Internet DRAFT - draft-irtf-nmrg-green-ps
draft-irtf-nmrg-green-ps
Network Working Group A. Clemm, Ed.
Internet-Draft C. Westphal
Intended status: Informational Futurewei
Expires: 2 July 2024 J. Tantsura
Nvidia
L. Ciavaglia
Nokia
C. Pignataro, Ed.
NC State University
M-P. Odini
30 December 2023
Challenges and Opportunities in Management for Green Networking
draft-irtf-nmrg-green-ps-02
Abstract
Reducing humankind's environmental footprint and making technology
more sustainable are among the biggest challenges of our age.
Networks play an important part in this challenge. On one hand, they
enable applications that help to reduce this footprint. On the other
hand, they contribute to this footprint themselves in no
insignificant way. Methods to make networking technology itself
"greener" and to manage and operate networks in ways that reduces
their environmental footprint without impacting their utility
therefore need to be explored. This document outlines a
corresponding set of opportunities, along with associated research
challenges, for networking technology in general and management
technology in particular to become "greener", i.e., more sustainable,
with reduced greenhouse gas emissions and less negative impact on the
environment.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
Clemm, et al. Expires 2 July 2024 [Page 1]
Internet-Draft Management for Green Networking December 2023
This Internet-Draft will expire on 2 July 2024.
Copyright Notice
Copyright (c) 2023 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document. Code Components
extracted from this document must include Revised BSD License text as
described in Section 4.e of the Trust Legal Provisions and are
provided without warranty as described in the Revised BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1. Motivation . . . . . . . . . . . . . . . . . . . . . . . 3
1.2. Approaching the Problem . . . . . . . . . . . . . . . . . 5
1.3. Structuring the Problem Space . . . . . . . . . . . . . . 6
2. Definitions and Acronyms . . . . . . . . . . . . . . . . . . 8
3. Network Energy Consumption Characteristics and
Implications . . . . . . . . . . . . . . . . . . . . . . 9
4. Challenges and Opportunities - Equipment Level . . . . . . . 12
4.1. Hardware and Manufacturing . . . . . . . . . . . . . . . 12
4.2. Visibility and Instrumentation . . . . . . . . . . . . . 13
5. Challenges and Opportunities - Protocol Level . . . . . . . . 15
5.1. Protocol Enablers for Carbon Optimization Mechanisms . . 15
5.2. Protocol Optimization . . . . . . . . . . . . . . . . . . 16
5.3. Data Volume Reduction . . . . . . . . . . . . . . . . . . 17
5.4. Network Addressing . . . . . . . . . . . . . . . . . . . 19
6. Challenges and Opportunities - Network Level . . . . . . . . 20
6.1. Network Optimization and Energy/Carbon/Pollution-Aware
Networking . . . . . . . . . . . . . . . . . . . . . . . 20
6.2. Assessing Carbon Footprint and Network-Level
Instrumentation . . . . . . . . . . . . . . . . . . . . . 21
6.3. Dimensioning and Peak Shaving . . . . . . . . . . . . . . 22
6.4. Convergence Schemes . . . . . . . . . . . . . . . . . . . 23
6.5. The Role of Topology . . . . . . . . . . . . . . . . . . 24
7. Challenges and Opportunities - Architecture Level . . . . . . 25
8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . 27
9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 28
10. Security Considerations . . . . . . . . . . . . . . . . . . . 28
11. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 28
12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 29
13. Informative References . . . . . . . . . . . . . . . . . . . 29
Clemm, et al. Expires 2 July 2024 [Page 2]
Internet-Draft Management for Green Networking December 2023
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 32
1. Introduction
1.1. Motivation
Climate change and the need to curb greenhouse gas (GHG) emissions
have been recognized by the United Nations and by most governments as
one of the big challenges of our time. As a result, curbing those
emissions is becoming of increasing importance for society and for
many industries. The networking industry is no exception.
The science behind greenhouse gas emissions and their relationship
with climate change is complex. However, there is overwhelming
scientific consensus pointing towards a clear correlation between
climate change and a rising amount of greenhouse gases in the
atmosphere. One greenhouse gas of particular concern, but by no
means the only one, is carbon dioxide (CO2). Carbon dioxide is
emitted in the process of burning fuels to generate energy that is
used, for example, to power electrical devices such as networking
equipment. Notable here is the use of fossil fuels, such as oil,
which releases CO2 that had long been removed from the earth's
atmosphere, as opposed to the use of renewable or sustainable fuels
that do not "add" to the amount of carbon in the atmosphere.
Greenhouse gas emissions are in turn correlated with the need to
power technology, including networks. Reducing those emissions can
be achieved by reducing the amount of fossil fuels needed to generate
the energy that is needed to power those networks. This can be
achieved by improving the energy mix to include increasing amounts of
renewable (and hence sustainable) energy sources such as wind or
solar. It can also be achieved by increasing energy savings and
improving energy efficiency so that the same outcomes can be achieved
while consuming less energy in the first place.
The amount of CO2 that is emitted in burning fossil fuels to generate
energy is also referred to as carbon footprint. Reducing this
footprint to net-zero is hence a major sustainability goal. However,
sustainability encompasses also other factors beyond carbon, such as
sustainable use of other natural resources, the preservation of
natural habitats and biodiversity, and the avoidance of any form of
pollution.
Clemm, et al. Expires 2 July 2024 [Page 3]
Internet-Draft Management for Green Networking December 2023
In the context of this document, we refer to networking technology
that helps to improve its own networking sustainability as "green".
Green, in that sense, includes technology that helps to lower
networking's greenhouse gas emissions including carbon footprint,
which turn includes technology that helps to increase efficiency and
realize energy savings as well as facilitating managing networks
towards stronger use of renewables.
Arguably, networks can already be considered a "green" technology in
that networks enable many applications that allow users and whole
industries to save energy and become more sustainable in a
significant way. For example, it allows (at least to an extent) to
replace travel with teleconferencing; it enables many employees to
work from home and "telecommute," thus reducing the need for actual
commute; IoT applications that facilitate automated monitoring and
control from remote sites help make agriculture more sustainable by
minimizing the application of resources such as water and fertilizer;
networked smart buildings allow for greater energy optimization and
sparser use of lighting and HVAC (heating, ventilation, air
conditioning) than their non-networked not-so-smart counterparts.
The IETF has recently initiated a reflection on the energy cost of
hosting meetings three times a year (see for instance [IETF-Net0]).
It conducted a study of the carbon emissions of a typical meeting and
found out that 99% of the emissions were due to the air travel. In
the same vein, [Framework] compared an in-person with a virtual
meeting and found a reduction in energy of 66% for a virtual meeting.
These findings confirm that networking technology can reduce
emissions when acting as virtual substitution for physical events.
That said, networks themselves consume significant amounts of energy.
Therefore, the networking industry has an important role to play in
meeting sustainability goals not just by enabling others to reduce
their reliance on energy, but by also reducing its own. Future
networking advances will increasingly need to focus on becoming more
energy-efficient and reducing carbon footprint, both for economic
reasons and for reasons of corporate responsibility. This shift has
already begun, and sustainability is already becoming an important
concern for network providers. In some cases such as in the context
of networked data centers, the ability to procure enough energy
becomes a bottleneck prohibiting further growth and greater
sustainability thus becomes a business necessity.
For example, in its annual report, Telefónica reports that in 2021,
its network's energy consumption per PB of data amounted to 54MWh
[Telefonica2021]. This rate has been dramatically decreasing (a
seven-fold factor over six years) although gains in efficiency are
being offset by simultaneous growth in data volume. In the same
Clemm, et al. Expires 2 July 2024 [Page 4]
Internet-Draft Management for Green Networking December 2023
report, it is stated as an important corporate goal to continue on
that trajectory and aggressively reduce overall carbon emissions
further.
