Unit I: Introduction to Fog and Edge Computing high fault tolerance.

high fault tolerance. After distributed computing, these increased the processing EC2, VMware vCloud, etc work on. Hardware virtualization is still one of the most
capabilities of the system. But these were very expensive. To reduce this cost, common types of virtualization.
Evolution of Cloud Computing – Limitations of Cloud for IoT – Need for Fog and Edge cluster computing came as an alternative to mainframe technology.
Computing – Definitions and Characteristics – Fog vs Edge vs Cloud Computing – Reference 7. Web 2.0
Architectures – Use cases and applications. 2. Distributed Systems(1970-1980) Web 2.0 is the interface through which the cloud computing services interact with
Distributed System is a composition of multiple independent systems but all of the clients. It is because of Web 2.0 that we have interactive and dynamic web
them are depicted as a single entity to the users. The purpose of distributed pages. It also increases flexibility among web pages. Popular examples of web 2.0
systems is to share resources and also use them effectively and include Google Maps, Facebook, Twitter, etc. Needless to say, social media is
efficiently. Distributed systems possess characteristics such as scalability, possible because of this technology only. It gained major popularity in 2004.
concurrency, continuous availability, heterogeneity, and independence in failures. 8. Service Orientation
Evolution of Cloud Computing But the main problem with this system was that all the systems were required to
The phrase "Cloud Computing" was first introduced in the 1950s to describe A service orientation acts as a reference model for cloud computing. It supports
be present at the same geographical location. Thus to solve this problem, low-cost, flexible, and evolvable applications. Two important concepts were
internet-related services, and it evolved from distributed computing to the modern distributed computing led to three more types of computing and they were-
technology known as cloud computing. Cloud services include those provided by introduced in this computing model. These were Quality of Service (QoS) which also
Mainframe computing, cluster computing, and grid computing. includes the SLA (Service Level Agreement) and Software as a Service (SaaS).
Amazon, Google, and Microsoft. Cloud computing allows users to access a wide
range of services stored in the cloud or on the Internet. Cloud computing services 3. Cluster Computing(1980-1990) Cloud Computing
include computer resources, data storage, apps, servers, development tools, and In 1980s, cluster computing came as an alternative to mainframe computing. Each Cloud Computing means storing and accessing the data and programs on remote
networking protocols. machine in the cluster was connected to each other by a network with high servers that are hosted on the internet instead of the computer’s hard drive or local
bandwidth. These were way cheaper than those mainframe systems. These were server. Cloud computing is also referred to as Internet-based computing, it is a
equally capable of high computations. Also, new nodes could easily be added to the technology where the resource is provided as a service through the Internet to the
cluster if it was required. Thus, the problem of the cost was solved to some extent user. The data that is stored can be files, images, documents, or any other storable
but the problem related to geographical restrictions still pertained. To solve this, document.
the concept of grid computing was introduced.
4. Grid Computing(1990-2000) Cloud Computing Models
In 1990s, the concept of grid computing was introduced. It means that different
systems were placed at entirely different geographical locations and these all were Cloud computing is categorized into three main models based on accessibility and
connected via the internet. These systems belonged to different organizations and ownership:
thus the grid consisted of heterogeneous nodes. Although it solved some problems 1. Public Cloud
but new problems emerged as the distance between the nodes increased. The main
 Hosted by third-party providers like AWS, Google Cloud, and Microsoft
problem which was encountered was the low availability of high bandwidth
Azure.
connectivity and with it other network associated issues. Thus. cloud computing is
 Cost-effective but may pose security concerns.
often referred to as "Successor of grid computing".
2. Private Cloud
5. Utility Computing(Late 1990-2000)  Used exclusively by a single organization for better security and
Utility Computing is a computing model that defines service provisioning
customization.
techniques for services such as compute services along with other major services
 More expensive but provides greater control.
such as storage, infrastructure, etc which are provisioned on a pay-per-use basis.
3. Hybrid Cloud
6. Virtualization(1980-Present)  A combination of public and private clouds, balancing cost-efficiency
1. Mainframe Computing(1950-1970) Virtualization was introduced nearly 40 years back. It refers to the process of with security.
Mainframes which first came into existence in 1951 are highly powerful and creating a virtual layer over the hardware which allows the user to run multiple  Ideal for businesses needing scalability and data protection.
reliable computing machines. These are responsible for handling large data such as instances simultaneously on the hardware. It is a key technology used in cloud
massive input-output operations. Even today these are used for bulk processing computing. It is the base on which major cloud computing services such as Amazon 4. Community Cloud
tasks such as online transactions etc. These systems have almost no downtime with

 A community cloud is a shared cloud infrastructure which is designed to  Reliability Cloud-based IoT faces critical limitations due to high latency, dependency on
meet the shared needs of multiple organizations like security,  High Accessibility continuous internet connectivity, and significant security/privacy risks when
compliance or performance.  Scalable transmitting sensitive data. The high cost of bandwidth, data storage, and the
 Ideal for Healthcare services, Educational Collaboration, Government Disadvantages of Cloud Computing complexity of securing distributed, diverse devices often challenge scalability. These
Agencies, Financial Institutions etc.  Internet Dependency bottlenecks often lead to unreliable real-time performance and potential system
Types of Cloud Computing Services  Issues in Security and Privacy failures.
 Data Breaches
MDPI +5
 Limitations on Control
Key limitations of using the cloud for IoT include:
 High Latency and Bandwidth Issues: Real-time processing is difficult
because data must travel to a remote server and back, which is unsuitable for
time-sensitive applications. Large amounts of data generated by devices can
overwhelm network capacity.

 Security and Privacy Risks: Data transferred to the cloud is susceptible to

interception and hacking. IoT devices often have limited security, making
them vulnerable to malware and Distributed Denial of Service (DDoS) attacks.

 Dependence on Internet Connectivity: If the internet connection fails, IoT

devices may become inoperable, causing operational failures.

 Cost and Scalability Constraints: While cloud is flexible, long-term costs for
data storage, processing, and network bandwidth can become expensive as
devices scale.

Cloud services are divided into three categories based on functionality:  Interoperability Challenges: Different IoT devices may use varying protocols
and standards, making seamless integration with a central cloud platform
1. Infrastructure as a Service (IaaS) difficult.
 Provides virtualized computing resources over the internet.
 Example: AWS EC2, Google Compute Engine, Microsoft Azure Virtual  Regulatory and Compliance Issues: Storing sensitive data in the cloud
Machines. raises compliance issues with regulations like GDPR, particularly regarding
2. Platform as a Service (PaaS) data location and privacy.
 Offers a managed platform with built-in tools for developers. MDPI +8
 Example: AWS Elastic Beanstalk, Google App Engine, Heroku.
3. Software as a Service (SaaS)
 Delivers ready-to-use software applications over the internet.
 Example: Google Drive, Dropbox, Salesforce.
Advantages of Cloud Computing
 Cost Saving
 Data Redundancy and Replication
 Ransomware/Malware Protection
 Flexibility
Need for Fog and Edge Computing – Definitions and 2⃣ Definition of Fog Computing 5. Resource-Constrained Environment – Limited storage and processing compared to
cloud.
Characteristics 6. Decentralized Architecture – Distributed intelligence across devices.
Fog Computing is a distributed computing paradigm that extends cloud computing services
With the rapid growth of the Internet of Things (IoT), billions of devices generate massive to the edge of the network. It acts as an intermediate layer between cloud data centers and
volumes of data. Traditional Cloud Computing processes this data in centralized data centres, edge devices.
which leads to challenges such as high latency, bandwidth congestion, and limited real-time
responsiveness. Introduced by Cisco Systems. 6⃣ Difference Between Fog and Edge (Short Comparison)
Feature Fog Computing Edge Computing
To overcome these limitations, Fog Computing and Edge Computing were introduced. Key Idea:
Location Between cloud and edge At the data source
1⃣ Need for Fog and Edge Computing Cloud → Fog Layer → Edge Devices
Layer Intermediate layer Device layer
Fog nodes may include routers, switches, gateways, or micro data centers.
1. Real-Time Processing Processing Power Moderate Limited

Applications such as autonomous vehicles, healthcare monitoring, and smart grids require Example Smart gateway Sensor, IoT device
immediate data processing. Sending data to distant cloud servers increases latency. Fog and
Edge computing bring computation closer to the data source. 3⃣ Characteristics of Fog Computing

2. Reduced Latency 1. Geographical Distribution – Fog nodes are distributed across multiple locations.
2. Low Latency – Processes data closer to users than cloud. 7⃣ Conclusion
Time-sensitive applications (e.g., industrial automation, smart traffic systems) need 3. Location Awareness – Supports mobility and location-based services.
millisecond-level responses. Edge and Fog nodes process data locally to minimize delay. 4. Real-Time Interaction – Suitable for time-sensitive applications. Fog and Edge computing complement Cloud computing by enabling low-latency,
5. Interoperability – Works with heterogeneous devices and networks. distributed, scalable, and real-time data processing. They are essential for modern IoT-
3. Bandwidth Optimization 6. Hierarchical Architecture – Multi-layer (Cloud–Fog–Edge). driven applications such as smart cities, Industry 4.0, autonomous systems, and healthcare
monitoring.
IoT devices generate huge data streams. Transmitting all raw data to the cloud increases
network congestion and cost. Local processing reduces unnecessary data transmission.
4⃣ Definition of Edge Computing
4. Improved Reliability
Edge Computing is a computing model where data processing is performed directly at or
If internet connectivity fails, local edge devices can continue operating independently without very near the data source (IoT devices, sensors, smart devices) instead of sending it to
complete reliance on cloud servers. centralized cloud servers.

5. Enhanced Security & Privacy Key Idea:

Sensitive data (health, finance, surveillance) can be processed locally instead of being sent to Processing happens at the edge of the network, closest to data generation.
remote cloud servers, reducing exposure risk.

6. Scalability for IoT Ecosystem

With the exponential growth of IoT devices, decentralized architectures provide better 5⃣ Characteristics of Edge Computing
scalability compared to centralized cloud systems.
1. Ultra-Low Latency – Immediate processing at device level.
2. Reduced Bandwidth Usage – Only processed data is sent to cloud.
3. Autonomous Operation – Can function without continuous cloud connectivity.
4. Enhanced Privacy – Sensitive data remains local.

