ARCHIVED - Transcript of Proceeding
This page has been archived on the Web
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.
Providing Content in Canada's Official Languages
Please note that the Official Languages Act requires that government publications be available in both official languages.
In order to meet some of the requirements under this Act, the Commission's transcripts will therefore be bilingual as to their covers, the listing of CRTC members and staff attending the hearings, and the table of contents.
However, the aforementioned publication is the recorded verbatim transcript and, as such, is transcribed in either of the official languages, depending on the language spoken by the participant at the hearing.
TRANSCRIPT OF PROCEEDINGS BEFORE
THE CANADIAN RADIO-TELEVISION AND
Review of the Internet traffic management practices of Internet service providers
140 Promenade du Portage
July 10, 2009
In order to meet the requirements of the Official Languages
Act, transcripts of proceedings before the Commission will be
bilingual as to their covers, the listing of the CRTC members
and staff attending the public hearings, and the Table of
However, the aforementioned publication is the recorded
verbatim transcript and, as such, is taped and transcribed in
either of the official languages, depending on the language
spoken by the participant at the public hearing.
Canadian Radio-television and
Review of the Internet traffic management practices of Internet service providers
Konrad von Finckenstein Chairperson
Len Katz Commissioner
Suzanne Lamarre Commissioner
Candice Molnar Commissioner
Timothy Denton Commissioner
Sylvie Bouffard Secretary
Regan Morris Legal Counsel /
Chris Seidl Hearing Managers
140 Promenade du Portage
July 10, 2009
- iv -
TABLE OF CONTENTS
PAGE / PARA
Telus Communications Company 711 / 4065
Cogeco Cable Inc. 770 / 4401
Barrett Xplore Inc. 801 / 4576
--- Upon resuming on Friday, July 10, 2009 at 0900
4059 LA PRÉSIDENT: Bonjour. Good morning everybody.
4060 Madame la Secrétaire, commençons.
4061 LA SECRÉTAIRE: Merci, Monsieur le Président.
4062 We will begin day 5 with Item 20 on our Agenda.
4063 I would now invite TELUS Telecommunications Company to make its presentation.
4064 Appearing for TELUS is Mr. Michael Hennessy. Please introduce your colleagues and proceed with your presentation.
4065 MR. HENNESSY: Good morning, Mr. Chairman, Vice-Chair, Commissioners and staff. My name is Michael Hennessy and I am Senior Vice President, Regulatory & Government Affairs at TELUS.
4066 With me this morning are, to my right, David Neale, Senior Vice President, Products and Services of our Consumer Solutions Branch as well as our Head of Mobile Industry Leadership; and, to my immediate left, Craig McTaggart, Director, Broadband Policy; and to Craig's left, Colin Lachance, Director, Access Solutions. As well, we would like to say greetings to all our friends at #CRTC that are following this hearing on the Tweet Universe.
4067 So TELUS appreciates the opportunity to offer its perspective on the subject of Internet traffic management practices and, as requested by the Commission, our comments apply only to practices related to the public Internet and not to managed wireline or wireless services like our IPTV that may use an IP platform.
4068 As has been stated, TELUS does not currently employ Internet traffic management practices like traffic shaping. However, that should not be taken to imply that TELUS would not consider adopting traffic management techniques if congestion becomes more of a problem.
4070 MR. NEALE: So, for example, peer-to-peer applications, as well as new video-rich applications, are challenging many of the engineering and economic assumptions on which Internet access facilities have historically been built and priced.
4071 Bandwidth costs money, so it makes sense to manage capacity and investment efficiently. We take the investment part very seriously. In 2009, in the midst of one of the worst economic crises that we have actually faced, TELUS will spend $950 million on next-generation hybrid fibre and wireless HSPA networks because we need to be competitive. However, simply pouring more investment into the ground or the airwaves is not the only solution to capacity challenges. Traffic management, innovation, application-based management techniques, and consumption-based pricing are other ways that we can manage capacity.
4072 We fully expect that Internet traffic will continue to show dramatic growth on wireline and wireless networks -- think high-definition "everything" -- and consider that all carriers must continue to have the flexibility to adopt reasonable measures to manage that traffic in any manner that is neither unjustly discriminatory or unduly preferential.
4073 Canada should be proud of the Internet facilities that we have and that we are building. According to the OECD, we still lead the G8, even before we count in the next-generation investments in HSPA wireless, DOCSIS 3 cable service and next-generation fibre-based services.
4074 TELUS has invested and continues to invest enormous amounts in physical infrastructure so that its customers can access the Internet and other services faster and in more places than ever. But we will only invest where we can see either a return on that investment or a need to compete to protect our business.
4075 That's why we are troubled by the statements of some parties that assume that all we need to do to keep up with traffic is to just keep investing with no guarantee of a return. Or, even more ludicrously, that if the Commission adopts even more wholesale arbitrage, then somehow facilities-based ISPs will be encouraged to invest more.
4076 Its simply nonsense to insist on the one hand that ISPs should invest more in capacity, and then suggest that that capacity must be shared, at a discount, with other ISPs that are not incented to invest and, worse still, that don't think there should be any restriction on the amount of someone else's bandwidth that they consume.
4078 MR. HENNESSY: The prospect of any new regulation that could restrict our ability to reasonably manage traffic or to fairly discriminate causes TELUS serious concern. Unlike in the monopoly era in telecommunications, regulation no longer shields carriers from risks, but it can increase them.
4079 Vigorous competition and rapidly changing technology already make our investments in next-generation networks risky, and the threat of new regulation at either at the network management, retail or wholesale level that could constrain our ability to earn a reasonable return on those investments only increases risk.
4080 In our opinion the basic issues in this proceeding are not difficult to identify and address. We agree that access to an open Internet should be an overarching policy goal. We just don't believe in absolutes or bright line tests. There are lots of examples of carriers blocking or managing traffic in ways that cause no concerns. No one disputes that it's a good thing that ISPs block spam and no one expects the CRTC to intervene to examine our practices when it comes to cyber security in terms of denial of service attacks.
4082 MR. McTAGGART: The principal concern that we have heard at these hearings is about the potential for anti-competitive behaviour by carriers that have affiliated content properties or business arrangements with particular content providers. We agree that ISPs should not engage in undue preferences that favour affiliated businesses or that materially degrade the ability of independent content providers and users to use the Internet. And we understand why non-affiliated providers of content worry about undue preference.
4083 Luckily, we already have robust non-discrimination rules in Canada, the very rules that net neutrality advocates in the U.S. continue to seek. That to us is a critical point of departure from the net neutrality debate that is raging south of the border.
4084 Under sections 27(2) and 36 of the Telecommunications Act, the Commission can already address specific ISP practices that it considers might contravene the Act, but in doing so we argue that it should take care not to adopt overly-broad ex ante rules of general application that could do any of these four things:
4085 - it could replace the fact-based/case-law approach that the Commission has employed successfully as the telecommunications industry has evolved over the past 30 years;
4086 - that could unnecessarily restrict the scope for the engineering and business judgment of each ISP to prevail in their individual responses to the rapidly changing and highly competitive Internet access marketplace;
4087 - that could prevent Canada's competing ISPs, often using different network architectures, from differentiating themselves by offering different types of services to meet the needs of different types of users; or, finally,
4088 - that could reduce the incentives of carriers to make or unduly constrain their ability to recover the enormous capital investments required to build next-generation telecommunications networks.
4089 Canadian law already protects Internet users' ability to access and use the legal Internet content and applications of their choice, subject to exceptions for compliance with legal requirements, enforcement of applicable contracts with customers, and reasonable network management.
4090 MR. NEALE: In our reply comments, TELUS noted the relatively small record on the issue of traffic management practices on wireless data networks and we urge the Commission not to establish rules applicable to those networks at this time. While we believe that the wireless Internet should be open, we reject a one-size-fits-all approach to traffic management on the different networks.
4091 Wireless ISPs RipNET and Barrett Xplore, as well as soon-to-be wireless provider Quebecor, have explained that wireless networks are very different from wireline, and there cannot be a one-size-fits-all rule. We agree. Because wireless networks use radio spectrum, capacity that is -- and will be for the foreseeable future -- limited relative to the wireline.
4092 The constraints imposed by limited spectrum make a higher degree of network management necessary and, as Quebecor put it in their initial comments, the imposition of a symmetrical regulatory regime would be too restrictive and it would risk restraining wireless networks from reaching their full potential.
4093 The good news is that the advent of next-generation wireless networks -- networks that TELUS and other carriers are deploying at a cost in the billions of dollars -- promise to bring the wireless Internet access experience more in line with that of the wireline.
4094 Perhaps because wireless customers have been more exposed to usage-based pricing, it may be easier to condition consumption through pricing, but there will still be significant requirements to manage the traffic loads and the nature of the network will continue to demand different management responses.
4095 MR. HENNESSY: Colin..?
4096 MR. LACHANCE: With respect to wholesale issues, TELUS notes that some parties to this proceeding have attempted to use it to further their agenda of seeking more mandated access to carrier facilities, but we appreciate that the Commission remains focused on the narrow issue of whether it is appropriate for carrier traffic management practices to be applied to wholesale Internet services.
4097 TELUS does not currently impose any traffic management practices on its wholesale Internet ADSL access services, but if it adopted such practices in the network used to serve its retail customers, then any ISPs using that platform to serve their own retail customers would be subject to those technical measures to the same extent, as section 27(2) of the Act would require. ISPs desiring a greater degree of control over traffic management would have to invest in more of their own facilities or subscribe to other dedicated or managed services to avoid reliance on a shared platform.
4099 MR. McTAGGART: TELUS does not object to the Commission adopting broad guidelines that describe values that should generally be reflected in ITMPs, but we oppose the calls of some parties for the Commission to develop ex ante tests or rules of general application governing any aspect of the provision of Internet access services. We believe that would be a recipe for unintended consequences.
4100 But in terms of ITMPs, we are of the opinion that section 27, not section 36, is the right tool to address the kinds of concerns that have been expressed in this hearing, because Section 36 is not engaged when content arrives in the same form in which it was sent.
4101 TELUS continues to take no position on whether the specific technical traffic management practices under examination in this proceeding are "acceptable" or not, but would like to contribute to the process of identifying principles that the Commission could use in the future when considering complaints from content and application providers or retail customers under section 27.
4102 TELUS would take the Open Internet Coalition's three-step proposal as a starting-point, but modify it as follows.
4103 The first step: Is the traffic management practice in question designed to enhance the security, efficiency, and/or availability of the network for users?
4104 This is intended to replace the subjective standard of "pressing and substantial" with more objective purposes. These purposes exclude the kinds of nefarious objectives that some parties have speculated about this week.
4105 The second step: Is the traffic management practice narrowly tailored to address one or more of these this objectives?
4106 We agree that the concepts of proportionality between problem and response and of a "tight fit" between the objective and the measure adopted are worthwhile elements of these guidelines.
4107 The third step: Is the traffic management practice reasonable?
4108 In TELUS' view, if the measure is narrowly tailored to serve a legitimate objective, then a requirement that it be the perfect solution should not be necessary. Rather, the ISP applying these guidelines must satisfy itself as a legal matter that the measure could ultimately pass muster under section 27(2), the standard for which, in practice, is ultimately reasonableness.
4109 Finally, we believe that the Commission should confirm that section 36 of the Act is meant to deal with the actual blocking or censorship of content, rather than traffic management, and only to require ex ante approval for measures that result in the blocking or modification of content.
4110 MR. HENNESSY: So in conclusion, Mr. Chairman, we have heard much innuendo this week but seen little evidence to support claims of "bad" discrimination. If there is a problem of undue preference, let someone bring it forward, then we can develop solutions. The CRTC has all the tools you need to address the concerns that have been raised, let's not create more rules or more committees without evidence of harm.
4111 Every year there is more content, choice, diversity and innovation online than the previous year. The simple reality that all competing ISPs face is that Canadian Internet users expect to be able to go where they want and do what they want on the Internet and any ISP, or now any wireless provider, that fails to meet that expectation will suffer in the marketplace.
4112 We do not believe that it's in the interests of ISPs to block or degrade content or traffic, but we agree that the Commission should be vigilant when it comes to undue preference.
4113 Equally, the Commission should be as vigilant in ensuring an environment that allows for fair forms of discrimination. That, we submit, forms the basis of an innovation society.
4114 Thank you for the opportunity to offer our perspectives and we look forward to any questions you may have.
4115 THE CHAIRPERSON: Thank you very much for your presentation.
4116 You deal with two subjects, wholesale and direct retail. If we develop guidelines, do you think we should develop guidelines for both separate ones or can the same guidelines serve for both?
