Scribd
Upload
4 views20 pages

BCSE313L Module4

The document discusses optimization problems in fog and edge computing, emphasizing the importance of minimizing latency and energy consumption while maximizing security and reliability. It outlines a formal modeling framework for fog computing, including a three-layer architecture consisting of cloud, edge resources, and end devices, and highlights various optimization opportunities throughout the fog architecture and service life cycle. Additionally, it addresses metrics for performance, resource usage, energy consumption, and financial costs, along with further quality attributes that enhance reliability and security.

Uploaded by

22f3003147
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
4 views20 pages

BCSE313L Module4

The document discusses optimization problems in fog and edge computing, emphasizing the importance of minimizing latency and energy consumption while maximizing security and reliability. It outlines a formal modeling framework for fog computing, including a three-layer architecture consisting of cloud, edge resources, and end devices, and highlights various optimization opportunities throughout the fog architecture and service life cycle. Additionally, it addresses metrics for performance, resource usage, energy consumption, and financial costs, along with further quality attributes that enhance reliability and security.

Uploaded by

22f3003147
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 20

BCSE313L– Fundamentals of Fog and Edge Computing

MODULE - 4
Optimization Problems in Fog and Edge
Computing
Preliminaries - The Case for Optimization in Fog Computing-Formal
Modeling Framework for Fog Computing – Metrics - Further Quality
Attributes - Optimization Opportunities along the Fog Architecture -
Optimization Opportunities along the Service Life Cycle - Toward a

Taxonomy of Optimization Problems in Fog Computing

Dr. V. Sakthivel , AsP/ SCOPE, VIT, Chennai


Optimization

• Optimization plays a crucial role in fog


computing.
• For example, minimizing latency and energy
consumption are just as important as maximizing
security and reliability.
• Because of the high complexity of typical fog
deployments (many different types of devices,
with many different types of interactions) and
their dynamic nature (mobile devices coming
and going, devices or network connections
failing permanently or temporarily etc.
Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 2
Preliminaries
• An optimization problem is generally defined by :

A tuple (v1, … , vn) that satisfies (1) and (2) is called a solution
of the problem. Thus, the goal is to find the solution with
highest
Winter Semester f value Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai
2024-2025 3
Formal Modeling Framework for Fog Computing

Three-layer model
of fog computing
Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 4
Contd…
• Higher layers represent higher computational capacity, but at the
same time also higher distance – and thus higher latency – from the
end devices.
• On the highest layer is the cloud with its virtually unlimited, high-
performance, and cost- and energy-efficient resources.
• The middle layer consists of a set of edge resources: machines
offering compute services near the network edge, e.g. in base
stations, routers, or small, geographically distributed data centers of
telecommunication providers.
• The edge resources are all connected to the cloud. Finally, the
lowest layer contains the end devices like mobile phones or IoT
devices. Each end device is connected to one of the edge
resources.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 5


Contd…
• let c denote the cloud, E the set of edge resources, De the set of
end devices connected to edge resource e ∈ E, and D = ⋃e∈E De
the set of all end devices. The set of all resources is R = {c} ∪ E ∪
D.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 6


Contd…

• Each resource r ∈ R is associated with a


compute capacity a(r) ∈ ℝ+ and a
compute speed s(r) ∈ ℝ+.
• Moreover, each resource has some power
consumption, which depends on its
computational load.
• Specifically, the power consumption of
resource r increases by w(r) ∈ ℝ+ for
every instruction to be carried out by r.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 7


Metrics

Performance

Total execution time of an example


computation offloading scenario
Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 8
Resource Usage

• End devices, which typically have very limited


CPU and memory capacity.
• Edge resources typically offer higher capacities,
but also those capacities can be limited, given
that edge resources may include machines like
routers that do not offer exhaustive
computational capabilities.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 9


Energy Consumption
• Energy can also be seen as a scarce resource, but it is quite
different from the other resource types already considered.
• Energy is consumed by all resources as well as the network.
• Even idle resources and unused network elements consume
energy, but their energy consumption increases with usage.
• Generally, assuming that the power consumption of a resource
depends linearly on its CPU load is a good approximation.
• It is important, though, to highlight the difference between power
consumption and energy consumption, since energy consumption
also depends on the amount of time during which power is
consumed.
• Thus, it is, for instance, beneficial in terms of overall energy
consumption to move a compute task from one resource to a
significantly faster one, even if the faster machine has slightly higher
power consumption.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 10


Financial Costs

• use of the cloud or edge infrastructure may incur


costs.
• These costs can be fixed or usage-based, or
some combination.
• Similarly, the use of the network for transferring
data may incur costs.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 11


Further Quality Attributes

• To increase reliability, it is beneficial to let


multiple resources perform the same critical
computations in parallel, so that the result is
available even if some of the resources stop
working or become unreachable, and also to
compare the results with each other to filter out
flawed results.
• Both security and privacy concerns may be
mitigated by preferring trusted resources.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 12


Contd…

• Co-location of computational tasks belonging to


different users / tenants may increase the
likelihood of tenant-on-tenant attacks.
• Co-location of tasks belonging to the same user
decreases the need for exchanging data over
the network, which in turn decreases the
likelihood of eavesdropping, man-in-the-middle,
and other network-based attacks.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 13


Optimization Opportunities along the Fog
Architecture
• Problems involving the cloud and the edge
resources.
• This is a meaningful setting, which allows for
example to optimize overall energy consumption
of cloud and edge resources, subject to capacity
and latency constraints.
• This setup shows some similarity to distributed
cloud computing; a potential difference is that
the number of edge resources can be several
orders of magnitude higher than the number of
data centers in a distributed cloud.
Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 14
Contd…

• Problems involving edge resources and end devices.


• The collaboration of end devices with edge resources
(e.g., offloading computations) is a typical fog computing
problem, and because of the limited resources of end
devices, optimization plays a vital role in such cases.
• many different technical issues would have to be
integrated Optimization Problems in Fog and Edge
Computing into a single optimization problem to capture
the different optimization concerns of the cloud, the edge
resources, and the end devices, which is challenging in
itself

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 15


Optimization Opportunities along the Service Life
Cycle
• Design-time optimization
– optimization will be constrained mostly to the cloud
and edge layers of the architecture, where more
information may already be available at design time.
• Deployment-time optimization
– When the deployment of the service on specific
resources is planned, the available information of the
resources can be used to make further optimization
decisions.

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 16


Contd…

• Run-time optimization
– some aspects of a fog system may be optimized in
advance (i.e., during design time or deployment time),
many important aspects become clear only when the
system is running and used.
– This requires continuous monitoring of important
system parameters, analysis of whether the system
still operates with acceptable effectiveness and
efficiency, and re-optimization whenever necessary

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 17


Toward a
Taxonomy of Optimization Problems in Fog Computing

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 18


Contd…

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 19


Contd…

Winter Semester 2024-2025 Dr. V. Sakthivel, AsP/ SCOPE, VIT, Chennai 20

You might also like