Fog vs Edge vs Cloud Computing Cloud Use Cases & Applications  Industrial robotics
 Smart surveillance systems
 Wearable health monitoring
 Big data analytics
Reference Architectures – Use Cases and Applications  AI/ML model training
 AR/VR applications
 Video streaming platforms
Modern distributed systems combine Cloud, Fog, and Edge computing to meet the  Enterprise applications
demands of the Internet of Things (IoT), AI, smart infrastructure, and Industry 4.0. Below is a  Backup and disaster recovery
structured explanation suitable for academic or university exam preparation. 3⃣ Fog Computing

Definition
2⃣ Edge Computing
1⃣ Cloud Computing Fog Computing extends cloud services closer to the edge of the network, acting as an
intermediate layer between edge devices and cloud data centers.
Definition
Definition
It was introduced by Cisco Systems.
Edge Computing is a decentralized computing paradigm where data processing occurs at or
Cloud Computing is a centralized computing model where data storage, processing, and very near the data source (sensors, IoT devices, gateways).
services are delivered over the internet from large-scale data centers.

Reference Architecture of Fog Computing

Reference Architecture of Edge Computing
Reference Architecture of Cloud Computing Layers:
Layers:
Layers: 1. Edge Layer
o IoT devices
1. Device Layer (Edge Nodes)
1. Client Layer o Sensors
o Sensors
o End users, web/mobile applications 2. Fog Layer (Intermediate Nodes)
o IoT devices
2. Service Layer o Routers
o Smart cameras
o IaaS (Infrastructure as a Service) o Switches
o Embedded systems
o PaaS (Platform as a Service) o Gateways
2. Edge Processing Layer
o SaaS (Software as a Service) o Micro data centers
o Local analytics
3. Resource Layer 3. Cloud Layer
o AI inference
o Virtual machines o Centralized storage and analytics
o Filtering and preprocessing
o Storage systems 3. Cloud Integration Layer (Optional)
o Databases o Sends summarized/filtered data to cloud
Architecture Nature:
4. Physical Layer
o Data centers  Distributed
Architecture Nature:
o Servers  Hierarchical (Cloud–Fog–Edge)
o Networking infrastructure  Supports mobility
 Highly decentralized
 Moderate computing capability
 Ultra-low latency
Architecture Nature:  Limited computing resources
 Real-time processing
 Centralized
 High computational power
 High storage capacity
Fog Use Cases & Applications
 Higher latency compared to edge/fog
Edge Use Cases & Applications 

Smart traffic management
Smart grid systems
 Connected healthcare
 Autonomous vehicles
  Industrial automation Together, they form the backbone of next-generation IoT, AI-driven systems, and Industry Tool / Platform Description Real-Life Use
 Smart city infrastructure 4.0 applications.
Industrial IoT edge-fog platform Factory machinery sensors monitor
FogHorn Lightning
1⃣ Cloud Computing Tools & Software for real-time analytics and control equipment

4⃣ Comparison: Fog vs Edge vs Cloud Purpose: Centralized storage, computing, and application hosting Edge/fog gateway for IoT data Smart buildings monitor energy
Dell Edge Gateway
Feature Cloud Fog Edge aggregation usage locally before cloud upload
Popular Cloud Platforms:
Location Central data center Between cloud & edge At data source OpenFog Frameworks for fog-based IoT
Research and enterprise IoT testbeds
Tool / Platform Description Real-Life Use Consortium Tools deployments
Latency High Low Very Low
Amazon Web Full cloud platform with compute, Netflix streams movies via
Processing Power Very High Moderate Limited Services (AWS) storage, AI, and databases AWS servers Key Features:
Storage Massive Moderate Small Cloud platform for apps, storage, Enterprise apps, university  Local processing reduces latency
Microsoft Azure  Aggregates IoT data before sending to cloud
analytics LMS hosting
Scalability Very High High Moderate  Supports real-time analytics
Google Cloud Cloud services for data storage, AI, and
Real-Time Support Limited Good Excellent Gmail, Google Drive
Platform (GCP) analytics
Architecture Centralized Distributed Decentralized
IBM Cloud
Enterprise cloud services including AI Large-scale enterprise 3⃣ Edge Computing Tools & Software
and blockchain applications
Purpose: On-device or very near-device computing for real-time processing
5⃣ Integrated Architecture (How They Work Together)
Key Features: Popular Edge Tools:
In modern systems:
 Centralized compute & storage Tool / Platform Description Real-Life Use
 Scalability and global availability
 Edge performs real-time data filtering and immediate response.
 Managed services for databases, AI, analytics NVIDIA Jetson / Tesla processes autonomous driving
 Fog performs aggregation, short-term analytics, and coordination. AI computing on edge devices
 Cloud performs deep analytics, AI model training, long-term storage. JetPack SDK sensor data on-board

Example: Smart City Runs local compute and Smart home IoT devices process
AWS IoT Greengrass
2⃣ Fog Computing Tools & Software messaging on edge devices actions locally
 Edge → Traffic sensors detect congestion
 Fog → Local traffic control optimization Purpose: Intermediate layer for processing data closer to IoT devices before sending it to the Microsoft Azure IoT Edge modules to run AI, Azure Industrial sensors controlling
 Cloud → Long-term traffic pattern analysis cloud Edge services on devices equipment locally

Popular Fog Platforms: Edge TPU devices for fast AI Security cameras detect motion
Google Coral
inference instantly
6⃣ Conclusion Tool / Platform Description Real-Life Use

Cloud, Fog, and Edge computing are complementary rather than competitive technologies. Cisco IOx / Edge Runs applications on routers or Smart city traffic lights process Key Features:
Fog Nodes gateways vehicle data locally
 Cloud is best for heavy computation and storage.  Ultra-low latency decision-making
 Fog bridges edge and cloud for distributed intelligence.  Processes data without full cloud dependency
 Edge enables real-time, low-latency processing.  Often uses AI/ML locally on device

📌 Quick Summary Table

Computing
Example Software / Tool Real-Life Use
Type

Cloud AWS, Azure, GCP Netflix, Gmail

Fog Cisco IOx, FogHorn, Dell Edge Gateway Smart traffic lights, factories

NVIDIA Jetson, AWS Greengrass, Azure IoT Tesla self-driving, smart

Edge
Edge cameras
Fog Computing machine. After this gained a little popularity, IBM, in 2015, coined a similar
term called "Edge Computing".
 It improves the overall security of the system as the data resides
close to the host.
Types of Fog Computing  It provides better privacy as industries can perform analysis on
Fog computing also known as fog networking or fogging, is a decentralized their data locally.
 Device-level Fog Computing: Device-level fog computing
computing architecture that brings cloud computing capabilities to the
network's edge. This method intends to increase efficiency, minimize
utilizes low-power technology, including sensors, switches, and Disadvantages of Fog Computing
routers. It can be used to collect data from these devices and  Congestion may occur between the host and the fog node due to
latency, and improve data processing capabilities. In this article, we will see
upload it to the cloud for analysis. increased traffic (heavy data flow).
concepts of fog computing in detail.
 Edge-level Fog Computing: Edge-level fog computing utilizes  Power consumption increases when another layer is placed
What is Fog Computing? network-connected servers or appliances. These devices can be between the host and the cloud.
Fog Computing is the term introduced by Cisco that refers to used to process data before it is uploaded to the cloud.  Scheduling tasks between host and fog nodes along with fog
extending cloud computing to an edge of the enterprise's network. Thus, it is  Gateway-level Fog Computing: Fog computing at the gateway nodes and the cloud is difficult.
also known as Edge Computing or Fogging. It facilitates the operation of level uses devices to connect the edge to the cloud. These  Data management becomes tedious as along with the data
computing, storage, and networking services between end devices and devices can be used to control traffic and send only relevant data stored and computed, the transmission of data
computing data centers. to the cloud. involves encryption-decryption too which in turn release data.
 Cloud-level Fog Computing: Cloud-level fog computing uses Applications of Fog Computing
cloud-based servers or appliances. These devices can be used  It can be used to monitor and analyze the patients' condition. In
to process data before it is sent to end users. case of emergency, doctors can be alerted.
Components of Fog Computing  It can be used for real-time rail monitoring as for high-speed
 Edge devices: Edge devices are the network devices nearest to trains we want as little latency as possible.
the data source. Edge devices consist of sensors, PLCs  It can be used for gas and oils pipeline optimization. It generates
(programmable logic controllers), and gateway routers. a huge amount of data and it is inefficient to store all data into the
 Data Processing: Data processing occurs locally on edge cloud for analysis.
devices rather than being routed to a central location for Difference Between Edge Computing and Fog Computing
processing. The end effect is greater performance and
lower latency. Edge Computing Fog Computing
 Data Storage: in Data storage. Instead of transferring data to a
central place, edge devices can keep information locally. This Highly scalable when compared to edge
Less scalable than fog computing.
increases security and privacy while lowering latency. computing.
 Connectivity: For fog computing to work, edge devices must be
connected to the rest of the network at high speeds. This can be
done using wired or wireless methods. Millions of nodes are present. Billions of nodes are present.
Fog Computing
When to Use Fog Computing?
 It is used when only selected data is required to send to the
cloud. This selected data is chosen for long-term storage and is Nodes in this computing are installed closer to
 The devices comprising the fog infrastructure are known as fog Nodes are installed far away from
less frequently accessed by the host. the cloud(remote database where data is
nodes. the cloud.
 It is used when the data should be analyzed within a fraction of stored).
 In fog computing, all the storage capabilities, computation
capabilities, data along with the applications are placed between seconds i.e Latency should be low.
the cloud and the physical host.  It is used whenever a large number of services need to be
 All these functionalities are placed more towards the host. This provided over a large area at different geographical locations. Edge computing is a subdivision of Fog computing is a subdivision of cloud
makes processing faster as it is done almost at the place where  Devices that are subjected to rigorous computations and fog computing. computing.
data is created. processings must use fog computing.
 It improves the efficiency of the system and is also used to  Real-world examples where fog computing is used are in IoT
ensure increased security. devices Devices with Sensors, Cameras (IIoT-Industrial Internet of The bandwidth requirement is very The bandwidth requirement is high. Data
Things), etc. low. Because data comes from the originating from edge nodes is transferred to
History of Fog Computing
The term fog computing was coined by Cisco in January 2014. This was Advantages of Fog Computing edge nodes themselves. the cloud.
because fog is referred to as clouds that are close to the ground in the same  This approach reduces the amount of data that needs to be sent
way fog computing was related to the nodes which are present near the to the cloud.
nodes somewhere in between the host and the cloud. It was intended to  Since the distance to be traveled by the data is reduced, it results Operational cost is higher. Operational cost is comparatively lower.
bring the computational capabilities of the system close to the host in saving network bandwidth.
 Reduces the response time of the system.