1.2. Approaching the Problem
An often considered gain in networking sustainability can be made
with regards to improving the efficiency with which networks utilize
power during their use phase, reducing the amount of energy that is
required to provide communication services. However, for a holistic
approach other aspects need to be considered as well.
Environmental footprint is determined not by power consumption alone.
The sustainability of power sources needs to be taken into account as
well. A deployment that includes devices that are less energy-
efficient but that are powered by a sustainable energy source can
arguably be considered "greener" than a deployment that includes
highly efficient device that are powered by Diesel generators. In
fact, in the same Telefónica report mentioned earlier, extensive
reliance on renewable energy sources is emphasized.
Similarly, deployments can take other environmental factors into
account that affect carbon footprint. For example, deployments in
which factors such as the need for cooling are reduced, or where
excessive heat that is generated by equipment can be put to
productive use, will be considered greener than deployments where
this is not the case. Examples include deployments in cooler natural
surroundings (e.g., in colder climates) where that is an option.
Likewise, manufacturing and recycling of networking equipment are
also part of the sustainability equation, as the production itself
consumes energy and results in a carbon cost embedded as part of the
device itself. Extending the lifetime of equipment may in many cases
be preferable over replacing it earlier with equipment that is
slightly more energy-efficient but that requires the embedded carbon
cost to be amortized over a much shorter period of time.
Management has an outsized role to play in approaching those
problems. To reduce the amount of energy used, network providers
need to maximize ways in which they make use of scarce resources and
eliminate use of resources which are not needed. They need to
optimize the way in which networks are deployed, which resources are
placed where, how equipment lifecycles and upgrades are being managed
- all of which constitute classic operational problems. As best
practices, methods, and algorithms are developed, they need to be
automated to the greatest extent possible and migrated over time into
the network and performed on increasingly short time scales,
transcending management and control planes.
Clemm, et al. Expires 2 July 2024 [Page 5]
Internet-Draft Management for Green Networking December 2023
1.3. Structuring the Problem Space
From a technical perspective, multiple vectors along which networks
can be made "greener" should be considered:
* Equipment level:
Perhaps the most promising vector for improving networking
sustainability concerns the network equipment itself. At the most
fundamental level, networks (even softwarized ones) involve
appliances, i.e., equipment that relies on electrical power to
perform its function. There are two distinct layers with
different opportunities for improvement:
- Hardware: Reducing embedded carbon during material extration
and manufacturing, improving energy and power efficiency during
operations, and reuse, repurpose, and recycle motions.
- Software: Improving sofware energy efficiency, maximizing
utilization of processing devices, allowing for software to
interact with hardware to improve sustainability.
Beyond making network appliances merely more energy-efficient,
there are other important ways in which equipment can help
networks become greener. This includes aspects such as support
for port power saving modes or downspeeding of links to reduce
power consumption for resources that are not fully utilized. To
fully tap into the potential of such features requires
accompanying management functionality, for example in order to
determine when it is "safe" to downspeed a link or enter a power
saving mode, and manage the network in such a way that conditions
to do so are maximized.
Most importantly from a management perspective, improving
sustainability at the equipment level involves providing
management instrumentation that allows to precisely monitor and
manage power usage and doing so at different levels of
granularity, for example accounting separately for the
contributions of CPU, memory, and different ports. This enables
(for example) controller applications to optimize energy usage
across the network and that leverage control loops to assess the
effectiveness (e.g. in terms of reduction in power use) of
measures that are taken.
As a side note, the terms "device" and "equipment", as used in the
context of this draft, are used to refer to networking equipment.
We are not taking into consideration end-user devices and
endpoints such as mobile phones or computing equipment.
Clemm, et al. Expires 2 July 2024 [Page 6]
Internet-Draft Management for Green Networking December 2023
* Protocol level:
Energy-efficiency and greenness are aspects that are rarely
considered when designing network protocols. This suggests that
there may be plenty of untapped potential. Some aspects involve
designing protocols in ways that reduce the need for redundant or
wasteful transmission of data to allow not only for better network
utilization, but greater goodput per unit of energy being
consumed. Techniques might include approaches that reduce the
"header tax" incurred by payloads as well as methods resulting in
the reduction of wasteful retransmissions. Similarly, there may
be cases where chattiness of protocols may be preventing equipment
from going into sleep mode. Designing protocols that reduce
chattiness in such scenarios, for example, that reduce dependence
on periodic updates or heartbeats, may result in greener outcomes.
Likewise, aspects such as restructuring addresses in ways that
allow to minimize the size of lookup tables and associated memory
sizes and their energy use can play a role as well.
Another role of protocols concerns the enabling of management
functionality to improve energy efficiency at the network level,
such as discovery protocols that allow for quick adaptation to
network components being taken dynamically into and out of service
depending on network conditions, as well as protocols that can
assist with functions such as the collection of energy telemetry
data from the network.
* Network level
Perhaps the greatest opportunities to realize power savings exist
at the level of the network as whole. Many of these opportunities
are directly related to management functionaliy. For example,
optimizing energy efficiency may involve directing traffic in such
a way that it allows for isolation of equipment that may at the
moment not be needed so that it can be powered down or brought
into power-saving mode. By the same token, traffic should be
directed in a way that requires bringing additional equipment
online or out of power-saving mode in cases where alternative
traffic paths are available for which the incremental energy cost
would amount to zero. Likewise, some networking devices may be
rated less "green" and more power-intensive than others or powered
by less-sustainable energy sources. Their use might be avoided
unless during periods of peak capacity demands. Generally,
incremental carbon emissions can be viewed as a cost metric that
networks should strive to minimize and consider as part of routing
and of network path optimization.
Clemm, et al. Expires 2 July 2024 [Page 7]
Internet-Draft Management for Green Networking December 2023
* Architecture level
The current network architecture supports a wide range of
applications, but does not take into account energy efficiency as
one of its design parameters. One can argue that the most energy
efficient shift of the last two decades has been the deployment of
Content Delivery Network overlays: while these were set up to
reduce latency and minimize bandwidth consumption, from a network
perspective, retrieving the content from a local cache is also
much greener. What other architectural shifts can produce energy
consumption reduction?
We believe that network standardization organizations in general, and
IETF in particular, can make important contributions to each of these
vectors. In this document, we will therefore explore each of those
vectors in further detail and for each point out specific challenges
for IETF. As our starting point, we borrow some material from a
prior paper, [GreenNet22]. For this document, this material has been
both expanded (for example, in terms of some of the opportunities)
and pruned (for example, in terms of background on prior scholarly
work). In addition, this document focuses on and attempts to
articulate specific challenges relating to work that could be
championed by the IETF to make a difference.
2. Definitions and Acronyms
Below you find acronyms used in this draft:
Carbon Footprint:
As used in this document, the amount of carbon emissions
associated with the use or deployment of technology, usually
correlated with the amount of energy consumption
CDN: Content Delivery Network
CPU: Central Processing Unit, that is the main processor in a
server
DC: Data Center
FCT: Flow Completion Time
GHG: Greenhouse Gas
GPU: Graphical Processing Unit
HVAC: Heating, Ventilation, Air Conditioning
Clemm, et al. Expires 2 July 2024 [Page 8]
Internet-Draft Management for Green Networking December 2023
ICN: Information Centric Network
IGP: Interior Gateway Protocol
IoT: Internet of Things
IPU: Infrastructure Processing Units
LEED: Leadership in Energy and Environmental Design, a green
building rating system
LEO: Low Earth Orbit
LPM: Longest Prefix Match, a method to look up prefixes in a
forwarding element
MPLS: Multi-Path Label Switchin
MTU: Maximum Transmission Unit, the largest packet size that can be
transmitted over a network
NIC: Network Interface Card
QoS, QoE: Quality of Service, Quality of Experience
QUIC: Quick UDP Internet Connections
SNIC: Smart NIC
SDN: Software-Defined Networking
TCP: Transport Control Protocol
TE: Traffic Engineering
TPU: Tensor Processing Unit
WAN: Wide Area Network
3. Network Energy Consumption Characteristics and Implications
Carbon footprint and, with it, greenhouse gas emissions are
determined by a number of factors. A main factor is network energy
consumption, as the energy consumed can be considered a proxy for the
burning of fuels required for corresponding power generation.
