BCSE313L_FUNDAMENTALS OF FOG AND EDGE COMPUTING

CHALLENGES IN Dr. B.V. Baiju,

FEDERATING SCOPE,
Assistant Professor
EDGE RESOURCES
VIT, Vellore Baiju B V, VIT, SCOPE
 INTRODUCTION
• Cisco estimates that there are IoE, with estimates of 8–10 billion
connected today.
• Cloud computing has been a key enabling technology for the IoT, a small
increase in the percentage of connected or cyber-physical objects
represents dramatic change in the feature space of computing and a
potential tsunami of computation and hyper-connectivity, which today’s
infrastructure will struggle to accommodate at historic levels of quality of
service (QoS).
• Traditional cloud computing architectures were simply not designed with
an IoT, characterized by extreme geographic distribution, heterogeneity
and dynamism, in mind.
• A novel approach is required to meet the requirements of IoT including
transversal requirements (scalability, interoperability, flexibility,
reliability, efficiency, availability, and security) as well as cloud-to-thing
(C2T)-specific computation, storage and communication needs
Baiju B V, VIT, SCOPE
 METHODOLOGY
• The objective of this systematic literature review is to present an
overview of academic literature on
(i) Use of modeling
techniques to
represent and
evaluate an
integrated
C2F2T system
(ii) Main scenarios
modeled
(iii) Metrics used to
evaluate models.

Baiju B V, VIT, SCOPE

• Fog computing is a technology that acts as a middle layer between the
cloud and smart devices (like sensors and smartphones).
• It uses fog nodes (intermediate devices) to manage data and
communication, making it easier to run Internet of Things (IoT)
applications.
• In essence, fog computing helps process data closer to where it's
generated, reducing the need to send everything to the cloud.
• This makes IoT applications faster and more efficient.

Baiju B V, VIT, SCOPE

 Integrated C2F2T Literature by Modeling Technique
• The analysis presented suggests that analytical models followed by
Petri Nets and Integer Linear Programs are the most common
techniques used for modeling an integrated C2F2T system.

Baiju B V, VIT, SCOPE

1.1 Analytical Models
• Analytical models are mathematical models that have a closed-form
solution, i.e. the solution to the equations used to describe changes in a
given system can be expressed as a mathematical analytic function.
• Analytical models can be used to predict computing resource requirements
related to
– workload behavior
– content
– volume changes
and to measure effects of hardware and software changes.
• In this systematic review 16 articles are using analytical models.
(1) Wei Li et al. ( System modelling and performance evaluation of a
three-tier cloud of things) defined the architecture, where each physical
and virtual component of layers are described as a vector of associated
features.
• A set of equations is defined to calculate metrics such as power
consumed and time latency of scenarios.
Baiju B V, VIT, SCOPE
(2) K. E. Desikan et al. (A Novel Distributed Latency-Aware Data
Processing in Fog Computing-Enabled IoT Networks) , considered the
– number of gateways (devices that receive data from IoT devices and
forward to cloud infrastructure or fog devices) present in an
architecture.
– total data received for each gateway.
– time spent by each gateway to process the data, in their modeling.
Imagine we have a fog computing network with 3 gateways (G1, G2, G3) that receive data
from IoT devices and process it before sending it to the cloud.

Gateway 1 (G1)
Total Data Received: 100 MB
Processing Time: 10 seconds
Gateway 2 (G2) G1 processes 100 MB of data in 10 seconds.
Total Data Received: 150 MB G2 processes 150 MB of data in 15 seconds.
Processing Time: 15 seconds G3 processes 200 MB of data in 20 seconds.
Gateway 3 (G3)
Total Data Received: 200 MB
Processing Time: 20 seconds Baiju B V, VIT, SCOPE
• Equations are proposed to represent the available buffer and the occupancy
efficiency of gateway buffers.
• Depending on the available space on the gateways, the data are transferred to
higher layers, increasing the delay for data processing.
• Several equations are defined for the calculation of the delay and the authors
propose an optimization of the gateways’ efficiency, improving the occupancy
and the response time efficiency for all gateways in the system.

Baiju B V, VIT, SCOPE

(3) J.F. Colom et al. (Collaborative building of behavioural models based
on Internet of Things) proposed an analytical model to represent a
scheduling mechanism in IoT environment.
• An equation is defined to represent the addition of a load to be
processed in a device.
• This process takes into account several variables, such as processor
load, free memory, and free bandwidth, among others.