Edge Computing Fog Computing

architecture. Fog architecture involves the distribution of functions at different layers, 2. Fog Layer
the types and the number of protocols used, and the constraints imposed at various  Fog layer includes devices like routers, gateways, access points, base stations, specific
High privacy. Attacks on data are
The probability of data attacks is higher.
very low. layers. fog servers, etc., called as Fog nodes.

 Fog nodes are located at the edge of a network. An edge can be a hop distance from
Edge devices are the inclusion of
the IoT devices or client’s network.
Fog is an extended layer of cloud. Hierarchical Fog Computing Architecture the end device. The Fog nodes are situated in-between end devices and cloud data
The hierarchical fog architecture comprises of following three layers:
centers.
The power consumption of nodes filter
The power consumption of nodes is important information from the massive  Fog nodes can be static, e.g., located in a bus terminal or coffee shop, or they can be
low. amount of data collected from the device and
moving, e.g., fitted inside in a moving vehicle.
saves it in the filter high.
 Fog nodes ensure services to the end devices. Fog nodes can compute, transfer and

Edge computing helps devices to get Fog computing helps in filtering important store the data temporarily.
faster results by processing the data information from the massive amount of data
simultaneously received from the collected from the device and saves it in the  Fog nodes and cloud data center connections are enabled by the IP core networks,
devices. cloud by sending the filtered data.
providing interaction and cooperation with the cloud for enhancing processing and

Conclusion storage capabilities.

Finally, fog computing delivers cloud capabilities to the edge of networks,
increasing efficiency, lowering latency, and improving data processing
capabilities. It is perfect for real-time data analysis, low-latency applications 3. Cloud Layer
such as IoT, and situations where data privacy and security are critical.
While it provides scalability and lower bandwidth usage, it also has issues  This layer consists of devices that can provide large storage and machines (servers)
in managing data congestion and increasing power consumption. Fog 1. Terminal Layer
computing is making progress in applications such as healthcare with high performance.
monitoring, industrial IoT, and real-time analytics across a variety of
 The terminal layer is the basic layer in fog architecture, this layer includes devices like  This layer performs computation analysis and stores data permanently, for back-up
industries.
Comment
mobile phones, sensors, smart vehicles, readers, smartcards, etc. and permanent access to the users.
Introduction to Fog Computing  The devices which can sense and capture data are present in this layer. Devices are  This layer has high storage and powerful computing capabilities.
Architecture distributed across a number of locations separated far apart from each other.  Enormous data centers with high computing abilities form a cloud layer. The data
Fog architecture involves using services of end devices (switches, routers,  The layer mostly deals with data sensing and capturing. Devices from different centers provide all the basic characteristics of cloud computing to the users. The

multiplexers, etc) for computational, storage and processing purposes. Fog platforms and different architectures are mainly found in this layer. data centers are both scalable and provide compute resources on-demand basis.

computing architecture consists of physical as well as logical elements of the  The devices have the property of working in a heterogeneous environment, with  The cloud layer lies at the extreme end of the overall fog architecture. It acts as a

other devices from separate technologies and separate modes of communication. back-up as well as provides permanent storage for data in a fog architecture. Usually,
network, software, and hardware to form a complete network of a large number of

interconnecting devices. Fog node distribution (physical as well as geographical, data that isn’t required at the user proximity is stored in a cloud layer.

along with the topology and protocols used form key architectural features of a fog
Layered Fog Computing Architecture
The layered fog architecture consists of six layers:  The fog nodes are checked for their energy consumption, the amount of battery 6. Transport Layer
power they consume while performing their tasks.
 The primary function of this layer is to upload partly-processed and fine-grained

secure data to the cloud layer for permanent storage. For efficiency purposes, the
3. Pre-processing Layer
portion of data is collected and uploaded.
 This layer performs various data operations mainly related to analysis.
 The data is passed through smart-gateways before uploading onto the cloud. The
 Data is cleaned and checked for any unwanted data present. Data impurity is
communication protocols used are chosen to be lightweight, and efficient, because
removed and only useful data is collected.
of the limited resources of fog computing.
 Data analysis at this layer can involve mining meaningful and relevant information

from a vast amount of data collected by the end devices.

Conclusion
 Data analysis is one of the essential features that should be taken into consideration Fog architecture is a model which comprises of a number of different layers. The

before data is used for a specific purpose. model serves as an ideal for the construction of a fog network. The architecture

provides an overview of different functionalities that are performed by respective

4. Temporary Storage
1. Physical and Virtualization Layer layers. The protocols used at different layers, the particular devices that are used at
 This layer is associated with non-permanent distribution and replication of data.
 This layer comprises nodes (Physical and virtual). The nodes perform the primary different layers and their functionalities, specifications are identified by going
 Storage virtualization like VSAN is used in this layer.
task of capturing data and are located at different locations. Nodes usually involve through the fog architecture.
 Data is removed from the temporary layer once data is moved to the cloud, from this
sensing technology to capture their surroundings.
layer.
 Sensors used at this node collect data from the surroundings and collect data which What is Edge Architecture?
 5. Security Layer
is then sent to upper layers via gateways for further processing. Edge architecture is a computing paradigm that processes data close to
its source. This reduces latency, enhances efficiency, and improves data
 A node can be a stand-alone device like a mobile phone or it can be a part of a large security by minimizing the need for long-distance data transmission.
 This layer is involved with the privacy of data, the integrity of data, encryption, and
Unlike traditional cloud computing, edge architecture decentralizes
device like a temperature sensor fitted inside a vehicle.
decryption of data. processing tasks. It is particularly beneficial in scenarios requiring real-
time data analysis. This article explores edge computing, its architecture,
 Privacy in the case of fog computing data can include use-based privacy, data-based components, benefits, and applications. We'll also discuss challenges,
2. Monitoring Layer security considerations, and comparisons with other architectures.
privacy, and location-based privacy.
 In this layer, we perform node monitoring related to various tasks.
 The security layer ensures secure and preservation of privacy for the data which is Important Topics to Understand Edge Architecture
 Nodes can be monitored for the amount of time they work, the temperature and  Understanding Edge Computing
outsourced to the fog nodes.
 Overview of Edge Architecture
other physical properties they are possessing, the maximum battery life of the  Key Components and Design Principles
 Benefits of Edge Architecture
device, etc.
 Use Cases and Applications of Edge Architecture
 Challenges and Solutions of Edge Architecture
 The performance of applications as well as their present state is also monitored.
 Security Considerations for Edge Architecture
 Comparison with Cloud and Other Architectures

Understanding Edge Computing Key Components of Edge Architecture: to central servers, lowering the expenses associated with data
Edge computing is a distributed computing paradigm that brings transfer and storage in the cloud.
computation and data storage closer to the location where it is needed.  Improved Reliability: Edge architecture ensures continuous
 IoT Devices and Sensors: Learn how to work with IoT devices
This proximity to data sources reduces latency, enhances performance, operation even when network connectivity is intermittent. Local
and sensors that generate data at the edge. Understand their
and ensures faster decision-making. Edge computing is especially useful processing allows critical applications to function without relying
capabilities, data generation methods, and integration
in applications requiring real-time data processing and low-latency on a constant connection to central servers.
techniques.
responses, such as autonomous vehicles, smart cities, and industrial  Scalability: Edge architecture can easily scale to accommodate
 Edge Nodes: Study the role of edge nodes in local data
automation. more devices and increased data volumes. New edge nodes can
processing. Learn about their hardware specifications,
be added as needed, without overwhelming the central
 Local Processing: Data is processed locally on edge devices or processing capabilities, and deployment scenarios.
infrastructure.
nearby servers, reducing the need to send data to a centralized  Edge Servers: Understand the function of edge servers in
cloud. storing and processing data locally. Familiarize yourself with Use Cases and Applications of Edge Architecture
different types of edge servers and their configurations. Here are the key use cases and applications of edge architecture:
 Real-Time Analysis: Immediate data analysis is possible,
enabling quick responses to changing conditions.  Edge Gateways: Learn how edge gateways manage data traffic  Healthcare: Edge computing supports real-time monitoring and
 Reduced Latency: By processing data closer to its source, edge between edge devices and servers. Understand their importance analysis of patient data through wearable devices and sensors.
computing minimizes latency, ensuring faster data retrieval and in ensuring efficient data flow and communication. This allows for immediate medical interventions and continuous
processing.  Communication Networks: Study the various communication health monitoring without relying on cloud connectivity.
 Bandwidth Optimization: Decreases the amount of data networks that connect edge devices, nodes, and servers. Learn  Manufacturing: In smart factories, edge computing enables
transmitted over networks, conserving bandwidth and reducing about network protocols, data transfer methods, and network predictive maintenance and real-time quality control. By
costs. topology. analyzing data from sensors on the production line,
 Scalability: Easily scales to accommodate numerous devices manufacturers can detect issues early and reduce downtime.
and vast amounts of data without overloading central servers. Design Principles of Edge Architecture:  Smart Cities: Edge architecture helps manage urban
 Enhanced Security: By keeping data local, edge computing infrastructure efficiently. Traffic management systems use local
reduces the risk of data breaches during transmission, improving  Decentralization: Understand the importance of decentralizing data to optimize signal timings and reduce congestion. Similarly,
overall security. processing tasks across multiple nodes. Learn how to design smart grids leverage edge computing for efficient energy
Overview of Edge Architecture systems that avoid single points of failure. distribution and monitoring.
Edge architecture involves distributing data processing tasks closer to  Scalability: Study how to design scalable edge architectures.  Retail: Retailers use edge computing for personalized shopping
where data is generated, rather than relying on centralized cloud servers. Learn about techniques to handle increasing data volumes and experiences and efficient inventory management. In-store
This decentralization helps in reducing latency, enhancing real-time data the addition of new devices. sensors track customer behavior and preferences, enabling
processing capabilities, and improving overall system efficiency. Edge  Interoperability: Learn how to ensure interoperability among targeted promotions and optimized stock levels.
architecture comprises various components, each playing a crucial role in diverse devices and communication protocols. Study standards  Autonomous Vehicles: Edge computing is crucial for the
managing and processing data locally. and best practices for seamless integration. operation of autonomous vehicles. It processes data from various
 Resilience: Understand the principles of designing resilient edge sensors and cameras in real-time, ensuring safe navigation and
 Edge Devices: These include IoT devices, sensors, and other quick response to environmental changes.
systems. Learn how to achieve high availability and fault
hardware that generate data at the edge of the network.  Agriculture: Edge devices in agriculture monitor soil conditions,
tolerance.
 Edge Nodes: Local processing units that handle initial data weather, and crop health. Farmers receive real-time insights and
 Security: Study the security considerations specific to edge
processing and analysis before sending relevant information to can make informed decisions to enhance crop yield and reduce
computing. Learn about data encryption, access control, and
central servers. resource wastage.
anomaly detection.
 Edge Servers: Servers located near the edge devices to store
and process data locally, reducing the need for data to travel to a Benefits of Edge Architecture Challenges and Solutions of Edge Architecture
Here are the key benefits of Edge architecture: Implementing edge architecture comes with its own set of challenges.
distant data center.
These challenges need to be addressed to ensure the effective
 Edge Gateways: Devices that manage data traffic between edge  Reduced Latency: Processing data locally reduces the time it functioning of edge systems. Here are the key challenges and their
devices and edge servers, ensuring efficient data flow and takes for data to travel, resulting in faster response times. This is solutions:
processing. critical for applications such as autonomous vehicles and
 Communication Networks: Networks that connect edge industrial automation, where real-time data processing is  Data Management: Managing large volumes of data locally can
devices, edge nodes, and edge servers, ensuring seamless data essential. be complex and resource-intensive.
transfer and integration.  Enhanced Security: Local data processing minimizes the risk of o Solution: Implement robust data management