Network energy consumption by itself does not tell the whole story,
as it does not take the sustainability of energy sources and energy
mix into account. Likewise, there are other factors such as hidden
Clemm, et al. Expires 2 July 2024 [Page 9]
Internet-Draft Management for Green Networking December 2023
carbon cost reflecting the carbon footprint expended in manufacturing
of networking hardware. Nonetheless, network energy consumption is
an excellent predictor for carbon footprint and its reduction key to
sustainable solutions. Exploring possibilities to improve energy
efficiency is hence a key factor for greener, more sustainable, less
carbon-intensive networks.
For this, it is important to understand some of the characteristics
of power consumption by networks and which aspects contribute the
most. This helps to identify where the greatest potential not just
for power savings but also for sustainability improvements lies.
Power is ultimately drawn by devices. Devices are not monoliths but
are composed of multiple components. The power consumption of the
device can be divided into the consumption of the core device - the
backplane and CPU, if you will - as well as additional consumption
incurred per port and line card. In addition, GPU and TPU may be
used as well in the network and may have different power consumption
profiles. Furthermore, it is important to understand the difference
between power consumption when a resource is idling versus when it is
under load. This helps to understand the incremental cost of
additional transmission versus the initial cost of transmission.
In typical networking devices, only roughly half of the energy
consumption is associated with the data plane [Bolla2011energy]. An
idle base system typically consumes more than half of the power over
the same system running at full load [Chabarek08], [Cervero19].
Generally, the cost of sending the first bit is very high, as it
requires powering up a device, port, etc. The incremental cost of
transmission of additional bits (beyond the first) is many orders of
magnitude lower. Likewise, the incremental cost of incremental CPU
and memory needed to process additional packets becomes fairly
negligible.
This means that a device's power consumption does not increase
linearly with the volume of forwarded traffic. Instead, it resembles
more of a step function in which power consumption stays roughly the
same up to a certain volume of traffic, followed by a sudden jump
when additional resources need to be procured to support a higher
volume of traffic.
By the same token, generally speaking it is more energy-efficient to
transmit a large volume of data in one burst (and subsequently
turning off or downspeeding the interface when idling), instead of
continuously transmitting at a lower rate. In that sense it can be
the duration of the transmission that dominates the energy
consumption, not the actual data rate.
Clemm, et al. Expires 2 July 2024 [Page 10]
Internet-Draft Management for Green Networking December 2023
The implications on green networking from an energy-savings
standpoint are significant. Of utmost importance are schemes that
allow for "peak shaving": networks are typically dimensioned for
periods of peak demand and usage, yet any excess capacity during
periods of non-peak usage does not result in corresponding energy
savings. Peak shaving techniques that allow to reduce peak traffic
spikes and thus waste during non-peak periods may result in outsize
sustainability gains. Peak shaving could be accomplished by
techniques such as spreading spikes out over geographies (e.g.
routing traffic across more costly but less utilized routes) or over
time (e.g. postponing and buffering non-urgent traffic).
Likewise, large gains can be made whenever network resources can
effectively be taken offline for at least some of the time, managing
networks in a way that enables resources to be removed from service
so they can be powered down (or put into a more energy-saving state,
such as when downspeeding ports) while not needed. Of course, any
such methods need to take into account the overhead of taking
resources offline and bringing them back online. This typically
takes some amount of time, requiring accurate predictive capabilities
to avoid situations in which network resources are not available at
times when they would be needed. In addition, there is additional
overhead such as synchronization of state to be accounted for.
At the same time, any non-idle resources should be utilized to the
greatest extent possible as the incremental energy cost is
negligible. Of course, this needs to occur while still taking other
operational goals into consideration, such as protection against
failures (allowing for readily available redundancy and spare
capacity in case of failure) and load balancing (for increased
operational robustness). As data transmission needs tend to
fluctuate wildly and occur in bursts, any optimization schemes need
to be highly adaptable and allow for control loops at very fast time
scales.
Similarly, for applications where this is possible, it may be
desirable to replace continuous traffic at low data rates with
traffic that is sent in burst at high data rates, in order to
potentially maximize the time during which resources can be idled.
As a result, emphasis needs to be given to technology that allows for
example to (at the device level) exercise very efficient and rapid
discovery, monitoring, and control of networking resources so that
they can be dynamically be taken offline or back into service,
without (at the network level) requiring extensive convergence of
state across the network or recalculation of routes and other
optimization problems, and (at the network equipment level) support
rapid power cycle and initialization schemes. There may be some
Clemm, et al. Expires 2 July 2024 [Page 11]
Internet-Draft Management for Green Networking December 2023
lessons that can be applied here from IoT, which has long had to
contend with power-constrained end devices that need to spend much of
their time in power saving states to conserve battery.
4. Challenges and Opportunities - Equipment Level
We are categorizing challenges and opportunities to improve
sustainability at the network equipment level along the following
lines:
* Hardware and manufacturing. Related opportunities are arguably
among the most obvious and perhaps "largest". However, solutions
here may lie largely outside IETF's scope.
* Visibility and instrumentation. Instrumenting equipment to
provide visibility into how they consume energy is key to
management solutions and control loops to facilitate optimization
schemes.
4.1. Hardware and Manufacturing
Perhaps the most obvious opportunities to make networking technology
more energy efficient exist at the equipment level. After all,
networking involves physical equipment to receive and transmit data.
Making such equipment more power efficient, have it dissipate less
heat to consume less energy and reduce the need for cooling, making
it eco-friendly to deploy, sourcing sustainable materials and
facilitating recycling of equipment at the end of its lifecycle all
contribute to making networks greener. More specific and unique to
networking are schemes to reduce energy usage of transmission
technology from wireless (antennas) to optical (lasers).
One critical aspect of the energy cost of networking is the cost to
manufacture and deploy the networking equipment. In addition, even
the development process itself comes with its own carbon footprint.
This is outside of the scope of this document: we only consider the
energy cost of running the network, as this is where the IETF can
play a role. However, a holistic approach would include into this
the embedded energy that is included in the networking equipment.
One aspect for the IETF may be to consider impact of deploying new
protocols on the rate of obsolescence of the equipment. For
instance, incremental approaches that do not require to replace
equipment right away - or even extend the lifetime of deployed
equipment - would have a lower energy footprint. This is one
important benefit also of technologies such as Software-Defined
Networking and Network Function Virtualization, as they may allow
support of new networking features through software updates without
requiring hardware replacements.
Clemm, et al. Expires 2 July 2024 [Page 12]
Internet-Draft Management for Green Networking December 2023
An attempt to compute not only the energy of running a network, but
also the energy embedded into manufacturing the equipment is
described in [Emergy] . This is denoted by "emergy", a portmanteau
for embedded energy. Likewise, an approach to recycling equipment
and a proof of concept using old cell phones recycled into a
"junkyard" data center are described in [Junkyard].
One trade-off to consider at this level is the selection of a
platform that can be hardware-optimized for energy efficiency vs a
platform that is versatile and can run multiple functions. For
instance, a switch could run on an efficient hardware platform, or
run as a software module (container) over some multi-purpose
platform. While the first one is operationally more energy
efficient, it may have a higher embedded energy from a smaller scale,
less efficient production process, as well as a shorter shelf life
once new functions need to be added to the platform.
