Fog Computing: Issues
While fog computing offers significant advantages in processing data close to the source for
lower latency and improved responsiveness, it faces several technical and operational
challenges:
Security and Privacy Risks: Since fog nodes are distributed and operate at the edge, they
are more susceptible to attacks. Authentication and privacy are critical concerns, as
maintaining secure communication between numerous heterogeneous devices and the cloud
becomes complicated [1] [2] [3] .
High Energy Consumption: The presence of many fog nodes increases overall energy
usage. Efficient power management is essential to avoid excessive operational costs and
support scalability [2] [4] .
Resource Constraints: Fog nodes often have limited computational, memory, and storage
resources compared to cloud data centers, potentially leading to bottlenecks for large-scale
or heavy computational tasks [4] .
Data Management: Ensuring consistent, synchronized, and secure data sharing among
multiple nodes is challenging in a distributed environment [3] .
Node Placement and Network Management: Effective placement of fog servers or nodes is
essential to optimize performance and minimize latency. Poor deployment may lead to
network congestion and reduced service quality [2] [3] .
Scalability and Maintenance: Managing a vast number of decentralized fog nodes across
geographic regions introduces complexity in updates, monitoring, fault tolerance, and
overall maintenance [1] [3] .
Heterogeneity: Supporting various device types, platforms, and protocols complicates
integration and increases interoperability issues [1] [3] .
Cost: Deploying millions of nodes, as required for comprehensive fog setups, can be costly
and introduce financial barriers for some organizations [5] .
Cloud vs. Fog Computing: Key Differences
Aspect Cloud Computing Fog Computing
Architecture Centralized (major data centers) Distributed (many small, local nodes)
Higher (data travels to distant data Lower (processing occurs near the data
Latency
centers) source)
Resource
Virtually unlimited compute/storage Limited, depends on node capability
Power
Scalability Highly scalable Less scalable, deployment-intensive
Aspect Cloud Computing Fog Computing
Location Far from data sources, typically central Close to end-users and IoT devices
Bandwidth Consumes more, as raw data is sent to More efficient, as only processed data may be
Usage the cloud sent
Mature security protocols (centralized Challenging due to decentralization, more
Security
control) attack surfaces
Big data analytics, storage, enterprise IT,
Use Cases Real-time analytics, industrial IoT, smart grids
SaaS, etc.
Mobility
Limited Designed for mobile and moving devices
Support
Complex, requires managing many distributed
Maintenance Centralized, easier to manage
nodes
Cloud excels at large-scale storage, processing, historical analytics, and handling
enterprise-wide applications but can be limited by network latency and bandwidth issues
for real-time or edge applications [6] [5] [7] [8] .
Fog supplements the cloud by handling real-time, latency-sensitive, and local data
processing tasks, reducing the load and latency on core infrastructure while empowering IoT
and mobile environments [6] [7] [8] [5] .
In summary, fog computing is well-suited for applications where immediate processing,
bandwidth optimization, and proximity to users/devices are crucial, while cloud computing
remains ideal for centralized, large-scale storage, and processing needs. Most robust
architectures blend both paradigms for optimal results.
⁂
1. https://www.sciencedirect.com/science/article/pii/S2772503023000099
2. https://yourtechdiet.com/blogs/fog-computing-issues/
3. https://wseas.com/journals/computers/2020/a245105-057.pdf
4. https://pmc.ncbi.nlm.nih.gov/articles/PMC8708798/
5. https://arxiv.org/pdf/1904.04026.pdf
6. https://www.scalecomputing.com/resources/edge-computing-vs-fog-computing-vs-cloud-computing
7. https://www.tutorialspoint.com/differences-between-fog-computing-and-cloud-computing
8. https://sam-solutions.com/blog/fog-computing-vs-cloud-computing-for-iot-projects/