Key Components and Design Principles data breaches during transmission. Sensitive data remains within frameworks and local storage solutions. Use data
Understanding edge architecture requires knowledge of its key the local network, reducing exposure to potential cyber-attacks compression and efficient data handling techniques to
components and design principles. These elements are crucial for and enhancing overall security. optimize storage and processing.
implementing efficient and scalable edge computing systems. Here are  Cost Efficiency: By processing data at the edge, organizations  Interoperability: Integrating diverse devices and communication
the requirements to learn these components and principles: can reduce bandwidth costs. Less data needs to be transmitted protocols can be difficult.
 o Solution: Adopt standardized communication protocols o Solution: Use network segmentation to isolate sensitive One of the main benefits of edge computing is the reduction in latency due to local
and interoperability frameworks. Ensure that all edge data and limit potential breaches. Deploy intrusion processing. Since data does not have to be sent to the cloud for processing, response
devices can communicate seamlessly with each other detection and prevention systems (IDPS) to monitor and times are significantly faster. These near-instant results are crucial for time-sensitive
and with central systems. block malicious activities. applications like autonomous vehicles or medical monitors, where delayed responses
can put human lives at risk.
 Scalability: Ensuring that edge architecture can scale with  Anomaly Detection: Identifying and responding to unusual
growing data volumes and devices. activities in real-time.
o Solution: Design scalable edge infrastructure that can o Solution: Implement AI and machine learning Lower Bandwidth Cost
handle increased data loads. Use modular and flexible algorithms to detect anomalies. Use automated
systems that can be easily expanded as needed. response systems to quickly address potential security Edge computing also reduces network congestion and operating costs because very
 Latency: While edge computing reduces latency, some threats. little data needs to be sent to the cloud. This also improves performance in
applications require extremely low latency that can be hard to environments where bandwidth is limited and cuts down on the cost of constant high-
achieve. What Is Edge Computing Architecture? speed connections.
o Solution: Optimize network infrastructure and use high-
speed communication channels. Implement edge nodes Edge computing architecture is the overall framework that is used to carry out edge
Improved Reliability
strategically to minimize data travel distance. computing. Unlike cloud computing, which uses remote processing, edge computing
uses localized processing and takes place on or near the device gathering data.
 Security: Protecting distributed data from cyber threats is a Because edge computing uses local processing, it has far fewer dependencies. For
major concern. example, edge computing is not as reliant on a stable internet connection and can
o Solution: Implement strong encryption for data at rest Core Components of Edge Architecture continue operations even during outages. This means significantly less downtime risk
and in transit. Use advanced authentication and access for mission-critical applications. Edge computing is also preferred in extreme situations
control mechanisms to secure edge devices and where internet connectivity is unreliable, such as monitoring coral reef health in the
There are several components that must come together to build a scalable edge middle of the ocean.
networks. architecture. Let’s take a closer look:
 Maintenance: Regular maintenance of numerous edge devices
can be challenging and costly. Heightened Security
o Solution: Develop automated maintenance and
monitoring systems. Use predictive maintenance Edge computing keeps data on-site, which makes it easier to control. Local processing
techniques to identify potential issues before they occur. eliminates the need for data transmission, reducing possible exposure. Additionally,
edge computing supports local encryption. All of this keeps regulated or sensitive data
Security Considerations for Edge Architecture exceptionally protected, making compliance with data regulations much smoother.
Security is a critical aspect of edge architecture due to its decentralized
nature. Protecting data across multiple edge devices and nodes requires
a comprehensive security strategy. Here are the primary security  Edge Device: The first piece is the actual edge device, such as an IoT device, Increased Flexibility
considerations and measures for edge architecture: camera, or sensor. This is where the raw data is generated, and is the first point
of contact in the system. Only minimal processing, like data filtering, is The relatively isolated processing of edge computing makes it much more flexible than
 Data Encryption: Protect data both at rest and in transit to performed here. traditional solutions. Edge computing facilitates growth in connected devices without
prevent unauthorized access.  Edge Gateways: These are where data from multiple devices is aggregated. Basic adding to central servers. It’s also possible to use hybrid scaling with cloud resources or
o Solution: Use robust encryption protocols like AES for analytics and preprocessing, such as aggregation and format conversion, also to deploy new edge nodes as demand increases.
data at rest. Implement TLS or IPSec for data in transit take place here.
to ensure secure communication between edge devices Edge Servers: Local processing for real-time applications happens here, along
and servers.

with running containerized workloads or AI inference models. This is also where
What Is an Edge Server?
 Authentication and Access Control: Ensuring only authorized critical data is temporarily held before being synced to the cloud.
users and devices can access the system.  Network Layer: Using LAN, 5G, Wi-Fi, or satellite, this is where edge An edge server is a server that carries out processing close to the data generation
components are connected to each other and the cloud. source. The fact that edge servers process and store data locally is especially important
o Solution: Implement multi-factor authentication (MFA)
 Cloud or Data Center: Provides long-term storage, in-depth analytics, and in remote or bandwidth-constrained environments, as it removes the need for sending
to verify user identities. Use role-based access control data to the cloud.
model training, along with centralized management and orchestration.
(RBAC) to manage permissions and restrict access
based on user roles. For example, edge servers can be used on offshore oil rigs to check pressure levels,
 Device Security: Protecting edge devices from physical and Benefits of Edge Computing Architecture detect leaks, and monitor equipment performance. In remote environments like this,
cyber threats. traditional cloud-based infrastructure does not provide the same level of reliability.
o Solution: Deploy tamper-resistant hardware and secure Edge computing offers distinct benefits when compared to traditional cloud-based
boot processes. Regularly update device firmware to computing. The reduced latency, lower costs, improved reliability, heightened security,
patch vulnerabilities and enhance security. and increased flexibility all make edge computing a powerful option for many modern
Edge vs. Cloud Computing: What’s the Difference?
 Network Security: Securing the communication channels workloads and emerging use cases.
between edge devices and central servers. Traditional cloud computing centralizes data processing in data centers, while edge
computing keeps processing closer to the actual data source. Aside from this core
Reduced Latency

difference, there are several other factors that differentiate edge computing and cloud 🔹 3⃣.1 Three-Layer Model
computing:
🔹 3⃣.2 Four-Layer Model (More Detailed)
1⃣ Device Layer (Perception Layer)
Feature Edge Computing Cloud Computing
1⃣ Device Layer
Components:
2⃣ Edge Node Layer
3⃣ Edge Cloud / Fog Layer
 Sensors
Processing Local, near device Centralized in a data center 4⃣ Central Cloud Layer
 IoT devices
 Smart cameras
 Wearables
Latency Extremely low, near-instant Higher, dependent on distance
Functions: Edge Node Layer
 Data collection
Bandwidth Reduced, selective data transfer Heavy, continuous transfer  Routers
 Initial signal processing  Industrial PCs
 On-premise mini servers
Example:
Smart temperature sensor measuring room temperature. Handles:
Reliability Works offline or intermittently Needs stable connectivity

 Data aggregation
 Protocol conversion
Security Local data storage improves control More exposure due to transit
2⃣ Edge Layer (Processing Layer)  Local caching