(4) H.Wu et al. (Deviation-based neighborhood model for context-aware

QoS prediction of cloud and IoT Services) proposed a method to predict
the QoS of IoT services.
• The model is based on a neighborhood collaborative filter and allows an
efficient global optimization scheme.
• A set of equations is defined to predict correctly the QoS of a service,
based on aspects such as latency, response time, and user network
condition.

Baiju B V, VIT, SCOPE

Thermostat Let's say we want to add a new load that requires:
Processor Load (PL): 20% Processor Load: 15%
Free Memory (FM): 50MB Memory: 20MB
Free Bandwidth (FB): 10Mbps
Bandwidth: 5Mbps
Security Camera Thermostat
Processor Load (PL): 50% New Processor Load: ( 20% + 15% = 35% )
Free Memory (FM): 100MB New Free Memory: ( 50MB - 20MB = 30MB )
Free Bandwidth (FB): 20Mbps New Free Bandwidth: ( 10Mbps - 5Mbps = 5Mbps )

Smart Light Security Camera

Processor Load (PL): 10% New Processor Load: ( 50% + 15% = 65% )
Free Memory (FM): 30MB New Free Memory: ( 100MB - 20MB = 80MB )
Free Bandwidth (FB): 5Mbps New Free Bandwidth: ( 20Mbps - 5Mbps = 15Mbps )
Smart Light
New Processor Load: ( 10% + 15% = 25% )
New Free Memory: ( 30MB - 20MB = 10MB )
New Free Bandwidth: ( 5Mbps - 5Mbps = 0Mbps )

Thermostat: New load is manageable.

Security Camera: New load is manageable.
Smart Light: New load might be too high due to limited bandwidth. Baiju B V, VIT, SCOPE
(5) Y. Lyu, F. Yan et.al (High-performance scheduling model for
multisensor gateway of cloud sensor system-based smart-living)
proposed a scheduling model to manage sensor applications by a gateway.

• The requirement resources of applications are taken into account in

scheduling problem.
Baiju B V, VIT, SCOPE
• In a scenario with n applications
A = {ai (Ri, ri, wi (t), si (t))
where Ri is a resource requirement vector
ri ∈ [0, 1] is a priority
wi(t) is a required work status
si(t) is an actual working status of ai
• The integral calculates the total waiting time of
the tasks, where
• T is evaluation time.
• uc is total CPU utilization
• cm is total memory capacity
• bl input bandwidth
• bO output bandwidth
• Vector [𝛼c, 𝛼m, −𝛼I, 𝛼O] represents the
overloading factors for the CPU, memory,
input and output bandwidth, respectively,
and indicates that the system can run
smoothly to some extent. Baiju B V, VIT, SCOPE
(6) F. Renna et al. (Query processing for the Internet-of-Things:
Coupling of device energy consumption and cloud infrastructure billing),
proposed a set of analytic expressions for representing the expected energy
consumption of devices, as well as a cloud billing method for a group of
devices on an IoT aggregator.
• Considering n devices the below equation computes the energy
consumption of each device over the monitoring period, T .

where E [ Ψe] is a query data volume of devices,

ge is consumption rate (joule-per-bit)
ie is “idle” energy consumption by each device.
• The integral of the second term represents the expected energy
consumption of a device in idle mode.
– ceE [ Ψe] is a threshold where the application activates “idle” mode
– Pe(𝜔e) represent the probability density function of Ψe.
Baiju B V, VIT, SCOPE
𝐸[Ψ𝑒]: Let's assume 𝐸[Ψ𝑒]=2
𝑔𝑒,𝑖𝑒,𝑐𝑒​ : Let 𝑔𝑒=3, 𝑖𝑒=2 and 𝑐𝑒=4

𝑃𝑒(𝜔𝑒): Assume 𝑃𝑒(𝜔𝑒) is uniform between [0,𝑐𝑒𝐸[Ψ𝑒]]

Compute the first term: E[Ψe​]ge​ =2⋅3=6

Compute the second term:

Substituting 𝑃𝑒(𝜔𝑒)=1/8 and 𝑐𝑒𝐸[Ψ𝑒]=8

Integral part split into 2

The integral of a constant 𝑎 over the range [0,𝑏] is:

The integral of 𝜔𝑒​ is:

Multiply by 𝑖𝑒=2: 2⋅ 4 = 8
Total 𝐸exp​ : Eexp​ = 6+8=14 Baiju B V, VIT, SCOPE
• The Expected cloud billing cost can be calculated by , taking into
account n aggregated query volumes from n devices.
– E[Ψb]gb represents the transfer/storage costs.
– the first integral represents the idle billing cost
– the second integral corresponds to the active billing cost

ib is the cost of transferring one bit (dollars-per-query-bit),

cb is a coupling point that define the expect billing cost.