4117 MR. HENNESSY: I think that the test that Mr. McTaggart put forward would cover both.
4118 THE CHAIRPERSON: The other thing, in your opening remarks you make the very large qualifier:
"... TELUS does not currently employ specialized Internet traffic management practices like traffic shaping. However, that should not be taken to imply that TELUS would not consider adopting traffic management techniques if congestion becomes more of a problem."
4119 Obviously on the face of it that's logical, but how far does this go?
4120 You have heard this week a lot of tough talk about application-specific ITMPs and essentially fairly strong voices saying application-specific ITMPs should not be allowed because they invariably discriminate, they have an unequal impact, it depends on who -- (a) they may not work, you may not be able to detect the application because of encryption or something, but even if you do, depending on who are your clientele, they may impact on one ISP different than another, one wholesale customer different than the other, and therefore that should not be allowed.
4121 When I spoke to OIC I said do we have to go that far or could we just say if you are doing something that's application-specific where there is such a great likelihood there might be any uneven impact it should at least require ex ante, but everything else you would do on an ex post basis.
4122 How do you feel about that? Do you feel ex ante application should at all be allowed and, if so -- application-specific should be allowed at all and, if so, should they have a specific treatment of ex ante authorization or not?
4123 MR. HENNESSY: Yes. I am firmly of the view that the Commission's jurisprudence in telecommunications when it comes to discrimination has been based basically on the rules -- developing rules that are ex post, dealing with a fact-based situation and I think that has served us very well.
4124 The thing with the Internet is that many of the applications that we are going to deal with in two years don't necessarily exist today. Everything on the Internet is fairly new.
4125 THE CHAIRPERSON: But by that same token, because it's such a vibrant and dynamic industry we don't know what's coming down, should we rule out automatically any kind of ex ante requirement for anything?
4126 MR. HENNESSY: In a proceeding like this I would say you wouldn't want to develop ex ante rules.
4127 But I think if you were to -- and let's face it, generally --
4128 THE CHAIRPERSON: What is the flip side?
4129 Normally our procedure is ex post, complaint driven, somebody comes, et cetera.
4130 MR. HENNESSY: Right. Yes.
4131 THE CHAIRPERSON: But should we rule out at all using ex-ante and, as I said, the specific case of application-specific ITMPs, should we say, you know, there the overwhelming possibility of this discrimination is so large that we feel we have to be vigilant and apply an ex ante rule?
4132 MR. HENNESSY: I would say no, because there isn't an overwhelming record of evidence to justify that.
4133 THE CHAIRPERSON: So you said to stay with your ex post approach, but leave open the ability to --
4134 MR. HENNESSY: Yes. You might --
4135 THE CHAIRPERSON: -- adopt ex ante if the need would be proven in future?
4136 MR. HENNESSY: Yes. Because I think in the future -- let's say, I mean if you really boil it down, most of your concern is going to deal with anti-competitive behaviour and probably anti-competitive behaviour in the sense of undue preference, when you have some sort of vertical integration. If you see serial behaviour, evidence of serial behaviour in that, then it is appropriate to develop an ex ante rule to deal with that. I think there is probably lots of evidence in how CRTC policies have developed to do exactly that kind of thing.
4137 But I think that it's worth leaving things open enough to find out if that's the fact.
4138 THE CHAIRPERSON: Okay. Thank you.
4140 CONSEILLÈRE LAMARRE: Merci, Monsieur le Président.
4141 You have clarified a few of the issues that were a little bit blurry when I read your submissions, but I still have a few questions. So I will be turning a little bit to your submission and your presentation from this morning for the questions.
4142 The first one that I would like to touch on is within your submission, and even to a certain extent in your presentation this morning, you are disputing the fact that Internet service providers are not a public utility and hence that Internet is not a public utility. Looking into it, you know, when I look at the French translation for a public utility, in French it's called un service public, so I have trouble reconciling the fact that you don't consider Internet as a service public as a utility and the type of service you are providing right now.
4143 Would you like to expand on why you don't consider that Internet service providers are actually providing the equivalent of utility?
4144 MR. HENNESSY: Well, we are providing the equivalent of a public service I think clearly in the public carrier service, but public utility regulation was designed to deal with monopoly providers and monopoly practices and in that sense the fact that you have multiple providers, even if a limited number, I think doesn't make the application of a utility regulation appropriate which is why it is forborne in the first place.
4145 COMMISSIONER LAMARRE: Okay. So it's in the sense of the competition that is present right now to provide those services, not in the sense of the nature of the service that you are providing?
4146 MR. HENNESSY: Yes.
4147 COMMISSIONER LAMARRE: Okay. You also seem to be worried to a certain extent that regulation would be placing difficulties that would be difficult to surmount with regards to making engineering decisions. I would like to understand just where your difficulty is, because throughout history engineering has always had to deal with regulation or laws in any way, shape or form and that has always been one of the factors to consider when you do engineering designs, it is not something that you take into consideration after you have done the design. So why would it become a problem if rules were extended to certain types of Internet traffic management practices?
4148 MR. HENNESSY: I think I will give that to Dave, since I brought him all the way from Toronto to talk about the dynamic nature of the Internet and unintended consequences.
4149 MR. NEALE: Thank you, Michael.
4150 So there are two reasons why we make our observations.
4151 One is that we believe that the management of the wireless network will continue to be quite different to the management of the wireline network and what we would be concerned about is essentially blanket regulation that might apply quite well to one but not be well matched to the other. So that's one key point.
4152 I think the other thing is that the whole science of traffic management, and particularly in the field of IP, is undergoing so much development at the moment that one would be concerned about adopting a framework now that in fact was being corrected by engineering developments over the coming months.
4153 So for example, one of the examples we were discussing was the advent of clean-slate network architecture and any of these new design initiatives that are looking at newer and better ways to manage traffic congestion and the way that networks behave, and we would be concerned that, maybe, broad-brush regulation now might inhibit some of those developments later.
4154 COMMISSIONER LAMARRE: If I understand you correctly, what you are really concerned about is that we don't make the mistake, in your own words, of not distinguishing between wireless and wireline, which have different technologies to work on, and, also, that there would be unintended consequences.
4155 But you are not disputing the fact that rules and regulations are part of what you need to consider when you do engineering design.
4156 MR. NEALE: No, we are not discounting the fact that we need to take those into account. All we are saying is, we have to be careful about applying too much rigidity at this stage.
4157 COMMISSIONER LAMARRE: Okay. Also in your submission, at one point, I was puzzled about the way you were either differentiating or not differentiating between section 27(2) and section 36 of the Telecom Act.
4158 In your presentation this morning you provided me with part of my answer, but I am still left with one question.
4159 On page 5 of your presentation you say: "Finally, we believe that the Commission should confirm that section 36 of the Act is meant to deal with the actual blocking or censorship of content."
4160 When I read section 36, it says that a Canadian carrier shall not control the content or influence the meaning or purpose of telecommunications carried by it to the public.
4161 So it seems that the element of influencing the purpose of the telecommunication carried is missing from what you stated in your presentation this morning.
4162 MR. HENNESSY: I don't think so, but I will give that to Craig. I think that our fundamental point is that we are talking about the content of the message in the very traditional sense of content or the ability of people to communicate ideas to each other.
4163 MR. McTAGGART: Right, that is very much the focus, the idea of content, and I think these rules were adopted at a time when telecommunications was characterized by a single monopoly network.
4164 And, of course, the historic common carriage rule was that the carrier simply didn't involve itself in the content, or alter the content.
4165 Over time, the rule was used in different circumstances, as new uses for the telecommunications network became apparent, and because of the monopoly market power of the network operator, legislatures and regulators decided to use those rules to keep the telephone company out of new and emerging businesses, like broadcasting and data processing and that kind of thing. The idea was that the carrier would simply carry.
4166 With respect to the subject matter of this proceeding and the internet traffic management practices that we have talked about, our view is that they don't generally involve the alteration of content. Rather, the concern is about how they are being handled, and section 36 is a binary rule --
4167 COMMISSIONER LAMARRE: I'm sorry; you just said that the concern is how it is being handled. Wouldn't the concern be that they are being handled at all?
4168 MR. McTAGGART: I think we might be using different meanings of the word.
4169 What I am trying to get at is that section 36 is an absolute rule that speaks to either altering content or not; whereas 27(2) inherently contains flexibility, in terms of the concepts of reasonable or undue or unjust.
4170 So when we talk about what I mean by how traffic is being carried, disagreement about whether it is being carried in a discriminatory fashion or not really falls better under 27(2) than 36.
4171 COMMISSIONER LAMARRE: Okay. You also, in your answer, made the distinction -- you talked about the rules in the Act being adopted in a monopoly situation, and now there is competition.
4172 If I could flip the question back to you, why should the result be different for the end user about how the telecommunication is being carried or handled now that there is competition?
4173 Shouldn't there be the same result, or at least a better result, now that there is competition?
4174 MR. McTAGGART: This is a concept that I note MTS Allstream put forward to you, as well; that is, certainly with respect to section 27(2) analysis, the degree of competition in both the ISP marketplace and also the marketplace for the content or application in question is relevant to the analysis of what is unjust, what is undue, what is unreasonable.
4175 And the Commission has, indeed, adopted that approach previously with respect to mobile wireless services, as we discuss in our initial comments --
4176 COMMISSIONER LAMARRE: But what about section 36?
4177 MR. McTAGGART: Exactly. Once you get to 36, it's less clear. That is why I say it's an absolute rule.
4178 I think in our initial comments we offered the view that there are implicit exceptions to section 36. Several parties said that it doesn't look that way, the way it reads, and I can appreciate that argument.
4179 That is why I focus on 27(2) instead of 36.
4180 COMMISSIONER LAMARRE: Okay. Thank you.
4181 In your submission, and you touched on it, also, in your presentation, you make a distinction between unjust discrimination and fair discrimination.
4182 To quote section 9 of your reply comments, you say that the question of whether an ISP should be prevented from discriminating in favour of internet content and applications according to their users' preference, or improve the ability of users to access and use legal internet content and applications of their choice is not -- that it should not be prevented.
4183 Then you go on to state that TELUS recognizes that sometimes there may be a fine line between the two.
4184 If I were to argue that it's not just sometimes, it's always a fine line between the two, how do you distinguish that fine line, and where do you draw it?
4185 MR. HENNESSY: I think, again, it gets back to issues -- usually the easiest way to look at it is to start at the anti-competitive end of things.
4186 But, clearly, by having a rule that says "no unjust discrimination" and "no undue preference" -- there is nothing wrong with having either preference or discrimination.
4187 In fact, the classic example, obviously, is price discrimination. There is nothing wrong, on its face, with discriminating between users on the internet, based on how much capacity they consume and charging them a different price.
4188 In fact, that is probably a very reasonable thing to do, but it is discriminatory, in that everybody is not treated equally from the perspective of paying the same amount to access the internet.
4189 COMMISSIONER LAMARRE: Okay. Touching on an issue that, I realize, you didn't have to talk too much about, because currently you have not put into place internet management practices, I would still like to get some precision from you.
4190 In your submission, at paragraph 34, you state that most customers have no interest in the traffic management decisions of telecom providers, but that practices that may affect the ability of customers to access legal internet content should be disclosed.
4191 First off, what evidence do you have that customers have no interest in those practices, because what we have seen so far, in preparing for this proceeding and in the proceeding, is that customers do have an interest in it.
4192 MR. HENNESSY: I would say that I think most customers might have an interest in the Michael Jackson funeral, but I think we would stick to our guns that most people couldn't care less about traffic management practices.
4193 Even when you actually measure the people who are following this proceeding online or through Twitter, it's a very insignificant number of the internet universe.
4194 I think that is the point we are making.
4195 People don't really care how things work. If you tell them that it's working in a way that's not fair, or that is impairing their ability to enjoy something, then they may become frustrated, but they still don't really care as to what is going on.
4196 I think Craig's point was that some degree of disclosure, when you are doing things, is still reasonable, but the detail isn't necessary.
4197 COMMISSIONER LAMARRE: Do you think they would care whether or not the practices that may be put in place are infringing on their privacy?
4198 I am asking the question because it is a topic that you have carefully avoided throughout your submission, your reply, and even your presentation this morning.
4199 Do you think that, if an ISP were using a practice that actually had an impact on privacy, that may raise more concerns?
4200 MR. HENNESSY: Yes, I absolutely agree.
4201 We didn't actually carefully avoid the subject, I guess it starts from the premise that, since we couldn't sniff into what people are doing, including, in another proceeding, measure whether or not they are actually going to Canadian internet sites or using Canadian content, it's kind of an irrelevant question for us.
4202 Dave probably, in terms of the information that we use about our customers -- I think at this point it would be --
4203 MR. NEALE: Yes, to Michael's point, the information that we have is not about an individual customer, it's about the aggregate behaviour of the customers on given segments of the network.
4204 To your point, we are not actually looking at the packets, because we don't need to do that. What we are using is traditional IP network management.
