Regis University

ePublications at Regis University

Regis University Student Publications Regis University Student Publications

(comprehensive collection)

Fall 2010

Deep Packet Inspection and its Effects On Net Neutrality

Michael A. DeRose
Regis University

Follow this and additional works at: https://epublications.regis.edu/theses

Part of the Computer Sciences Commons

Recommended Citation
DeRose, Michael A., "Deep Packet Inspection and its Effects On Net Neutrality" (2010). Regis University
Student Publications (comprehensive collection). 355.
https://epublications.regis.edu/theses/355

This Thesis - Open Access is brought to you for free and open access by the Regis University Student Publications
at ePublications at Regis University. It has been accepted for inclusion in Regis University Student Publications
(comprehensive collection) by an authorized administrator of ePublications at Regis University. For more
information, please contact epublications@regis.edu.
 Regis University
College for Professional Studies Graduate Programs
Final Project/Thesis

Disclaimer
Use of the materials available in the Regis University Thesis Collection
(“Collection”) is limited and restricted to those users who agree to comply with
the following terms of use. Regis University reserves the right to deny access to
the Collection to any person who violates these terms of use or who seeks to or
does alter, avoid or supersede the functional conditions, restrictions and
limitations of the Collection.

The site may be used only for lawful purposes. The user is solely responsible for
knowing and adhering to any and all applicable laws, rules, and regulations
relating or pertaining to use of the Collection.

All content in this Collection is owned by and subject to the exclusive control of
Regis University and the authors of the materials. It is available only for research
purposes and may not be used in violation of copyright laws or for unlawful
purposes. The materials may not be downloaded in whole or in part without
permission of the copyright holder or as otherwise authorized in the “fair use”
standards of the U.S. copyright laws and regulations.
 DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY

A THESIS

SUBMITTED ON TWENTY SECOND OF OCTOBER, 2010

TO THE DEPARTMENT OF INFORMATION TECHNOLOGY

OF THE SCHOOL OF COMPUTER & INFORMATION SCIENCES

OF REGIS UNIVERSITY

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS OF MASTER OF SCIENCE IN

INFORMATION TECHNOLOGY MANAGEMENT

BY

Michael A. DeRose

APPROVALS

Ernest Eugster, Ph.D., Thesis Advisor

Shari Plantz-Masters

Stephen D. Barnes
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY i

Abstract

Deep packet inspection (DPI) is becoming increasingly important as a means to classify and

control Internet traffic based on the content, applications, and users. Rather than just using

packet header information, Internet Service Providers are using DPI for traffic management,

routing and security. But by being able to control traffic by content, a growing number of public

policy makers and users fear ISPs may discriminately charge more for faster delivery of their

data, slow down applications or even deny access. They cite such practices as endangering the

principle of net neutrality; the premise that all data on the Internet should be treated equally.

The existing literature on DPI and net neutrality is sizeable, but little exists on the relationship

between DPI and net neutrality. This study examines the literature, develops a research

methodology and presents results from a study on the challenges of DPI in regards to privacy and

net neutrality. The findings show that although most users are unaware of DPI technology, they

feel strongly that it places their privacy at risk.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY ii

Acknowledgements

I would like to thank my fellow students, the faculty of Regis University and all those

that have made this culmination of my education possible. I have been privileged to have had

some great cohorts on this journey and I am thankful for the contributions, discussions,

arguments and revelations that would not have been possible without your support.

Additionally, I would like to thank my Thesis Advisor, Dr. Ernest Eugster whose tireless efforts

kept me on target and track.

I would also like to thank my friends, family and coworkers that stepped up and

supported me by participating in the longest survey ever composed. In particular, I would like to

thank Timmy Harper for his tireless efforts in taking the survey. My sincere thanks also go out

to my children, Sofia and Conner for putting up with Daddy being grumpy, busy and generally

unavailable afterhours for the past six years. Well, Daddy is done and I promise that it was all

worth it! Most importantly, I would like to thank my wife, Jodee, for her patience, her wisdom,

her guidance and her never ending support on my journey to complete my education. Without

her help and innumerable sacrifices, this would not have been remotely possible.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY iii

Table of Contents

Abstract ...............................................................................................................................i

Acknowledgements ............................................................................................................ii

Table of Contents ..............................................................................................................iii

List of Tables ....................................................................................................................v

Chapter 1 - Introduction.....................................................................................................1

1.1 Deep Packet Inspection, Technology and Net Neutrality....................................1

1.2 Statement of Problem..........................................................................................4

1.3 Statement of Goals and Objectives ....................................................................4

1.4 Project Description.............................................................................................5

1.5 Summary............................................................................................................6

Chapter 2 - Literature Review............................................................................................7

2.1 Deep Packet Inspection......................................................................................7

2.2 Traffic Shaping..................................................................................................9

2.3 Deep Packet Inspection Devices.....................................................................10

2.4 Benefits of Deep Packet Inspection................................................................11

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY iv

2.5 Net Neutrality..................................................................................................12

2.6 Summary.........................................................................................................14

Chapter 3 - Methodology....................................................................................................15

3.1 Ontology…………………….........................................................................15

3.2 Questions and Survey.....................................................................................16

3.3 Summary.....…………………………............................................................17

Chapter 4 – Data Analysis and Findings...........................................................................18