(7) F. Jalali et al. (Interconnecting fog computing and microgrids for

greening IoT) the utilization of a strategy using fog computing with
microgrids to reduce the energy consumption of IoT applications is
considered.
• Two equations are proposed to evaluate energy consumption.
Baiju B V, VIT, SCOPE
• Energy consumption between IoT and cloud infrastructure is calculated
using

– (EGW−r) energy consumed by IoT gateways when receiving data from IoT
devices and sensors
– (EGW−t) energy consumed by IoT gateways to transmit data to the cloud
data center
– (Enet) energy consumed by transport network between IoT gateways and
cloud
– (EDC) the energy consumed by components of data center (EDC).

Given Data:
Energy consumed by the Gateway Receiver: 𝐸GW-r = 5 units
Energy consumed by the Gateway Transmitter: 𝐸GW-t = 3 units
Energy consumed in the Network: 𝐸net =2 units
Energy consumed by the Data Center: 𝐸DC =10 units
EIoT-cloud​ =5+3+2+10 = 20 units
Baiju B V, VIT, SCOPE
• Energy consumption between IoT and fog infrastructure is calculated
using

• This equation takes into account the same components of previous

equation plus two other components:
– EGW-c, the energy consumed by IoT gateways for local computation and
processing
– 𝛽, a ratio of the number of updates from the fog to the cloud for
synchronization.

Let’s assume:
𝐸GW-r = 4units
𝐸GW-c = 6 units
𝐸GW-t = 3units
E IoT-fog​ =4+6+0.5(3+2+8)
𝐸net = 2 units
𝐸DC = 8 units E IoT-fog​ =16.5 units
𝛽=0.5(50% of the data is offloaded to the Cloud)

Baiju B V, VIT, SCOPE

(8) R. K. Behera et al. (Reliability modelling of service-oriented Internet of
Things) proposed an analytical modeling for estimating the reliability of an IoT
scenario for a smart home context.
• An algorithm is proposed to estimate the reliability of the IoT service, which is
formed by n subsystems.
• Reliability of IoT system is calculated using the below equation

(Ppr) the availability of the all k programs running on the virtual machines
(Pf) the availability of the f input files for the programs
(ISR) the reliability of each subsystem , i.e. the reliability of the VM being
executed.
(9) W. Zeng et al. (Opacity in Internet of Things with Cloud Computing) , the
authors presented a model to evaluate the security level of C2T systems.
• The focus of the model is the flow of information, where an initial state of the
system is defined, and a set of operations are performed.
• Thus, after performing these actions, the system reaches new states.
Baiju B V, VIT, SCOPE
Let’s assume:
N=3 (There are 3 components).
ISR1=0.9, ISR2=0.85, ISR3=0.95 (Reliabilities of individual components).
Pf(1)=0.1, Pf(2)=0.15, Pf(3)=0.05 (Failure probabilities of the components).
Ppr(1)=0.8, Ppr(2)=0.9, Ppr(3)=0.85(Performance probabilities of the components).