Clearly, there are distinct differences between edge computing and cloud computing. Components:
However, it is not necessary to choose one or the other; hybrid models can give you the
best of both worlds. A hybrid model, or the edge cloud, uses the local processing of edge  Edge servers Fog Layer
computing while relying on the cloud for scalability and analytics. This creates a  Gateways
combination of real-time responsiveness and long-term intelligence, which helps  Local micro data centers
Intermediate layer between edge and cloud.
optimize cost, performance, and reliability.
Functions: Provides:
Top Enterprise Applications of Edge Computing  Real-time data processing  Load balancing
 Filtering  Distributed analytics
Enterprises have many unique applications of edge computing; here are the top edge  Analytics
compute use cases:  AI inference Concept introduced by Cisco as "Fog Computing".
 Autonomous Vehicles: Low-latency edge computing is used for processing Example:
camera inputs. This is key for both safety and navigation processing. Smart camera performing face detection locally.
 Healthcare: Edge computing is used for analyzing patient monitoring devices,
where every second counts. Local processing can also help monitor hospital 4⃣ Hardware Components in Edge Computing
equipment analytics to maintain performance standards.
 Manufacturing: Manufacturing equipment logs can be processed locally in order
to enhance maintenance routines. Real-time monitoring helps with production 3⃣ Cloud Layer
and can reduce errors. 4.1 Edge Devices
 Retail Analytics: Edge computing can provide real-time insights on consumer Components:
behavior in-store and streamline inventory management.  IoT sensors
 Smart Cities: Real-time traffic management, energy optimization, and public  Centralized data centers  Raspberry Pi
safety monitoring all rely on edge computing for quick and local processing.  Big data storage  Arduino
 AI model training systems  Smart meters
 Industrial controllers
3⃣ Layered Model of Edge Computing Functions:
Features:
Edge architecture is generally divided into 3-layer or 4-layer models.  Long-term storage ✔ Low power
 Model training ✔ Embedded processors
 Large-scale analytics ✔ Limited memory
  Docker containers Component Function
 Kubernetes orchestration
4.2 Edge Gateways Storage Temporary caching
Cloud providers support lightweight edge deployments.
Communication Send to cloud
 Collect data from multiple devices
 Provide protocol translation
Security Protect data
 Secure data transmission
5.3 Edge Middleware
Examples:
Industrial IoT gateways. Middleware provides:
What is Virtualization?
✔ Device management Virtualization is a technology that creates virtual versions of computer resources
✔ Data management such as hardware platforms, operating systems, storage devices, and network
4.3 Edge Servers / Micro Data Centers ✔ Communication resources. It’s like creating a software-based replica of a physical machine,
✔ Security allowing you to run multiple isolated environments on the same hardware or
 High-performance CPUs across a distributed system.
 GPUs for AI inference Examples:
 SSD storage MQTT brokers, API gateways.  Imagine you have a powerful computer but you only use a small portion
of its resources.
Used for:  Virtualization allows you to split that computer into several virtual
machines (VMs), each acting like a separate computer with its operating
 Real-time analytics
 Video processing
5.4 AI & Analytics Software system and applications.
 Smart manufacturing  Each virtual machine is isolated from the others, meaning issues in one
 Local inference engines virtual machine won’t affect others.
 TensorFlow Lite  This allows you to optimize resource utilization, run multiple
 ONNX Runtime
applications on a single machine, and improve scalability by easily
4.4 Network Infrastructure Used for: adding or removing virtual machines as needed.

 5G routers  Face detection

 Ethernet switches  Anomaly detection What is Containerization?
 Wi-Fi access points  Predictive maintenance Containerization is a lightweight form of virtualization that allows you to run
applications and their dependencies in isolated containers. Each container shares
5G significantly improves edge performance.
the same operating system kernel but is isolated from other containers, providing
a portable and consistent runtime environment for applications.
5.5 Security Software
 Containers provide process isolation, ensuring that applications running
5⃣ Software Components in Edge Computing  Encryption in one container do not affect applications running in other containers.
 Firewall  Containers encapsulate all dependencies and configuration required to
 Identity management run an application, making them portable across different
 Secure boot
5.1 Operating Systems environments.
Edge security is critical due to distributed nature.  Containers are lightweight compared to traditional virtual machines
 Linux-based OS (VMs) because they share the host operating system kernel.
 Windows IoT  Containers are designed to be scalable, allowing you to quickly scale up
 Real-Time Operating Systems (RTOS) or down based on demand.
6⃣ Functional Components of Edge Architecture  Containers enable developers to build, test, and deploy applications
Component Function more efficiently, leading to faster release cycles and improved
collaboration between development and operations teams.
5.2 Virtualization & Containerization Data Acquisition Collect sensor data
Used for deploying applications efficiently. Data Filtering Remove unnecessary data Virtualization Vs. Containerization
Examples: Below are the differences between virtualization and Containerization.
Local Processing Real-time decisions

 Low latency  VirtualBox

Aspect Virtualization Containerization  Lightweight processing  VMware Workstation
 Efficient resource usage
 Scalability
Each VM runs its own guest Containers share the host
Isolation operating system operating system kernel To achieve this, virtualization and containerization technologies are used at edge nodes.
Virtualization at the Edge – Use Cases
 Running multiple IoT services on one gateway
Resource Each VM requires its own set Containers are lightweight and  Isolating industrial applications
2⃣ Virtualization at the Edge
Usage of resources share host resources  Secure multi-tenant deployments

Definition Cloud-edge platforms like:

May have higher overhead Lower overhead as containers
 Amazon Web Services (AWS IoT Greengrass)
Performance due to multiple OS instances share the host OS kernel Virtualization is the creation of multiple virtual machines (VMs) on a single physical
 Microsoft Azure (Azure IoT Edge)
hardware system using a hypervisor.
support VM-based deployments in edge environments.
It allows:
VMs are less portable due to Containers are highly portable
Portability varying guest OS across different systems  Isolation
 Multi-tenancy
 Resource sharing Advantages of Virtualization at Edge
Deployment Slower deployment times Faster deployment times as
Speed due to OS boot process containers start quickly ✔ Strong isolation
✔ OS-level flexibility
How Virtualization Works ✔ Supports legacy applications
Resource Requires more resources as More efficient resource utilization ✔ Improved security
Architecture:
Utilization each VM has its own OS with containerization
Applications
↓
Mature ecosystem with Growing ecosystem with tools like
Guest OS
↓ Limitations
Ecosystem various hypervisors and tools Docker and Kubernetes Hypervisor
↓ ❌ Heavy resource consumption
Physical Hardware
❌ High memory usage
Ideal for microservices ❌ Slower boot time
Ideal for running multiple ❌ Not ideal for low-power devices
architectures and cloud-native
Use Cases
applications on a single host Hypervisor Types
applications
1⃣ Type-1 (Bare Metal)
Best Scenarios for Virtualization and Containerization 3⃣ Containerization at the Edge
Runs directly on hardware.

Examples: Definition
 VMware ESXi
Virtualization and Containerization at the Edge  Microsoft Hyper-V
Containerization packages applications and their dependencies into lightweight containers
that share the host OS kernel.
(Suitable for [Link] / [Link] – Cloud Computing / Edge Computing / IoT syllabus) Unlike VMs:
Containers do NOT require full guest OS.
2⃣ Type-2 (Hosted)
1⃣ Introduction Runs on top of host OS.

Edge computing requires: Examples: Architecture

Applications
4⃣ Virtualization vs Containerization at Edge ✔ Secure boot
↓
Containers ✔ Encrypted containers
Feature Virtualization Containerization
↓ ✔ Role-based access
Container Engine (Docker)
↓ Isolation Strong Moderate ✔ Network isolation
Host OS
↓
Hardware
Resource Usage High Low

Boot Time Slow Fast 8⃣ Challenges at the Edge

Popular Container Technologies OS Required Separate OS per VM Shared OS
1⃣ Resource Constraints
 Docker Edge Suitability Medium High
Limited RAM and storage.
 Kubernetes
 containerd
2⃣ Connectivity Issues
 CRI-O
5⃣ Why Containerization is Preferred at the Edge? Intermittent cloud connectivity.

Because edge environments require: 3⃣ Device Heterogeneity

Why Containers are Ideal for Edge?
✔ Real-time performance Different hardware architectures (ARM, x86).
Edge devices usually have: ✔ Minimal resource usage
✔ Fast deployment 4⃣ Management Complexity
 Limited CPU ✔ Microservice architecture
 Limited RAM Updating thousands of edge nodes.
 Limited storage Hence, containers are widely adopted for edge AI and IoT.

Containers are:
9⃣ Hybrid Approach
✔ Lightweight 6⃣ Edge AI and Containers
✔ Fast startup Modern edge systems use:
✔ Low overhead Edge AI frameworks run inside containers:
✔ Efficient for microservices  Lightweight VMs + Containers
 TensorFlow Lite  MicroVMs (Firecracker)
 ONNX Runtime  Unikernels
 OpenVINO
Example:
Container Orchestration at Edge Containers allow: Serverless edge platforms.

For managing multiple containers:  Easy model updates

 Version control
 Kubernetes (lightweight versions like K3s)
 Edge-native orchestration tools
 Distributed deployment 🔟 Practical Deployment Example

Used for: Smart Factory:

 Automatic scaling 7⃣ Security Considerations Sensors → Edge Gateway → Containerized Analytics → Cloud
 Fault tolerance
 Rolling updates At the edge: Containers:

 Devices are geographically distributed  Data collector

 Physically exposed  Anomaly detection model
 Vulnerable to attacks  Dashboard API

Security measures:

Orchestration Platforms – Service Placement Strategies (in Edge & 3.1 Kubernetes Cloud Control Plane
↓
Cloud Computing) Kubernetes
Edge Cluster Manager
↓
Edge Nodes
↓
(Suitable for [Link] / [Link] – Edge Computing / Cloud / Distributed Systems syllabus)  Open-source container orchestration platform Containers / Services
 Originally developed by Google
 Manages containerized applications

1⃣ Introduction Key Features: Components

In modern distributed systems: ✔ Pod management 1⃣ Control Plane
✔ Auto-scaling
 Applications are broken into microservices ✔ Load balancing  Scheduler
 Services run on multiple edge nodes and cloud servers ✔ Self-healing  API server
 Resources are geographically distributed  Controller manager
✔ Rolling updates
To manage this complexity, we use: 2⃣ Worker Nodes

✔ Orchestration platforms  Run application containers

Lightweight Versions for Edge:  Provide local compute
✔ Service placement strategies
 K3s
 MicroK8s

2⃣ Orchestration Platforms Used for resource-constrained environments. 5⃣ Service Placement Strategies

Definition
Definition
An orchestration platform automates:
3.2 Docker Swarm
Service placement is the decision of where to deploy a service (edge node, fog node, or
 Deployment Docker cloud) to optimize:
 Scaling
 Monitoring  Built-in orchestration tool for Docker  Latency
 Networking  Simpler than Kubernetes  Cost
 Lifecycle management  Suitable for small deployments  Energy
 Resource utilization
of applications across distributed nodes (edge + cloud).

