CD325 - FOG AND EDGE COMPUTING

I. Answer ALL Questions

a) What does FEC stand for in the context of cloud computing?
Ans: FEC stands for Fog and Edge Computing.

b) Define SCANC.
Ans: SCANC is an acronym representing key features of FEC:
Security, Cognition, Agility, Notification, and Continuity.

c) What is the role of an IP Gateway Device?

Ans: An IP Gateway Device connects IoT sensors to the internet and forwards data
between local devices and cloud/fog servers.

d) List networking challenges in a federated edge environment.

Ans: Challenges include latency variation, resource discovery, mobility handling,
and interoperability.

e) What does EaaS stand for in the context of Edge Resource Management?
Ans: EaaS stands for Edge as a Service.

f) What does C2F2T stand for in the context of Fog Computing?

Ans: C2F2T stands for Cloud-to-Fog-to-Thing.

g) Explain network slicing.

Ans: Network slicing is the creation of multiple virtual networks on a shared 5G
infrastructure, each optimized for specific use cases like IoT or video streaming.

h) Name any two key metrics used for optimization in Fog Computing.
Ans: Latency and Energy Consumption.

i) Why is security management important in edge cloud architectures?

Ans: Security is crucial because data is processed near users, increasing the risk of
attacks and privacy breaches.

j) Discuss the primary role of middleware in fog and edge computing.

Ans: Middleware manages device communication, data processing, and service
orchestration across fog and cloud environments.

k) Name 2 state-of-the-art middleware infrastructures used in edge computing.

Ans: Google Fit and CoTWare.

l) How does fog computing benefit health monitoring applications?

Ans: Fog computing enables real-time data analysis, reduces latency, and improves
emergency alerts in health monitoring.

1 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

m) Why is fog computing essential for smart transportation applications?

Ans: It enables real-time traffic analysis, accident detection, and low-latency
decision-making at the edge.

n) Explain key challenge real-time human object tracking addresses in edge

computing.
Ans: It addresses the challenge of low-latency detection and tracking of human
movement for security and surveillance purposes.

2 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

UNIT I

2.a) Key Differences between Cloud Computing and Fog

& Edge Computing (FEC) (7 Marks)
Introduction

Cloud computing offers centralized services from remote data centers, while Fog and
Edge Computing (FEC) decentralize services by processing data closer to the devices
at the network’s edge. This reduces delay and improves performance in IoT-based
systems.

Feature Cloud Computing Fog and Edge Computing (FEC)

Data Processing Distributed at fog nodes or edge

Centralized cloud servers
Location devices
Proximity to Data
Far from IoT devices Close to IoT devices
Source

Higher latency due to Low latency because of local

Latency
transmission to cloud processing

High – raw data sent to the Lower – only filtered/processed data

Bandwidth Use
cloud sent

Real-Time Limited support for real- Ideal for real-time processing (e.g.,
Response time apps health monitoring)

Connectivity Requires constant Internet

Works with intermittent connectivity
Needs connection

Energy High, as data travels long

Lower, due to local computing
Consumption distances

Slower due to central Quicker, localized updates via FEC

Security Updates
update system nodes

Easier to scale up with Horizontal scaling with distributed

Scalability
centralized infra nodes is more complex

Local decision-making is possible

Decision Making Cloud makes all decisions
(fog/edge cognition)

Centralized failure can Local nodes reduce risk of single

Failure Risk
affect all systems point of failure

Cloud storage, web apps, Smart cities, autonomous cars, IoT

Use Cases
big data analytics control systems

3 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Conclusion

While cloud computing remains essential for large-scale storage and analytics, FEC
improves responsiveness, efficiency, and reliability for IoT and real-time applications.

2.b) Role of Service-Centric Model, Reliability, and

Service Mobility (7 Marks)
1. Service-Centric Model

Focuses on what the service provides instead of where it is located.

Aims to decouple services from specific server locations.

Services are accessed using unique identifiers, allowing dynamic redirection

to optimal nodes.

Enhances load balancing, service availability, and scalability.

Essential in federated edge systems where services are distributed across

different locations.

Benefits:

Seamless user experience.

Easier service replication and migration.

Better suited for dynamic, mobile environments

2. Reliability

Ensures high availability of services even with node failure or network

fluctuations.

Key methods:

Replication: Having multiple copies of services across edge locations.

Redundancy: Backup nodes ready to take over.

Fault detection & self-healing: Systems automatically detect and fix

errors.