The overall reliability or

performance Rs(tb) is
approximately:
Rs(tb)≈0.00033

Rs(tb) = 0.72375×0.00075×0.612=0.00033290625
Baiju B V, VIT, SCOPE
(10) V. B. Souza et al. (Towards Distributed Service Allocation in Fog-to-
Cloud (F2C) Scenarios) used a Multidimensional Knapsack Problem (MKP) to
find the optimal service allocation in C2F2T scenarios.
• For this, they consider a number of application aspects:
– load balancing
– delay
– energy consumption.
• The service allocation is defined as an MKP problem, where the objective is
threefold:
– minimizing the energy consumption by devices
– minimizing the overload in terms of processing capacity
– minimize the overall allocation of services in infrastructure.
(11) Y. Liu et al. (Incentive mechanism for computation offloading using
edge computing: a Stackelberg game approach) address the interactions
among cloud operator and IoT service provide as optimization problem.
• They formulated analytically the problem and maximize the utilities of cloud
service with the purpose of obtaining optimal payment and computation
offloading. Baiju B V, VIT, SCOPE
(12) Z. Ming et al. (QoS-aware computational method for IoT composite
service) proposed an analytical modeling to represent QoS of IoT composite
services, taking into account such metrics as availability, reliability, and
response time.
• An optimization algorithm is proposed to find optimal cost with QoS
constraints.
(13) Y. Li et al. (Leveraging Renewable Energy in Edge Clouds for Data
Stream Analysis in IoT ) proposed an analytical models that is used to decide if
offloading computing will be processed in IoT devices or in cloud, taking
into account the desired QoS and energy level available in IoT devices.
(14) S. Sarkar et al. (Theoretical modelling of fog computing: a green
computing paradigm to support IoT applications) proposed an analytical
models to compare the performance between the fog architecture against
traditional cloud computing.
• They consider several aspects, such as
– location of devices
– operation mode
– hardware details
– type of events. Baiju B V, VIT, SCOPE
(15,16) L. Skorin-Kapov et al. (Energy efficient and quality-driven continuous
sensor management for mobile IoT applications, Collaborative Computing:
Networking, Applications and Worksharing (CollaborateCom)) and T. Dinh et
al. (A location-based interactive model for Internet of Things and cloud (IoT-
cloud). Ubiquitous and Future Networks) presented analytical models to
represent mobile nodes connected to cloud computing. In both articles, the
proposed models consider the movement of the devices that are connected to the
cloud.
• In first, the model details the architecture in a selection of components:
wireless sensor network (WSN), the cloud infrastructure, applications,
and mobile users.
• In second, authors consider that the mobile devices are connected to the
cells to send data to the cloud.

Baiju B V, VIT, SCOPE

1.2 Petri Net Models
• A Petri Net is a well-known model to represent systems with respect to
evaluate performance and dependability.
• To solve a Petri Net, one can use two options:
(i) analytic solution by using Markov chain (in which case all
transitions must follow exponential distributions); or
(ii) simulations using theory of discrete event simulation.
• Petri Nets allows a more fine-grained representation of the system, utilizing
Markovian and non-Markovian distributions, and represents the system
behavior with a fewer number of states.

Baiju B V, VIT, SCOPE

• Petri Net models used two works
• G. Merlino proposed a mobile crowd sensing framework, integrating mobile
devices into services hosted in the cloud.
• This model enables efficient data collection and processing while optimizing
resource utilization.
• The places represent four possible states of a contributing node:
• contributing node availability (Av/ NAv)
• position on the interest area (In/Out).

Available (Av): The node is available to contribute data.

Not Available (NAv): The node is not available to contribute data.
In: The node is within the area of interest.
Out: The node is outside the area of interest

Baiju B V, VIT, SCOPE

 The transitions
represent the
probability of the
devices entering or
exiting these states.

State Description
The node is available and located within the area of interest, ready to
Av/In
contribute data.
The node is available but outside the area of interest, thus not
Av/Out
contributing.
The node is in the area of interest but unavailable due to constraints
NAv/In
(e.g., low battery).
NAv/Out The node is both unavailable and outside the area of interest. Baiju B V, VIT, SCOPE
• R. K. Lomotey (Wearable IoT data stream traceability in a distributed
health information system) evaluated the problem of managing the
traceability of data in C2T scenario.
• A Petri Net model is proposed to map and match device data to users that
assists tracking a transparent data trace route, and possible detection of data
compromises.
• The PetriNet represents the behavior of a proposed wearable IoT architecture.
• The places represent the different sources where data are generated or
collected.
• Transitions represent events that may occur, such as new medical data
readings, for example vitals, from the wearable IoT device.

Baiju B V, VIT, SCOPE

Imagine a health monitoring system where wearable devices track
patients' vital signs, such as heart rate and blood pressure.

Places:
Wearable Device: The source where data is generated (e.g., a smartwatch).
Health Database: Where the collected data is stored.
User Profile: The patient's profile that receives and displays the data.
Transitions:
New Reading: An event where the wearable device records a new heart rate
or blood pressure reading.
Data Upload: The event of transferring the new reading from the wearable
device to the health database.
Profile Update: The event of updating the patient's profile with the new
data.