4.1 Background Information..................................................................................18

4.2 Privacy and Content.........................................................................................22

4.3 Regulation........................................................................................................26

4.4 Trust and Moral Obligations ...........................................................................28

4.5 Privacy of Communications.............................................................................30

4.6 Summary..........................................................................................................31

Chapter 5 – Discussion and Conclusions............................................................................32

References .........................................................................................................................34

Appendix A – Survey Questions .......................................................................................36

Appendix B – Survey Results ...........................................................................................44

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY v

List of Tables

Table 1. Which Best Describes Your Background?..……….............................................. 19

Table 2. Background Knowledge Questions (All Participants)...........................................19

Table 3. Background Knowledge Questions (IT Field Only).............................................20

Table 4. Background Knowledge Questions (Non-IT Field Only).....................................22

Table 5. Examination of Network and Telephone Communications...................................23

Table 6. ISPs Should Be Allowed to Deny or Delay Communications...............................24

Table 7. Telephone Deny or Delay Communications........................................................24

Table 8. Premium Content, Premium Fees? ....................................................................25

Table 9. To Regulate or Not to Regulate? ......................................................................26

Table 10. Government Regulation? ..............................................................................27

Table 11. Standards Agency Regulation? ......................................................................28

Table 12. Moral Obligations? ......................................................................................29

Table 13. Trusted Providers. .......................................................................................29

Table 14. Privacy of Communications...........................................................................30

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 1

Chapter 1 – Introduction

For decades, ISPs have been managing and routing traffic based on packet header

information. DPI, however, has emerged as a critical traffic management tool and is the subject

of this research. By being able to control traffic based on content, in addition to header

information, DPI offers ISPs improved traffic management and routing as well as security. But

having ISPs controlling traffic content has lead to increasing public debate. The opposing sides

of the net neutrality debate argue that without safeguards in place, ISPs would cut lucrative deals

to prioritize some kinds of content and throttle others, turning themselves into the unofficial

gatekeepers of the world's best leveling force. ISPs argue that practices such as tiered-pricing

are needed to ensure continued investments in Internet infrastructure.

This chapter introduces DPI and the concept of net neutrality. It identifies how the

technology works and the challenges it can present to net neutrality. The chapter also describes

the goals of the research and the research methodology used.

1.1 Deep Packet Inspection Technology and Net Neutrality

Deep packet inspection takes the process of inspecting the origin and destination of

packets and expands it to examine the actual data being sent. This technology allows Internet

Service Providers (ISPs) and network administrators the ability to filter data transmissions by

denying, delaying or giving precedence to certain types of data.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 2

Historically, packet inspection occurred at the network layer which was dependant upon

the header and footer information contained in packets to determine routing and filtering options.

Deep packet inspection came about as a means to enhance the process to include the data that

was once only accessible in the application and presentation layers of the OSI model. Access to

the presentation and application layer information at the network layer, allows for packet

filtering to be implemented based on the actual data, not header information. The ability to filter

data at the lower layers allows ISPs to circumvent any header information and route packets

based on what the payload contains.

As the type, contents, the destination and any digital signatures of the data can be

identified, deep packet inspection can provide many benefits for ISP environments. Kumar

(2006) identified three popular applications of DPI:

• Content-based traffic management and routing, where packets are classified and

processed based upon content.

• Network intrusion detection systems (NIDS) generally scan the packet header and

payload to identify a given set of signatures of well known security threats.

• Layer 7 switches and firewalls provide content-based filtering, load-balancing,

authentication and monitoring.

By applying content-based traffic management, ISPs can give priority to traffic based on the

type of content being sent, guaranteeing enhanced delivery for premium content providers.

Network intrusion detection systems (NIDS) utilize DPI to examine the content of packets and

compare the digital signatures to a database of known threats, discarding transmissions that pose

security risks. DPI also enables layer 7 switches to become application aware. This awareness
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 3

gives application layer switches the ability to filter traffic based on the type of application. This

provides ISPs with a scalable and efficient traffic management option.

Such applications and benefits can provide ISPs an important competitive advantage.

Stallings (2007) observed that many opportunities exist for companies to use differentiation as a

strategy to create competitive advantages. Today’s networks can offer differing levels of quality

of service (QoS), which include specifications for maximum delay and minimum throughput.

They also provide a variety of customizable services in the network management and security

arenas. Deep packet inspection can provide the differentiation that will help ISPs offer a wide

array of services with guaranteed QoS levels. These services and guarantees are possible

because DPI allows network traffic to be manipulated with intimate knowledge of the data being

sent.

Net neutrality enters the picture as the movement to keep the flow of information free.

Free access to content, the ability to connect with any device or run any application is what has

made the Internet so popular and useful. This openness is the heart of innovation that has driven

the rapid growth of the Internet and to deny any content, applications or devices would be

stifling. Net neutrality is the process of keeping the Internet open and freely available to all

users, devices and content.

Net neutrality began as an information revolution to help guide Internet policies as they

were being formed. Since 1930, the United States communication networks have been governed

by non-discrimination policies like net neutrality. Older regulations focusing on telephone

communications still have their place, but updated regulation aimed at Internet communications

should be addressed (Schahczenski, 2008).

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 4

Congressional debate on net neutrality heated up in 2005 and the debate still continues.

Initially, Congress had focused on video franchising, attempting to revamp the laws to allow

phone companies to compete in the video space. Once the legislation progressed, net neutrality

as a whole was addressed and some minimal progress was achieved. Although a recent bill was

withdrawn, Congress continues the debate.