4 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Load balancing: Distributes tasks to avoid overload and failure.

3. Service Mobility

Enables on-the-fly migration of services based on:

User movement.

Load conditions.

Network performance.

Ensures low-latency and uninterrupted service delivery.

Critical in mobile scenarios like connected vehicles and mobile health

monitoring.

➤Challenges:

Maintaining session continuity.

Avoiding data loss during migration.

Ensuring compatibility of hardware/software during transfer.

Conclusion:

These three roles are fundamental in building a flexible, reliable, and efficient edge
infrastructure, enabling real-time services that adapt to user and network behavior
dynamically.

5 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

3.a) Concepts of “X as a Service,” Support Services, and

Application Services (7 Marks)
1. X as a Service (XaaS)

XaaS refers to delivering computing resources as services (e.g., IaaS, PaaS,

SaaS).

In FEC, it includes:

Infrastructure (IaaS) – renting edge devices, fog nodes.

Platform (PaaS) – middleware platforms on fog nodes.

Software (SaaS) – analytics tools deployed at the edge.

IndieFog is a model where individuals contribute personal devices (e.g.,

Raspberry Pi) to form decentralized FEC services.

Examples:

Fon: Wi-Fi sharing network by individuals.

MQL5: People offer unused computing power.

2. Support Services

These include:

Installation and setup of fog nodes.

Device configuration and integration with IoT networks.

Firmware and security updates.

Remote diagnostics and monitoring.

Providers like Cisco offer such services with their fog-enabled routers (e.g.,
Cisco IR829).

Future Trend:

Vendor-independent support similar to Red Hat for Linux.

6 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

3. Application Services

Delivers specific edge-based software solutions for real-time data processing

and analysis.

Helps in decision-making at the edge level without relying on the cloud.

Examples:

Digital Twinning – Real-time virtual representation of systems.

Traffic Monitoring – Real-time analysis of smart city traffic.

AAL (Ambient Assisted Living) – Support services for elderly through edge-
based sensors and analytics.

Apache Edgent – Platform for edge analytics.

Conclusion:

These business models ensure a sustainable and scalable ecosystem for Fog and
Edge Computing by encouraging innovation, participation, and monetization across
all levels—from individuals to large enterprises.

7 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

3.b) SCALE Advantages of Fog and Edge Computing (7

Marks)
SCALE = Security, Cognition, Agility, Latency, Efficiency

1. Security

Offers localized and faster software updates, reducing risk of attacks.

Handles authentication and encryption near data sources.

Reduces attack surface area by isolating processes at the edge.

2. Cognition

FEC systems are context-aware and capable of:

Self-management (e.g., reconfiguration based on environment).

Local decision-making without cloud dependency.

Examples: Smart thermostats adjusting based on real-time room occupancy.

3. Agility

Enables faster deployment of services.

Supports scalable microservices for IoT.

Allows startups and individuals to deploy localized services.

4. Latency

Processes data close to the source, eliminating delays.

Essential for mission-critical apps: self-driving cars, remote surgeries, etc.

Examples:

Connected car systems using roadside FEC nodes.

Smart factories requiring millisecond-level response.

8 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

5. Efficiency

Reduces need to transmit large volumes of raw data to the cloud.

Saves network bandwidth and power consumption.

Allows edge analytics (e.g., anomaly detection at sensor level).

Conclusion:

SCALE advantages make FEC a powerful alternative to centralized computing in

scenarios requiring real-time processing, local autonomy, and operational efficiency.

9 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

UNIT II

4.a) Modeling Techniques in C2F2T Integration (7

Marks)
🔹 Introduction:

C2F2T stands for Cloud–Fog–IoT integration. Since it involves distributed layers, it

needs modeling techniques to analyze system behavior (e.g., delay, cost, power).
Modeling helps design efficient, scalable, and predictable infrastructure.

1. Analytical Models

Use mathematical equations (closed-form expressions).

Predict latency, power, bandwidth usage using workload parameters.

Evaluate “what-if” scenarios.

⚠️ May include approximations, which affect real-time accuracy.

Example:
Power consumption (P) = Active Power × Time + Idle Power × Time

2. Petri Net Models

Graph-based + mathematical system modeling.

Shows concurrency, synchronization, and parallelism.

Can analyze using:

Markov chains (probabilistic)

Graph-based transitions

Enables fine-grained control with fewer states.

Use Case:
Designing dependable real-time traffic control systems.