3.3 Cloud Edge Orchestration Platforms

6⃣ Types of Service Placement Strategies
 Amazon Web Services – AWS IoT Greengrass
Why Orchestration is Required at the Edge?  Microsoft Azure – Azure IoT Edge
 Google Cloud – Anthos
Edge systems have: 6.1 Latency-Aware Placement
Provide:
 Heterogeneous devices
Goal:
 Limited resources
✔ Centralized management Minimize response time.
 Intermittent connectivity
 Dynamic workloads ✔ Hybrid cloud-edge orchestration
✔ Security & monitoring Strategy:
Deploy service near end user.
Manual management is impossible.
Example:
Real-time video analytics placed on edge.
4⃣ Architecture of Orchestration at Edge
3⃣ Popular Orchestration Platforms
Typical Architecture:
6.2 Resource-Aware Placement 7⃣ Static vs Dynamic Placement 🔟 Challenges in Service Placement
Static Placement Dynamic Placement
Goal: 1⃣ Resource heterogeneity
Match service requirements with node capacity. Fixed deployment Real-time decision 2⃣ Network variability
3⃣ Security concerns
If: Simple Complex 4⃣ Scalability
 High CPU service → deploy on high-performance edge server 5⃣ Real-time decision making
Less flexible Adaptive
 Lightweight service → deploy on gateway

Dynamic placement is preferred in edge systems.

1⃣1⃣ Example Scenario
6.3 Energy-Aware Placement
Smart City:
Important in battery-powered devices. 8⃣ Optimization Models for Placement  Traffic camera analytics → Edge
Goal:  Historical traffic analysis → Cloud
Placement problem can be formulated as:  Alert generation → Nearby fog node
Minimize energy consumption.
Minimize: Orchestrator decides optimal placement.

Latency+Energy+CostLatency + Energy + CostLatency+Energy+Cost

6.4 Cost-Aware Placement
Subject to: 1⃣2⃣ Benefits of Orchestration & Smart Placement
Goal:
Reduce operational cost.  CPU constraint
 Memory constraint ✔ Reduced latency
Edge processing reduces cloud bandwidth cost.  Bandwidth constraint ✔ Efficient resource usage
✔ Improved reliability
Often solved using: ✔ Automatic scaling
✔ Better user experience
 Heuristic algorithms
6.5 Load-Balancing Placement  Genetic algorithms
 Reinforcement learning
Goal:
Distribute services evenly across nodes. 1⃣3⃣ Limitations
Prevents:
9⃣ Service Migration ❌ Complex algorithms
 Overloaded edge nodes ❌ High management overhead
 Performance degradation Sometimes service must move from: ❌ Security risks
❌ Synchronization issues
Edge Node A → Edge Node B

Reasons:
6.6 Mobility-Aware Placement
✔ Node failure
🔹 Interconnection
Used in:
✔ Overload Concept Purpose
 Vehicular networks ✔ User mobility
✔ Network congestion Orchestration Platform Manages deployment
 Mobile edge computing

Orchestrator handles migration. Scheduler Decides where to run services

Service migrates when user moves.
Service Placement Strategy Optimization logic

Migration Dynamic adjustment

Unit III: Resource Management and Data Processing 1.5 Challenges
Resource Allocation and Scheduling – Load Balancing – Task Offloading Techniques – 1.3 Types of Resources  Resource heterogeneity
 Mobility of devices
Data Processing Models – Stream Processing at the Edge – Latency-aware and Energy-  Scalability issues
 Computation resources (CPU, GPU)
efficient Computing – Quality of Service (QoS) issues.
 Storage resources  Network variability
 Network bandwidth
Fog and Edge Computing: Resource Allocation, Scheduling, Load  Energy resources (battery-powered devices)
Balancing, and Task Offloading Techniques
1.6 Importance
Introduction 1.4 Resource Allocation Strategies  Maximizes resource utilization
 Reduces latency
1. Static Allocation  Improves system efficiency
Fog and Edge Computing extend cloud computing by bringing computation, storage, and
networking closer to end devices such as IoT sensors, smartphones, and embedded systems.
Unlike centralized cloud systems, fog and edge environments are distributed, resource-  Resources are assigned before execution
constrained, and latency-sensitive, making efficient resource allocation, scheduling, load  Simple but not flexible
balancing, and task offloading critical for performance. 2. Scheduling in Fog and Edge Computing
These mechanisms ensure:
2. Dynamic Allocation
 Low latency 2.1 Definition
 Efficient utilization of limited resources  Resources are allocated based on real-time conditions
 Energy efficiency  Adapts to workload changes Scheduling determines when and where tasks are executed in the system.
 Improved Quality of Service (QoS)

3. Priority-Based Allocation 2.2 Objectives

1. Resource Allocation in Fog and Edge Computing
 Tasks are assigned resources based on priority levels  Minimize execution time
 Reduce latency
 Improve throughput
1.1 Definition  Balance resource usage
4. Energy-Aware Allocation
Resource allocation is the process of assigning available resources (CPU, memory,
bandwidth, storage) to tasks or applications in an efficient manner.  Minimizes energy consumption in devices
2.3 Types of Scheduling
1. Task Scheduling
1.2 Characteristics in Fog/Edge Environment 5. QoS-Aware Allocation
 Assign tasks to appropriate nodes
 Limited computational resources  Ensures performance metrics such as latency and throughput
 Distributed nodes
 Dynamic workload
 Heterogeneous devices
2. Workflow Scheduling

 Manage dependent tasks in sequence  Dynamic network conditions

 Task dependencies
 Resource constraints 3. Distributed Load Balancing
 Real-time requirements
3. Real-Time Scheduling  Each node participates in decision-making

 Meet strict deadlines

2.6 Importance
 Efficient task execution 3.4 Techniques
 Reduced response time
2.4 Scheduling Algorithms  Improved QoS 1. Round Robin

1. First Come First Serve (FCFS)  Tasks assigned sequentially

 Tasks executed in order of arrival 3. Load Balancing in Fog and Edge Computing

2. Least Loaded Node

2. Shortest Job First (SJF) 3.1 Definition  Tasks assigned to node with minimum load

 Tasks with smallest execution time first Load balancing is the process of distributing workloads across multiple nodes to avoid
overloading any single node.

3. Resource-Aware Load Balancing

3. Priority Scheduling  Considers CPU, memory, bandwidth

3.2 Objectives
 High-priority tasks executed first
 Prevent system overload
 Improve performance
 Ensure reliability
4. Heuristic-Based Scheduling
3.5 Challenges
 Node heterogeneity
 Uses rules to optimize performance
 Communication overhead
3.3 Types of Load Balancing  Real-time decision making

1. Static Load Balancing

5. Metaheuristic Algorithms
 Decisions made in advance
 Genetic Algorithm 3.6 Benefits
 Particle Swarm Optimization
 Better resource utilization
2. Dynamic Load Balancing  Reduced response time
 Increased system reliability
2.5 Challenges  Adjusts based on current system state
4. Task Offloading Techniques  Tasks sent to nearby edge servers 5. Integration of Concepts

In fog and edge computing:

4.1 Definition 2. Fog Offloading  Resource allocation assigns resources
 Scheduling determines execution order
Task offloading is the process of transferring computation from resource-constrained  Tasks processed at intermediate fog nodes  Load balancing distributes workload
devices to more powerful nodes (fog or cloud).
 Task offloading shifts computation

Together, they ensure:

3. Cloud Offloading
4.2 Need for Task Offloading  Efficient system performance
 Tasks sent to centralized cloud  Reduced latency
 Limited device resources  Optimal resource utilization
 High computational requirements
 Energy constraints
 Latency reduction
4.5 Decision Factors Conclusion

 Network latency Fog and Edge Computing require efficient management of distributed resources. Techniques
 Energy consumption such as resource allocation, scheduling, load balancing, and task offloading play a crucial
4.3 Types of Offloading  Task complexity role in achieving low latency, high performance, and energy efficiency.
 Resource availability
These mechanisms enable modern applications such as:
1. Full Offloading
 Smart cities
 Entire task moved to another node  Autonomous vehicles
 Healthcare monitoring
4.6 Challenges  IoT systems

2. Partial Offloading  Network delays

 Security concerns
 Task divided; some parts executed locally  Task partitioning complexity
 Mobility of devices

3. Adaptive Offloading

 Decision based on network and system conditions

4.7 Benefits
 Reduced device workload
 Improved performance
 Energy savings
4.4 Offloading Strategies  Faster response time

1. Edge Offloading

Fog and Edge Computing: Data Processing Models, Stream Processing 1. Centralized Processing (Cloud-Based) Limitations

at the Edge, Latency-aware & Energy-efficient Computing, Quality of  Data is sent to centralized cloud servers  Limited resources
Service (QoS) Issues  Processing is done in data centers

Advantages

Introduction  High computational power 4. Fog Processing

 Scalable
Fog and Edge Computing bring computation closer to data sources, enabling real-time  Intermediate processing between edge and cloud
processing, reduced latency, and efficient resource utilization. Unlike traditional cloud Limitations
computing, where data is sent to centralized servers, edge and fog systems process data Features
locally or near the source, which is essential for time-sensitive and data-intensive  High latency
applications such as IoT, smart cities, healthcare, and autonomous systems.  Bandwidth consumption  Data filtering and aggregation
 Not suitable for real-time systems  Partial processing
To achieve efficient performance, several key aspects must be addressed:

 Data Processing Models

 Stream Processing at the Edge
 Latency-aware and Energy-efficient Computing
 Quality of Service (QoS) Issues 2. Distributed Processing 5. Hybrid Processing Model
 Data is processed across multiple nodes (edge/fog/cloud)  Combines edge, fog, and cloud processing

Features
1. Data Processing Models in Fog and Edge Computing Working

 Parallel processing  Critical tasks → edge/fog

 Data sharing among nodes  Heavy computation → cloud
1.1 Definition
Advantages
Data processing models define how data is collected, processed, and delivered across edge,
fog, and cloud layers.  Faster processing
 Reduced latency 1.4 Challenges
 Data distribution complexity
 Resource heterogeneity
1.2 Need for Data Processing Models  Synchronization issues