Baiju B V, VIT, SCOPE

1.3 Integer Linear Programming
• In this systematic review, two works modeled their problems using ILP and
MILP
• E. Sturzinger (Application-aware resource provisioning in a
heterogeneous Internet of Things) proposed an ILP model for calculating
the financial cost of a fog-cloud architecture located in a metropolitan area
network (MAN).
• The authors represent application profiles and characteristics of each node in
the MAN as two vectors.
• The ILP model minimizes the operational cost necessary to support the traffic
in network topology while satisfy the application constraints.
• Objective function used in ILP model presented using equation

Costt represents the total cost

Costp is the processing cost
Costs is the storage cost
Costu are the total upstream and downstream costs respectively
Costc is the total MAN link capacity. Baiju B V, VIT, SCOPE
Imagine a smart city network with three nodes: Node A, Node B, and Node C.
Each node has different roles and associated costs.
Costp (Processing Cost): The cost of Costd (Download Cost): The cost of
processing data at the nodes. downloading data from the cloud.
Node A processes traffic data: $50 Node A downloads traffic data: $5
Node B processes air quality data: $30 Node B downloads air quality data: $10
Node C processes energy usage data: $20 Node C downloads energy usage data: $5

Costs (Storage Cost): The cost of storing Costc (Communication Cost): The cost of
data at the nodes. communication between nodes.
Node A stores traffic data: $10 Node A communicates with Node B: $10
Node B stores air quality data: $15 Node B communicates with Node C: $10
Node C stores energy usage data: $5 Node C communicates with Node A: $10

Costu (Upload Cost): The cost of

uploading data to the cloud.
Node A uploads traffic data: $25
Node B uploads air quality data: $20
Node C uploads energy usage data: $15

Costt=(50+30+20)+(10+15+5)+(25+20+15)+(5+10+5)+(10+10+10)= 240 Baiju B V, VIT, SCOPE

Integrated C2F2T Literature by Use-Case Scenarios
• The use case scenarios modeled are described in the identified papers
e.g. resource management, smart cities, WSN, health and
other/generic are summarized in table

Baiju B V, VIT, SCOPE

• W. Zeng (Opacity in Internet of Things with Cloud Computing ) used a
medical application as a case study for their proposed model that analyzes
the security of the information flow in IoT systems integrated with cloud
infrastructures.

In a smart hospital, wearable IoT devices like fitness bands and smartwatches
collect health data such as heart rate, oxygen levels, and blood pressure from
patients. This data is sent to a shared edge server, where it is processed for
different uses. For example, one application can monitor the vitals and send
emergency alerts if an issue like an irregular heartbeat is detected. Another
application can track long-term health trends, while a third can analyze sleep
patterns or stress levels. By sharing the same IoT system, these applications
work together efficiently, reducing costs and providing real-time, scalable
health monitoring.
Baiju B V, VIT, SCOPE
• J.F. Colom (Collaborative building of behavioural models based on
internet of things) proposed a framework that enables multiple applications
to share IoT computational devices for health monitoring.

Baiju B V, VIT, SCOPE

• R. K. Lomotey (Wearable IoT data stream traceability in a distributed
health information system) proposed a wearable IoT architecture for
healthcare systems.

Baiju B V, VIT, SCOPE

• Y Benazzouz (A fault fuzzy-ontology for large scale fault-tolerant
wireless sensor networks) proposed a fault fuzzy-ontology that can be used
to verify fault tolerance in large-scale WSN using service-oriented
applications

Baiju B V, VIT, SCOPE

• T. Dinh (A location-based
interactive model for Internet of
Things and cloud (IoT-cloud))
proposed a WSN model for sensing
as a service that integrates IoT and
cloud infrastructures.

In a traditional model, these sensors

would continuously collect and
transmit data to a central server,
regardless of whether anyone needs the
data at that moment. This leads to
unnecessary energy consumption and
data transmission.