4205 So the parameters that we are measuring, we are looking at traffic engineering and the optimization of congestion management.
4206 Our argument is that a lot of these practices are reasonably arcane and, from the customer's perspective, what we are attempting to do is provide persistence of service, so they wouldn't see that.
4207 The example you gave of collecting information about an individual, yes, we agree that a customer would be concerned and should be aware of it, hence our position on transparency.
4208 MR. HENNESSY: I think that Craig has one point on the law itself.
4209 MR. McTAGGART: Yes. Lest it appear that we are not concerned at all about privacy issues, I would rather echo something that Mr. Glick said on behalf of the Open Internet Coalition; that is, we already have federal private sector privacy legislation and there is a process for dealing with problems.
4210 And you heard several other parties, particularly those who operate networks, say that they are subject to privacy law.
4211 So it's not a situation that carriers can violate privacy law willy-nilly in their internet traffic management practices, they have to consider the requirements of the existing law in those practices.
4212 So it's not a case that they are unregulated with respect to privacy.
4213 COMMISSIONER LAMARRE: But are you also in agreement that the Commission itself has to be concerned with the privacy issue, as per section 7(i) of the Act, which requires that whatever policies or guidelines we come up with, that we craft, have to contribute to the protection of privacy?
4214 MR. HENNESSY: I would put it a little differently. I think it goes back to our fundamental principle, where we say that everybody should have the right to access anything they want on the internet, as long as it is legal.
4215 And I think that, in a funny way, it goes without saying.
4216 I get very concerned when the Commission starts to get involved in an area, even though they did very much in the past, that is now covered by statute under a different regulatory body, because what you end up with, then, are duelling forms of regulation, and that usually leads to no good.
4217 COMMISSIONER LAMARRE: I am assuming that you are referring to PIPEDA. PIPEDA is actually a general rule that applies to several industries, and there might be something specific about telecommunications that the Commission needs to be concerned about.
4218 MR. HENNESSY: Again, I wish they wouldn't.
4219 COMMISSIONER LAMARRE: Okay. Going back to the test that you proposed on page 5 of your presentation this morning, which is -- you took over the Open Internet Coalition's three-step proposal, with your modifications.
4220 At (c) you have modified the test. You would rather see it being, "Is the traffic management practice reasonable," period, instead of, "the least restrictive means to reach the objective."
4221 Would you be concerned, or would you agree, that if a specific traffic management practice was identified as infringing on privacy, that, in this particular case, the test should be, "Is that practice the least restrictive means to reach the objective," because it is infringing on privacy?
4222 MR. HENNESSY: I think, if it's infringing on privacy, then it's something that should really be dealt with by the Privacy Commissioner.
4223 I would hope that between Kevin Doyle, our privacy head, and our regulatory team, and rational people like David Neale here, we would put the boots to anybody in our company who came up with a traffic management idea that invaded somebody's privacy.
4224 But at the end of the day, I think, again, the critical issue is really who has overarching jurisdiction in the area.
4225 COMMISSIONER LAMARRE: And it is your contention that because of PIPEDA, a statute that was adopted after the Telecommunications Act, and which makes no provision for overriding what is in the Telecommunications Act, we should, in fact, ignore our duty under section 7(i) of the Act?
4226 MR. HENNESSY: I will let Craig answer that. I often take positions that the Commission disagrees with, and usually they prevail.
4227 MR. McTAGGART: To be frank, Commissioner Lamarre, as you pointed out at the start of your questioning, privacy is not something that we have devoted a great deal of attention to in our submissions, primarily because we don't currently use the technologies. So I have to admit that I don't have a solid answer to that question, although I would note that the issue of overlapping jurisdiction over privacy is something that is a live issue in several areas, and I would actually defer to others who might appear before you who are more familiar with privacy law to deal with that specifically.
4228 MR. HENNESSY: Just to get over that hump, let me be clear. If you come across -- you know, if TELUS adopts a traffic management practice, or a more nefarious company like, let's say, MTS Allstream -- I'm just kidding, Chris -- if you come across a carrier doing something in traffic management and it raises privacy concerns, I think that it's totally reasonable for you to, at a minimum, address it, express your concern, suggest that it shouldn't happen.
4229 And if there is no response, then I think that the Privacy Commissioner kicks in.
4230 In practice, I don't think that you should ignore that kind of stuff. I am not being really redneck about the whole thing.
4231 COMMISSIONER LAMARRE: And I do note the point you made, that currently you are not engaging in any such practices.
4232 Those are my questions, Mr. Chair.
4233 THE CHAIRPERSON: Len...
4234 COMMISSIONER KATZ: Thank you, Mr. Chairman.
4235 Good morning. I have a couple of questions stemming from your opening comments this morning.
4236 On the first page you reinforced, I guess, the Commission's view that this hearing is related to the public internet and not to managed wireline or wireless, like your IPTV. I just want to confirm that any growth at all in your managed business on the IP side does not restrict or narrow the public internet capacity that is out there today.
4237 MR. HENNESSY: Dave could talk to that, but I think that, absolutely, it does.
4238 COMMISSIONER KATZ: Someone was shaking their head the other way on your left, so you may want to --
4239 MR. HENNESSY: Let me give you an example. It's the same as in the cable business; right? Part of your network is used for internet, part of it is used for television.
4240 In fact, I think, under the -- you know, we have a licensed broadcast undertaking using the same facilities that we use to provide internet. Under the Commission's rules, under the broadcasting rules, you would argue that preference has to be given to the broadcasting content that we deliver under licence over the internet service that we deliver on an unregulated basis.
4241 COMMISSIONER KATZ: Okay, but you have a pipe, and it is so big --
4242 MR. NEALE: If I could add to that, what Michael is describing is, obviously, if you have a service like TTV, which cannot tolerate outage, similarly, on cable TV, you can't tolerate flow restrictions because it results in the destruction of the image.
4243 What we are doing is, we are actually increasing the capacity all of the time, because not only do we have to provide capacity for our own managed service, but we have to cope with the capacity demands on the public internet.
4244 So when we talk about the broadband expansion activities that we are doing, that is actually to address that which you describe, Commissioner Katz. It is to accommodate the fact that, although, as Michael says, one has to provide quality of service on a linear broadcast model, we also can't afford to have that negatively impact the public internet demand, so we are always increasing our internal capacity.
4245 COMMISSIONER KATZ: Yes, but let's assume that your capacity is fixed for six months. You have to ramp up, you have to do lots of stuff to order facilities, and whatever else.
4246 It's fixed. As your managed IP business grows, for whatever reason, does that take away from the capacity on the public internet side?
4247 MR. NEALE: Not really, because, essentially, the amount of resources you have available determines your ability to sell wholesale.
4248 So, essentially, a lot of the managed services -- there is your own and there are managed services which are provided by third parties. But, essentially, if we had a constraint based on exhaust of resource, then, obviously, we would not be pushing and selling those wholesale resources.
4249 But the way that the network is set up is that we do not disadvantage the public internet. In fact, we were just discussing last night that, even in an IP world, you still groom traffic within your network.
4250 COMMISSIONER KATZ: Let me ask the question in a different way.
4251 I think you stated, both in your comments and in your document here, that you support the notion of oversight, particularly where it is related to self-dealing, where there is an undue preference because of treating yourself -- the perception of treating yourself differently. As that pipe that we talked about gets more and more congested with managed IP, which is the other half of your business, and complaints come in, do you see that as being within the purview of the CRTC to say: You are consuming more and more bandwidth for your managed services, why aren't you building out capacity rather than taking it from the public internet?
4252 MR. HENNESSY: Just before I give it back to Dave -- and I think that his point, just to get me under control, was to point out that on our access facility, unlike cable --
4253 It is probably much more of a problem for cable.
4254 On our access facilities, you tend to be deriving one stream down per TV set, so it is more in your backhaul network and everything that you see different traffic loads.
4255 Let's take the cable example -- and Dave, having been with Rogers, can explain that a little, but what I am trying to get at is, you kind of have two conflicting problems here.
4256 If you say to somebody, "You are a broadcast undertaking and you must give preference to X," all other things being equal, in a static world, that can mean that less preference is given to Y.
4257 So now you are talking about the Y being the internet. If, in a telecom world, you say, "You must give preference to the public internet over the regulated broadcasting system," then, clearly, you ultimately may run into a conflict. The more the network is shared -- and I think that maybe maybe you could describe a bit on the cable side why that becomes a bigger problem and then, you know, follow up with the Rogers guys with it.
4258 MR. NEALE: So, to that point, the primary difference between Internet provision from Telus using DSL and comparing, for instance, a cable provider, on the DSL link between the customer and the central office there is no traffic at all on the customer's link until they begin a service.
4259 So, for instance, the copper is lying there quiet until they begin a CO to the Internet; whereas, as you are familiar, in the cable world essentially that cable is carrying the broadcast traffic all of the time and it's being shared between the houses within that hub.
4260 So, essentially the difference in the access network there is more constraint, there is stuff that's in there all the time. They have more band width, but they have to manage more because each of the customer's pool resource within the hub; whereas in our own particular case in DSL you're nailing an individual circuit between each house and the central office.
4261 So, that does have some huge impact.
4262 COMMISSIONER KATZ: I understand that, but I want to come back to what Mr. Hennessy said and, that is, the classical dilemma that suddenly there's a restraint on dollars in your business and it's your business, and Mr. Entwistle decides that he wants to get a higher margin out of his IP management network, out of the cable network rather than investing more money and, so, he basically says, let's use more of the capacity that's out there for the IP managed business.
4263 And the result of that is your margin goes up on that part of your business because you're not investing the capital but you're getting the revenues coming out of it, but it's at the expense of the public Internet where that network gets squeezed, perhaps because the margins are lower than the cable or vice versa, I don't know.
4264 MR. HENNESSY: Yeah, I --
4265 COMMISSIONER KATZ: These are business decisions, these are classical decisions that you make.
4266 MR. HENNESSY: I don't mind, you know, addressing that theoretically. I would make the point that we could only wish that there was any margin in our TV business as opposed to where, you know -- when we put a dollar into the Internet and give people more capacity in the Internet we're more likely to benefit than...
4267 COMMISSIONER KATZ: But it's exactly for that reason because the margin's not there now on the TV business.
4268 MR. HENNESSY: Yeah.
4269 COMMISSIONER KATZ: My belief is you may not invest as much -- any more money into it, you want to get out whatever you can and, therefore, you'll use the pipe and you'll continue to expand it at the exclusion of the public interest.
4270 That's exactly the point I was trying to make.
4271 MR. HENNESSY: You know, I don't think so. I think that, you know, you look at the satellite guys, right, who've been added and are a lot bigger than we are and they're still not making any money in terms of positive net income.
4272 COMMISSIONER KATZ: Okay.
4273 MR. HENNESSY: So, I hear it -- you know, that's why I'm saying, it's a theoretical question.
4274 I hear your point, where do you -- you know, do you favour the public Internet or the -- I mean, another way I guess of asking it is, should -- if push comes to shove and we have to make a decision, which is more important the public Internet or the TV network.
4275 You know, I would say personally the public Internet.
4276 MR. NEALE: Yeah. If I could just add to that as well, Commissioner Katz.
4277 The dynamic is obviously we're in a commercial world, so if we had made decisions that ended up degrading the public Internet service, our customers are in a situation where they can get alternative services from other people.
4278 So, there is no wisdom at all in us degrading a service that actually causes churn because, as you know, that is very damaging.
4279 So, we actually have these correcting, moderating factors all the time anyway. The fact that there is competition out there, it stands to reason the last thing we could afford to do is degrade the service that in terms of customer volume is by far the majority in terms of individuals that use the service.
4280 MR. HENNESSY: Yeah. I guess at the end of the day, right, I mean, like in our territory Shaw has about 50 percent more Internet customers than we do and I think they probably have -- well, I wouldn't say a hundred percent but, you know, they have 87 percent of the market and the satellite guys have the rest and we have 100,000 subscribers.
4281 So, it's a bit of a -- it's not even a "Sophie's Choice" for us. We've got to pour money in at both ends and that's why we have the highest CAPEX ratio of any carrier in North America.
4282 COMMISSIONER KATZ: Okay. I want to move on.
4283 We heard yesterday that, I guess it was Dr. Odlyzko indicate that technological progress is evolving at a pace that is in concert or at par with demand. That was a general statement he made.
4284 Is that the case that you're finding in Canada as well?
4285 MR. NEALE: In many cases I would say that statement is fair. Quite often a problem that manifests itself today is already being approached quite often by some Internet task force anyway.
4286 So, one of the areas that we enjoy is that the very things that turn up as problems are quite often being addressed by the open source community or the Internet community in general.
4287 So, it's our belief that a lot of the things that we describe are -- essentially, they're all works in progress and the evolution and the rate at which the catch-up and the fixes are taking place is actually really working I think in our mutual favour.