 Handle large volumes of IoT data 3. Edge Processing

 Reduce network congestion
 Enable real-time analytics  Data processed directly at edge devices (sensors, gateways)
 Optimize resource utilization 1.5 Importance
Advantages
 Improves efficiency
 Very low latency  Reduces latency
 Reduced data transmission  Optimizes bandwidth usage
1.3 Types of Data Processing Models
2. Stream Processing at the Edge Techniques

1. Task Placement
2.5 Applications
2.1 Definition  Smart traffic systems  Place tasks closer to users
 Healthcare monitoring
Stream processing refers to continuous, real-time processing of data streams generated by
 Industrial automation
devices.
2. Edge Processing

 Process data locally

2.2 Characteristics 2.6 Challenges

 High data velocity 3. Efficient Scheduling
 Real-time data flow
 Limited processing power
 Continuous processing
 Data consistency  Prioritize time-sensitive tasks
 Low latency requirements

2.3 Working 2.7 Benefits

3.2 Energy-Efficient Computing
 Real-time decision making
1. Data generated by sensors
 Reduced network load
2. Processed immediately at edge nodes
 Faster response
3. Results transmitted or stored
Definition

Energy-efficient computing aims to reduce power consumption while maintaining

3. Latency-Aware and Energy-Efficient Computing performance.
2.4 Techniques
1. Filtering
3.1 Latency-Aware Computing Importance
 Remove unnecessary data
 Edge devices are battery-powered
 Reduces operational cost
Definition
2. Aggregation
Latency-aware computing focuses on minimizing delay in processing and communication.
 Combine multiple data points Techniques

1. Dynamic Voltage and Frequency Scaling (DVFS)

Importance
3. Event Detection  Adjust power based on workload
 Critical for real-time applications
 Identify important events  Improves user experience

2. Task Offloading 3. Reliability

 Shift heavy tasks to fog/cloud  System’s ability to perform consistently 3. Task Scheduling

 Prioritize tasks
3. Sleep Modes
4. Availability
 Turn off idle components
 System uptime 4. Monitoring

 Track system performance

5. Scalability
3.3 Trade-off Between Latency and Energy
 Ability to handle increased workload
 Reducing latency may increase energy consumption
 Energy saving may increase delay 4.5 Importance
Balancing both is a key challenge.  Ensures user satisfaction
6. Energy Efficiency
 Improves system performance
 Supports real-time applications
 Power consumption

4. Quality of Service (QoS) Issues 5. Integration of Concepts

4.3 QoS Challenges in Fog/Edge In fog and edge computing:
4.1 Definition  Resource constraints  Data processing models define where processing happens
 Network variability  Stream processing handles real-time data
QoS refers to the performance level of a system as perceived by users.  Device heterogeneity  Latency-aware computing reduces delays
 Mobility of devices  Energy-efficient computing saves power
 QoS ensures performance and reliability

4.2 QoS Parameters

1. Latency 4.4 QoS Management Techniques Conclusion
 Time delay in processing
1. Resource Allocation Fog and Edge Computing rely on efficient data processing and system optimization
techniques to meet the demands of real-time, large-scale applications. By implementing
 Assign resources based on demand advanced data processing models, stream processing techniques, and balancing latency
and energy requirements, systems can achieve high performance.
2. Throughput
Quality of Service (QoS) plays a critical role in ensuring that these systems meet user
 Amount of data processed per unit time expectations and operate reliably.
2. Load Balancing

 Distribute tasks evenly

Unit IV: Security, Privacy, and Reliability  Increased attack surface 1.5 Data Security and Privacy
 Higher risk of node compromise
Security Challenges in Fog and Edge Computing – Authentication and Authorization – Sensitive data is processed at edge nodes.
Data Privacy and Trust Management – Secure Data Storage and Transmission – Fault Threats
Tolerance – Reliability and Availability – Case studies. 1.2 Resource Constraints
 Data leakage
Security Challenges in Fog and Edge Edge devices often have: 

Unauthorized access
Data modification
Computing – Authentication and 

Limited CPU power
Limited memory
Authorization  Limited battery life

Impact
1.6 Network Security Threats
 Man-in-the-middle attacks
 Cannot implement heavy security algorithms
 Eavesdropping
Introduction  Vulnerable to attacks due to weak protection
 Denial-of-Service (DoS) attacks

Fog and Edge Computing extend cloud services to the network edge, enabling real-time
processing and reduced latency. However, due to their distributed, heterogeneous, and
resource-constrained nature, they introduce significant security challenges. 1.3 Heterogeneity of Devices 1.7 Mobility and Dynamic Topology
Unlike centralized cloud systems, fog and edge environments involve: Devices in fog/edge environments vary in:
Devices frequently move and connect/disconnect.
 Multiple devices (IoT sensors, gateways, edge servers)  Hardware
 Dynamic network topologies  Operating systems Challenges
 Limited computational and storage capabilities  Communication protocols
 Maintaining secure connections
 Continuous authentication
This makes security, authentication, and authorization mechanisms more complex and Challenges
critical.
 Difficult to standardize security mechanisms
 Compatibility issues
1.8 Lack of Centralized Control
1. Security Challenges in Fog and Edge Unlike cloud systems, fog/edge systems lack a central authority.

Computing 1.4 Physical Security Risks

Impact
Edge devices are often deployed in public or remote locations.
 Difficult to enforce uniform security policies
 Increased complexity in access management
Risks
1.1 Distributed Architecture
 Physical tampering
Fog and edge systems consist of numerous distributed nodes, often deployed in  Device theft
uncontrolled environments.  Unauthorized access

Issues
2. Authentication in Fog and Edge
 Difficult to manage and monitor all nodes
Computing

 Uses digital certificates

 Based on Public Key Infrastructure (PKI)
2.1 Definition Benefits:
Authentication is the process of verifying the identity of users, devices, or systems before
3. Authorization in Fog and Edge
Secure and reliable
granting access.

Computing

2.2 Importance 4. Device Authentication 3.1 Definition

 Prevents unauthorized access  Verifies identity of devices (IoT nodes) Authorization determines what actions an authenticated user or device is allowed to
 Ensures system integrity
perform.
 Protects sensitive data Techniques:

 Device IDs
 Hardware-based authentication
2.3 Types of Authentication 3.2 Importance
 Controls access to resources
 Prevents misuse of system
5. Lightweight Authentication  Ensures data security
1. Password-Based Authentication

 Uses username and password  Designed for resource-constrained devices

Limitations: Features: 3.3 Access Control Models

 Vulnerable to attacks (brute force, phishing)  Low computation
 Efficient protocols
1. Role-Based Access Control (RBAC)

 Access based on user roles

2. Multi-Factor Authentication (MFA)
2.4 Authentication Challenges in Fog/Edge Example:
Uses multiple factors: Admin, User, Guest
 Resource limitations
 Something you know (password)  Large number of devices
 Something you have (OTP, token)  Dynamic network conditions
 Something you are (biometrics)  Scalability issues
2. Attribute-Based Access Control (ABAC)
Advantages:
 Access based on attributes:
o User attributes
 Strong security 2.5 Authentication Techniques o Resource attributes
o Environment conditions
 Token-based authentication
 Biometric authentication
 Blockchain-based authentication (emerging)
3. Certificate-Based Authentication
3. Policy-Based Access Control 4. Access granted or denied  Industrial IoT

 Access decisions based on predefined policies

5. Security Solutions and Best Practices Conclusion

4. Capability-Based Access Control
Security in fog and edge computing is challenging due to distributed architecture, resource
 Users are given tokens/capabilities for access constraints, and dynamic environments. Authentication and authorization play a critical
1. Lightweight Cryptography role in ensuring secure access and data protection.

By implementing robust and lightweight security mechanisms, systems can achieve:

 Suitable for resource-constrained devices
 Secure communication
3.4 Authorization Challenges  Controlled access
 Protection against cyber threats
 Dynamic environments 2. Zero Trust Model
 Device mobility
 Scalability  “Never trust, always verify”
 Policy management complexity

3. Blockchain-Based Security
3.5 Fine-Grained Access Control
 Decentralized authentication
 Provides detailed control over resources
 Ensures users access only what is necessary

4. Continuous Monitoring
 Detect suspicious activities
4. Integration of Authentication and
Authorization
5. Secure Key Management
Authentication and authorization work together:
 Protect cryptographic keys
 Authentication → Who are you?
 Authorization → What can you do?

Process Flow 6. Applications

1. User/device requests access  Smart cities
2. Authentication verifies identity  Healthcare IoT systems
3. Authorization checks permissions  Autonomous vehicles

1.3 Techniques for Data Privacy 4. Tamper-Proof Logs: Blockchain or append-only logs for auditability.
Fog and Edge Computing: Data Privacy
and Trust Management, Secure Data  Encryption at Rest and In Transit: Protects data stored on devices or transmitted
across networks.
Storage and Transmission, Fault Tolerance, 

Anonymization and Pseudonymization: Hides sensitive information.
Differential Privacy: Adds noise to datasets to protect individual data points.
2.2 Secure Data Transmission
Reliability and Availability, Case Studies  Policy-Based Data Access: Ensures only authorized entities can access specific data. Challenges

 Data travels across wireless, often insecure networks.