4.2. Visibility and Instrumentation
Beyond "first-order" opportunities as outlined in the previous
subsection, network equipment just as importantly plays an important
role to enable and support green networking at other levels. Of
prime importance is the equipment's ability to provide visibility to
management and control plane into its current energy usage. Such
visibility enables control loops for energy optimization schemes,
allowing applications to obtain feedback regarding the energy
implications of their actions, from setting up paths across the
network that require the least incremental amount of energy to
quantifying metrics related to energy cost used to optimize
forwarding decisions. Absent an actual measurement of energy usage
(and until such measurement is put in place), the network equipment
could advertise some proxy of its power consumption (say, a labelling
scheme as silver, gold, platinum similar to the LEED sustainability
metric in building codes or the Energy Star label in home appliances;
or a description of the type of the device as using CPU vs GPU vs TPU
processors with different power profiles).
One prerequisite to such schemes is to have proper instrumentation in
place that allows to monitor current power consumption at the level
of networking devices as a whole, line cards, and individual ports.
Such instrumentation should also allow to assess the energy
efficiency and carbon footprint of the device as a whole. In
addition, it will be desirable to relate this power consumption to
data rates as well as to current traffic, for example, to indicate
current energy consumption relative to interface speeds, as well as
for incremental energy consumption that is expected for incremental
traffic (to aid control schemes that aim to "shave" power off current
services or to minimize the incremental use of power for additional
Clemm, et al. Expires 2 July 2024 [Page 13]
Internet-Draft Management for Green Networking December 2023
traffic). This is an area where the current state of the art is
sorely lacking and standardization lags behind. For example, as of
today, standardized YANG data models [RFC7950] for network energy
consumption that can be used in conjunction with management and
control protocols have yet to be defined.
To remedy this situation, an effort to define sets of green
networking metrics is currently under way
[I.D.draft-cx-green-metrics]. An agreed set of such metrics will
provide the basis for further steps such as the implementation of
corresponding data models as part of management and control
instrumentation.
Instrumentation should also take into account the possibility of
virtualization, introducing layers of indirection to assess the
actual energy usage. For example, virtualized networking functions
could be hosted on containers or virtual machines which are hosted on
a CPU in a data center instead of a regular network appliance such as
a router or a switch, leading to very different power consumption
characteristics. For example, a data center CPU could be more power
efficient and consume power more proportionally to actual CPU load.
Instrumentation needs to reflect these facts and facilitate
attributing power consumption in a correct manner.
Beyond monitoring and providing visibility into power consumption,
control knobs are needed to configure energy saving policies. For
instance, power saving modes are common in endpoints (such as mobile
phones or notebook computers) but sorely lacking in networking
equipment.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances:
* Basic equipment categorization as "energy-efficient" (or not) as a
first step to identify immediate potential improvements, akin to
the EnergyStar program from the US's Environmental Protection
Agency.
* Equipment instrumentation advances for improved energy-awareness;
definition and standardization of granular management information.
* Virtualized energy and carbon metrics and assessment of their
effectiveness in solutions that optimize carbon footprint also in
virtualized environments (including SDN, network slicing, network
function virtualization, etc.).
Clemm, et al. Expires 2 July 2024 [Page 14]
Internet-Draft Management for Green Networking December 2023
* Certification and compliance assessment methods that ensure that
green instrumentation cannot be manipulated to give false and
misleading data.
* Methods that allow to account for energy mix powering equipment,
to facilitate solutions that optimize carbon footprint and
minimize pollution beyond mere energy efficiency [Hossain2019].
5. Challenges and Opportunities - Protocol Level
There are several opportunities to improve network sustainability at
the protocol level. We characterize them along several categories.
The first and arguably most impactful category concerns protocols
that enable carbon footprint optimization schemes at the network
level and management towards those goals. Other categories concern
protocols designed to optimize data transmission rates under energy
considerations, protocols designed to reduce the volume of data to be
transmitted, and protocl aspects related to network addressing
schemes. While those categories may be less impactful, even areas
with smaller gains should not be left unexplored.
There is also substantial work in the area of IoT, which has had to
contend with energy-constrained devices for a long time. Much of
that work was motivated not by sustainability concerns but practical
concerns such as battery life. However, many aspects appear to also
apply in the context of sustainability, such as reducing chattiness
to allow IoT equipment to go into low-power mode. Accordingly, there
is opportunity to extend IoT work to more generalized scenarios. The
use of power-constrained protocols into the wider Internet happens
regularly. For instance, ARM-based chipsets initially designed for
energy-efficiency in battery-operated mobile devices have been
embraced in data centers for a similar trajectory.
5.1. Protocol Enablers for Carbon Optimization Mechanisms
As will be discussed in Section 6, energy-aware and pollution-aware
schemes can help improve network sustainability but require awareness
of related data. To facilitate such schemes, protocols are needed
that are able to discover what links are available along with their
energy efficiency. For instance, links may be turned off in order to
save energy and turned back on based upon the elasticity of the
demand. Protocols should be devised to discover when this happens,
and to have a view of the topology that is consistent with frequent
topology updates due to power cycling of the network resources.
Also, protocols are required to quickly converge onto an energy-
efficient path once a new topology is created by turning links on/
off. Current routing protocols may provide for fast recovery in the
Clemm, et al. Expires 2 July 2024 [Page 15]
Internet-Draft Management for Green Networking December 2023
case of failure. However, failures are hopefully relatively rare
events, while we expect an energy efficient network to aggressively
try to turn off links.
Some mechanism is needed to present to the management layer a view of
the network that identifies opportunities to turn resources off
(routers/links) while still providing an acceptable level of Quality
of Experience (QoE) to the users. This gets more complex as the
level of QoE shifts from the current Best Effort delivery model to
more sophisticated mechanisms with, for instance, latency, bandwidth
or reliability guarantees.
Similarly, schemes might be devised in which links across paths with
a favorable energy mix are preferred over other paths. This implies
that the discovery of topology should be able support corresponding
parameters. More generally speaking, any mechanism that provides
applications with network visibility is a candidate for
scrutinization as to whether it should be extended to provide support
for sustainability-related parameters.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances:
* Protocol advances to enable rapidly taking down, bring back
online, and discover availability and power saving status of
networking resources while minimizing the need for reconvergence
and propagation of state.
* Assess which protocols could be extended with energy- and
sustainability-related parameters in ways that would enable
"greener" networking solutions, and exploring those solutions.
5.2. Protocol Optimization
The second category involves designing protocols in such a way that
the rate of transmission is chosen to maximize energy efficiency.
For example, Traffic Engineering (TE) can be manipulated to impact
the rate adaptation mechanism [Ren2018jordan]. By choosing where to
send the traffic, TE can artificially congest links so as to trigger
rate adaptation and therefore reduce the total amount of traffic.
Most TE systems attempt to minimize Maximal Link Utilization (MLU)
but energy saving mechanisms could decide to do the opposite
(maximize minimal link utilization) and attempt to turn off some
resources to save power.
Another example is to set up the proper rate of transmission to
minimize the flow completion time (FCT) so as to enable opportunities
to turn off links. In a wireless context, [TradeOff] studies how
Clemm, et al. Expires 2 July 2024 [Page 16]
Internet-Draft Management for Green Networking December 2023
setting the proper initial value for the congestion window can reduce
the FCT and therefore allow the equipment to go faster into a low-
energy mode. By sending the data faster, the energy cost can be
significantly reduced. This is a simple proof of concept, but
protocols that allow for turning links into a low-power mode by
transmitting the data over shorter periods could be designed for
other types of networks beyond Wi-Fi access. This should be done
carefully: in the limit, a high rate of transmission over a short
period of time may create bursts that the network would need to
accommodate, with all attendant complications of bursty traffic. We
conjecture there is a sweet spot between trying to complete flows
faster while controlling for burstiness in the network. It is
probably advisable to attempt to send traffic paced yet in bulk
rather than spread out over multiple round trips. This is an area of
worthwhile exploration.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances:
* Protocol advances that allow greater control over traffic pacing
to account for fluctuations in carbon cost, i.e., control knobs to
"bulk up" transmission over short periods or to smoothen it out
over longer periods.