By giving ISPs the ability to discriminate packets based on content or premium access,

only ideas or services backed by a significant funding or offered by ISPs would flourish.

Although the technology allows for safer, more scalable and more controllable networks, DPI

raises serious concerns about privacy and net neutrality.

1.2 Statement of Problem

The problem under investigation in this study focuses on the threats that deep packet

inspection can pose to net neutrality.

DPI gives ISPs the ability to manipulate and inspect every bit of data sent over their

networks and this does not sit well with pundits for net neutrality. The ability to throttle

services, divvy out bandwidth and even reject contents of information traversing the network

goes against the philosophy of a Free Internet.

1.3 Statement of Goals and Objectives

Using data from a survey of Internet users, this study has three goals. The first objective

of the study is to determine the risks that DPI poses to net neutrality. To reach this objective, a

survey was conducted of Internet users. The results of the study revealed that most users are not

aware their privacy may even be compromised and that IT industry insiders appear largely in the

dark with regards to network monitoring capabilities. ISPs have always been able to challenge
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 5

the free flow of information, but deep packet inspection allows them to do so with frighteningly

intrusive means.

A second and more general objective is to add to the body of knowledge. Considerable

literature on DPI exists, but it is largely focused on technical issues and case studies. The

existing literature on net neutrality is also sizable. But there exists little technical description on

how ISPs use DPI for monitoring and control. The study has emerged from this researcher’s

belief that the coupling of DPI technology and its net neutrality context has not been well

explored.

Third, this researcher hopes that this study will result in informed public debate. By

discussing the technology behind deep packet inspection and mapping that to the affects it can

have on privacy and a free Internet, this study attempts to educate the general user on challenges

of DPI to network neutrality.

1.4 Project Description

To reach these goals, the research proceeded in three phases. In the first phase, this

researcher reviewed the existing literature, both in DPI and in net neutrality. The literature is

vast, yet weighs heavily to technical descriptions and public policy debate. This paper attempts

to redress this imbalance by examining and integrating DPI’s role in endangering net neutrality.

The literature review also provided a better definition of guidelines for exploration, especially in

phases of question generation and survey phases.

In the second phase, specific questions were generated. Should ISPs be allowed to

examine the content of network communications? Should they be allowed to deny or delay

communications based on the type of content being sent (e.g., music, text, video)? Should they

be allowed to deny or delay communications if the content is business versus personal? To

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 6

answer these questions, a survey represented the most appropriate research method to verify the

knowledge and attitudes of the general Internet population. These questions were grouped to

determine specific types of information. The first group was participant background

information, followed by Internet communication questions, phone communication questions

and finally privacy questions.

In the third phase, the survey targeted general Internet users. The sole criterion was that

the participant had to have a working Facebook profile. This tactic was selected for a few

reasons. First, through the popularity of Facebook, the rate of response was hoped to be

adequate for the study (67, or 33% out of 200 responded to the invitation posted to Facebook).

Second, by only querying active Facebook users, the respondents were guaranteed to be Internet

users. Lastly, Facebook provided a medium other than email to communicate the invitations to

participate in the study. It is this researcher’s belief that had email been employed, the rate of

response would have been drastically reduced.

The researcher felt this approach would be more rewarding than interviewing ISPs. One

of the fundamental questions in the net neutrality debate is how much ISPs should be allowed to

discriminate packets traversing their networks. As the debate rages on, the argument gets

sidetracked with semantics. ISPs argue that they have the right to manage traffic on their

networks, users and content providers argue that while it is acceptable for ISPs to manage traffic,

it is not within their rights to block or delay any traffic. By focusing the study on users, the

researcher hoped to avoid the black hole of the semantic argument.

1.5 Summary

Deep packet inspection provides many benefits to ISPs. But it can also adversely affect

privacy. After laying the foundation for the technology and its benefits, the problem of net
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 7

neutrality and a free Internet were discussed. Also, due to the lack of substantial literature on the

specific impacts of DPI on privacy and net neutrality, it was necessary to generate a study

surrounding the issue from the end user’s point of view. It is these competing factors that

combine to formulate the need for further research and education on the topic. Finally, the

chapter concluded by showing research method used.

The next chapter presents a review of the literature on DPI and net neutrality.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 8

Chapter 2 –Literature Review

This chapter delves deeper into literature on deep packet inspection and net neutrality.

There is considerable literature on DPI, but it is limited to engineering issues and case studies.

The literature on net neutrality is also vast, but little exists on how DPI poses a threat to privacy.

This research attempts to fill this gap.

2.1 Deep Packet Inspection

A number of researchers have examined DPI technology. Hills (2006), for example,

stated that DPI devices can operate on layers two through seven of the Open Systems

Interconnect (OSI) model.

Deep packet inspection is packet filtering that inspects the data payload of an IP packet.

DPI devices take deep looks into the data of each packet and either allows or denies passage

according to some set of predetermined rules. Smith (2008) noted that as networks incorporate

increasingly sophisticated services into their infrastructure, DPI uses application-specific data

found in packet payloads to make routing decisions, to block or rate-limit unwanted traffic, to

perform intrusion detection, and to provide quality of service.

The DPI inspection engine parses each packet and compares the contents against its rule

set. This rule set is comprised of known electronic signatures of content which allow
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 9

identification of the packet’s data. In the past, network packets were classified by their headers,

but DPI now allows them to be classified by the actual content of their payloads.