Baiju B V, VIT, SCOPE

• Y. Li . (Leveraging Renewable Energy in Edge Clouds for Data Stream
Analysis in IoT) presented a vehicle-to-cloud monitoring application that
interacts with edge computing; their analytic model defines where the data will
be processed – e.g. in the edge or in the cloud infrastructure.
• They use a video-streaming application to validate their model

Baiju B V, VIT, SCOPE

• B. Karakostas (Event prediction in an IoT Environment using Naïve
Bayesian models) proposed an airplane monitoring scenario, where data
generated by C2T application are analyzed through a model in order to
estimate flights delay.
– Suppose an inbound flight is delayed due to bad weather.
– The model collects this data and uses the Naïve Bayesian algorithm to
estimate the probability of a delay in the corresponding outbound flight.

• Y. Lyu (High-performance scheduling model for multisensor gateway

of cloud sensor system-based smart-living ) used a smart living space
in their scheduling model with gateways to access cloud resources.

• Benazzouz and I. Parissis (A fault fuzzy-ontology for large scale fault-

tolerant wireless sensor networks) presents a model to improve the
energy efficiency of the IoT devices.
– Suppose a sensor monitoring air quality fails.
– The fuzzy-ontology model detects this fault by analyzing the data
patterns and comparing them with expected behavior.
Baiju B V, VIT, SCOPE
• Renna et al. (Query processing for the Internet-of-Things: Coupling of
device energy consumption and cloud infrastructure billing) analyzes
the relation between energy consumption and cloud infrastructure
billing cost.

• H.Wu et al. (Deviation-based neighborhood model for context-aware

QoS prediction of cloud and IoT Services) used a model to analyze the
energy consumption of a WSN.
Baiju B V, VIT, SCOPE
• F. Jalali (Interconnecting fog computing and microgrids for greening
IoT) studies energy consumption but, in this case, the authors analyze
how the integration of IoT and fog can reduce this consumption.

• Desikan, Srinivasan, and Murthy (A Novel Distributed Latency-Aware

Data Processing in Fog Computing-Enabled IoT Networks) proposed a
model to improve the response time of an IoT system that uses gateways
to communicate to the cloud, using fog computing as an intermediary
platform. Baiju B V, VIT, SCOPE
• Li et al. (System
modelling and
performance
evaluation of a
three-tier cloud of
things) presents a
three-tier model
that address the
C2F2T scenario.

Baiju B V, VIT, SCOPE

 Integrated C2F2T Literature by Metrics
• Metrics observed in articles

Baiju B V, VIT, SCOPE

1.1 Energy Consumption
• IoT devices are limited by power source, typically a battery, which
comprises application performance.
• W. Li, I. Santos et al. (System modelling and performance evaluation
of a three-tier cloud of things) proposed a model to represent a C2F2T
architecture.
• They evaluated the
energy consumption in
each of the application
layers (IoT, fog, and
cloud), identified the
main sources of
consumption of each,
and verified the energy
consumption in relation
to the increase in the
number of devices in the
architecture. Baiju B V, VIT, SCOPE
• F. Renna (Query processing for the Internet-of-Things: Coupling of
device energy consumption and cloud infrastructure billing), presented
energy consumption of devices in an IoT infrastructure in a different way.
• The model represents the device energy cost in three modes:
– Active
– Idle
– when the application switches from idle to active state.

Smart Light:
Active Mode: The light is on,
illuminating the room.
Idle Mode: The light is off but
ready to turn on when
needed.
Switching: The light turns on
when someone enters
the room.
Baiju B V, VIT, SCOPE
• L. Skorin-Kapov (Energy efficient and quality-driven continuous
sensor management for mobile IoT applications) address the energy
consumption of idle and active devices sending data to a cloud
environment, but in this work, the energy consumption for transmitting,
receiving, listening, sensing, and computing are taken into account.

Traffic Monitoring:

Transmitting: Traffic sensors send data about vehicle flow to the cloud.
Receiving: Sensors receive updates or commands from the cloud to adjust
their monitoring frequency.
Listening: Sensors stay alert for any emergency commands.
Sensing: Actively collecting data on traffic conditions.
Computing: Processing data locally to detect traffic anomalies before sending
it to the cloud.

Baiju B V, VIT, SCOPE

• T. Gomes (An IoT-based system
for collision detection on
guardrails) , considered the
operational mode of devices,
however, in this work the energy
consumption depends on the
distance between the
communicating nodes and the
number of bits that will be
transmitted.
• In addition, the authors take into
account the medium where the
communication will occur – for
instance, open space.