* Protocol advances that allow to optimize link utilization
according to different goals and strategies (including maximizing
minimal link utilization vs minimizing maximal link utilization,
etc.)
* Assessments of the carbon impact of such strategies.
5.3. Data Volume Reduction
The first category involves designing protocols in such a way that
they reduce the volume of data that needs to be transmitted for any
given purpose. Loosely speaking, by reducing this volume, more
traffic can be served by the same amount of networking
infrastructure, hence reducing overall energy consumption.
Possibilities here include protocols that avoid unnecessary
retransmissions. At the application layer, protocols may also use
coding mechanisms that encode information close to the Shannon limit.
Currently, most of the traffic over the Internet consists of video
streaming and encoders for video are already quite efficient and keep
improving all the time, resulting in energy savings as one of many
advantages (of course being offset by increasingly higher
resolution). However, it is not clear that the extra work to achieve
higher compression ratios for the payloads results in a net energy
gain: what is saved over the network may be offset by the
Clemm, et al. Expires 2 July 2024 [Page 17]
Internet-Draft Management for Green Networking December 2023
compression/decompression effort. Further research on this aspect is
necessary.
At the transport protocol layer, TCP and to some extent QUIC react to
congestion by dropping packets. This is a highly energy inefficient
method to signal congestion, since the network has to wait one RTT to
be aware that the congestion has occurred, and since the effort to
transmit the packet from the source up until it is dropped ends up
being wasted. This calls for new transport protocols that react to
congestion without dropping packets. ECN [RFC2481] is a possible
solution, however not widely deployed. DC-TCP [Alizadeh2010DCTCP] is
tuned for Data Centers, L4S is an attempt to port similar
functionality to the Internet [RFC9330]. Qualitative Communication
[QUAL] [Westphal2021qualitative] allows the nodes to react to
congestion by dropping only some of the data in the packet, thereby
only partially wasting the resource consumed by transmitted the
packet up to this point. Novel transport protocols for the WAN can
ensure that no energy is wasted transmitting packets that will be
eventually dropped.
Another solution to reduce the bandwidth of network protocols by
reducing their header tax, for example applying header compression.
An example in IETF is [RFC3095]. Again, reducing protocol header
size saves energy to forward packets, but at the cost of maintaining
a state for compression/decompression, plus computing these
operations. The gain from such protocol optimization further depends
on the application and whether it sends packets with large payloads
close to the MTU (the header tax and any savings here are very
limited), or whether it sends packets with very small payload size
(making the header tax more pronounced and savings more significant).
An alternative to reducing the amount of protocol data is to design
routing protocols that are more efficient to process at each node.
For instance, path based forwarding/labels such as MPLS [RFC3031]
facilitate the next hop look-up, thereby reducing the energy
consumption. It is unclear if some state at router to speed up look
up is more energy efficient that "no state + lookup" that is more
computationally intensive. Other methods to speed up a next-hop
lookup include geographic routing (e.g., [Herzen2011PIE]). Some
network protocols could be designed to reduce the next hop look-up
computation at a router. It is unclear if Longest Prefix Match (LPM)
is efficient from an energy point of view or if constitutes a
significant energy burden for the operation of a router.
Clemm, et al. Expires 2 July 2024 [Page 18]
Internet-Draft Management for Green Networking December 2023
Beyond the volume of data itself, another consideration is the number
of messages and chattiness of the protocol. Some protocols rely on
frequent periodic updates or heartbeats, which may prevent equipment
to go into sleep mode. In such cases, it makes sense to explore the
the use of feasible alternatives that rely on different communication
patterns and fewer messages.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
* Assessments of energy-related tradeoffs regarding protocol design
space and tradeoffs, such as maintaining state versus more compact
encodings or extra computation for transcoding operations versus
larger data volume.
* Protocol advances for improving the ratio of goodput to throughput
and to reduce waste: reduction in header tax, in protocol
verbosity, in need for retransmissions, improvements in coding,
etc.
* Protocols that allow to manage transmission patterns in ways that
facilitate periods of link inactivity, such as burstiness and
chattiness.
5.4. Network Addressing
There may be other ways to shave off energy usage from networks. One
example concerns network addressing. Address tables can get very
large, resulting in large forwarding tables that require considerable
amount of memory, in addition to large amounts of state needing to be
maintained and synchronized. From an energy footprint perspective,
both can be considered wasteful and offer opportunities for
improvement. At the protocol level, rethinking how addresses are
structured can allow for flexible addressing schemes that can be
exploited in network deployments that are less energy-intensive by
design. This can be complemented by supporting clever address
allocation schemes that minimize the number of required forwarding
entries as part of deployments.
Alternatively, the address could be designed so as to allow for more
efficient processing than LPM. For instance, a geographic type of
addressing (where the next hop is computed as a simple distance
calculation based on the respective position of the current node, of
its neighbors and of the destination) [Herzen2011PIE] could be
potentially more energy-efficient.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
Clemm, et al. Expires 2 July 2024 [Page 19]
Internet-Draft Management for Green Networking December 2023
* Devise methods to assess the magnitude of the carbon footprint
that is associated with addressing schemes.
* Devise methods to improve addressing schemes, as well as address
assignment schemes, to minimize their footprint.
6. Challenges and Opportunities - Network Level
6.1. Network Optimization and Energy/Carbon/Pollution-Aware Networking
Networks have been optimized for many years under many criteria, for
example to optimize (maximize) network utilization and to optimize
(minimize) cost. Hence, it is straightforward to add optimization
for "greenness" (including energy efficiency, power consumption,
carbon footprint) as important criteria.
This includes assessing the carbon footprints of paths and optimizing
those paths so that overall footprint is minimized, then applying
techniques such as path-aware networking or segment routing [RFC8402]
to steer traffic along those paths. (As mentioned earlier, other
proxy measures could be used for carbon footprint, such as an energy-
efficiency ratings of traversed equipment.) It also includes aspects
such as considering the incremental carbon footprint in routing
decisions. Optimizing cost has a long tradition in networking; many
of the existing mechanisms can be leveraged for greener networking
simply by introducing carbon footprint as a cost factor. Low-hanging
fruit include the inclusion of carbon-related parameters as a cost
parameter in control planes, whether distributed (e.g., IGP) or
conceptually centralized via SDN controllers. Likewise, there are
opportunities in right-placing functionality in the network. An
example concerns placement of virtualized network functions in
carbon-optimized ways - for example, cohosted on fewer servers in
close proximity to each other in order to avoid unnecessary overhead
in long-distance control traffic.
Other opportunities concern adding carbon-awareness to dynamic path
selection schemes. This is sometimes also referred to as "energy-
aware networking" (respectively "pollution-aware networking"
[Hossain2019] or "carbon-aware networking", when carbon footprint
related parameters beyond pure energy consumption are taken into
account). Again, considerable energy savings can potentially be
realized by taking resources offline (e.g., putting them into power-
saving or hibernation mode) when they are not currently needed under
current network demand and load conditions. Therefore, weaning such
resources from traffic becomes an important consideration for energy-
efficient traffic steering. This contrasts and indeed conflicts with
existing schemes that typically aim to create redundancy and load-
balance traffic across a network to achieve even resource
Clemm, et al. Expires 2 July 2024 [Page 20]
Internet-Draft Management for Green Networking December 2023
utilization. This usually occurs for important reasons, such as
making networks more resilient, optimizing service levels, and
increasing fairness. One of the big challenges hence concerns how
resource weaning schemes to realize energy savings can be
accommodated while preventing the cannibalization of other important
goals, counteracting other established mechanisms, and avoiding
destabilization of the network.