 Vulnerable to eavesdropping, man-in-the-middle attacks.
1.4 Trust Management
Introduction Techniques
Trust management ensures that nodes, devices, or services behave reliably and securely.
Fog and Edge Computing bring computation, storage, and networking closer to end devices  End-to-End Encryption: TLS/SSL protocols
like IoT sensors, smartphones, and gateways. While this architecture improves latency and  Trust Metrics: Include reliability, historical behavior, security compliance.  VPNs and Secure Tunnels: Protect data over public networks
real-time responsiveness, it introduces significant security, privacy, and reliability  Reputation Systems: Track past behavior of devices or nodes.  Message Authentication Codes (MACs): Ensure data integrity
challenges.  Blockchain-Based Trust: Decentralized verification of transactions and interactions.  Lightweight Cryptography: For low-power devices

Data is processed in a distributed, heterogeneous, and often resource-constrained

environment, making data privacy, trust management, secure storage, fault tolerance,
and reliability critical for maintaining system integrity and user confidence.
1.5 Importance 3. Fault Tolerance
 Protects sensitive IoT and user data
 Builds confidence in edge and fog systems
1. Data Privacy and Trust Management  Reduces insider and outsider threats
3.1 Definition
Fault tolerance is the ability of a system to continue functioning correctly even when some
components fail.
1.1 Data Privacy Definition 2. Secure Data Storage and Transmission
Data privacy in fog and edge computing ensures that sensitive user or device data is
protected from unauthorized access, disclosure, or misuse, especially when data is processed 3.2 Faults in Fog/Edge Systems
close to its source.
2.1 Secure Data Storage
 Node failure (device or gateway crash)
Fog and edge devices often store temporary or cached data for processing.  Network link failure
 Software or hardware malfunction
1.2 Challenges Challenges  Power failure

1. Distributed Processing: Data is processed at multiple nodes, increasing exposure.  Limited storage capacity
2. Heterogeneous Devices: IoT devices differ in capability and security measures.  Exposure to physical tampering
3. Dynamic Environments: Devices frequently join and leave networks, complicating  Potential insider attacks 3.3 Techniques for Fault Tolerance
data protection.
4. Limited Resources: Lightweight devices cannot implement heavy encryption or Techniques 1. Redundancy: Replicating tasks or data on multiple nodes.
privacy mechanisms. 2. Checkpointing: Saving intermediate states to resume tasks after failure.
1. Data Encryption: AES, RSA, ECC for secure storage. 3. Task Migration: Moving tasks from failing nodes to healthy nodes.
2. Redundant Storage: Store copies on multiple nodes to prevent data loss. 4. Self-Healing Systems: Automatically detect and correct failures.
3. Access Control: Fine-grained policies to prevent unauthorized access.
 4.3 Challenges  Data Privacy and Trust → Ensure sensitive data is protected and devices behave
reliably.
3.4 Importance  Secure Storage and Transmission → Prevent data breaches and tampering.
 Network partitioning
 Fault Tolerance → Maintain functionality during failures.
 Resource-constrained edge devices
 Reliability and Availability → Ensure consistent and uninterrupted service.
 Ensures uninterrupted service  Dynamic workloads
 Maintains data integrity  Security incidents affecting uptime
Together, these measures ensure robust, secure, and trustworthy edge and fog computing
 Improves system reliability
environments.

5. Case Studies
4. Reliability and Availability Conclusion
Fog and edge computing require holistic approaches to security, reliability, and data
5.1 Smart Healthcare IoT privacy. Implementing:
4.1 Reliability
 Wearable devices and health monitors process data at edge gateways.  Strong encryption and access control
Reliability refers to the probability that a system performs its intended function without  Privacy: Patient data encrypted before storage.  Trust and reputation-based management
failure over a specified period.  Trust: Gateway devices maintain reputational trust metrics.  Redundancy, failover, and checkpointing for fault tolerance
 Fault Tolerance: Redundant edge nodes handle device failures.  Mechanisms to ensure high reliability and availability
 Metrics include Mean Time Between Failures (MTBF) and failure rate.  Reliability & Availability: Critical for real-time health alerts.
enables efficient, resilient, and trustworthy distributed systems that can support critical
Techniques to Improve Reliability applications in healthcare, smart cities, and industry.

 Redundant nodes and storage 5.2 Smart Cities

 Error detection and correction
 Continuous monitoring  Edge sensors monitor traffic, pollution, and energy usage.
 Data Privacy: Citizen information anonymized.
 Secure Transmission: Encrypted communication over public networks.
 Fault Tolerance: Multiple fog nodes process traffic data.
4.2 Availability  High Availability: Continuous traffic control without downtime.

Availability is the percentage of time a system is operational and accessible.

 High availability ensures minimal downtime. 5.3 Industrial IoT (IIoT)

 Measured as:
 Factory machines send data to edge gateways for predictive maintenance.
Availability=UptimeUptime+Downtime×100Availability = \frac{Uptime}{Uptime +  Trust Management: Edge nodes verify machine identity.
Downtime} \times 100Availability=Uptime+DowntimeUptime×100  Fault Tolerance: Tasks migrated from failing nodes to backups.
 Secure Storage & Transmission: Encrypted logs for compliance.
Techniques to Improve Availability

 Load balancing across multiple nodes

 Failover mechanisms
 Distributed architecture with backup nodes 6. Integration of Concepts
In fog and edge computing:
Unit V: Applications, Case Studies, and Emerging Trends  Reduces latency by processing data near the source. 2.4 Industrial IoT (IIoT)
 Minimizes bandwidth usage by filtering and aggregating data locally.
Fog and Edge Computing in IoT – Smart Cities – Healthcare – Autonomous Vehicles –  Supports real-time decision-making.
 Sensors monitor machinery, production lines, and energy usage.
Industrial IoT – AI at the Edge – Edge Analytics – Integration with 5G – Research challenges  Enhances privacy and security by keeping sensitive data local.
 Edge nodes process sensor data for predictive maintenance.
and future directions.  Fog layer aggregates plant-wide analytics for operational optimization.
 Benefits:
Fog and Edge Computing in IoT: 2. Applications
o Reduced downtime
o Enhanced productivity

Applications, AI at the Edge, Edge o Energy optimization

Analytics, 5G Integration, Research

Challenges, and Future Directions 2.1 Smart Cities
3. AI at the Edge
 Edge devices monitor traffic, pollution, energy usage, and public safety.
 Fog nodes process aggregated data for urban planning and emergency response.
 Benefits:
Introduction o Real-time traffic management
o Energy-efficient lighting and utilities
3.1 Definition
o Quick emergency alerts
The Internet of Things (IoT) has transformed everyday environments into smart, Artificial Intelligence models deployed at edge nodes or gateways to analyze data in real-
 Example: Smart street lighting adapts to pedestrian and vehicle presence using fog-
interconnected systems. Devices ranging from sensors to vehicles continuously generate time.
enabled analytics.
massive data streams. While cloud computing provides centralized processing, it struggles
with latency, bandwidth, and real-time requirements.

Fog and Edge Computing address these challenges by bringing computation, storage, and
2.2 Healthcare 3.2 Advantages
analytics closer to the data source, enabling real-time decision-making, reduced latency,
and improved efficiency.  Real-time inference: Immediate decisions without cloud delays.
 IoT devices monitor vital signs (heart rate, blood pressure, glucose).
 Bandwidth efficiency: Raw data does not need to be transmitted to the cloud.
Applications span smart cities, healthcare, autonomous vehicles, industrial IoT, and AI-  Edge gateways perform local analytics for anomaly detection.
 Privacy: Sensitive data remains local.
enabled edge analytics. Integration with 5G networks further enhances performance, while  Fog layer aggregates patient data and interfaces with hospital systems.
ongoing research aims to tackle challenges like security, heterogeneity, and scalability.  Benefits:
o Real-time monitoring and alerts
o Reduced network congestion
o Data privacy (sensitive patient information stays local) 3.3 Use Cases
1. Fog and Edge Computing in IoT  Smart cameras for anomaly detection
 Predictive maintenance in factories
2.3 Autonomous Vehicles  Voice recognition in smart assistants

1.1 Definition  Vehicles generate massive sensor data (LIDAR, camera, GPS).
 Edge computing enables real-time collision detection and navigation.
 Edge Computing: Data processing directly at IoT devices or nearby gateways. 

Fog computing supports fleet coordination and traffic prediction.
Benefits:
4. Edge Analytics
 Fog Computing: Intermediate layer between edge devices and the cloud for
processing, storage, and networking. o Ultra-low latency decision-making
o Cooperative driving and vehicle-to-everything (V2X) communication

4.1 Definition
1.2 Role in IoT

Processing and analyzing data streams at the edge for actionable insights. 6.1 Scalability  AI-enhanced resource management: Machine learning models to dynamically
allocate resources and tasks
 Blockchain-based trust and security: Decentralized authentication and audit for IoT
 Millions of heterogeneous IoT devices
nodes
 Dynamic network topology
4.2 Techniques  Green Edge Computing: Energy-efficient architectures for sustainable IoT
 5G and beyond integration: Support ultra-low latency, high reliability applications
 Federated Learning at the Edge: Train AI models locally without transferring
 Stream processing
sensitive data to the cloud
 Event detection 6.2 Security and Privacy  Edge-as-a-Service Platforms: Cloud-like services available at edge nodes for
 Local aggregation and filtering
developers
 Machine learning inference on edge devices
 Distributed architecture increases attack surface
 Sensitive data at edge requires strong encryption and trust mechanisms

4.3 Benefits Conclusion

 Faster decision-making
6.3 Resource Management Fog and Edge Computing are transformative technologies for IoT, enabling real-time
 Reduced network traffic processing, reduced latency, enhanced privacy, and efficient resource utilization.
 Limited CPU, memory, and energy at edge devices
 Scalability across distributed IoT nodes Applications in smart cities, healthcare, autonomous vehicles, and industrial IoT
 Efficient task offloading, scheduling, and load balancing are critical
demonstrate their potential.

AI at the edge, edge analytics, and 5G integration further expand capabilities, while
ongoing research addresses challenges in security, scalability, reliability, and energy
5. Integration with 5G Networks 6.4 Interoperability efficiency.

 Devices use diverse protocols, platforms, and data formats These systems are crucial for the next generation of smart, connected, and autonomous
 Requires standardization for seamless integration applications.
5.1 Role of 5G
 High-speed connectivity with low latency
 Massive device connectivity 6.5 Fault Tolerance and Reliability
 Supports real-time edge analytics and AI applications
 Edge nodes may fail or disconnect
 Critical applications (healthcare, autonomous vehicles) require robust mechanisms

5.2 Benefits for Fog and Edge Computing

 Ultra-reliable low-latency communication (URLLC) for autonomous vehicles and 6.6 Energy Efficiency
healthcare
 Enhanced mobile edge computing (MEC) for smart cities  Edge devices are battery-powered
 Supports bandwidth-heavy applications like AR/VR  Need energy-efficient AI and analytics methods

6. Research Challenges 7. Future Directions