Baiju B V, VIT, SCOPE

• Zaineb T (Virtualization framework for energy efficient IoT networks )
addresses that the energy consumption is also impacted by processing-
induced VMs located in the networking elements at the upper three layers.

Baiju B V, VIT, SCOPE

• Jalali et al. (Interconnecting fog computing and microgrids for
greening IoT) evaluate the energy consumption with respect to two
perspectives:
– the synchronization between IoT devices and the fog
– the synchronization between IoT devices and the cloud.

Baiju B V, VIT, SCOPE

Synchronization between IoT devices and the fog

Consider a smart traffic management system where IoT sensors on traffic

lights and vehicles communicate with nearby fog nodes. These fog nodes
process data locally to make real-time decisions, such as adjusting traffic
light timings to reduce congestion. The synchronization ensures that data
from multiple sensors is processed simultaneously, enabling quick and
efficient traffic management1.

Synchronization between IoT devices and the cloud

Imagine a smart home system where various IoT devices like
thermostats, security cameras, and smart locks send data to a cloud
server. The cloud server processes this data to provide insights and
control options via a mobile app. Synchronization ensures that all devices
are updated with the latest settings and data, allowing users to control
their home environment seamlessly from anywhere

Baiju B V, VIT, SCOPE

• Y. Li et. al. (Leveraging Renewable Energy in Edge Clouds for Data
Stream Analysis in IoT assesses) the energy consumption of fog devices
and cloud equipment.
• The fog devices are more similar to the IoT devices with respect to
computational capacity.
• It presents a renewable energy source where the fog devices are powered
by photovoltaic panels.
• They evaluate the consumption of energy produced for photovoltaic panels,
and other energy sources, such as batteries.

Baiju B V, VIT, SCOPE

1.2 Performance
• W. Li et al. (System modelling and performance evaluation of a three-tier
cloud of things) and S. Sarkar et al. (Theoretical modelling of fog
computing: a green computing paradigm to support IoT applications)
evaluated the system performance with respect to the latency of the
application.
• This latency can be divided into
(i) Processing latencies
(ii) Transmission latencies.
• The processing latency is the time taken by an application to process all tasks.
This is the time your email application takes to prepare the email, attach the
file, and get it ready to send. For instance, if your computer takes 2 seconds to
attach a file and another 1 second to encrypt the email, the total processing
latency is 3 seconds.
• The transmission latency is the communication delay to send a unit of data to
the destination.
This is the time it takes for the email to travel from your computer to the
recipient's email server. If your internet connection takes 5 seconds to send the
email to the server, then the transmission latency is 5 seconds. Baiju B V, VIT, SCOPE
• T. Dinh (A location-based interactive model for Internet of Things and
cloud (IoT-cloud)), evaluated the packet delivery latency
• This is calculated by
– the number of hops to the destination
– the sleep interval of a node
– the time to transmit a data packet
Consider a small IoT network where a data packet needs to be sent from a
sensor node to the cloud server. The packet must pass through 3 intermediate
nodes (3 hops). Each intermediate node has a sleep interval of 20 ms, and the
time to transmit a data packet per hop is 10 ms.
Sleep interval for each hop = 20 ms
Transmission time for each hop = 10 ms
Total hops = 3
Total Latency=Number of Hops × (Sleep Interval + Transmission Time)

Total Latency = 3×(20ms+10ms)

Total Latency = 3×30 ms
Total Latency = 90 ms
Baiju B V, VIT, SCOPE
• E. Sturzinger (Application-aware resource provisioning in a
heterogeneous Internet of Things ) presented the delay of processing
applications requests which is defined analytically as a combination of
computational complexity of devices and the average flow size of
application.
• Lyu et al. (High-performance scheduling model for multisensor gateway
of cloud sensor system-based smart-living ) evaluate the performance of a
scheduling mechanism.
– For this, they assess the scalability of the scheduling and caching
solver and add several sensors connected to cloud servers.
– In addition, the average waiting time and throughput applications are
analyzed.
• G. Merlino et al. (Mobile crowdsensing as a service: a platform for
applications on top of sensing clouds ) proposed a mobile crowdsensing
scenario is used to evaluate the performance and the effectiveness of the
system using a Petri Net model.