An opportunity may lie in making a distinction between "energy modes"
of different domains. For instance, in a highly trafficked core, the
energy challenge is to transmit the traffic efficiently. The amount
of traffic is relatively fluid (due to multiplexing of multiple
sessions) and the traffic is predictable. In this case, there is no
need to optimize on a per session basis nor even at a short time
scale. In the access networks connecting to that core, though, there
are opportunities for this fast convergence: traffic is much more
bursty, less predictable and the network should be able to be more
reactive. Other domains such as DCs may have also more variable
workloads and different traffic patterns.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
* Devise methods for carbon-aware traffic steering and routing;
treat carbon footprint as a traffic cost metric to optimize.
* Apply ML and AI methods to optimize networks for carbon footprint;
assess applicability of game theoretic approaches.
* Articulate and, as applicable, moderate tradeoffs between carbon
awareness and other operational goals such as robustness and
redundancy.
* Extend control-plane protocols with carbon-related parameters.
* Consider security issues imposed by greater energy awareness, to
minimize the new attack surfaces that would allow an adversary to
turn off resources or to waste energy.
6.2. Assessing Carbon Footprint and Network-Level Instrumentation
As an important prerequisite to capture many of the opportunities
outlined in Section 6.1, good abstractions (and corresponding
instrumentation) that allow to easily assess energy cost and carbon
footprint will be required. These abstractions need to account for
not only for the energy cost associated with packet forwarding across
a given path, but related cost for processing, for memory, for
maintaining of state, to result in a holistic picture.
Clemm, et al. Expires 2 July 2024 [Page 21]
Internet-Draft Management for Green Networking December 2023
Optimization of carbon footprint involves in many cases trade-offs
that involve not only packet forwarding but also aspects such as
keeping state, caching data, or running computations at the edge
instead of elsewhere. (Note: there may be a differential in running
a computation at an edge server vs. at an hyperscale DC. The latter
is often better optimized than the latter.) Likewise, other aspects
of carbon footprint beyond mere energy-intensity should be
considered. For instance, some network segments may be powered by
more sustainable energy sources than others, and some network
equipment may be more environmentally-friendly to build, deploy and
recycle, all of which can be reflected in abstractions to consider.
Assessing carbon footprint at the network level requires
instrumentation that associates that footprint not just with
individual devices (as outline in Section 4.2 but relates it also to
concepts that are meaningful at the network level, i.e., to flows and
to paths. For example, it will be useful to provide visibility into
the carbon intensity of a path: Can the carbon cost of traffic
transmitted over the path be aggregated? Does the path include
outliers, i.e., segments with equipment with a particularly poor
carbon footprint?
Similarly, how can the carbon cost of a flow be assessed? That might
serve many purposes beyond network optimization, from the option to
introduce green billing and charging schemes to the ability to raise
carbon awareness by end users.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
* Devise methods to assess, to estimate, to predict carbon-intensity
of paths.
* Devise methods to account for carbon footprint of flows and
networking services.
6.3. Dimensioning and Peak Shaving
As mentioned in Section 3, the overall energy usage of a network is
in large part determined by how the network is dimensioned,
specifically: which and how many pieces of network equipment are
deployed and turned on. A significant portion of energy is drawn
even when simply in idle state. Minimizing the amount of equipment
that needs to be turned on in the first place presents hence one of
the biggest energy saving opportunities.
Clemm, et al. Expires 2 July 2024 [Page 22]
Internet-Draft Management for Green Networking December 2023
Network deployments are generally dimensioned for periods of peak
traffic, resulting in excess capacity during periods of non-peak
usage that nonetheless consumes power. Shaving peak usage may thus
result in outsized sustainability gains, as it reduces not only
energy usage during peak traffic, but more importantly waste during
non-peak periods.
While traffic volume is largely a function of demand traffic that
network providers have little influence over, some peak shaving cand
nevertheless be accomplished by techniques such as spreading spikes
out over geographies (e.g. redirecting some traffic across more
costly but less utilized routes, particular in cases when traffic
spikes are of a more local or reginal nature) or over time (e.g.
postponing non-urgent traffic, storing or buffering using edge clouds
or extra storage where feasible).
To make techniques effective, accurate learning and prediction of
traffic patterns is required. This includes the ability to perform
forecasting to ensure that additional resources can be spun up in
time should it be needed. Clearly, this presents interesting
challenges, yet also opportunities for technical advances to make a
difference.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
* Support for methods that allow to monitor and forecast traffic
demand, involving new mechanisms and/or performance improvements
of existing mechanisms to support the collection of telemetry and
generation of traffic matrices at very high velocity and scale
* Additional methods that allow for even traffic load distribution
across the network, i.e. load balancing on a network scale, and
enablement of those methods through control protocol extensions as
needed.
6.4. Convergence Schemes
One set of challenges of carbon-aware networking concerns the fact
that many schemes result in much greater dynamicity and continuous
change in the network as resources may be getting steered away from
(when possible) and then leveraged again (when necessary) in rapid
succession. This imposes significant stress on convergence schemes
that results in challenges to the scalability of solutions and their
ability to perform in a fast-enough manner. Network-wide convergence
imposes high cost and incurs significant delay and is hence not
susceptible to such schemes. In order to mitigate this problem,
mechanisms should be investigated that do not require convergence
Clemm, et al. Expires 2 July 2024 [Page 23]
Internet-Draft Management for Green Networking December 2023
beyond the vicinity of the affected network device. Especially in
cases where central network controllers are involved that are
responsible for aspects such as configuration of paths and the
positioning of network functions and that aim for global
optimization, the impact of churn needs to be minimized. This means
that, for example, (re-) discovery and update schemes need to be
simplified and extensive recalculation e.g., of routes and paths
based on the current energy state of the network needs to be avoided.
Challenges and opportunities for IETF-led advances in this space
include:
* Protocols that facilitate rapid convergence (per Section 5.1).
* Investigate methods that mitigate effects of churn, including
methods that maintain memory or state as well as methods relying
on prediction, inference, and interpolation.
6.5. The Role of Topology
One of the most important network management constructs is that of
the network topology. A network topology can usually be represented
as a database or as a mathematical graph, with vertices or nodes,
edges or links, representing networking nodes, links connecting their
interfaces, and all their characteristics. Examples of these network
topology representations include routing protocols link-state
databases, and service function chaining graphs.
As we desire to add carbon and energy awareness into networks, the
energy proportionality of topologies directly supports sustainability
visibility and improvements via automation.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
* Embedding carbon and energy awareness into the representation of
topologies, wheather considering IGP LSDBs (link-state databases)
and their advertisements, BGP-LS (BGP Link-State), or metadata for
the rendering of service function paths in a service chain.
* Use of those carbon-aware attributes to optimize topology as a
whole under end-to-end energy and carbon considerations.
Clemm, et al. Expires 2 July 2024 [Page 24]
Internet-Draft Management for Green Networking December 2023
7. Challenges and Opportunities - Architecture Level
Another possibility to improve network energy efficiency is to
organize networks in a way that they can best serve important
applications so as to minimize energy consumption. Examples include
retrieval of content or remote computation. This allows to minimize
the amount of communication that needs to take place in the first
place, although energy savings within the network may at least in
part be offset by additional energy consumption elsewhere. The
following are some examples that suggest that it may be worthwhile
reconsidering the ways in which networks are architected to minimize
their carbon footprint.
For example, Content Delivery Networks (CDNs) have reduced the energy
expenditure of the Internet by downloading content near the users.
The content is sent only a few times over the WAN, and then is served
locally. This shifts the energy consumption from networking to
storage. Further methods can reduce the energy usage even more
[Bianco2016energy] [Mathew2011energy] [Islam2012evaluating]. Whether
overall energy savings are net positive depends on the actual
deployment, but from the network operator's perspective, at least it
shifts the energy bill away from the network to the CDN operator.