3. Integer Linear Programming (ILP) / MILP

Optimization under linear constraints.

ILP: Variables are integers (e.g., number of fog nodes).

10 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

MILP: Combines continuous + integer variables.

Solve for:

Cost minimization

Energy use

Task offloading

Example:
MILP used to reduce energy in a smart city fog setup.

4. Simulation-Based Models

Simulate dynamic environments using tools like:

iFogSim

CloudSim

Measures QoS, load balancing, migration, etc.

More realistic but needs computational effort.

5. Queuing Theory Models

Used to model traffic behavior and response time.

Common in gateway, fog, and task scheduling simulations.

Conclusion:

Each modeling method supports different levels of analysis, from cost optimization
(ILP) to real-time behavior (Petri nets), helping researchers optimize integrated
C2F2T infrastructures.

11 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

4.b)Concept of Network Slicing in 5G Framework (7

Marks)
Definition:

Network Slicing is a 5G architecture technique that allows multiple virtual networks

(slices) to operate over a common physical infrastructure, each optimized for a
specific application or use case.

Need for Network Slicing:

Different 5G applications have diverse requirements:

eMBB: High bandwidth (e.g., 4K video)

mMTC: Connects many IoT sensors

URLLC: Ultra-low latency (e.g., remote surgery)

Traditional networks can’t serve all at once, so slicing isolates traffic and resources
for each service.

12 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Key Technologies:

SDN (Software Defined Networking):

Centralized control

Separates control and data plane

NFV (Network Function Virtualization):

Turns routers/firewalls into virtual functions

Can be placed on cloud or fog nodes

Core Elements of 5G Slicing:

Component Function
UPF Handles user data packets
SMF Session control
AMF Access/mobility management
PCF Policy enforcement
UDM User data management

Benefits of Network Slicing:

Customization: Each slice meets unique latency/bandwidth needs

Isolation: Fault in one slice won’t affect others

Security: Independent encryption policies

Edge Integration: Combines edge/fog resources efficiently

Cost Reduction: Optimized resource use

Challenges:

Dynamic resource allocation

Orchestration across infrastructure

Maintaining QoS and SLAs

Handling mobility and handovers

13 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Conclusion:

Network slicing is the backbone of 5G for supporting multi-domain, multi-usecase

environments and is tightly integrated with fog and edge systems to ensure
responsive and secure services.

14 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

5.a) Optimization Challenges in Fog Computing (7

Marks)
Introduction:

Fog computing aims to support low-latency, localized data processing near IoT
devices.
However, managing resources efficiently in dynamic and heterogeneous fog
environments creates optimization challenges.

Key Optimization Challenges:

Latency Minimization

Must process data locally for real-time response.

Delay increases if processing is pushed to cloud.

Energy Efficiency

IoT devices and fog nodes often battery-powered.

High computing = high power use → trade-offs required.

Resource Allocation

How to distribute tasks across cloud, fog, edge?

Need load balancing and efficient task scheduling.

Cost Efficiency

Higher cloud usage = higher billing

Optimizing task placement saves cost.

Mobility & Scalability

IoT devices frequently move → need dynamic reassignment.

Network topology changes → affects performance.

Reliability & Fault Tolerance

15 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Nodes may fail → must ensure fail-safe designs.

Requires backup planning and redundancy.

Security Constraints

Fog layer closer to user = more vulnerable.

Must balance encryption overhead with performance.

Conclusion:

Fog environments require real-time, adaptive optimization strategies to ensure

performance, availability, and security without overloading the system.

16 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

5.b) Evaluation Metrics in C2F2T Systems (7 Marks)

Introduction:

Evaluating integrated Cloud-Fog-IoT (C2F2T) systems requires a set of performance

and reliability metrics to analyze system effectiveness and trade-offs.

1. Performance

Measured by:

Latency: Total processing + transmission delay

Execution Time: Time taken to complete all tasks

Throughput: No. of successful operations per second

2. Energy Consumption

Critical for IoT & edge devices

Measured in 3 modes:

Active Mode

Idle Mode

Transition Time

Goal: Maximize battery life without loss of performance

3. Resource Usage

CPU usage

Memory consumption

Bandwidth utilization

Important to avoid overload and failures

4. Cost

Based on:

Cloud usage time

17 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Bandwidth consumption

Energy cost

Higher processing = higher cost

5. Quality of Service (QoS)

Measures ability to meet:

Deadlines

Bandwidth limits

Task success rate

Often broken into smaller service modules and measured separately

6. Security

Ensures:

Authentication of device

Confidentiality & integrity of data

Must comply with privacy regulations (e.g., GDPR)

7. Reliability

Measured using:

Failure rate

Redundancy level

Recovery time

High reliability is needed for critical applications (e.g., healthcare)

Conclusion:

A holistic evaluation using these metrics helps ensure that C2F2T systems meet the
performance, security, and economic expectations of users and providers.