To perform this functionality, DPI devices rely on a database of application signatures

that are crosschecked against to determine the nature of the packet traversing the network. The

DPI device groups the packets by protocol and security levels then processes the packets by

“performing application level checks as well as stateful inspection” (Ranum, 2005). DPI devices

look for any anomalies in the packets based on their know application signatures. If any packet

is deemed out of the ordinary, it is not allowed to pass. An example would be the order in which

commands are given for a certain protocol. An application will always order commands in the

same sequence, where as a human attempting to hack into a system might issue the commands in

some random order. By inspecting the packet for its application signature, then comparing that

signature against known parameters, DPI devices can thwart attacks that would have gotten past

traditional packet inspection principles.

Normally, DPI is used to monitor and shape IP traffic. ISPs can use DPI to monitor the

type of traffic on their networks and give priority to the protocols they deem more important.

This type of traffic shaping can slow down less important protocols, while not entirely cutting of

access to the particular service.

Deep packet inspection does come with a heavy cost on the processing and bandwidth

sides of the equation. To perform such a thorough look at each individual packet traversing a

network, while keeping the throughput speed at normal levels, is quite a challenge. Becchi

(2007) and Kumar (2006) showed that advanced algorithms are needed to enhance DPI’s ability

to meet the challenge. They argued the processing bottleneck is a result of the speed in which

comparisons between known electronic signatures and quickly moving data occurs. As the data
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 10

is moving at ever increasing speeds, it is necessary for the processing to increase at the same rate

as the bandwidth. Not until recently have DPI device vendors been able to come close to real-

time DPI for mass market consumption. Even then, the financial costs can be prohibitive.

Anderson (2007) noted that “...top-of-the-line products can set you back several hundred

thousand dollars, but some of them can inspect and shape every single packet—in real time—

for nearly a million simultaneous connections while handling 10-gigabit Ethernet speeds and

above.” The processing power needed to make DPI successful at real time speeds has been the

major roadblock to widespread adaptation of the technology.

Along with throughput concerns comes the fact that DPI devices depend upon software to

match, categorize, interpret and finally decide which packets are allowed to pass. As with any

software, there are bound to be some cases of vulnerability. According to Porter (2005), Remote

Procedure Call (RPC) attacks, stack overflow attacks, buffer overflow attacks, VoIP command

processing vulnerabilities and H.225 messages over TCP are all cases of known DPI

vulnerabilities. Even though DPI provides a robust manner in which to monitor network

communications, it is an evolving technology.

2.2 Traffic Shaping

Traffic shaping is defined as, “the appropriate allocation of bandwidth to support

application requirements” (Goldman, 2004). This boils down to ISPs ranking network traffic

and throttling bandwidth in accordance to the ranking. The rankings can be based on several

different factors, such as service level agreements, types of traffic, application requirements and

performance considerations.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 11

Shaping network traffic is accomplished by either controlling the pace of the data or by

queuing the data for a timed release. These two methodologies can also be combined to further

control or shape the traffic as it travels over a network. Georgiadis (1996) stated that “Reshaping

makes the traffic at each node more predictable and, therefore, simplifies the task of

guaranteeing performance to individual connections; when used with a particular scheduling

policy, it allows the specification of worst case delay bounds at each node. End-to-end delay

bounds can then be computed as the sum of the worst case delay bounds at each node along the

path.” This is analogous to an interstate highway at rush hour. Traffic flows in the main lanes,

while stop lights queue on ramp traffic to enter the interstate in a staggered fashion. All the

while, the carpool lane is comfortably cruising at above average speeds.

2.3 Deep Packet Inspection Devices

DPI devices are a combination of previous filtering technology with deep packet

inspection functionality. Stateful inspection, packet sniffing and firewall technology work in

concert with the DPI database to perform network monitoring and traffic shaping on a level that

has not previously been attainable. As speeds and functionality improve, deep packet inspection

will become embedded within the network core.

Several flavors of deep packet inspection devices exist, but devices are mainly found in

the form of hardware firewalls. The main consideration with DPI devices is the throughput in

which they can operate. Deep inspection of packets is expensive in terms of processing and

memory overhead. Porter (2005) stated that “searching through the payload for multiple string

patterns within the data stream is a computationally expensive task. The requirement that these

searches be performed at wire speed adds to the cost.”

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 12

2.4 Benefits of Deep Packet Inspection

Deep packet inspection can provide many benefits for corporate and ISP environments.

The addition of a DPI to a network monitoring portfolio can help to bolster the services and

security provided by their respective networks. Increased service levels can be attained by

utilizing content-based traffic management while increased understanding and control of their

networks will help to cut operating and capital expenditures. Application aware switches can

provide increased load balancing, authentication and monitoring capabilities. In addition to

increased service levels and more control, DPI can help ISPs secure their networks by

implementing network intrusion detection systems based on electronic signatures of well known

threats.