4288 COMMISSIONER KATZ: Yes. Thank you.
4289 Page 2 of your comments this morning, just before the Issues in the Law, there's a statement that says:
"It is simply nonsense to insist on the one hand that ISPs should invest more in capacity and then suggest that capacity must be shared..." (As read)
4290 COMMISSIONER KATZ: And you added the words:
"...at a discount with other ISPs." (As read),
4291 COMMISSIONER KATZ: So, you put a caveat in there. So, is the issue financial as opposed political, if I can call it that?
4292 MR. HENNESSY: It's -- I would say fundamental issue in the legacy network is financial is the, you know, the Commission I think quite deliberately sets rates on the wholesale basis relative to retail to create a margin for other reselling ISPs to compete and the idea I guess at one point was to ensure that the Internet was characterized by competition.
4293 I think that's a really kind of archaic model and it's really I think the essence of what this proceeding is about is a recognition that competition is no longer driven by the number of suppliers at the access point but by the number of people that can freely innovate at the edge of the network which is, you know, the Google premise about innovation without permission.
4294 And to the extent that you have competition at the edge of the network, you've actually solved what the artificial arbitrage model you had created in a past decade was intended to do, that you now have more competition in the downstream and more choice than you can ever deal with.
4295 So, we deliberately --
4296 COMMISSIONER KATZ: I think the question though is, and it's a dilemma that we'll all face, is where is that edge technology, how far back does it go on the network or how far out does it go as well?
4297 MR. HENNESSY: Well, I think --
4298 COMMISSIONER KATZ: It's not going to be debated here...
4299 MR. HENNESSY: All you have to do today is look at the degree of competition at the edge, which is not measurable because there is so much.
4300 To get your answer, I don't think that you really have to deal any more -- you've already -- you've created so much competition in choice and value added service -- not you but the world generally through the Internet -- that the idea that you need some form of regulated arbitrage competition in the access and to create competition downstream has become absurd because it already exists beyond anything a regulator could possibly --
4301 COMMISSIONER KATZ: Well, I'm not looking at that end of it, I'm looking at the other end of it.
4302 But I've got one more question before I get yanked here and, that is, it's just a yes or no answer.
4303 On page 4 under your wholesale issues you say that:
"Telus currently does not impose any traffic management practices but, if adopted, such practices in the network used to serve its retail customers would be subject to the same technical measures on the wholesale side." (As read)
4304 COMMISSIONER KATZ: Is that a technical reason for that or is that a business reason?
4305 Just a yes or no answer.
4306 MR. HENNESSY: Colin.
4307 MR. LACHANCE: It's a technical reason, yes.
4308 COMMISSIONER KATZ: Okay. Then you have to explain the technical reason.
4309 MR. LACHANCE: As we've indicated in our comments throughout, the wholesale Internat ADSL service that we provide is provided on a shared platform with retail, so often the links throughout the network are carrying traffic for both retail and wholesale customers.
4310 So, application of various management practices, however defined, would apply equally to both, just as a natural consequence of the arrangement of the network.
4311 COMMISSIONER KATZ: And there's no way you can differentiate them?
4312 MR. LACHANCE: We can't say a hundred percent. There are small elements of the network that --
4313 COMMISSIONER KATZ: Well, let me ask this question this way. Last year -- or actually it was yesterday, someone mentioned that when Bell Canada entertained their throttling exercise last year with CAIP they did the retail first and the wholesale second, which basically implies they're able to segregate their retail from the wholesale business.
4314 Are you not able to do the same thing?
4315 MR. HENNESSY: Different networks, right.
4316 COMMISSIONER KATZ: I just want to know if you can do it.
4317 MR. HENNESSY: It would be -- for a large portion of the network we would not be able to segregate the difference between wholesale and retail.
4318 The network currently -- the current broadband network is a consequence of what was grown up between Alberta and B.C. before the companies merged. There are different evolu -- different parts of the network are at different points of evolution.
4319 So, it may be that in one community somewhere it's possible to differentiate between retail and wholesale...
4320 COMMISSIONER KATZ: Can I ask you guys in your final submission to us to lay out exactly what the obstacles are and what would be required if, in fact, you were asked to separate them.
4321 MR. LACHANCE: I mean, one would argue, if you really wanted to segregate at that level you would really have to start doing inline high-speed deep packet inspection; wouldn't you?
4322 COMMISSIONER KATZ: I don't know. Bell Canada has done it, so whatever they're doing, they've invested the money.
4323 And what I want to know is what would be involved for Telus to --
4324 THE CHAIRPERSON: You've been asked a specific request. Do it in your written form, don't waste any more time, okay. We're not going to get to the bottom of it in this exchange.
4325 So, in the interest of time, let's move on, okay.
4327 COMMISSIONER DENTON: Thank you, Mr. Chairman.
4328 Good morning.
4329 Mr. Hennessy, you've frequently been guilty of telling the truth in public and I would be interested in your view on the following question.
4330 Is it your view that there is, in essence, no price, no wholesale price that could ever be established which would make retail competition legitimate in your view?
4331 Is it the entire idea of setting a whole price whereby competitors can lease access illegitimate; in your view?
4332 MR. HENNESSY: Yeah, I think at this point that's probably fundamentally true, although I would think a price that reflected the total cost that we have to bear when we take the investment risk would be appropriate, but it's not -- that's not usually a traditional regulatory form of costing.
4333 COMMISSIONER DENTON: So then, well that's -- you seem to have opened up a broad exception to the principle you just enunciated.
4334 There might be a price at which such leasing would be acceptable?
4335 MR. HENNESSY: Well, yeah. The reason I -- and again I'm not suggesting -- you know, part of my position on wholesale has always been that, you know, as the Telecom Review Panel suggested we should be, you know, not really dealing with NGNs and moving much more to an ex post kind of environment.
4336 So, I would maintain, the CRTC should retain the jurisdiction to deal with issues of unjust discrimination and undue preference when it comes to dealing in the wholesale market, and just as somebody may argue that they have a need for our facilities, I would say that there are many parts of the country where we have a need for other people's facilities and there is probably a line where some people would say, well, that's a reasonable commercial market-based price versus that is, you know, something that's being proposed that's anti-competitive to, you know, really in effect deny access.
4337 So, yeah, I mean there are shades of grey.
4338 I think my principle is that we have to move away from an arbitrage thing where you're stuck protecting the margins that were set for various resellers' business and move to an environment where rates are much more reflective of commercial circumstances and ensure that, you know, there are incentives to continue to invest.
4339 And I think that is best done by ex post regulation, as the Panel proposed
4340 COMMISSIONER DENTON: Thank you.
4341 You've said -- one of your four principles on page 3 was:
"The Commission should take care not to adopt overly broad ex ante rules of general application that could do any of these four things." (As read)
4342 COMMISSIONER DENTON: The last point was:
"To reduce the incentives of carriers to make or unduly constrain their ability to recover the enormous capital investments required to build next generation telecommunications networks." (As read)
4343 COMMISSIONER DENTON: In your view, is next generation telecommunications networks a term of art; does this refer to an open Internet where you can innovate from the edge, or does next generation network for you connote some specialized term of art that needs to be unpacked and explained to us?
4344 MR. HENNESSY: Probably the latter. I would think the latter.
4345 MR. NEALE: Yeah. And I think what we are referring to is cutting edge technology invariably to increase capacity and to increase speed.
4346 So, for example, if we're talking on the wireless side, we refer to next generation as being HSPA plus, which is defined as 3.5G technology and then LTE which we define as fourth generation technology.
4347 So, these are technologies that are about to be deployed in roll-out mode. So, what we're saying is they're enhancements, they're genuinely order of magnitude, if we're lucky, in terms of capacity or increasing speed, or...
4348 COMMISSIONER DENTON: Well, one of the principal differences between, say, the PSDN and the cable system on the one hand and the Internet on the other is, in one the standards are closed proprietary standards, in the other there's an open standards process.
4349 And obviously most of the parties appearing here have favoured an open Internet in the sense of that innovation from the edge.
4350 So, it would be of great concern to us, one way or another, if the next generation networks meant closed proprietary standards.
4351 MR. NEALE: No, in fact, they are all based on these very large open standards organizations.
4352 So, it would be safe to say that next generation genuinely moved and embraced the open concept we describe.
4353 COMMISSIONER DENTON: Thank you.
4354 My next question is that most of the parties -- well, many of the parties before us have expressed the very greatest of concerns for what might be called application-specific traffic management measures and they have -- when I've probed them it's really come down to a question of, you know, the balance of power between creators and carriers and consumers is at stake here.
4355 So, my question for you then is, in terms of your principles, do you see that application-specific measures of traffic management are legitimate?
4356 MR. HENNESSY: I think we suggested they were, I think that's our position.
4357 We certainly reserve -- you know, we reserve the right to do that. We don't do it today.
4358 But are they legitimate? Yeah, I think as long as they don't materially degrade someone's ability to use the Internet and don't result in anti-competitive self dealing, they shouldn't be assumed to be prima facie illegitimate.
4359 COMMISSIONER DENTON: So, as long as they -- can you say that again, please?
4360 MR. HENNESSY: I think the -- you know, if the result is to materially degrade someone's service or to result in an anti-competitive self dealing, you know, undue preference in terms of vertical integration and sort of in the principal example, that's a bad thing.
4361 But all application management including, you know, going forward into the future with applications we may not define, should not be assumed to be prime facie illegitimate.
4362 MR. NEALE: And if I might give an example. Let's assume that the ability to provide a public safety alert system goes forward, then we see traffic management also having positive opportunities.
4363 So, in that particular case you might use traffic management to ensure that those high priority public alerts are always given absolute priority. There's an example of traffic management for the common good.
4364 So, that would be the reason why we don't think it would be appropriate to just exclude in general.
4365 MR. HENNESSY: The other example I might give of that is in the new media proceeding, some parties had suggested the idea of finding ways to manage Canadian content in a way that gave it preference or priority over other, you know, non-Canadian applications and without, you know, sort of saying whether or not that's a good or a bad thing.
4366 I think there's, you know, another example of where, you know, the use of prioritization or application management might or might not be a good thing.
4367 COMMISSIONER DENTON: Thank you.
4368 My final question concerns the interpretation of section 36. As you know it was carried forward pretty well in tact from the previous Act in 1993, and Mr. McTaggart has proposed that the use of that section was to prevent carriers from engaging in broadcasting in new businesses to exert their market power beyond where it was properly -- to expand their businesses beyond the proper zone of the carrier.
4369 Is it not also possible to consider that what was intended there was to keep common carriers common carriers and that it was not so much to prevent business expansion as to keep the nature of a carrier one that practises the minimal levels of discrimination necessary for the sake of its profitability and the public benefit?
4370 MR. HENNESSY: Yes, I agree.
4371 COMMISSIONER DENTON: Thank you.
4372 THE CHAIRPERSON: Thank you.
4373 One last question.
4374 When you comment on the OIC test in No. 3 you water it down, for instance, instead of saying least restrictive means to reach the objective, you put in more reasonable.
4375 Why do you feel that is necessary? Surely the aim of our OIC was to make sure you have an open Internet and so if you intrude, you intrude in the least --
4376 MR. HENNESSY: Because, you know, the -- and Craig can get into it, but I think it's pretty simple.
4377 When you use terms like "least restrictive" they're highly subjective, they require detailed assessment of one network management approach versus another that may not matter if the rest of the test, as we've suggested, can be narrowed and whatever you choose to do is reasonable.
4378 I think that is much more appropriate --
4379 THE CHAIRPERSON: Mr. Hennessy --
4380 MR. HENNESSY: -- or in line with your own jurisdiction and how you apply your test.
4381 THE CHAIRPERSON: No, it's not.
4382 Surely you will want to be as objective as possible. There are two criterias here, economic and technological, and on both of them you put reasonable and you suggest, what's the trade-off. When you put least intrusive, you have to demonstrate on each one of them that there isn't another way of doing it that's less --
4383 MR. HENNESSY: Exactly. So, that is hugely interventionist from a regulatory perspective and from the perspective of people sitting around trying to argue what is the least restrictive thing.
4384 You know, I guess I could reach back in my many years of various committees to suggest when you get lots of people around with different interests at a table trying to argue over whose network management practices are best, there is no easy answer and you get bogged down in an incredible amount of detail, which isn't necessary if, you know, on its face either approach looks reasonable.
4385 I think that's -- I think it's just really --
4386 THE CHAIRPERSON: May I suggest you rethink this, and should you change your mind and least restrictive remains the criteria you suggest how one measures least restrictive other than a more obscure standard such as reasonable, because reasonable in this context can mean a whole host of things.
4387 I know it is --
4388 MR. HENNESSY: I understand, Chairman, and you know, for the record, we find that -- you know, we find it objectionable. So, you know, we will certainly debate what least restrictive is, but we clearly do not like it.