18 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

UNIT III

6. a) Design Issues in Middleware for Fog and Edge

Computing (7 Marks)
Introduction:

Middleware in Fog and Edge Computing acts as a software layer between IoT
devices, applications, and cloud/fog infrastructure. Its goal is to ensure seamless
communication, security, task scheduling, and data processing. However, designing
such middleware is challenging.

Key Design Issues:

Real-Time Context Awareness

Middleware must process data with minimum delay.

Need to monitor sensor environments in real-time.

Example: Smart traffic lights adjusting in real-time.

Limited Resources at the Edge

Edge devices have limited power, memory, and compute capacity.

Middleware must work efficiently without consuming much power or

bandwidth.

Dynamic Network Conditions

Edge networks often face disconnections, low bandwidth, and node

mobility.

Middleware must support adaptive routing and task rescheduling.

Mobility and Heterogeneity

Devices may move across locations.

Devices differ in hardware, OS, capabilities.

Middleware must handle device discovery and handovers.

Task Scheduling and Execution

Needs to decide which node executes which task.

19 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Must consider energy, load, latency, and availability.

Security and Privacy

Sensitive data from users and devices needs authentication and

encryption.

Middleware must support lightweight, secure communication

protocols.

Scalability

Middleware should handle large-scale deployment with thousands of

edge devices.

Load balancing and distributed architecture are necessary.

Fault Tolerance

Middleware should ensure service continuity despite failures.

Backup mechanisms and alternate task paths must be available.

Conclusion:

Middleware design in Fog and Edge Computing must ensure adaptability,

performance, and security in highly distributed, resource-constrained, and real-time
environments.

20 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

6. b) State-of-the-Art Middleware Infrastructures for

Fog and Edge Computing (7 Marks)
Introduction:

Several middleware platforms have been developed to handle task scheduling, data
analytics, context awareness, mobility, and security in Fog and Edge systems. These
infrastructures support real-time applications in smart cities, healthcare, and mobile
networks.

Notable Middleware Infrastructures:

Google Fit

Cloud-based IoT middleware.

Collects health & fitness data from smartphones and wearables.

Provides APIs for activity tracking and analytics.

M-Sense

Middleware for mobile sensing.

Offers APIs for sensor discovery and data collection.

Lightweight and optimized for smartphones.

Global Sensor Networks (GSN)

Middleware for distributed sensor networks.

Provides data abstraction, integration, and processing.

Uses SQL-like queries for ease of access.

FemtoCloud

Configures mobile edge devices to offer local computing services.

Smartphones share processing, storage, and networking.

PO³ (Process on Our Own)

Performs local data processing on devices.

Reduces cloud dependency.

21 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Suitable for bandwidth-constrained environments.

CoTWare

Integrates IoT–Fog–Cloud using cloud-hosted services.

Supports task offloading and service orchestration.

MobiPADs & MobiCon

Focus on context-aware computing.

Adapt middleware behavior based on user location or behavior.

Improve decision-making for mobile services.

CloudAware

Adaptive middleware for managing cloudlet connectivity.

Supports handover, service continuity, and reliability.

Features Supported:

Context monitoring

Mobility support

Security management

Dynamic task scheduling

Distributed execution

Conclusion:

These advanced middleware infrastructures provide efficient, scalable, and secure

computing for next-gen fog and edge systems, supporting real-time and location-
sensitive applications.

22 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

7. a) Role of Lightweight Container Middleware in Edge

Cloud (7 Marks)
Introduction:

Edge cloud systems consist of resource-constrained devices that require efficient

software execution. Traditional VMs are too heavy.
Lightweight containers (like Docker, LXC, Kata Containers) offer a portable, fast, and
resource-efficient middleware solution.

Role and Benefits:

Resource Optimization

Containers use shared OS kernel, reducing memory and CPU usage.

Ideal for edge devices like Raspberry Pi, gateways.

Fast Deployment

Containers start faster than VMs.

Allows real-time services to scale up or down instantly.