Previous network monitoring devices only gave the network administrators an overview

of bandwidth, services and the destination of network traffic. By adding DPI functionality,

network monitoring takes a step up to the next level. Companies will be able to stop many

network attacks in their tracks. Combining existing intrusion prevention technology with the

additional filtering of DPI devices will allow for networks to identify and prevent many attacks

that currently are able to bypass today’s prevention measures. ISPs will be able to shape their

bandwidth to better serve their customers most critical needs. DPI will allow them to throttle

services at peak usage times to better accommodate the needs of their customers. While a DPI

device is in use, ISPs and corporations will gain a deeper understanding of what their networks

are actually being used for. This greater level of knowledge will allow them to focus on specific

areas of need and not waste time or expense on areas that are performing up to par. As a result

of a more controlled network, companies can focus on additional services and offerings that will

help to generate additional opportunities for revenue generation. Hill (2006) wrote that “Once
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 13

ISPs have networks under better control, it is time to look into how to extract additional revenue

streams, and DPI can be leveraged to create additional tiers of service.”

2.5 Net Neutrality

The advent of inspecting the data payloads of packets traversing the network is a

powerful tool which can help better manage traffic, increase security, and shape usage. The

ability to inspect the data portion of IP packets, however, poses challenges to net neutrality.

What is net neutrality? Jordan (2009) defined net neutrality as, “the idea that Internet

users are entitled to service that does not discriminate on the basis of source, destination, or

ownership of Internet traffic.” The implications of net neutrality are vast and it is getting

attention. For example:

• Google and Verizon released their joint policy proposal aimed preserving the openness of

the Web, but exempting net neutrality regulation to the mobile industry (Krigman, 2010).

• The FCC now requires ISPs to treat lawful content, applications, and services in a

nondiscriminatory manner (Feldman, 2010).

• Congress was debating the Waxman Net Neutrality Bill, requiring ISPs to follow the

basic principles of the "open internet" advocated by net neutrality supporters, including

bans on the blocking of unreasonable interference with lawful content, applications,

services and devices. The bill has since been dropped, but it is evidence that net

neutrality continues to be a hot topic (Kennedy, 2010).

One of the main drivers for net neutrality was to prevent ISPs and backbone network

providers to charge more for specific types of network traffic (Economides, 2008). This

differentiation of network communications is greatly enhanced by devices that can peer deep into

the payload of network communications and report the type and content of the data being
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 14

transmitted. Jordan (2009) wrote that advocates of net neutrality want to prohibit discrimination

of Internet communication by ISPs based on dedicated bandwidth or improved QoS. He argues

that these discriminations could result in outright blocking of sites or content types which could

result in a decrease in innovation and development. Jordan contends that dedicated bandwidth

and improved QoS as a result of priority access offerings should not be allowed due to the

restrictive costs imposed by ISPs for these premium services.

Those against net neutrality argue that free market forces and competition will be

sufficient regulation of ISPs, as any official regulation will hinder ISPs ability to fund

infrastructure improvements (Jordan 2009). It is an age old argument, free market versus

regulation, but in this case DPI devices provide the ability that oversteps the boundaries of

network performance and encroaches on basic rights of free speech. As such, net neutrality is

one of the main arguments against increased use of DPI.

When asked about how concerned he was about Internet companies using his bandwidth,

AT&T CEO Ed Whitacre stated, “How do you think they’re going to get to customers? Through

a broadband pipe. Cable companies have them. We have them. Now what they would like to do

is use my pipes free, but I ain’t going to let them do that because we have spent this capital and

we have to have a return on it. So there’s going to have to be some mechanism for these people

who use these pipes to pay for the portion they’re using. Why should they be allowed to use my

pipes? The Internet can’t be free in that sense, because we and the cable companies have made

an investment and for a Google or Yahoo! or Vonage or anybody to expect to use these pipes

[for] free is nuts!” (Economedies 2008).

The ability of ISPs to track, monitor and thoroughly log the data being sent over their

networks, raises important ethical questions. Is it even ethical to parse through data being sent
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 15

via network communications? Surely a private corporation is well within its rights to monitor

data transmitted by its employees, but is it right for an ISP to log the data sent by its customers?

Proponents of net neutrality also raise issues with ISPs throttling back services or charging extra

for those that utilize higher bandwidth or services that are deemed less than critical. Deep packet

inspection and subsequent logging of the mined data are tantamount to bugging your phone line,

and in fact could be exactly that in the case of voice over IP (VoIP) communications (Renals,

2009). Throw in government monitoring of private citizens and DPI gives ISPs the technology

to comply with government surveillance initiatives. “Although the technology isn't yet common

knowledge among consumers, DPI already gives network neutrality backers nightmares and

enables American ISPs to comply with Commission on Accreditation for Law Enforcement

Agencies (CALEA) (government-ordered Internet wiretaps) reporting requirements” (Anderson,

2007).

2.6 Summary

This chapter focused on the technology of deep packet inspection, the devices that

employ the technology and the different ways it is used. It also examined the principles behind

net neutrality and how DPI is at odds with a free Internet. Deep packet inspection can come in

many forms, and provide great functionality, but it also gives ISPs the means to threaten net

neutrality.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 16

Chapter 3 – Methodology

This research proceeded in three phases. The first step was to gather the academic

literature on net neutrality and deep packet inspection to identify any gaps in knowledge. The

results of this phase formed the basis of Chapter 2. The second phase developed questions based

on the identified gaps in the knowledgebase. In phase three, the survey was conducted to

quantify the assertions made in the first two phases.

3.1 Ontology

While most of the technological facts about the benefits of deep packet inspection are

empirical, or fact based, the main focus of the argument for restraining the technology came

from personal perspectives with citations from non empirical sources. This maintained an

overarching approach to the study which encompassed the technological facts along with

affirming and deferring opinions, which were generated through a survey of general Internet

users.