While CDNs operate as an overlay, another architecture has been
proposed to provide the CDN features directly in the network, namely
Information Centric Networks [Ahlgren2012survey], studied as well in
the IRTF ICNRG. This however shifts the energy consumption back to
the network operator and requires some power-hungry hardware, such as
chips for larger name look-ups and memory for the in-network cache.
As a result, it is unclear if there is an actual energy gain from the
dissemination and retrieval of content within in-network caches.
Clemm, et al. Expires 2 July 2024 [Page 25]
Internet-Draft Management for Green Networking December 2023
Fog computing and placing intelligence at the edge are other
architectural directions for reducing the amount of energy that is
spent on packet forwarding and in the network. There again, the
trade-off is between performing computation in a an energy-optimized
data center at very large scale, but requiring transmission of
significant volumes of data across many nodes and long distances,
versus performing computational tasks at the edge where the energy
may not be used as efficiently (less multiplexing of resources, and
smaller sites are inherently less efficient due to their smaller
scale) but the amount of long-distance network traffic is
significantly reduced. Softwarization, containers, microservices are
direct enablers for such architectures, and the deployment of
programmable network infrastructure (as for instance Infrastructure
Processing Units - IPUs or SmartNICs that offload some computations
from the CPU onto the NIC) will help its realization. However, the
power consumption characteristics of CPUs are different from those of
NPUs, another aspect to be considered in conjunction with
virtualization.
Other possibilities concern taking economic aspects into
consideration impact, such as providing incentives to users of
networking services in order to minimize energy consumption and
emission impact. An example for this is given in
[Wolf2014choicenet], which could be expanded to include energy
incentives.
Other approaches consider performing a late binding of data and
functions to be performed on the data [Krol2017NFaaS]. The COIN
Research Group in IRTF focuses on similar issues. Jointly optimizing
for the total energy cost, taking into account networking and
computing (and the different energy cost of computing in an
hyperscale DC vs an edge node) is still an area of open research.
In summary, rethinking of the overall network (and networked
application) architecture can be an opportunity to significantly
reduce the energy cost at the network layer, for example by
performing tasks that involve massive communications closer to the
user. To what extend these shifts result in a net reduction of
carbon footprint is an important question that requires further
analysis on a case-by-case basis.
The following summarizes some challenges and opportunities that can
provide the basis for IETF-led advances in this space:
Clemm, et al. Expires 2 July 2024 [Page 26]
Internet-Draft Management for Green Networking December 2023
* Investigate organization of networking architecture for important
classes of applications (examples: content delivery, right-placing
of computational intelligence, industrial operations and control,
massively distributed machine learning and AI) to optimize green
foot print and holistic approaches to trade off carbon footprint
between forwarding, storage, and computation.
* Models to assess and compare alternatives in providing networked
services, e.g., assess carbon impact relative to alternatives
where as to where to perform compute, what information to cache,
and what communication exchanges to conduct.
8. Conclusions
How to make networks "greener" and reduce their carbon footprint is
an important problem for the networking industry to address, both for
societal and for economic reasons. This document has highlighted a
number of the technical challenges and opportunities in that regard.
Of those, perhaps the key challenge to address right away concerns
the ability to expose at a fine granularity the energy impact of any
networking actions. Providing visibility into this will enable many
approaches to come towards a solution. It will be key to
implementing optimization via control loops that allow to assess the
energy impact of decision taken. It will also help to answer
questions such as: is caching - with the associated storage energy -
better than retransmitting from a different server - with the
associated networking cost? Is compression more energy-efficient
once factoring the computation cost of compression vs transmitting
uncompressed data? Which compression scheme is more energy
efficient? Is energy saving of computing at an efficient hyperscale
DC compensated by the networking cost to reach that DC? Is the
overhead of gathering and transmitting fine-grained energy telemetry
data offset by the total energy gain by ways of better decisions that
this data enables? Is transmitting data to a Low Earth Orbit (LEO)
satellite constellation compensated by the fact that once in the
constellation, the networking is fueled on solar energy? Is the
energy cost of sending rockets to place routers in Low Earth Orbit
amortized over time?
Determining where the sweet spots are and optimizing networks along
those lines will be a key towards making networks "greener". We
expect to see significant advances across these areas and believe
that IETF has an important role to play in facilitating this.
Clemm, et al. Expires 2 July 2024 [Page 27]
Internet-Draft Management for Green Networking December 2023
9. IANA Considerations
This document does not have any IANA requests.
10. Security Considerations
Security considerations may appear to be orthogonal to green
networking considerations. However, there are a number of important
caveats.
Security vulnerabilities of networks may manifest themselves in
compromised energy efficiency. For example, attackers could aim at
increasing energy consumption in order to drive up attack victims'
energy bill. Specific vulnerabilities will depend on the particular
mechanisms. For example, in the case of monitoring energy
consumption data, tampering with such data might result in
compromised energy optimization control loops. Hence any mechanisms
to instrument and monitor the network for such data need to be
properly secured to ensure authenticity.
In some cases, there are inherent tradeoffs between security and
maximal energy efficiency that might otherwise be achieved. An
example is encryption, which requires additional computation for
encryption and decryption activities and security handshakes, in
addition to the need to send more traffic than necessitated by the
entropy of the actual data stream. Likewise, mechanisms that allow
to turn resources on or off could become a target for attackers.
Energy consumption can be used to create covert channels, which is a
security risk for information leakage. For instance, the temperature
of an element can be used to create a Thermal Covert Channel [TCC],
or the reading/sharing of the measured energy consumption can be
abused to create a covert channel (see for instance [DRAM] or
[NewClass]). Power information may be used to create side-channel
attacks. For instance, [SideChannel] provides a review of 20 years
of study on this topic. Any new parameters to consider in protocol
designs or in measurements is susceptible to create such covert or
side channel and this should be taken into account while designing
energy efficient protocols.
11. Contributors
Michael Welzl, University of Oslo, michawe@ifi.uio.no
Clemm, et al. Expires 2 July 2024 [Page 28]
Internet-Draft Management for Green Networking December 2023
12. Acknowledgments
We thank Dave Oran for providing the information regarding covert
channels using energy measurements. Additional acknowledgments will
be added at a later stage.
13. Informative References
[Ahlgren2012survey]
Ahlgren, B., Dannewitz, C., Imbrenda, C., Kutscher, D.,
and B. Ohlman, "A survey of information-centric
networking", IEEE Communications Magazine Vol.50 No.7,
2012.
[Alizadeh2010DCTCP]
Alizadeh, M., Greenberg, A., Maltz, D., Padhye, J., Patel,
P., Prabhakar, B., Sengupta, S., and M. Sridharan, "Data
Center TCP (DCTCP)", ACM SIGCOMM pp.63-74, 2010.
[Bianco2016energy]
Bianco, A., Mashayekhi, R., and M. Meo, "Energy
consumption for data distribution in content delivery
networks", IEEE International Conference on Communications
(ICC) pp.1-6, 2016.
[Bolla2011energy]
Bolla, R., Bruschi, R., Davoli, F., and F. Cucchietti,
"Energy Efficiency in the Future Internet: A Survey of
Existing Approaches and Trends in Energy-Aware Fixed
Network Infrastructures", IEEE Communications Surveys and
Tutorials Vol.13 No.2, pp.223-244, 2011.
[Cervero19]
Cervero, A. G., Chincoli, M., Dittmann, L., Fischer, A.,
and A. Garcia, "Green Wired Networks", Wiley Journal on
Large-Scale Distributed Systems and Energy
Efficiency pp.41-80, 2019.
[Chabarek08]
Chabarek, J., Sommers, J., Barford, P., Tsiang, D., and S.
Wright, "Power awareness in network design and routing",
IEEE Infocom pp.457-465, 2008.