Portability

Containers run identically across platforms.

Developers build once and deploy anywhere—laptop, cloud, fog node.

Application Isolation

Each container runs in its own sandbox.

Enhances security and prevents app conflicts.

Support for Orchestration

Tools like Kubernetes, Docker Swarm, or Cloudify manage:

Load balancing

Scaling

Updates

23 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Enables smooth service orchestration in distributed edge clouds.

Support for Microservices

Containers allow breaking apps into independent services.

Each service can be deployed, updated, and managed independently.

Edge AI Support

Containers run ML models locally for real-time decisions (e.g., object

detection in surveillance).

Middleware Capabilities

Middleware based on containers can:

Handle task offloading

Manage device discovery

Offer runtime environments

Use Cases:

IoT Gateways for sensor data processing

Video analytics at edge

Smart city data aggregation

Caching and streaming (e.g., YouTube edge caching)

Conclusion:

Lightweight container middleware forms the foundation of efficient, modular, and

scalable Edge Cloud Architectures, especially for real-time, low-power applications.

24 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

7. b) Security Challenges in Edge Cloud Architectures (7

Marks)
Introduction:

Edge cloud environments are highly distributed and involve multiple stakeholders,
including device owners, service providers, and network operators. This makes
security complex and essential.

✅Key Security Challenges:

Device Authentication

Verifying identity of newly connected or mobile devices.

Risk: Unauthorized devices may access sensitive data.

Data Privacy

Location and sensor data can expose user behavior.

Protection required against data leakage and identity tracking.

Data Integrity

Data must remain untampered during transmission.

Ensured through digital signatures, hashing, and blockchain.

Access Control

Devices must access only authorized services.

Policy-Based Access Control (PBAC) ensures permission management.

Secure Communication

Data should be encrypted using lightweight cryptography due to

limited power.

Protocols like MQTT + TLS or HTTPS used.

Lightweight Security

Security protocols must work efficiently on low-resource devices (e.g.,

sensors).

25 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Lightweight authentication and encryption algorithms needed.

VM and Container Security

Containers must be isolated to prevent intra-host attacks.

Tools like gVisor, Kata Containers, or AppArmor help secure

environments.

Blockchain-Based Security

Used for:

Device identity verification

Data provenance

Smart contracts for rule enforcement

Software Updates

Delayed or unauthorized updates may introduce malware or

vulnerabilities.

Role of Blockchain:

Immutability: No data tampering

Transparency: Every transaction is recorded

Smart Contracts: Automates service agreements and security policies

Identity Management: Prevents spoofing of edge nodes

Conclusion:

Edge cloud systems must implement lightweight, scalable, and decentralized

security mechanisms, including blockchain, encryption, and device identity checks,
to ensure safe operation across highly dynamic environments.

26 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

UNIT IV

8.a) Explain how Fog Computing improves Video

Stream Processing (7 Marks)
Introduction:

Traditional video surveillance systems send all video to the cloud, causing delays,
network congestion, and privacy issues.
Fog computing processes video locally at the edge, enabling real-time object
detection, tracking, and alert generation.

Fog-Based Video Stream Architecture:

Video Capture:

CCTV or IP cameras capture live video.

Fog Node/Edge Device:

Devices like NVIDIA Jetson, Raspberry Pi, or fog-enabled routers

analyze video frames locally.

Local Processing:

AI algorithms (YOLO, OpenCV, TensorFlow) detect humans, motion,

suspicious activity.

Decision Making:

Alerts generated in real time and forwarded to security personnel.

Cloud (Optional):

Stores long-term footage or performs detailed analytics.

27 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Benefits of Fog in Video Stream Processing:

Benefit Description
Real-Time Analytics Detects intrusions, loitering, or sudden movements instantly
Low Latency Immediate alerts without cloud dependency
Bandwidth Saving Only relevant events are sent to the cloud
Enhanced Security Keeps raw footage local, protecting privacy
Fast Object Tracking Tracks objects (e.g., people, vehicles) frame by frame using AI
On-Site AI Inference Allows facial recognition, people counting, gesture detection

Real-World Example:

In a smart city, edge cameras detect and track people near restricted zones.
Fog devices process video and alert authorities within seconds.

Conclusion:

Fog computing transforms video surveillance by enabling distributed, efficient, and

secure stream processing with real-time responsiveness.