By employing qualitative ontology to the thesis, the goal was to discover how deep

packet inspection technology impacts privacy and net neutrality. This qualitative approach

allowed for the research to be measured against general principles of right and wrong and how

those principles change based on the environment. Once the study was complete, it was apparent

the privacy concerns due to monitoring of telephone service, were more clearly understood and
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 17

held in higher regard than the privacy concerns of monitoring Internet communications. This

further proved that the perceived gap in the body of knowledge existed and needed to be

accounted for.

3.2 Questions and Survey

As noted, Chapter 2 contains the results of the first phase of the research. Gaps in the

literature were discovered and areas of improvement were apparent. The literature review also

provided a foundation for the remaining two phases: question development and survey.

In phase two, questions were developed. The survey consisted of 33 questions divided

into four sections: a) background information, such as field of work (IT vs. non-IT), knowledge

of packet inspection and traffic shaping, b) details on Internet Service Providers with regards to

how they treat network communications, regulation and pricing, c) descriptions of Telephone

Service Providers such as phone tapping, conversation shaping and regulation, and d) privacy of

communications. For a list of the survey questions, see Appendix A.

The participants were asked to rank each survey statement on a scale of 1 – 5, with an

answer of 1 meaning strongly disagree with the statement and an answer of 5 meaning strongly

agree with the statement. These rankings were combined to tally a final score for each survey

question. For example, a score of 4.45 would mean that the majority of respondents strongly

agree with the statement, while a score of .25 would indicate that most respondents strongly

disagree with the statement. An average score of 3 would show that the average respondent fell

somewhere in the middle (an average of 3 could also mean half strongly agreed and half strongly

disagreed, but that scenario was not present in the result set).
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 18

In phase three, the researcher posted the survey to Facebook. The survey was available

for two weeks, August 28 to September 11 2010. Once the allotted timeframe expired, the

survey was closed and the results were tallied.

Of the 200 invitations extended, 68 (33%) responses were received. As expected, each

respondent was a technology user with at least a passing knowledge of Internet usage (only those

with Facebook accounts were surveyed). The general goal was to extract a theme or pattern to

display the knowledge level and attitudes of Internet users with regards to the topics covered in

this study.

3.3 Summary

Any endeavor worth attempting requires a plan. Through research of the academic

literature involving deep packet inspection, net neutrality and their relationships, a gap in the

knowledge base was ascertained. This study embraced the missing components and presented

new research, accompanied by empirical data, to fill the void. The plan became a framework

which allowed the research to be quantified. Once the research was in place, deliverables were

created to support the study. These deliverables included the academic research and the survey

of Internet users. The combination of deliverables formed the basis of what is hoped to be a

solid contribution to the field by directly addressing the identified gaps in the body of

knowledge. By specifically targeting an audience that stood to gain the most from the study, the

methodology was doubly effective as it was able to educate while it collected data. This research

framework allowed the study to uncover real concerns about privacy while detailing the

relationship between deep packet inspection and net neutrality.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 19

Chapter 4 – Data Analysis and Findings

The survey suggested several findings that seem to support the investigation into how

DPI threatens net neutrality. The results of the study revealed that most users are not aware of

the dangers posed by DPI and that many IT insiders were only slightly more informed. The

second finding showed that even though considerable literature on DPI exists, it is largely

targeted to the engineering side of the technology. The last finding displayed that although

literature on net neutrality is vast, little research has been documented on how damaging DPI

technology can be to net neutrality.

4.1 Background Information

Table 1 provides an overview on the technical background of survey respondents. Of the

68 respondents, 14 or 21% work in the IT field, while 53 or 79% do not (1 abstained from

answering). This question was posed to decipher the demographics of the audience and to

provide a means to compare the results between IT and non-IT workers.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 20

Table 1

Which Best Describes Your Background?

Table 2 shows the lack of awareness of DPI. The majority of respondents 46, or 68%

were not aware of what packet inspection was. Even fewer know about deep packet inspection,

55 or 81% or traffic shaping 52 or 77%. These questions set the baseline for the group as a

whole. When viewing the results, they begin to show that a gap exists in the general Internet

user’s knowledge of network monitoring and filtering.

Table 2

Background Knowledge Questions (All Participants).

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 21

To explore these findings further, the researcher looked those working in the IT field.

The data in Table 3 shows that all respondents knew what packet inspection was, while 9, or

64% knew about deep packet inspection and 10, or 71% were familiar with traffic shaping. It

was no surprise that greater knowledge of network monitoring and filtering technologies was

expected of those in the IT field.

Table 3

Background Knowledge Questions (IT Field Only).

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 22

Inversely, by looking at the same questions when asked only to non-IT workers, the

results speak loudly. As shown in Table 4, only 8 respondents, or 15% knew what packet

inspection was, while only 4 or 7% knew about deep packet inspection and only 6, or 11% were

familiar with traffic shaping. When comparing IT workers against non-IT workers, there was a

decrease in background knowledge of 85%, 57%, and 60% respectively. A decrease was

expected, but this radical fall off provided solid evidence that serious gaps in network monitoring

knowledge exists and a reminder that what may seem like common knowledge to industry

insiders, could be foreign to laymen.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 23

Table 4

Background Knowledge Questions (Non-IT Field Only).