4389 THE CHAIRPERSON: Okay.
4390 But -- I think we both would feel comfortable if we could have a measurable standard rather than a subjective standard, and least intrusive or reasonable or something. We hope it is objective, but if you could do it on a quantitative basis would probably be the easiest way of resolving it.
4391 MR. HENNESSY: With due respect, you know, I want to disagree on that because I just -- I don't think that the Commission wants to go down these measurability practices when it comes to engineering, I don't think it's a -- it's, you know, your choice at the end of the day but, you know, my warning that it's a very deep and complex rabbit hole that we would slide down.
4392 THE CHAIRPERSON: Well, whichever way you go, whatever cautions you have, why don't you elaborate on those in your follow up.
4393 MR. HENNESSY: We will put warning stickers on our final reply.
4394 THE CHAIRPERSON: Thank you.
4395 We'll take a five-minute break.
--- Upon recessing at 1013
--- Upon resuming at 1022
4396 LE PRÉSIDENT : O.K. Commençons, Madame.
4397 LA SECRÉTAIRE : Merci.
4398 Tel que mentionné hier, Cybersurf ne comparaîtra pas à l'audience aujourd'hui.
4399 J'inviterais donc maintenant Cogeco Cable Inc. à faire sa présentation.
4400 M. Yves Mayrand comparaît pour Cogeco Cable Inc. Veuillez nous présenter vos collègues, après quoi vous aurez 15 minutes pour faire votre présentation.
4401 M. MAYRAND : Merci, Madame la Secrétaire.
4402 Monsieur le Président, mesdames et messieurs les Conseillers, merci de nous donner l'occasion de vous présenter le point de vue de Cogeco Cable Canada à l'occasion de cette audience publique en matière de télécommunications concernant les pratiques de gestion du trafic Internet.
4403 Je suis Yves Mayrand, vice-président, Affaires d'entreprise.
4404 Chris MacFarlane, vice-président, Ingénierie corporative, est assis à ma gauche.
4405 Et à sa gauche, Mario Garceau, Directeur principal, Ingénierie corporative.
4406 Enfin, Michel Messier, Directeur, Affaires réglementaires, Télécommunications, est assis à ma droite.
4407 First, let me remind you how Cogeco is involved in the provision of Internet services.
4408 Cogeco now offers broadband connections to both residential and business customers in more than 94 percent of its cable footprint. We currently offer four residential packages with download speeds of up to 640 Kilobits, 3 Megabits, 10 and 16 Megabits per second.
4409 In addition, as required by the Commission under the authority of the Telecommunications Act, we offer a Third-Party Internet Access, or TPIA, service with equivalent speed tiers of Cogeco's High-Speed Internet service. This wholesale service enables Internet Service Providers to use our access facilities so that they may provide competitive retail Internet services of their own.
4410 Finally, as mentioned at the New Media hearing, we do not produce, aggregate or market programs for viewing through the Internet connection.
4411 In short, as an ISP, we strictly provide telecommunications services and act in all respects as a telecommunications common carrier for the provision of both retail and wholesale Internet access service.
4412 To this end, as any other ISP, Cogeco periodically upgrades its Internet cable network in order to meet the growth in bandwidth demand and to offer an Internet access service with a level of quality of service expected by consumers.
4413 Needless to say, the competitive pressures in the Internet market are such that no ISP can avoid upgrading its Internet network without running the risk of creating customer dissatisfaction and losing market share.
4414 That said, at a certain point in time, we faced no choice other than implementing traffic-shaping measures as part of our network management activities.
4415 The implementation of these measures was not an abitrary decision. Because peer-to-peer, or P2P, applications consume a disproportionate amount of bandwidth, mainly on the upstream, traffic-shaping measures were and are still required to relieve the specific network congestion problem created by this type of application and to ensure fair and sustainable usage of the available bandwidth among the end users on our Internet network.
4416 In fact, given the finite bandwidth available and the fact that we operate a shared network, it is impossible in practice to run the Internet access network embedded in the DOCSIS network without controlling the upstream side. This is a fundamental technical constraint that cannot be ignored.
4417 The key issue of this proceeding is to determine if the Internet Traffic Management practices used by the Canadian ISPs, mainly the traffic-shaping measures, are employed in accordance with the Act.
4418 In this regard, there is no doubt in our view that traffic-shaping or throttling measures implemented by Cogeco are used fully in accordance with the Act.
4419 These measures do not result in any manner in unjust discrimination, undue or unreasonable preference or advantage, nor do they have the effect of controlling the content or influencing the meaning of the telecommunications that we transmit.
4420 To be clear, the traffic-shaping tool used by Cogeco:
4421 - is not applied on the downstream but only on the upstream;
4422 - is carried out solely on P2P applications that are based on bulk transfer protocols, such as BitTorrent applications;
4423 - is not interfering with the content transmitted, which is always delivered without change to the recipient;
4424 - is not applied in a manner to block or disrupt the telecommunication but has only the effect of slowing down the transmission of P2P traffic;
4425 - is inspecting the header and payload of each packet to the minimum extent required in order to identify the specific signature of P2P protocols; and
4426 - is agnostic to any personal identifiers.
4427 In short, by limiting the bandwidth available on the upstream for P2P traffic through the use of traffic-throttling measures, Cogeco is merely slowing P2P traffic by reducing the data rates on all virtual access links established for the exchange of this type of traffic.
4428 Finally, while we recently implemented usage billing for additional usage in excess of the bandwidth limitations of Cogeco's Internet packages, we see no reason to remove traffic-shaping measures used within our network at this time.
4429 Of course, we hope that the implementation of a usage-based billing approach will change usage patterns of our customers in a manner that will allow better sharing of available network capacity in the near future. However, we are not convinced at this time that this pricing approach will be sufficient to alleviate the specific congestion problem created by P2P traffic or to prevent any other network problem of the same nature.
4430 As frontline network operators, we need to rely on equipment that enables us to intervene in a timely manner if such a congestion problem occurs.
4431 So for the foreseeable future, we plan to continue to rely on technical management tools to protect our network and ensure fair and manageable use among all Cogeco's retail and wholesale end users of our network.
4432 Let me now turn to the assumptions and key issues you have asked us to address in our oral remarks.
4433 The specific objective of the present hearing is the establishment of guidelines to determine what are acceptable ITM practices in Canada.
4434 As a general comment, we would simply say that acceptable ITM practices are those that do not infringe the statutory prohibitions under sections 27(2) and 36 of the Act, while unacceptable ITM practices are those that do, as determined by the Commission, however, through an ex post case-by-case examination process.
4435 We submit that the specific circumstances of the network operational context of each ISP should be taken into account before concluding that a given ITM practice used by a Canadian ISP is not acceptable.
4436 In Telecom Decision 2008-108, we note that the Commission has elaborated and used several criteria to deal with issues raised by the Bell ITM practice and, at the same time, has defined what is acceptable in the area of ITM practices.
4437 This set of criteria provides not only an appropriate analytical framework for the Commission to deal with issues raised by ITM practices but also clear guidelines for Canadian ISPs in this matter.
4438 In summary, the Commission indicated that traffic-shaping measures:
4439 - should be implemented in response to a "need;"
4440 - should represent "the only practical option that is technologically and economically suitable, at this time, for addressing congestion;"
4441 - should not provide an undue preference to the ISP applying these measures, such as securing sufficient bandwidth for its own services;
4442 - should be implemented in a manner such that equivalent treatment is applied to both retail and wholesale Internet service end users;
4443 - should not result in a reduction of competition;
4444 - should not involve blocking of any telecommunications;
4445 - should not impede these telecommunications to reach their intended recipients with their contents unchanged;
4446 - should not involve controlling the content while controlling the speed of telecommunications;
4447 - should not involve any editorial control over the content of the telecommunications; and
4448 - should not influence or alter the meaning or purpose of telecommunications.
4449 With respect to usage-based measures, the Commission also determined in Telecom Decision 2006-77 that cable carriers must provide equivalent treatment to both retail and wholesale Internet access service end users.
4450 In the absence of evidence on the record of this proceeding demonstrating the existence of problems with the ITM practices used by Canadian ISPs, we submit that these criteria remain appropriate and sufficient. They should be reaffirmed by the Commission as a result of this proceeding.
4451 Several alarmist speculations have been made with respect to the privacy issue related to the use of deep packet inspection or the DPI tool. Based on our own experience and the practical limitations of this tool, we can affirm that these privacy concerns are completely unfounded.
4452 As already mentioned, the traffic-shaping measures applied by Cogeco are agnostic to any personal identifiers. To be clear, we have no idea of the identity of customers subject to traffic-shaping. Since IP addresses are changed frequently through dynamic addressing, external intelligence systems would need to be implemented to obtain such information. We do not have such systems.
4453 While we capture some information through traffic-shaping, the information collected is stored for a very short period and is only used to validate overall traffic distribution on a sporadic basis.
4454 In other words, it is impossible for Cogeco to use this information or generate personal information to deploy personalized advertising or targeted marketing campaigns.
4455 In any event, if that ever became possible, we submit that such use of personal information would clearly be subject to the requirements of the PIPED Act. In short, we see no reason for additional rules to ensure that the protection of personal privacy at the present time are in order.
4456 By arguing that the current approach to ITM practices is anti-competitive, many ISPs took this proceeding as an opportunity to request unbundling of the Internet access service without the use of any traffic management technologies or usage-based measures.
4457 This self-serving request is not only beyond the scope of this proceeding but simply unrealistic with respect to the wholesale Internet access service provided by cable companies.
4458 The cable distribution network is a shared infrastructure. This means that there are no unique, physical, permanent or virtual transmission paths between a TPIA customer and its end users.
4459 To be clear, without the application of the same ITM practices on both TPIA and Cogeco's end users in the same node, there is no way to guarantee that the consumption of the end users of the TPIA customers would not create congestion or degrade the quality of service provided to Cogeco's end users and vice versa.
4460 Given this fundamental technical constraint, it would be counterproductive and even discriminatory with respect to Cogeco's HSI customers to not apply traffic shaping or usage-based measures to the TPIA end user traffic in the same manner as for Cogeco's retail customers.
4461 The equal treatment requirement adopted by the Commission remains the only appropriate way to apply section 27(2) of the Act in this context.
4462 With respect to the disclosure issue, we remain of the view that regulatory guidance is unnecessary in this matter.
4463 On the retail side, we share the view that market forces will continue to dictate the appropriate level of disclosure. Currently, as other ISPs, we are informing our customers through Cogeco's Acceptable Use Policy that we are monitoring, managing and protecting our Internet network against users' abuse that could affect other customers.
4464 Nevertheless, given the attention captured by this question throughout North America, we plan to revise and include in Cogeco's Acceptable Use Policy a dedicated section outlining our ITM practices by the end of this year.
4465 On the wholesale side, we see no reason to extend the notification requirements imposed on Bell to cable ISPs. The circumstances are indeed quite different.
4466 First, TPIA customers have never had to make adjustments to their networks as a result of the implementation of Cogeco's traffic-shaping measures.
4467 Secondly, Cogeco never received any complaints about these measures from its TPIA customers nor any requests for the purpose of informing their end customers.
4468 In conclusion, we maintain that no regulatory measures designed to regulate Canadian ITM practices are required at this time.
4469 We are also of the view that the criteria already elaborated by the Commission to deal with this matter provide sufficient guidelines to the Canadian ISPs to implement acceptable ITM practices.
4470 And we submit that the Commission should continue to rely on a case-by-case approach to ensure that ITM practices will, going forward, continue to be used by Canadian ISPs in accordance with the Act.
4471 Thank you for hearing us and it will be our pleasure to answer your questions.
4472 THE CHAIRPERSON: Thank you for your presentation. You should come and write our decision. Your summary of our decision is much more precise than the original decision, the way you outline it on page 4.
4473 But coming back to the substance, it seems to me that you are saying you are employing application-specific ITMP to deal with P2P but you are only applying it on the upstream side?
4474 MR. MAYRAND: That is correct.
4475 THE CHAIRPERSON: But surely if you apply it on the upstream side, it has an effect on the downstream because it is being uploaded for the use of some other users on the downstream. So surely by applying it, you are affecting both the upstream and downstream of that P2P application?
4476 MR. MacFARLANE: Formerly BitTorrent used to have a token-based system where it limited the amount of bandwidth that a customer could use in the downstream based on what it was uploading to the network but that is no longer the case in most applications of peer-to-peer.
4477 And in the case of cable, it has to be recognized that the upstream is an extremely limited resource that needs to be tightly managed and therefore --
4478 THE CHAIRPERSON: I understand that, but in your first statement you are saying that now limiting upstream does not have an effect on the downstream of BitTorrent or whatever P2P you were talking about.