28 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

8.b) Discuss key issues such as Latency, Bandwidth, and

Computational Efficiency (7 Marks)
Introduction:

Fog computing aims to bring computation closer to data sources. However, several
performance-critical issues must be managed to ensure system effectiveness.

1. Latency

Definition: Time taken from input (sensor event) to system response.

Issue: Cloud dependency causes delay.

Solution: Fog processing reduces this delay by executing tasks closer to

sensors.

Use Case: In health monitoring, delays in fall detection can be fatal.

2. Bandwidth

Definition: Volume of data transmitted over a network.

Issue: Sending raw data to the cloud increases bandwidth usage and costs.

Solution: Fog filters data locally, sending only important info.

Example: Video surveillance sends only alerts, not full video, to the cloud.

3. Computational Efficiency

Definition: Effective use of limited computing resources.

Issue: Edge devices are resource-constrained (battery, CPU, RAM).

Solution: Use of lightweight containers, smart scheduling, and edge-friendly

AI.

Example: Docker containers run efficient microservices on fog nodes.

29 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Additional Considerations:

Issue Description

Energy Efficiency IoT devices must conserve battery life

Security Overhead Encryption can slow down low-powered devices

Scalability Managing thousands of devices requires dynamic load balancing

Task Scheduling Assigning tasks optimally to fog/cloud nodes is complex

Conclusion:

Latency, bandwidth, and computational efficiency are critical factors in Fog

Computing. Their proper management ensures real-time, reliable, and scalable
systems for smart environments.

30 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

9.a) Major Issues and Challenges in Implementing Fog

Computing for Real-World Applications (7 Marks)
Introduction:

Fog computing promises real-time services, low latency, and efficiency, but real-
world deployments face several technical, economic, and operational challenges.

1. Resource Constraints

Fog nodes have limited compute power compared to cloud.

Challenge: Running advanced AI/ML models on small devices.

2. Heterogeneity

Different devices, OS, protocols create interoperability issues.

Difficult to build one-size-fits-all fog infrastructure.

3. Security & Privacy

Data is closer to users → increased attack surface.

Authentication, encryption, and access control become critical.

4. Mobility and Reliability

IoT devices and fog nodes (e.g., in vehicles) are mobile.

Challenge: Maintaining service continuity during handovers.

5. Real-Time Processing

Time-sensitive applications like smart healthcare or traffic control require

low delay.

Scheduling and load balancing become complex.

6. Scalability

As IoT grows, managing thousands of fog nodes is hard.

Needs self-configuring and adaptive orchestration systems.

7. Deployment Cost

31 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Initial setup for hardware, maintenance, training is high.

Economic feasibility for small orgs is a concern.

8. Standardization

No unified global framework or protocol for fog computing yet.

Leads to vendor lock-in or integration failures.

Conclusion:

Fog computing faces real-world issues in performance, integration, security, and

economics. Overcoming these will enable its widespread adoption across industries.

32 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

9.b) Explain how Fog Computing Enhances Decision

Making in Transportation (7 Marks)
🔹 Introduction:

Transportation systems require fast, context-aware, and decentralized decision-

making. Fog computing provides this by enabling real-time analytics at the edge.

Key Ways Fog Improves Decision-Making in Transportation:

Local Processing of Vehicle Data

Analyzes data from GPS, traffic signals, and sensors on-site.

Detects traffic congestion and redirects vehicles instantly.

Accident Detection & Emergency Alerts

Fog nodes in traffic poles detect collisions and send instant alerts to
ambulances and control centers.

Smart Traffic Light Control

Fog-enabled signals dynamically adjust timing based on vehicle

density.

Low-Latency Vehicle Communication (V2X)

Vehicles communicate with roadside fog nodes for coordinated

driving, improving safety.

Adaptive Parking Systems

Fog nodes detect available slots and guide drivers to the nearest
parking area.

CEAL Benefits in Transportation:

Feature Explanation
C - Computing Efficiency Local data analysis → faster routing
E - Energy Saving Saves energy on both vehicles and networks
A - Adaptive Scalability Fog nodes added as city expands
L - Low Latency Quick decisions in emergencies

33 Karthik kodali
 CD325 - FOG AND EDGE COMPUTING

Example Scenario:

A traffic fog node detects an accident.

Notifies nearby vehicles to reroute.

Updates traffic lights to reduce congestion.

Logs incident to cloud for reporting and analytics.

Conclusion:

Fog computing makes transportation smarter, safer, and faster by allowing

distributed decision-making based on local, real-time data processing.

34 Karthik kodali