[DRAM] Paiva, T. B., Navaridas, J., and R. Terada, "Robust Covert
Channels Based on DRAM Power Consumption", In book:
Information Security (pp.319-338) , 2019.
Clemm, et al. Expires 2 July 2024 [Page 29]
Internet-Draft Management for Green Networking December 2023
[Emergy] Raghavan, B. and J. Ma, "The Energy and Emergy of the
Internet", ACM HotNets , 2011.
[Framework]
Faber, G., "A framework to estimate emissions from virtual
conferences", International Journal of Environmental
Studies, 78:4, 608-623 , 2021.
[GreenNet22]
Clemm, A. and C. Westphal, "Challenges and Opportunities
in Green Networking", 1st International Workshop on
Network Energy Efficiency in the Softwarization Era IEEE
NetSoft 2022, June 2022.
[Herzen2011PIE]
Herzen, J., Westphal, C., and P. Thiran, "Scalable routing
easy as PIE: A practical isometric embedding protocol",
19th IEEE International Conference on Network Protocols
(ICNP) pp.49-58, 2011.
[Hossain2019]
Hossain, M., Georges, J., Rondeau, E., and T. Divoux,
"Energy, Carbon and Renewable Energy: Candidate Metrics
for Green-Aware Routing?", DOI 10.3390/s19132901,
Sensors Vol. 19 No. 3, June 2019,
<https://ieeexplore.ieee.org/document/6779082>.
[I.D.draft-cx-green-metrics]
Clemm, A., Dong, L., Mirsky, G., Ciavaglia, L., Tantsura,
J., Odini, M., Schooler, E., and A. Rezaki, "Green
Networking Metrics", June 2023.
[IETF-Net0]
Daley, J., "Towards a net zero IETF", IETF News , 6 May
2022,
<https://www.ietf.org/blog/towards-a-net-zero-ietf/>.
[Islam2012evaluating]
Islam, S. U. and J. Pierson, "Evaluating Energy
Consumption in CDN Servers", Proceedings of the Second
International Conference on ICT as Key Technology against
Global Warming pp.64-78, 2012.
[Junkyard] Switzer, J., Kastner, R., and P. Pannuto, "Architecture of
a Junkyard Datacenter", arXiv:2110.06870v1, October 2021 ,
2021.
Clemm, et al. Expires 2 July 2024 [Page 30]
Internet-Draft Management for Green Networking December 2023
[Krol2017NFaaS]
Krol, M. and I. Psaras, "NFaaS: Named Function as a
Service", ACM SIGCOMM ICN Conference , 2017.
[Mathew2011energy]
Mathew, V., Sitaraman, R., and P. Shenoy, "Energy-Aware
Load Balancing in Content Delivery Networks", CoRR
http://arxiv.org/abs/1109.5641 , 2011.
[NewClass] Khatamifard, S. K., Wang, L., Kose, S., and U. R.
Karpuzcu, "A New Class of Covert Channels Exploiting Power
Management Vulnerabilities", IEEE Computer Architecture
Letters , 2018.
[QUAL] Li, R., Makhijani, K., Yousefi, H., Westphal, C., Xong,
L., Wauters, T., and F. D. Turck, "A Framework for
Qualitative Communications using Big Packet Protocol",
Proceedings ACM Sigcomm Workshop On Networking For
Emerging Applications And Technologies pp.22-28, 2019.
[Ren2018jordan]
Ren, J., Ren, K., Westphal, C., Wang, J., Wang, J., Song,
T., Liu, S., and J. Wang, "JORDAN: A Novel Traffic
Engineering Algorithm for Dynamic Adaptive Streaming over
HTTP", IEEE International Conference on Computing,
Networking and Communications (ICNC) pp.581-587, 2018.
[RFC2481] Ramakrishnan, K. and S. Floyd, "A Proposal to add Explicit
Congestion Notification (ECN) to IP", RFC 2481,
DOI 10.17487/RFC2481, January 1999,
<https://www.rfc-editor.org/info/rfc2481>.
[RFC3031] Rosen, E., Viswanathan, A., and R. Callon, "Multiprotocol
Label Switching Architecture", RFC 3031,
DOI 10.17487/RFC3031, January 2001,
<https://www.rfc-editor.org/info/rfc3031>.
[RFC3095] Bormann, C., Burmeister, C., Degermark, M., Fukushima, H.,
Hannu, H., Jonsson, L., Hakenberg, R., Koren, T., Le, K.,
Liu, Z., Martensson, A., Miyazaki, A., Svanbro, K.,
Wiebke, T., Yoshimura, T., and H. Zheng, "RObust Header
Compression (ROHC): Framework and four profiles: RTP, UDP,
ESP, and uncompressed", RFC 3095, DOI 10.17487/RFC3095,
July 2001, <https://www.rfc-editor.org/info/rfc3095>.
[RFC7950] Bjorklund, M., Ed., "The YANG 1.1 Data Modeling Language",
RFC 7950, DOI 10.17487/RFC7950, August 2016,
<https://www.rfc-editor.org/info/rfc7950>.
Clemm, et al. Expires 2 July 2024 [Page 31]
Internet-Draft Management for Green Networking December 2023
[RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L.,
Decraene, B., Litkowski, S., and R. Shakir, "Segment
Routing Architecture", RFC 8402, DOI 10.17487/RFC8402,
July 2018, <https://www.rfc-editor.org/info/rfc8402>.
[RFC9330] Briscoe, B., Ed., De Schepper, K., Bagnulo, M., and G.
White, "Low Latency, Low Loss, and Scalable Throughput
(L4S) Internet Service: Architecture", RFC 9330,
DOI 10.17487/RFC9330, January 2023,
<https://www.rfc-editor.org/info/rfc9330>.
[SideChannel]
Randolph, M. and W. Diehl, "Power Side-Channel Attack
Analysis: A Review of 20 Years of Study for the Layman",
Cryptography 2020, 4, 15 , 2020.
[TCC] Rahimi, P., Singh, A. K., and X. Wang, "Selective Noise
Based Power Efficient and Effective Countermeasure Against
Thermal Covert Channel Attacks in Multi-Core Systems",
Journal on Low Power Electronics and Applications , 2022.
[Telefonica2021]
Telefonica, "Telefonica Consolidated Annual Report 2021.",
2021.
[TradeOff] Welzl, M., "Not a Trade-Off: On the Wi-Fi Energy
Efficiency of Effective Internet Congestion Control",
IEEE/IFIP WONS , 2022.
[Westphal2021qualitative]
Westphal, C., He, D., Makhijani, K., and R. Li,
"Qualitative Communications for Augmented Reality and
Virtual Reality", 22nd IEEE International Conference on
High Performance Switching and Routing (HPSR) pp.1-6,
2021.
[Wolf2014choicenet]
Tilman, W., Griffioen, J., Calvert, L., Dutta, R.,
Rouskas, G., Baldin, I., and A. Nagurney, "ChoiceNet:
Toward an Economy Plane for the Internet", SIGCOMM
Computer Communciations Review Vol.44 No.3, July 2014.
Authors' Addresses
Clemm, et al. Expires 2 July 2024 [Page 32]
Internet-Draft Management for Green Networking December 2023
Alexander Clemm (editor)
Futurewei
2330 Central Expressway
Santa Clara, CA 95050
United States of America
Email: ludwig@clemm.org
Cedric Westphal
Futurewei
Email: cedric.westphal@futurewei.com
Jeff Tantsura
Nvidia
Email: jefftant.ietf@gmail.com
Laurent Ciavaglia
Nokia
Email: laurent.ciavaglia@nokia.com
Carlos Pignataro (editor)
North Carolina State University
United States of America
Email: cpignata@gmail.com, cmpignat@ncsu.edu
Marie-Paule Odini
Email: mp.odini@orange.fr
Clemm, et al. Expires 2 July 2024 [Page 33]