4479 MR. MacFARLANE: It has a marginal effect. If another one of our customers happens to be downloading from one of our customers that they are currently uploading from, it can have a marginal impact but not a great impact.
4480 THE CHAIRPERSON: The implication being he may be downloading from people who are not your customers and therefore --
4481 MR. MacFARLANE: That is correct.
4482 THE CHAIRPERSON: I see. And are you doing this to all P2P or only to some P2P? I mean there are P2Ps which are extremely latency- or jitter-sensitive such as Skype or so, if I understand, or at least that is the evidence we heard. So are you doing this to all P2Ps or do you distinguish between P2Ps which you manage and which ones you don't manage?
4483 MR. MacFARLANE: We distinguish between file transfer peer-to-peer only and we leave some of the real-time peer-to-peer transfers alone, so Skype and others.
4484 THE CHAIRPERSON: And what do you say to the argument that we have heard from lots of people that by doing that, you in effect will have a discriminatory effect, that depending on where the customers are and what the mix is of the usage, et cetera, this may have an effect on certain applications or may certainly have an effect on certain ISPs who get the traffic from you wholesale because they will not be able to serve their customers as well if they have a preponderance of P2P customers or something like that?
4485 That is an argument we have heard here all week long, that the moment you are application-specific, whether you intend to or not, you are going to have a discriminatory effect.
4486 MR. MAYRAND: That goes, I think, to an issue that was briefly canvassed in the earlier presentation and the question of whether the telecom environment is an environment where there is an absolute bar on any discriminating factor in the way the network is managed in practical terms and I think the discussion led to the conclusion that certainly there have to be some distinctions drawn at some point in time in the operations of the network.
4487 Now, what we are talking about here is identifying certain specific situations where an application, by essence, or a set of applications, by essence, are designed to hoard capacity -- let's not kid ourselves, that is what they are about -- to hoard capacity on our network, which is a shared network, ought to be identified so that we don't have a situation that goes completely offhand and we have a huge spike or hockey stick congestion problem occurring on the network, to the detriment of all other users.
4488 THE CHAIRPERSON: We understand all that and of course there can be discrimination. Price discrimination is a perfect example. The Act talks about unjust discrimination and the question is how you determine unjust.
4489 Do you look only on -- in order to determine whether it is unjust, surely you have to look at the impact and the effect and not only as a way to measure. The measure of its face may be totally non-discriminatory but if the impact is overwhelmingly negative for certain small ISPs or so, then clearly don't we have to take that into consideration?
4490 MR. MAYRAND: Well, if I may --
4491 THE CHAIRPERSON: I mean, I am stipulating, I don't know whether it is obviously.
4492 MR. MAYRAND: I would submit to you that, you know, you have a body of practice under the section 27 requirement of the Act that you must be aware of, that all ISPs must be aware of, and there is obviously a process under that provision that may lead to findings of unjust discrimination.
4493 I am not going to agree with you that there is just a sole criterion of potential impact on some users as the sole driving criterion to conclude, irrespective of the network configuration, the management constraints and the particular facts of the case, that there is unjust discrimination.
4494 THE CHAIRPERSON: Well, that is not what I suggested. You know that.
4495 MR. MAYRAND: Okay.
4496 THE CHAIRPERSON: What I suggested was don't we have to take -- if there is evidence that in effect your measure which you are applying right now, peer-to-peer upstream you are throttling or you are slowing it down, has a disproportionate effect on your wholesale customers, then surely -- I am stipulating, I am not saying it does.
4497 But if evidence is produced on a certain category of your wholesale customers, would that not be something that we have to take into account in order to determine whether it is unjust discrimination or not?
4498 MR. MAYRAND: In the case of a specific complaint raising, among others, this issue, of course, the Commission would look at it and, of course, we would look at all the circumstances and not just that criterion.
4499 THE CHAIRPERSON: Okay.
4500 MR. MAYRAND: And I should add, by the way, because you referred to the evidentiary question, certainly it is absolutely clear to us that as part of this proceeding there is no evidence to suggest that our management practice constitutes, ipso facto, a discriminatory measure that violates the requirements of the Act. We content that it is absolutely not the case.
4501 THE CHAIRPERSON: One last question before I pass you on to my colleague.
4502 As I understand it, you apply this slowdown or throttling on a 24/7 basis?
4503 MR. MAYRAND: That is correct.
4504 THE CHAIRPERSON: Everybody who has come before us said such measures should be as narrow, as tailored, as targeted as possible. Why would you not apply this only if there is actual congestion? Why do you have to apply it ex ante across the board 24/7?
4505 MR. MAYRAND: I think the concern -- I believe we have tried to explain that in our written submissions -- is that if we are deploying this tool, why should we provide any opportunity for bypassing the management tool that we have deployed in certain time periods?
4506 The concern is that if we limit it to specific time periods, there might be ways in which we find, you know, spikes in traffic and congestion problems occurring at time periods where we assume that there would be none.
4507 So unless there is an overriding concern that the measure in itself creates an issue within the four corners of the Act -- and I should add, within the four corners of the Act as interpreted under the Police Direction that you have -- there really wouldn't be any reason to say that it should be on at certain time periods of the day and off at others.
4508 THE CHAIRPERSON: But in your page 4, you say in your very first point:
"In summary, the Commission indicated that traffic-shaping measures:
- should be implemented in response to a 'need;' "
4509 You don't have a 24/7 need. You have a need to respond because there may be congestion at certain times.
4510 MR. MAYRAND: Well, how is the need defined? Is it defined on a granular period of time over the 24-hour schedule or are we talking about the need that arises out of experience with congestion and the risk of congestion?
4511 I would argue with you, sir, that the need can be looked at on a minute-by-minute basis in practical terms.
4512 THE CHAIRPERSON: We may actually have that argument should there be a complaint.
4513 Anyway, I will pass you on. Tim...?
4514 COMMISSIONER DENTON: Good morning, gentlemen. Thank you for a nice, tight presentation of your issues and concerns.
4515 I want to talk just briefly about economic measures. It is because you don't have enough evidence yet as to their effect on reducing congestion that you continue to rely on technical measures as much as you do or is there a state of dissatisfaction with economic measures to control congestion?
4516 MR. MAYRAND: The answer to your question is twofold.
4517 First of all, as we said, we are just beginning to apply usage billing. So we don't have obviously a dataset that allows us to measure the impact on that usage billing in terms of the traffic consumption patterns and the congestion issues that we face on the network.
4518 The second reason is that we are not convinced from a general and a specific technical standpoint that usage billing, or let's call them economic measures leading to self-discipline by users, will necessarily cover all situations, contingencies or new developments on the Internet.
4519 I think we are very specific in our filings in response to interrogs as to the history of what happened with our experience in dealing with unexpected and, frankly, very abnormal traffic usage patterns that led us to apply in succession various measures to maintain the integrity of our network and a reliable and fair service to all users.
4520 We are concerned that if we lose the ability to apply any measure other than economic control tools that we can still have very punctual but very hard frontline issues to deal with and that it would be highly, highly impractical to go back to the Commission, you know, to ask for oversight or an exception to deal with an emergency situation.
4521 I mean, frankly, this whole discussion turns around what is reasonable for people who have the responsibility of investing huge amounts of money in a shared network facility and have the day-to-day frontline responsibility to run it efficiently and reasonably in the interest of all its users. These people have to make frontline decisions quickly when a problem arises.
4522 I cannot stress enough the fact that, you know, I think the Commission would be going down the wrong track if it were headed towards some kind of ex ante management of the network. The Act provides the basic rules, you have provided additional guidance in already a prior decision, which, as we say, you can very well generalize in the circumstances, and that should be amply sufficient to deal with actual issues that arise.
4523 COMMISSIONER DENTON: So then you are suggesting that our previous guidance might be instituted as a policy and that would suffice for your consideration of the form in which ex ante guidance should take place?
4524 MR. MAYRAND: I suggest that the guidelines are what this whole proceeding is about and I suggest that then to see whether there is actually a real problem with a particular management practice adopted by a particular ISP for a particular network configuration, that is not something that you can deal with ex ante. That is something you have to look at because it is a question of mixed fact and law and you have to look at the specifics of the case on a complaint.
4525 COMMISSIONER DENTON: Point taken.
4526 In response to one of the interrogatories -- it may have been taken up, I just want to get your understanding of it -- you said that you apply Internet traffic management practices at all times of the day.
4527 Why are they implemented at all times of the day rather than in response to congestion as it actually occurs?
4528 MR. MAYRAND: Well, I thought I tried to cover that question from the Chair but I may restate it briefly.
4529 My point there was that in our view, if the management tool that we have deployed meets the requirements of the Act and eventually the requirements that are apparent from the guidelines that the Commission may adopt in applying these legal requirements of the Act, why should we provide further complexities and/or uncertainties in using the tool on and off, depending on the time of the day? Certainly we don't see what purpose would be served in making that distinction.
4530 COMMISSIONER DENTON: Thank you.
4531 You pointed out in one of your reply paragraphs as well as in your messages this morning that some of the data is retained for a "very short period" and you say it's mainly used to validate traffic distribution.
4532 How long is that very short period and what are the other uses for this information?
4533 MR. MacFARLANE: Generally we are keeping traffic patterns for no more than a period of probably 30 days. What we are using it for is we are just characterizing traffic patterns for traffic engineering overall. So in a matter of engineering our backbone or matter of engineering our network as a whole, we use this data to characterize what is on our network and what is the proportion of use and therefore the characteristics that we need to engineer.
4534 COMMISSIONER DENTON: Thank you.
4535 Perhaps you heard when Primus was testifying yesterday that it was using an application agnostic method of traffic management.
4537 MR. MAYRAND: Maybe I can start answering that question. Chris might have something to add.
4538 You know, it's very difficult for us to comment on specific ITMPs used by another ISP in another set of circumstances, particularly a very different network configuration. What we can tell you is that in the case of Cogeco Cable's plant it is a fully shared plant with a very specific constraint on the upstream that we have to watch with a hawk's eye.
4539 COMMISSIONER DENTON: This is the TPIA's technical capacities? Is that the source of this concern to watch it so carefully?
4540 MR. MAYRAND: No, although if you bring in the TPIA consideration I think we have been very, very clear that we don't propose to make any differences or to cause any hardship on third-party suppliers that we deal with through the TPIA, if that's where you're getting at.
4541 COMMISSIONER DENTON: No, actually my question is misplaced, it should be just more general. What is the need for you to watch with a hawk's eye the upstream traffic? I just need to understand better.
4542 MR. MAYRAND: Well, the need -- and Chris may expand on that -- but the need is directly and specifically related to the configuration, technical configuration and parameters of the shared broadband plant that we operate and the DOCSIS platform.
4543 Chris, I don't know if you want to add to that.
4544 MR. MacFARLANE: To clarify, so our upstream traffic -- our upstream network is far more limited than our downstream, number one.
4545 Number two, in our submission, in the confidential section of our submission we did elaborate that there are technologies that we are looking at. So you can refer to that submission to take a more detailed look at other things that we are looking at that will do much the same to what Primus was alluding to.
4546 The third issue I would like to put on the record is that we manage this application at this particular point in time because of its inordinate capability to consume the resources on that upstream. And note that we don't do it on the downstream because we don't view that resource to be as scarce.
4547 COMMISSIONER DENTON: Thank you. Those are my questions.
4548 THE CHAIRPERSON: Len...?
4549 COMMISSIONER KATZ: I have no questions, Mr. Chairman.
4550 THE CHAIRPERSON: Candice...?
4551 COMMISSIONER MOLNAR: Thank you.
4552 I would like to ask you about your disclosure requirements. You have said that on the wholesale side -- and this is simply in your opening remarks:
"We see no reason to extend the notification requirements imposed on Bell to cable ISPs. The circumstances are quite different." (As read)
4553 Could you just expand on that? I mean the fact that you have not received complaints, I'm trying to understand why your TPIA customers should not be made aware in a timely manner of what the activities you do or the ITMPs that you apply to your network, what impact that would have on their end customers so their end customers can be informed in a timely manner. How is that circumstance different?
4554 MR. MAYRAND: Well, first of all, let me just clarify that from the end-user customer perspective we provide whatever information, notification or terms we provide to our own customers. We don't do that for people using the TPIA platform. They deal with their customers and they provide whatever notification or requirements they feel is appropriate.
4555 I would venture to say that it is perfectly acceptable and it would be perfectly in order for TPIA service providers to indicate the same type of notification that we provide to our own customers, but in the end, you know, it's their responsibility.
4556 Now, in terms of the knowledge of the service provider that uses our platform under TPIA, first of all, it's very, very clear with the amount of publicity that this issue has had that those suppliers are very much aware of the tools that we use and there is, you know, no uncertainty as to what these tools are about.