4.2 Privacy and Content

Once the background information was cataloged, the study moved to understand the

attitudes of respondents towards privacy and content. As previously discussed, the format

changed from yes or no responses to a rating scale, where a score of 0 meant the respondent

strongly disagreed with the statement and a score of 5 meant they strongly agreed. The first

segment focused on Internet Service Providers.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 24

Should ISPs be allowed to monitor network communications? As Table 5 shows, answers

to this question scored a .9 making it clear that the respondents did not think ISPs should be

allowed to examine user network communications. This response was echoed when the subject

switched to telephone communications as well. These queries were included to validate the

necessity of privacy when communicating over any medium and the response overwhelmingly

confirmed the need.

Table 5

Examination of Network and Telephone Communications.

After establishing the need for privacy, the study continued and focused on ISPs ability to

delay or deny communications based on the content being sent. Table 6 presents the scores of

1.46 in response to deny communications and 1.49 in response to delay communications. While

not as lopsided as the answers in Table 5, the scores clearly illustrate disapproval of ISPs

denying or delaying communications based on the content.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 25

Table 6

ISPs Should Be Allowed to Deny or Delay Communications.

The same questions were posed regarding telephone communications. In response to

service provider’s ability to deny or delay communications, the subjects disagreed even more

stringently. As shown in Table 7, the scores tallied 1.14 and 1.2, clearly laying out a disdain for

any type of delaying or denying of communications based on content. This is a critical point that

must be disseminated; Internet and telephone users do not want their communications disrupted

based on the type of content they are utilizing.

Table 7

Telephone Deny or Delay Communications.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 26

To cap off the content based questions, the survey turned to the respondent’s attitudes

towards content based fees. For both network and telephone communications, the subjects

decreed a strong unwillingness to pay for service based on content. In this case, the scores for

both network communications (1.75) and telephone communications (1.61) were squarely placed

in the strongly disagree category. As evidenced in Table 8, the thought of paying a premium for

specific types of communications did not sit well.

Table 8

Premium Content, Premium Fees?

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 27

4.3 Regulation

The content and privacy attitudes were clearly displayed as evidenced by reviewing the

results in the previous segment, but the trend did not continue with the regulation results as they

were a bit more ambiguous.

After reviewing the data for the network monitoring topics (packet inspection, deep

packet inspection and traffic shaping), the majority of the answers were agreeable (final score of

~3) for regulation, but disagreeable (final score of ~2) for government regulation. The general

regulation results are defined in Table 9, which showed the respondents were middle of the road

when asked about general regulation of packet inspection, deep packet inspection and traffic

shaping. Table 10 illustrated that the respondents did not feel government regulation was the

answer. Following suite, the questions surrounding regulation by a standards agency scored

roughly on par with general regulation as shown in Table 11. These results detail the overall

uneasiness felt around government regulation, but still supported that some regulation would be

beneficial.

Table 9

To Regulate or Not To Regulate?

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 28

Table 10

Government Regulation?
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 29

Table 11

Standards Agency Regulation?

4.4 Trust and Moral Obligations

As deep packet inspection gives ISPs the ability to peer into the content of network

communications, the threat of abuse is present. The study attempted to gauge the respondent’s

feelings about whether or not service providers had moral obligations when utilizing network

monitoring technology. Table 12 shows that while the ratings were agreeable, with scores of

3.34 for ISPs and 3.47 for telephone service providers. This shows that even though the

respondents were lukewarm towards regulation, they felt that there still was a moral obligation

for service providers when examining communications.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 30

Table 12

Moral Obligations?

Continuing the theme of the previous statements, the next two survey questions dealt with

trust. As illustrated in Table 13, Internet users felt they could somewhat trust service providers

to examine only the content and not the information of their communications. It should be noted,

that the highest tallies were present in the strongly disagree (no trust) options on both statements

and the lowest tallies were in the strongly agree (trust) columns.

Table 13

Trusted Providers.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 31

4.5 Privacy of Communications

The loudest declaration from the respondents was far and away their privacy concerns.

These last two questions were included to put a statement on the survey, and the results came

through. Table 14 shows the direct privacy questions scored the highest of all questions. When

asked if Internet and telephone communications should be private, the audience resoundingly

agreed, with scores of 4.44 and 4.55. These final questions were further evidence that privacy is

a concern that is of utmost importance to general Internet users.

Table 14

Privacy of Communications.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 32

4.6 Summary

Clearly the results provided a deeper understanding of the technical knowledge and

attitudes of general Internet population. This understanding validated that the identified gap in

the literature would directly address the concerns of the public.

Once the demographics of the study were established, the results between IT and non-IT

workers showed a vast difference in network monitoring knowledge. This delta in the

knowledge adds to the evidence that more research and communication is needed to ensure the

general population is educated about threats to their privacy. When queried about regulation, the

audience felt it was a good idea, but they felt strongly that government regulation was not the

answer. That privacy was worth worrying about was validated as the respondents made it clear

they want their communications to remain private and they want service providers to stay away

from the information they are sharing.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 33

Chapter 5 – Discussion and Conclusions

This researcher believes this study contributes to the literature still dominated by

engineering issues and public policy debate. The findings help explain how a gap in the

literature can have far reaching consequences and the need for more education and awareness of

the general public with regards to privacy of communications is needed. Analyzing the results

clearly illustrated a desire for privacy, yet the knowledge of network monitoring technologies

was scant at best. Without raising the awareness of technologies like deep packet inspection and

communicating the dangers it carries, net neutrality will be difficult to achieve.