Baiju B V, VIT, SCOPE

• K. E. Desikan (A Novel Distributed Latency-Aware Data Processing in
Fog Computing-Enabled IoT Networks), examined the performance of
C2F2T applications using a variety of means.
• The response time of system is defined as the time elapsed from the data
was generated until it is processed by a gateway.
• Efficiency processing is formulated mathematically and takes account of
– the time to process the application data
– the occupancy buffer
– the response time of all gateways

Tp​ =30ms (Time to process the application data).

𝑇𝑏=20 ms (Buffer occupancy time).
𝑇𝑔=50 ms (Total response time of all gateways).
Total system response time (Ttotal​) = Tp​ +Tb​ +Tg​

Ttotal​ =30ms+20ms+50ms
𝑇𝑡𝑜𝑡𝑎𝑙=100 ms
Baiju B V, VIT, SCOPE
1.3 Resource Consumption
• Due of the limited capabilities of IoT devices, careful attention is required
when allocating tasks on these devices.
• Fog and cloud computing offer additional capabilities for IoT devices, it is
necessary to examine where such workloads should be processed.
• V. B. Souza et al. (Towards Distributed Service Allocation in Fog-to-
Cloud (F2C) Scenarios) aim to obtain the optimal allocation of services
on the available resources, taking into account resources of devices
available in C2F2T infrastructures.
– They present the advantages of using fog computing and evaluate the
number of allocated resources and the usage of resource devices.
• Y. Liu et al. (Incentive mechanism for computation offloading using
edge computing: a Stackelberg game approach) assessed the level
consumption in C2T scenario.
• They evaluated the consumption about two aspects:
– Cloud infrastructure consumption
– IoT devices consumption
Baiju B V, VIT, SCOPE
1.4 Cost
• Applications that require a lot of cloud computing resources can dramatically
increase the cost to the service provider and ultimately the consumer
• Renna et al. (Query processing for the Internet-of-Things: Coupling of
device energy consumption and cloud infrastructure billing) examine the
billing costs when computational resources are reserved to process queries
uploaded by IoT applications.
• Billing costs are directly proportional to the expected query volume generated
by IoT devices.

In a smart agriculture system, IoT sensors monitor soil moisture, temperature,

and humidity, continuously sending data to the cloud. If the number of sensors
increases or the frequency of queries rises, the cloud infrastructure usage also
increases. As a result, billing costs grow since they are directly tied to the
volume of queries processed. This highlights the importance of optimizing both
query frequency and resource allocation to control costs while maintaining
efficiency.

Baiju B V, VIT, SCOPE

• Sturzinger et al. (Application-aware resource provisioning in a
heterogeneous Internet of Things ) attempt to minimize the
operational cost of provisioning IoT traffic to the cloud, while satisfying
all application constraints.
• The total cost is composed of the sum of the cost of all devices including
as processing, storage, upstream and downstream cost, local and
global traffic, and the link capacity cost.

Baiju B V, VIT, SCOPE

1.5 Quality of Service
• Z. Ming (QoS-aware computational method for IoT composite service),
evaluated composite services in an IoT environment.
• R. K. Behera (Reliability modelling of service-oriented Internet of Things)
presented reliability of the IoT system depends on a variety of factors,
including the
– availability of the service
– availability of the input files to service
– reliability of each sub-system that composes the overall IoT system.
• R. K. Lomotey (Wearable IoT data stream traceability in a distributed
health information system) defined reliability as the ability of the
traceability model to detect and prevent attacks on applications.

Baiju B V, VIT, SCOPE

• Y. Li et al. (Leveraging Renewable Energy in Edge Clouds for Data Stream
Analysis in IoT ) presented the use case for analyses of video streams in the
cloud generated by vehicles on a road.
• Authors represented QoS as the detection accuracy of an object in video
images.
• They vary the resolution of video in order to decrease the CPU, memory and
bandwidth consumption.
• In addition,
considering the
scenario where there
are 2n+1 cars, the
accuracy is defined as
the probability of a
result (object detected
in image) appears
exceeds n + 1 times
among 2n + 1 results.

Baiju B V, VIT, SCOPE

1.6 Security
• R. K. Behera et al. (Reliability modelling of service-oriented Internet
of Things) evaluated the concept of opacity in data flow in the C2T
context.
• Opacity is a uniform approach to describe security properties expressed
as predicates.
• The authors used analytical models to verify the opacity in the medical
research application scenario.

Baiju B V, VIT, SCOPE