4557 Second, I would tend to be concerned along your line of questioning if there was a real risk of unfair or unjust treatment of our TPIA customers, but that is not the case. We provide tariffed service under standard tariff terms and we provide the service exactly transparently on the same basis as we provide service to our own customers.
4558 COMMISSIONER MOLNAR: Okay. Let me re-ask the question. One of the requirements that fell out of the CAIP decision was advance notification to wholesale customers such that they can understand and be prepared to respond to questions from their end customers and you are saying that is not appropriate. Why is that not appropriate?
4559 MR. MESSIER: The way that we understand the reason why Bell was required to provide notification to the ISP is because the ISP complained that the fact that Bell implemented -- actually being on their network has an impact on their network, the way that they provide Internet services, and they had a request from their customer to explain why, what has happened with that.
4560 So the correlation between the two reasons is not there in our case because the impact of the traffic shaping on our TPIA customers, there is no impact on their network. So that also explains why probably they received no request for information or received no complaint about the impact of the traffic shaping that we implement on our network.
4561 So given the fact that the TPIA customer was not impacted by the traffic shaping that we implement on our network, the two aspects are not there in our case.
4562 COMMISSIONER MOLNAR: The TPIA customer was not impacted by the traffic shaping you did on your network and yet you say you do the same traffic shaping on all traffic regardless of whether it's for your end customers or whether it's TPIA traffic.
4563 How is it that they were not impacted by the traffic shaping you did on your network? You know, if I'm misunderstanding, just let me know.
4564 MR. MAYRAND: I think they were not impacted any differently, okay, and there were, in fact -- and that is a huge practical difference -- there were no complaints or even, I think, any inquiries from our wholesale clients on our management tools.
4565 COMMISSIONER MOLNAR: Okay. Let's step back from history just to a matter of principle or policy.
4566 Would you think it is unreasonable to provide TPIA customers with advance notice such that they are ready and prepared to respond to their end customers when the traffic -- when the nature of the traffic is changed due to ITMP practices?
4567 MR. MAYRAND: I would think that, you know, whatever notification we provide to our end users, if that is the gist of your question, and that our wholesale customers for some reason don't get to read them, we can certainly forward them to them.
4568 As I mentioned, the communication between TPIA and their end customers is for them to manage, but certainly if the issue is --
4569 COMMISSIONER MOLNAR: Advance notification.
4570 MR. MAYRAND: -- the same ground-level information as we provide to our own customers, I have no problem with that.
4571 COMMISSIONER MOLNAR: Okay. That's all, thank you.
4572 THE CHAIRPERSON: Okay. Thank you very much for your presentation. I think we will take a five-minute break before we do the next one.
--- Upon recessing at 1103
--- Upon resuming at 1014
4573 LE PRÉSIDENT: Bon. Commençons, Madame la Secrétaire.
4574 LA SECRÉTAIRE: Merci, Monsieur le Président.
4575 I would now invite Barrett Xplore Inc. to make its presentation. Please introduce yourself for the record and proceed with your presentation.
4576 Mme PRUDHAM : Merci. Thank you.
4577 Good morning, Mr. Chairman and Commissioners. My name is C.J. Prudham, I am the Vice-President, General Counsel of Barrett Xplore Inc. and Barrett Broadband Networks Inc. Thank you for the opportunity to be here today. Mindful that we are the last presentation between you and your weekend I will try to be as brief as possible.
4578 MS PRUDHAM: I would like to take this time to focus on traffic management from the unique perspective of the smaller ISP that serves rural and remote parts of Canada. In doing so, I would like to highlight for the Commission the significant capacity issues and congestion problems that are particular to our service.
4579 I will begin by providing a brief overview of the ISP business in which we are engaged and then identify the reasons why an ISP like Barrett requires the ability to manage traffic on its network.
4580 I will close by addressing very briefly the six specific issues that were set out in the Commission's letter of June 5th, which were reiterated by the Chairman in his opening remarks.
4581 Barrett Xplore Inc. serve customers in over 500 rural and remote communities across Canada. We are in every province and every territory. We provide Internet access services under the Xplornet brand name in near-urban, rural and remote parts of Canada using fixed wireless and satellite technologies.
4582 Barrett has deployed fixed wireless networks in areas where the population density is sufficient to support the capital cost of building such a network. In less dense areas we use satellites to provide broadband Internet access to our customers at affordable prices.
4583 From its inception, Barrett has invested over $140 million of private capital in extending broadband satellite and fixed wireless networks to individual consumers and small businesses across Canada that are not served by terrestrial cable, wireline or fiber networks.
4584 While the evidence presented by ISPs in this proceeding have demonstrated the use of traffic management techniques is important to all ISPs to alleviate congestion and to ensure that all customers have the equivalent ability to access content throughout the Internet, traffic management is vital to a company like Barrett which relies on satellite and fixed wireless technologies to provide Internet across rural and remote communities.
4585 Some interveners have suggested that the issues of network congestion and capacity limitations can be addressed simply by increasing the amount of money that an ISP spends on network upgrades. In essence, the interveners are saying that an ISP should be required to spend whatever is necessary to expand capacity to accommodate bandwidth-hogging applications such as peer-to0peer file-sharing, applications like BitTorrent.
4586 Such a free-spending approach to network congestion is simply not realistic and perhaps even preposterous. If that approach were adopted by Barrett, our company would very quickly find itself in financial distress and it would threaten our ability to provide broadband services to Canadians living in rural communities.
4587 Successive federal and provincial governments have actively pursued a policy of encouraging the private sector to invest in broadband technologies that will extend the Internet to rural and unserved communities.
4588 Barrett has responded to this policy by creating a rural broadband network consisting of fixed wireless and satellite technologies that today, as I said, is in more than 500 communities across the country.
4589 To expand the capacity on a satellite-based or fixed wireless network is not simple or inexpensive. As our friends at TELUS very kindly acknowledged this morning, it is extremely expensive and logistically challenging to expand the capacity of a fixed wireless portion of our network.
4590 As just one example, the broadband facilities that Barrett uses to backhaul traffic in rural and remote communities is in short supply and barely sufficient to meet our current needs. As the amount of Internet traffic transmitted on our network has increased exponentially in recent years, the backhaul facilities that have been used to transmit the traffic have been stretched to the limit.
4591 The time and cost of expanding the backhaul capability through new fiber builds in rural and remote regions of the country is not something that a small ISP like Barrett can assume.
4592 As challenging as it is to expand fixed wireless networks, as I'm sure you can imagine, expanding satellite capacity is positively daunting. When you utilize satellites to deliver broadband and Internet services there is an enormous cost, both in terms of time and money associated with expanding that network.
4593 Launching a new satellite and building ground stations takes years to complete and hundreds of millions of dollars to construct and maintain. It's also, as you might guess, a little lumpy. You can't just add little bits and pieces, it is satellite by satellite.
4594 As Barrett has indicated previously, the capacity of a satellite is, in large measure, determined at launch. While it is possible to increase the capacity by enhancing the electronics on the ground, the satellite's design has inherent limitations, which means that the capacity cannot be significantly increased during the space of its life, which is typically 10 to 15 years. Therefore, any upgrade on the satellite network capacity is extremely expensive and a long-term process.
4595 In light of the enormous costs associated with expanding both satellite and fixed wireless networks, Barrett must constantly make trade-offs between growing its business geographically to unserved areas, expanding existing capacity and developing its backbone network in an orderly fashion, while maintaining a retail price that can reasonably be charged to consumers for their Internet services.
4596 If the retail price becomes too high because of the exorbitant costs associated with increasing capacity, then our business will fail, along with the government's objective of expanding broadband Internet access to rural and remote parts of Canada.
4597 We believe that each ISP should be permitted to adopt traffic management tools that are consistent with its particular circumstance, network configurations and business models. That way, all ISPs, but especially smaller ones like Barrett that serve rural and remote communities, will be able to employ appropriate tools to manage traffic at peak usage times to ensure the best possible quality of service for all customers. This will enable us to provide the rural and remote customer with broadband services in a cost-efficient manner without threatening our financial well-being.
4598 It is our submission, therefore, that the Commission should establish a traffic management regime that is flexible and adaptable and that recognizes the unique circumstances of each ISP.
4599 In our written comments we have outlined a principled approach to traffic management that emphasizes transparency, non-discriminatory treatment between like traffic and non-interference with content. By adopting these three principles the Commission will address many of the concerns identified by the interested parties in these proceedings without imposing undue hardship or excessive costs on the ISPs.
4600 Although some interveners have suggested that carriers are discriminating against certain types of applications such as BitTorrent or other peer-to-peer bulk file transfer applications by restricting use at peak times, in response to that, Barrett would offer the following points.
4601 First, since existing capacity is finite, the network cannot always carry all the traffic presented to it at an acceptable speed. Simple laws of physics.
4602 Second, from a business perspective, Barrett cannot cut off hundreds or thousands of its customers from its service in order to carry the excessive volumes of traffic that may be presented by a few of its customers at any given moment in time.
4603 To discriminate against bandwidth-hogging applications is not unjust or unreasonable, we would submit, under section 27(2) of the Telecommunications Act, it is an entirely justifiable response to the use of the shared networks by a large customer base.
4604 Moreover, traffic management technologies used in these circumstances are not directed at individual users or individual applications, they are directed at all excessive use of bandwidth at peak periods and are applied without discrimination to consumers who subscribe to a given service level.
4605 As long as customers are made aware of the service limitations when they subscribe to the service, these traffic management tools are both reasonable and necessary. Without these tools affordable services would not be viable for the rest of the consumers and there would be no business case for extending into rural and remote parts of Canada.
4606 If I may, now I will very briefly address the six issues that were highlighted in the June 5th letter.
4607 The first issue in that set of questions related to acceptable Internet traffic management practices.
4608 As noticed above, Barrett believes that each ISP should be permitted to adopt and implement traffic management practices that are appropriate in the circumstances.
4609 In our view there is no one-size-fits-all traffic management technique that would be appropriate to all ISPs. Each ISP has a different network configuration, distinct capacity issues and unique set of consumer demands. ISPs must therefore be given a very broad discretion to address traffic congestion problems in a manner that makes sense in their particular circumstance.
4610 In our view, the only criteria the Commission should use to determine whether a particular traffic management practice is acceptable is to assess whether the use of the technique complies with the three principles highlighted earlier: transparency, non-discriminatory treatment and non-interference with the meaning of content.
4611 The second issue asked us to address ISP disclosure.
4612 Throughout this preceding Barrett has been an advocate for transparency. We believe that each ISP should provide notice to its customers of traffic management practices. However, this does not necessarily mean setting forth in great detail in the contract what exactly those policies are.
4613 Barrett's approach is to advise its customers there are limitations on the use of their service and to direct them to a website where there is additional information about our acceptable use policies and our traffic management policies. Disclosing this information is consumer-friendly and will, we believe, provide customers with the tools they need to make informed choices and adjust their behaviour to limit the impact of traffic management practices.
4614 The third issue relates to privacy.
4615 In Barrett's view the traffic management practices do not raise privacy concerns. We do not collect, use or disclose personal information when employing traffic management techniques like deep packet inspection. As other ISPs have indicated in the materials they have filed, they have said the same thing. There is simply no evidence of personal privacy issues here.
4616 The wholesale services issue, which was the fourth issue raised by the Commission, we very simply suggest, in our view, traffic management is an issue that is best left to commercial negotiations between the parties.
4617 In respect of the mobile carrier, we would simply suggest that, as we have previously said, the Commission adopt a flexible approach that allows ISPs, including the mobile carriers, the ability to adopt traffic management practices that best fit the configuration of their network and their capacity demands.
4618 Finally, the Commission asked about the scope of section 36 of the Telecommunications Act.
4619 In our view, the analytical framework the Commission should adopt with respect to section 36 is twofold.
4620 First, a party that alleges that a certain traffic management practice contravenes section 36 should be required to demonstrate that the ISP has in fact exercised control over the content of the telecommunications or influenced the meaning or purpose of the telecommunications.
4621 It is only after the actual burden has been met that the ISP should have any obligation to defend its practice or demonstrate that controlling the content or influencing the meaning of the telecommunications is appropriate or necessary in the circumstances.
4622 The use of traffic management practices does not, as a general matter, offend section 36. The techniques that Barrett and other ISPs use, which include specialized technology like DPI and non-technical measures like bandwidth caps and usage charges, do not allow ISPs to control the nature of the content distributed over the networks, nor do these techniques enable ISPs to influence the meaning or purpose of the content.
4623 Barrett appreciates this opportunity to present our views here today and would be happy to address any questions you might have. Thank you.
4624 THE CHAIRPERSON: Thank you very much for your presentation.
4625 When I visited you in Wakefield, New Brunswick I was very impressed with your operation and your CEO told me that any place in Canada that can be served by Barrett. That still applies today, doesn't it?