In designing the research, a three phase methodology was adopted. The first phase was

to gather existing literature on deep packet inspection and net neutrality and how they related to

privacy of communications. The gap in the literature became evident. Once the empirical data

was garnered, the second phase focused on generating questions to help gauge the attitudes and

knowledge of Internet users. Once the questions were in hand, they were presented via survey to

help quantify the aforementioned literature gap.

The methodology used may limit the use and interpretation of the data. First, by relying

on qualitative means to demonstrate the missing links, the study will always be open to

interpretation. As the nature of qualitative research is to generate and quantify opinions, it can

only be as strong as the opinions and the analysis of them. This is especially difficult when the
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 34

subject matter is as controversial as net neutrality. Second, the number of respondents was

small. But this researcher believes that even with a greater number of respondents, the overall

trends would remain largely unchanged. The focus of the study was to fill a gap in the literature.

The validity of the gap was proven as the drastic differences in knowledge between IT and non-

IT workers was displayed in the results. The study’s ability to fill that gap will be judged as time

passes. This study should be considered an early attempt to investigate the linkages between

deep packet inspection and net neutrality. The literature could benefit from other studies;

notably - a survey of ISPs. Further research is necessary to ensure that while technology

advances, so does the ability to keep a balance between progress and privacy. This research

generally supports the belief that DPI endangers net neutrality and privacy. But more research

would, in the long run, produce a more complete picture of the challenges that DPI poses to net

neutrality.
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 35

References

Anderson, N. (2007). Deep packet inspection meets ‘Net Neutrality, CALEA’. Arstechnica.

Retrieved from http://arstechnica.com/articles/culture/Deep-packet-inspection-meets-net-

neutrality.ars

Becchi, M., Crowley, P. (2007). An improved algorithm to accelerate regular expression

evaluation. Proceedings of the 3rd ACM/IEEE Symposium on Architecture for

networking and communications systems.

Economedies, N. (2008). “Net Neutrality,” Non-Discrimination and digital distribution of

content through the Internet. I/S: A Journal of Law and Policy. 4 (2) 209-233.

Feldman, P. J. (2010). FCC releases net neutrality NPRM - Let the jousting begin.

CommLawBlog. Retrieved from

http://www.commlawblog.com/2009/10/articles/wireline-telephony/fcc-releases-net-

neutrality-nprm-let-the-jousting-begin/

Georgiadis, L. Guérin, R. Peris, V. & Sivarajan, K. N. (1996). Efficient network QoS

provisioning based on per node traffic shaping. IEEE/ACM Transactions on Networking

(TON), 4-n.(4), 482-501.

Goldman, J. E. & Rawls, P. T. (2004). Applied data communications (4th ed.). Hoboken, NJ:

Wiley.

Hills, T. (2006). Deep packet inspection. Lightreading. Retrieved from

http://www.lightreading.com/document.asp?doc_id=111404

Jordan, S. (2009). Implications of Internet architecture on net neutrality, ACM Transactions on

Internet Technology, 9, 1-28.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 36

Kennedy, J. (2010). Draft net neutrality bill to go before US Congress. Silicon Republic.

Retrieved from http://www.siliconrepublic.com/comms/item/18046-draft-net-neutrality-

bill

Kirgman, E. (2010). Net neutrality bill might be more about message than action. National

Journal. Retrieved from http://techdailydose.nationaljournal.com/2010/09/net-neutrality-

bill-might-be-m.php

Kumar, S., Turner, J., & Williams, J. (2006). Advanced algorithms for fast and scalable deep

packet inspection. Proceedings of the 2006 ACM/IEEE Symposium on Architecture for

Networking and Communications Systems.

Porter, T. (2005). The perils of deep packet inspection. Security Focus. Retrieved from

http://www.securityfocus.com/infocus/1817

Ranum, M. (2005). What is "Deep Inspection”?". Ranum. Retrieved from

http://www.ranum.com/security/computer_security/editorials/deepinspect/index.html

Renals, J. (2009). Blocking skype through deep packet inspection. Proceedings of the 42nd

Hawaii International Conference on System Sciences.

Schahczenski, C. (2008). Net neutrality, computing and social change. SIGCAS Computers and

Society, 38-2.

http://www.ranum.com/security/computer_security/editorials/deepinspect/index.html

Smith, R., Estan, C., Jha, S., & Kong,S. (2008) Deflating the big bang: fast and scalable deep

packet inspection with extended finite automata. Proceedings of the ACM SIGCOMM

2008 Conference on Data Communication.

Stallings, William. (2007). Data and computer communications (8th ed.). Upper Saddle River,

NJ: Pearson Education, Inc.

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 37
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 38

Appendix A – Survey Questions

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 39
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 40
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 41
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 42
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 43
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 44
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 45
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 46

Appendix B - Survey Results

Section 1 - BACKGROUND QUESTIONS (4 questions)

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 47

Section 2 - INTERNET SERVICE PROVIDERS (15 questions)

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 48
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 49
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 50
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 51
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 52
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 53
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 54
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 55

Section 3 - TELEPHONE SERVICE PROVIDERS (12 Questions)

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 56
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 57
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 58
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 59
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 60
DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 61

Section 4 - PRIVACY OF COMMUNICATIONS (2 questions)

DEEP PACKET INSPECTION AND ITS EFFECTS ON NET NEUTRALITY 62