance of resource sharing and optimization in edge environments. Multi-tenancy is highlighted as crucial in publicly accessible infrastructures.

Infrastructure Underpinning Resource Management

The survey underscores the importance of the hardware and software stack in facilitating fog/edge computing:

• Hardware: The paper suggests utilizing miniaturized computing resources such as single-board computers and network devices, which are critical given their lightweight characteristics and scalability.
• System Software: This includes virtualization technologies, both system-level (VMs and containers) and network (SDN and NFV), emphasizing the necessity for efficient infrastructure management and resource allocation. Notably, containers are favored for their lightweight nature, enhancing application density on resource-constrained devices.
• Middleware: The research speaks to orchestration technologies that ensure optimal deployment and execution of applications across hierarchically structured computing resources, facilitating effective edge-cloud synergy.

Algorithms Driving Fog/Edge Computing

The paper categorizes key algorithmic approaches into four areas:

1. Discovery Algorithms: Protocols for identifying viable edge resources, which is a foundational step in the network’s extended ecosystem.
2. Benchmarking Techniques: Essential for performance evaluation, these methods capture key metrics relevant to edge environments, accounting for the unique constraints of edge resources.
3. Load-Balancing Algorithms: The paper evaluates approaches designed to distribute workloads efficiently across the edge resources while considering latency and throughput requirements to maintain service quality.
4. Placement Strategies: These methods determine the optimal allocation of applications on edge resources to maximize performance, utilizing dynamic and iterative methodologies.

Implications and Future Directions

The implications of this survey extend to numerous practical applications, from latency-sensitive services like video streaming to real-time analytics in smart environments. Future research is speculated to focus on enhancing lightweight data processing tools, integrating GPU resources effectively in edge devices, and extending orchestration capabilities to address the dynamic nature of mobile edge environments.

This survey serves as a critical resource for researchers and practitioners seeking a detailed understanding of current trends and challenges in resource management in fog and edge computing, offering insights that are pivotal for advancing infrastructure and service provision in distributed computing paradigms.