4626 MS PRUDHAM: Yes, sir. We can reach you if you are anywhere in Canada, except perhaps under a rock. The satellite doesn't go through rock too well.
4627 THE CHAIRPERSON: Right. And is there a great difference in price in terms of the service, whether I -- I mean in terms of the installation and getting set up? I mean assuming that it's an individual service, like at my cottage, which is about 50 km from Ottawa, no big deal, I would think. If I am way up in Moosonee or something like that, it might be more expensive. How do you price this?
4628 MS PRUDHAM: We have three components to our price. One is the equipment, one is the actual cost of the service and the third component is the installation cost. So in some provinces we do have a flat fee for the installation regardless of where you are located in the province. In other areas, particularly in the Far North, for example, unfortunately, installation charges are higher.
4629 THE CHAIRPERSON: Now, we are not there and hopefully not soon, but I think it is foreseeable that at some point in the future there will be a basic service obligation saying that all Canadians are entitled to high-speed Internet access, et cetera.
4630 Would you have the capacity to deal with people in remote areas who are not reachable by present networks, be they cable or telephone?
4631 MS PRUDHAM: Yes. As I indicated, literally through our satellite network we have coverage in every part of Canada.
4632 THE CHAIRPERSON: I know the coverage. The capacity is what I was referring to.
4633 MS PRUDHAM: Capacity issues are a challenge for us. The Far North, which everyone would think would be the most difficult to reach, we actually have a fair amount of capacity because, needless to say, there aren't a lot of people up there.
4634 THE CHAIRPERSON: Yes.
4635 MS PRUDHAM: We have greater capacity issues obviously as you come much further south and hit the folks in like Southern Ontario and Southern Québec. So we are working towards achieving capacity that would allow us to serve at least double or more than what we have today but there are some constraints in the immediate future.
4636 THE CHAIRPERSON: Now, the various people appearing before us suggested we use an analytical approach to look at ITMPs on an ex post facto basis and determine in accordance with section 27(2) whether there would be unjust discrimination or not.
4637 Is there anything about providing access via satellite that is different or that requires different consideration or a different approach than if you look at it by wireless -- normal wireless or by fixed wireline?
4638 MS PRUDHAM: I think one of the key differences is the number of customers potentially impacted. In a wireline or fixed wireless situation, you have a finite number of people going either off that particular node or off the access point tower, versus in the satellite circumstance you are talking about how many people are within the beam of the satellite and there are significantly more people per beam than there are typically off of a fixed wireless tower, for example.
4639 So if you have one particular person who is using a peer-to-peer application that is consuming all of the available capacity at that precise second, the number of people who are impacted on the satellite could be much, much larger.
4640 THE CHAIRPERSON: Okay, thank you.
4641 Candice, you have some questions?
4642 COMMISSIONER MOLNAR: Yes. And welcome.
4643 MS PRUDHAM: Thank you.
4644 COMMISSIONER MOLNAR: I, too, am mindful that we are the last group here.
4645 I would like to begin just by understanding a little more about the ITMPs within your network. You had put in your evidence, in your submission, the fact that you have three different -- you use the Telesat satellite, HughesNet as well as fixed wireless.
4646 Can you tell me where within those three different scenarios that Barrett Xplore itself actually controls the ITMPs?
4647 MS PRUDHAM: At this precise moment, at no point. On the fixed wireless network at the present time we do not have, as we said in our submission, any form of traffic management. So that is obviously a concern, an issue for us at this time.
4648 Within the satellites it is controlled by the satellite operators. So Telesat, as we indicated, has their policies and procedures and they control those. Hughes, likewise, controls theirs.
4649 COMMISSIONER MOLNAR: And just to be clear, ITMPs is quite a broad term and we have talked here about there being economic measures and technology measures. So you control the economic measures but the technology measures are based on the satellite providers; is that right?
4650 MS PRUDHAM: Yes. I'm sorry, I focused on the technology side. Yes, we control the economics, obviously.
4651 COMMISSIONER MOLNAR: Okay. Are you aware if there are any particular technology measures that are unique to satellite or work best with the satellite technology that don't work in the wireless world -- or in the wireline world or otherwise? Anything particular related to the technology?
4652 MS PRUDHAM: I'm sorry, I'm really not qualified as an engineer and certainly not as a satellite engineer, unfortunately, to answer that question. I'm afraid I can't help you.
4653 COMMISSIONER MOLNAR: Okay. The ITMPs that are in place today, are they 24/7? Is there ongoing technology -- how do I use the word? Are the controls that are in place today applied on a 24/7 basis or are they based on actual congestion?
4654 MS PRUDHAM: They do exist throughout the 24-hour period. They are different, as we indicated in our response, at different times, based on the level of peak versus off-peak traffic, which is a little complicated based on where the gateways are located. So it's a moving time zone depending on the gateway that is serving at the relevant time period. So there are, if you like, slightly more stringent controls at peak times and less stringent at off-peak times.
4655 COMMISSIONER MOLNAR: Okay. And just to be clear, to what extent do you have influence over what those traffic management policies are?
4656 MS PRUDHAM: Again, as we indicated, we have worked very closely with both Telesat and Hughes on their policies and we have certainly had influence on them. Ultimately it is their technology, but they have been receptive to input.
4657 COMMISSIONER MOLNAR: So you input, but ultimately don't control the decisions as to at what periods of time ITMPs are deployed or what technologies are deployed?
4658 MS PRUDHAM: That's right, in large part because the satellites are shared. So for example if you look at the Telesat Anik F2 as an example, while we have the majority of the capacity available for Ka-band use on that satellite, it is not 100 percent, there are others. So in fairness I don't think they could give us the final say on those matters.
4659 COMMISSIONER MOLNAR: As essentially a reseller of their services, do you have any issues or concerns regarding disclosure that they provide or what your requirements are as an end-user provider as it regards disclosure of their ITMP practices to you?
4660 MS PRUDHAM: I think they have been very open with us. As we said, we think it is appropriate to leave these sort of things to commercial negotiation because that has obviously been part of our discussions with them. So to date they have been forthright in their disclosure.
4661 There are issues. As you may have seen in our original materials, we didn't have a lot of information at the time we filed the responses to the interrogatories on the Hughes satellite because it was a new satellite and the systems weren't fully deployed at that time but their engineers have certainly been very helpful.
4662 COMMISSIONER MOLNAR: So you have never been in an issue or problem where your end customers are asking about the effects on their end-user experience that have been caused by your upstream provider?
4663 MS PRUDHAM: No, not --
4664 COMMISSIONER MOLNAR: No, not a concern.
4665 MS PRUDHAM: -- not without our knowledge of what was going on.
4666 COMMISSIONER MOLNAR: No concerns about having advance notification of changes to their traffic management problems or practices or anything like that?
4667 MS PRUDHAM: They have always given us advance warning, so it hasn't been an issue.
4668 COMMISSIONER MOLNAR: It hasn't been an issue. Okay.
4669 I want to ask you about your statement -- and I'm just going to find it here.
4670 COMMISSIONER MOLNAR: Okay. Well, I know you had made it in your opening statements but I have tagged it in your response.
4671 MS PRUDHAM: Okay.
4672 COMMISSIONER MOLNAR: So I will go to there because I have it tagged. You talk about:
"... provided that traffic management tools are applied in an equivalent manner to all users and that all like traffic is treated alike and that there is no attempt to control the nature, there is no basis for suggesting that the use of traffic management tools violates the Act." (As read)
4673 MS PRUDHAM: M'hmm.
4674 COMMISSIONER MOLNAR: We have had some parties come before us to say that if you unjustly discriminate against an application or a protocol that too can violate the Act. Do you have any comments on that?
4675 MS PRUDHAM: Well, interesting. It is perhaps not a view I share in terms of the interpretation of section 27(2). It does specifically refer to unjust discrimination against people. Applications are not people. And if you look at any given user I'm sure all of you use multiple different applications every day. Our view is that we should treat all users the same, and we do.
4676 If there is a specific application that is causing a problem, then that may need to be addressed. So I guess the answer is I don't view the sections of the Act as applying to things, like to applications versus people.
4677 COMMISSIONER MOLNAR: To application providers?
4678 MS PRUDHAM: Well, not to providers. I'm not suggesting that. I am specifically talking about like a BitTorrent application which is fundamentally designed to consume the network.
4679 COMMISSIONER MOLNAR: Okay. One final question and it relates to disclosure.
4680 Once again, you brought up in your comments that you believe that transparency and disclosure are fundamentally important and your approach is to advise customers there are limitations on the use of their service and direct them to a website where there is additional information both on Barrett's acceptable use policy and its traffic management policies.
4681 I did go to your website to take a look at that and while it does say that customers may experience temporary throughput limitations, it doesn't say anything about the type of traffic that is affected or how it is affected.
4682 MS PRUDHAM: M'hmm.
4683 COMMISSIONER MOLNAR: Can you explain why those sorts of details are omitted? Don't you believe that that is something that customers should understand, what types of traffic is affected?
4684 MS PRUDHAM: I think that's primarily because it's not necessarily a specific type of traffic that is being affected. In our particular case, we are not being application-specific. It is about the total throughput that the customer is attempting to move through the pipe at that precise moment.
4685 COMMISSIONER MOLNAR: So you don't target any applications or protocols?
4686 MS PRUDHAM: No. It is total amount.
4687 COMMISSIONER MOLNAR: Okay. Thank you. Those are my questions.
4688 THE CHAIRPERSON: Thank you.
4689 Suzanne, you had a question?
4690 COMMISSIONER LAMARRE: Yes. Merci, Monsieur le Président.
4691 On the privacy issue, on page 6 of your presentation you do state that Barrett does not collect, use or disclose personal information when employing traffic management techniques like deep packet inspection.
4692 So unlike Cogeco this morning who said that they were keeping information for 30 days, am I to understand that you dispose of the information as you go, that the information is actually used to manage the traffic, but once the traffic has been managed whatever information was looked at is actually deleted from the system?
4693 MS PRUDHAM: I think there is a whole pile of things combined in that. A lot of our data or the vast majority of it is aggregated data, so there isn't anything specific. It is all those cute little charts and that that you may have seen that show that there is "X" number of users on the network and they are using the following types of applications or that the total traffic moving through is a specific amount.
4694 COMMISSIONER LAMARRE: Okay. So you do keep aggregated data though?
4695 MS PRUDHAM: We very definitely have aggregated data.
4696 COMMISSIONER LAMARRE: Okay.
4697 MS PRUDHAM: With respect to individual data, if the customer triggers the specific limit, that would be recorded on their file. So we will know they have triggered the limit.
4698 COMMISSIONER LAMARRE: Okay. Thank you.
4699 THE CHAIRPERSON: Before I let you go, I just looked at your website and you have something called a fair access policy --
4700 MS PRUDHAM: Yes.
4701 THE CHAIRPERSON: -- which basically says you will slow down excessive users up to 50 percent. The closing sentence says:
"The Fair Access Policy (FAP) is straightforward. Based on our analysis of usage data, Xplornet has established a download data usage threshold well above the maximum typical usage rate. When a customer exhibits patterns of system usage which exceed that threshold for an extended period of time, the FAP may temporarily limit that subscriber's throughput to ensure the integrity of the system for all subscribers."
4702 Which it seems to me is your -- unless I misunderstand it also, it's very clear, you will slow down -- if I am in a peer-to-peer download and go over the average usage you will slow me down until such time that the network has the capacity to accommodate my excessive use, if I understand that correctly.
4703 MS PRUDHAM: Right. It's a little different between the two different satellites, but to use the more -- most customers on the Telesat one, it resets at the end of the hour.
4704 THE CHAIRPERSON: It resets at the end of the hour.
4705 MS PRUDHAM: Yes. So it does slow you down so that everyone sort of evens out and at the end of the hour it resets you and you start over again.
4706 THE CHAIRPERSON: But the net effect is basically that I will be slowed down during the peak hours, I assume, and in non-peak hours there is no slowdown?
4707 MS PRUDHAM: As long as you are not over the -- there is a limit still in the non-peak hours, again because you can't use the whole beam, but a much higher threshold.
4708 THE CHAIRPERSON: Okay. Thank you very much for your presentation.
4709 MS PRUDHAM: Thank you.
4710 LE PRÉSIDENT : Madame la Secrétaire, je crois que c'est tout pour aujourd'hui. À quelle heure est-ce qu'on va recommencer lundi?
4711 LA SECRÉTAIRE : Nous allons reprendre l'audience lundi matin à 9 h 00.
4712 THE CHAIRPERSON: Thank you.
--- Whereupon the hearing adjourned at 1143, to resume on Monday, July 13, 2009 at 0900
Johanne Morin Jean Desaulniers
Sue Villeneuve Beverley Dillabough
Monique Mahoney Madeleine Matte
- Date modified: