Network Neutrality is Killing the Internet, A Telecom Engineer’s Perspective
Disclaimer: I have not only worked for a big Telco (about 5 years ago!), but have been part of design teams that designed entire VPN networks and Mobile networks, from scratch!
I am no longer ‘invested’ in that space. I am an entrepreneur, developing diagnostic products (muchiniwam.co.za). Check, check, check it out!
You Probably Do Not Know How the Internet Works
Most established telecommunications companies that own national telecom infrastructures, are old. They were mostly built during the heydays of the industrial era after the discovery of the telegraph; which was shortly followed by the introduction of the telephone.
For communications within a country to be possible, corporate investments had to be made, with some assistance from governments, to build a physical infrastructure that would support long distance call routing, switching, accounting and authentication. A lot of cables, copper lines, switches, boxes, routers and intelligent machines had to be installed to make this a possibility.
The costs increased even further when these companies started routing calls within a country to each other and also to some international destinations.
Telephone technology changed drastically throughout the 20th century. It started with copper lines coupled with a clumsy mechanical infrastructure that also required human operators to switch calls. After that it evolved into fully automated mechanical switching. With the arrival of computer electronics, the telephone infrastructure ultimately reached full maturity with the use of electronic switches that were smaller, quieter, cheaper, broke down less often and easier to maintain.
For as long as these companies designed, built, upgraded and maintained these telephone networks, they had full rights to make profits (in any way deemed feasible) from the telephone services, products and solutions they provided.
Then everything changed.
The Internet arrived.
With the arrival of Internet, a lot of these same companies had to do much more than an ‘upgrade’ to their infrastructure. They essentially had to redesign and rebuild.
ISDN/DSL
ISDN/DSL are technologies that allowed internet routing over the same public switched telephone network (PSTN) infrastructure that ‘big Telco’ had been building for the past century. But ‘big Telco’ had to buy equipment to enable IP routing for internet services.
ISP
With the arrived of the ISDN/DSL and the Internet, many rules had to change. Providing an internet service is a very complex task. The easiest way to grasp this is to understand how one connects to the Internet:
-> you have a computer, -> connected to a modem, -> connected to the ISDN/DSL port, -> the packets (smallest unit of internet data) sent from the port are routed via the PSTN network to the local exchange, - > the exchange is connected to a local ‘regional’ internet router (it is this router which authenticates, authorises and assigns an IP address to your computer), -> if you were authorised, the packets you sent will be routed, with the help of the Domain Name Server, DNS, to the appropriate destination (similar process applies for anything internet related, e-mail, websites, etc), -> if the destination of the packet is outside the country, the packet is sent to a ‘national’ router (several, usually!), which will then route either to another international DNS, or towards the destination (internet) address / server.
But this is not even half of the story. The other side is of the Internet ecosystem is Website services (hosting, SSL certificates), E-mail services, other so-called ‘Over the Top’ (OTT) services such as Voice over IP (VoIP), Instant Messaging (IM), and many others.
Given this detail, it will now be easier to understand the role of the Internet Service Provider.
Many Internet stakeholders, enthusiasts, OEMs and standardisation bodies quickly realised what lay before them: the threat of the entire Internet infrastructure, services, products, and systems being entirely owned by one conglomerate, the big ‘Telco’!
Because of the complexity of the Internet infrastructure, the reason for its existence (i.e. to create and share information) and the many innovations in services and products it could offer, it posed a great threat for all that responsibility to rest only on the shoulder of the big Telco. It was now clear that the Internet, is the new electricity that had to be carefully managed and regulated.
This, is what led to the Three Tier Layered Internet infrastructure as we know it today. The first Tier is the big Telco, who owns most of the physical routing infrastructure. The second and third Tiers are the Internet Service Providers (ISPs). The only difference between the second and third tiers is on the amount of physical routing infrastructure (and the extent of national reach thereof) that an ISP might have invested in.
That, then, is the story of how the ISP was born. The idea behind the ISP is that it would handle all the other Internet broadband routing services that include: IP address allocation (static and dynamic), AAA (Authorisation, Authentication, and Accounting of Internet packets), DNS, Website/E-mail hosting and many others. Basically, even though ‘big Telco’ can still provide the same services, he was gradually being turned into a big pipe. A big pipe, I said, and nothing more.
The VoIP Problem
Everything was good and well in the early days of the Internet. Everyone who founded an ISP (second or third Tier), or was running a Telco company, or founded an OTT service company was extremely happy and relatively rich.
Then, the Voice over IP (VoIP) OTTs arrived.
ISPs were happy, it meant more customers, even more Internet traffic (driven by audio traffic), and there was clearly more money to be made.
But, for the first time, big Telco felt seriously threatened. His voice revenues are now going to the ISP and the OTT service provider. And this is all happening on the big Telco’s infrastructure!
That meant that the increased demand for bandwidth both on the last mile (from the PSTN exchange to the home) and on the core broadband infrastructure had to be addressed by the ‘big Telco’ even though he is clearly not only losing customers on one of his core services (i.e. voice calls), but, in many cases, is also not making more from providing the ‘big (internet) pipe’ to the ISPs.
If Skype experiences a high demand of calls, all they have to worry about as whether their VoIP servers will handle the load. But most of the audio/video call traffic does not even have to traverse their servers. Which means they are transferring the network traffic problem caused by their own services to someone else. Big Telco naturally has to upgrade the network just for Skype users to stay happy with the quality of their calls. Skype does nothing. In fact, Skype should do nothing. And that, is the argument of Net neutrality.
The Speed Problem
So far, this article has only mentioned the provision of the Internet only from the wired (ISDN/DSL) perspective. A perspective that not only excludes later faster wired technologies such as Fibre (FTTx), but more importantly, wireless broadband technologies such as 3G and LTE. In most cases, the same principles apply. The wireless case is only interesting in that the packets are transmitted wirelessly and that the three Tier Internet framework is not enforced but possible. A Wireless Access Service Provider (WASP) is basically analogous to the ISP.
I mention this because, with the arrival of cellphones, especially smartphones, the Internet landscape has been drastically altered. Firstly, there are more devices that require connectivity, which means more data is required and/or consumed. Secondly, given the competing access speeds from wireless technologies such as 3G and LTE, the pressure for the wired broadband connections to also be much faster exists.
This has pushed for investment in faster and faster technologies which require more and more infrastructure investment on the side of big Telco. The big Telco clearly has a lot to do. But where is he to get the money?
He faces declining voice revenues, competition from mobile (if he’s fixed only) and vice-versa; he faces increased pressure from ISPs and regulators to drop the ‘big pipe’ data prices; and his infrastructure can barely keep up with the demand for both higher speeds as well as the higher amounts of data traversing the infrastructure due more users and devices increasingly joining the network; he has to pay even higher fees for international data traffic.
Big Telco is Dying
VoIP is here to stay. It is the primary way most of us make international calls. Increasing demand for data at cheaper rates (and the transferred pressure to the big Telco to drop the prices) is also here to stay. There is also increasing demand for video traffic from porn sites, youtube, facebook, Netflix and the like.
The ‘big telco’ is expected to carry the burden for all this traffic. Alone.
Now, with all these challenges facing him, big Telco has a choice to make, to leave the current infrastructure as it is, as he is clearly fast losing options for funding the upgrades, or to improve the infrastructure, within the confines of the law and the current ecosystems, using Quality of Service (QoS) enforcements.
Net Neutrality is Killing the Internet
It is this QoS enforcement option that has made so many companies, usually OTTs of some sort, extremely angry. From an OTT perspective, if he is providing a service like Skype, he should not be obliged to pay some fee to big Telco so that the video call is of the highest possible quality and given priority over other traffic.
Which may be fine. But the biggest problem that the entire Three Tier framework is facing right now is that someone has to pay for the infrastructure that carries so many services and users at decent speeds. If Skype experiences a high demand of calls, all they have to worry about as whether their VoIP servers will handle the load. But most of the audio/video call traffic does not even have to traverse their servers. Which means they are transferring the network traffic problem caused by their own services to someone else. Big Telco naturally has to upgrade the network just for Skype users to stay happy with the quality of their calls. Skype does nothing. In fact, Skype should do nothing. And that, is the argument of Net neutrality.
You Probably Got Lied To By Facebook, Amazon and Facebook, Don’t Believe Them
As I see it, there are three main social aspects to the Net neutrality debate:
- freedom of internet (which includes access to, and generation of, information)
- privacy, and
- the cost of internet (i.e. information) access
Freedom Of the Internet
The argument has been made, in articles like this and that, that Net neutrality is really about the ‘freedom of the Internet’. It is yet to be qualified whether this ‘freedom’ is from the view of the user, the OTT service provider, the ISP or even the ‘big Telco’. (Let alone the fact that mobile network providers tend to be suspiciously missing from the entire debate!!).
Definitions
Firstly, let us assume, for the sake of the argument, that ‘freedom’ refers to the user. That is, the ability of the user to access any site he would like to visit at a flat data rate, without the shackles of a ‘restricting’ internet package or a network (service) provider limiting content provided to the user.
For this definition, it would appear as if the Net neutrality advocates are clearly stating an economic position. Let us briefly examine the factors and impact of this argument.
In its essence and by the merit of prima-facie validity, the argument would result in the status quo being maintained. That is, it is a fear tactic. Nothing more. It does not suggest how the Internet infrastructure aught to be paid for moving forward, nor does it give the opportunity to big Telco and ISPs to explain how they will avoid bringing humanity to this particular ‘perilous’ turn of events.
When examined closely though, it is evident that this is not actually what the Net neutrality proponents are fearing. What they are fearing is the prospect of ‘big Telco’ and ISPs somehow handling their specific data accounts with differentiated prejudice.
This prejudice, will be based on the user’s internet data subscription package. The reader will hopefully now realise that, at this point, the debate has now turned into a philosophical economics one: should everyone be granted the same level of equality with respect to access to a (non-free/non-natural) resource, e.g. the quality and quantity of internet data. But this will not be argued here as the aim of this series is not to debate any philosophies; it is simply to state the facts of the Net neutrality debate.
To be fair, it is possible that on this point, the Net neutrality side is fundamentally arguing that all things being equal (in terms of the data connection speed the user may have applied for — or have been granted access to, with applicable, understood and existing contention ratios) there should be no manipulation of the type and nature of internet content the user has access to.
If and only if that is the argument the Net neutrality team is arguing, then I foresee no opposition from regulators, big Telco and ISPs in terms of addressing this concern. But if this is indeed the essence of the argument, it is yet to be made clear (and it has absolutely nothing to do with Net Neutrality)!
The second assumption that could be made is that ‘freedom’ here refers to the ease, cost and viability of small OTT service providers to launch services that can compete with existing big players that may already have pre-existing favourable agreements with ISPs and big Telco.
This threat is real and is a definite possibility. It is also, however, incorrect, but requires a more comprehensive rebuttal.
The Rebuttal — the Non-Net-neutrality Scene
Let us briefly imagine the world without Net neutrality, it would possibly look like the following:
A. Big Telco/ISPs have differentiated network traffic, grouping various types of data into their respective categories (audio, video, text and so on).
B. Big Telco/ISPs can then decide which traffic to prioritise, using existing QoS technoligies like the LTE Quality Class Identifiers (QCI), Metro Ethernet IEEE P802.1p standard for classifying traffic over the backhaul, as well the enforcement of the VPN/IP DiffServ architecture.
C. Big Telco/ISPs would have increased freedom to tag content which should be of the highest possible quality, e.g. a voice over IP call, to the extent that the delays and jitters that are now a common occurrence in VoIP services will be a thing of history!
D. The wide usage and application of Deep Packet Inspection (DPI) technologies that monitor traffic, for the following purposes:
- More innovative implementation of caching/data-compression technologies
- Lawful interception
- Zero rating of certain services
- Premium rating of certain services
- Higher prioritisation of certain services as per service agreements with OTTs
Points A and B pertain to the nature of Next Generation Networks (NGNs), this would include Intelligent Broadband Gateways (IBGs), IP Multimedia Services (IMS), and Rich Communication Services (RCS). A great example of a rich communication service is the ability of the network provider to route a telephone call to the user’s TV screen while the user is watching TV if the user’s phone is off or in another location. An exiting example of a multimedia service is the ability of the user to access his favourable TV shows on any device over any network (mobile or fixed) and in any location in the world. That means the death of different data packages: one from a mobile network, another from a fixed/DSL network and even another from a roaming/international network. What is even more important is how the user will win in terms of finally being able to reduce his data costs due to combined packages that are indifferent to the ‘telco’ network the user is using.
Now, to the extent that doing away with Net neutrality might threaten OTTs, point C also provides the clearest example of how, ultimately, internet users will be the biggest winners in this new world (with OTT’s being a major party in the ecosystem).
As for possible threats to smaller OTTs due to point D, imagine the increased Quality of service Experience (QoE) that the user will experience for the services that are highly in demand (and have prioritised traffic), e.g. Skype audio/video calling, Youtube, Facebook, Whatsapp, Netflix, Google drive, Microsoft updates and Office 365, and so on. And also the reduced data prices he will pay in the long run.
As it is currently, the current Net neutrality ecosystem cripples the big Telco, disadvantages the internet user and guarantees a lopsided amount of profits to the OTT. No wonder most of these OTTs are the ones fighting the change of the Net neutrality regime.
The Rebuttal — the little OTT guy
It is still however possible that the currently existing ‘big OTTs’ are altruistically protecting the next ‘little OTT’ still in its infancy; still in a garage! But this is a false concern. Firstly because it would cost these ‘big OTTs’ next to nothing to benefit from the new regime; as they can negotiate favourable agreements with Telcos/ISPs to drive more users to a particular network, e.g. the zero-rating techniques already being used by networks for services like facebook and Whatsapp! They can also afford to buy, at extremely low cost, caching services such as the google caching servers that are now installed in most networks throughout the world to reduce the amount of international traffic. In most cases, these OTTs currently do not pay anything for these caching servers to be installed on the big Telco network; it is clearly a win-win situation for both the Telco (=reduced international traffic) and for the big OTT (=faster user access to the OTT content). Secondly, even if these big OTTs have to pay for their content to be treated favourably, it will cost them next to nothing to pay the big Telcos, relative to how much they are already making (in many cases more than that big Telco!).
As for the little OTT guy, there is neither a requirement nor an economic incentive, in a non-Net neutrality environment for the big Telco to block any kind of traffic. The internet user, if that would happen, would quickly notice this and simply move to another network service provider.
To top it all, this would not make any significant impact on how NGN technologies are meant to handle all kinds and types of traffic, which is still superior to how internet content is currently treated in a Net neutrality environment. In other words, voice calls will necessarily always, regardless of the OTT, receive higher priority, because that’s how things should be. Video traffic usually follows this, depending on the relevant QoS architecture deployed, and e-mail, text-based and all other traffic is usually (or should be) last in the internet traffic queue. Basically, there is nothing for the little OTT to fear. If anything, the superiority of his service over the big OTT providing the same service will become evident much sooner as the priority level of his traffic is likely to be similar (if not the same) as the one given to the big OTT.
The Net neutrality discussion is a complex and heavy one. This is evident in the type of accusations that get thrown back and forth around the matters of privacy and the costs of internet data. I mentioned matters concerning the ‘freedom on the internet’ we will now focus on the privacy issue.
Internet Privacy
A concern that Net neutrality advocates raise is on how the use of technologies like Deep Packet Inspection (DPI) might threaten the privacy of users on the internet.
As it is currently, many networks have already deployed DPI technologies. The internet experience as we know it today, would simply not be the same without DPI. DPI allows for the most effective implementations of caching (i.e. temporary storage of transient data) on telecom networks. This is based on the paradigm that video content generates the most traffic on the internet. This video content is usually from specific sites which are often the most popular among internet users. So caching technologies naturally do a much better job when a DPI server is well-informed on which sites/content are the most frequented.
The Net neutrality argument, then, is that DPI and similar Next Generation Network (NGN) technologies are already invading on the user’s privacy, and the Telcos will do this at a much larger scale outside the Net neutrality framework.
The Rebuttal — OTTs have a bigger problem
The easiest way to settle this accusation, if it is brought by any of the current OTTs against big Telco, is to simply focus on the degree to which either side is better placed to disregard and/or exploit the internet users’ privacy.
The argument by the net neutrality advocate is the following: through the help of DPI or NGN technologies, the ISP/big Telco is always fully aware of the type of content the user is being exposed to. This means that the ISP or big Telco is more likely to exploit the user’s privacy. But this argument, obscures that in most cases, the user is actually accessing the OTT content/services directly. The OTT as a result ends up with more private information about the user than the big Telco will ever be able to, over encrypted network sessions. (The use of ‘end-to-end’ encryption, in most cases, renders any ‘Telco’ DPI inspections, largely useless. Not so much for the OTT’s, who need our session data to sell us targeted ads!).
Most current networks (from the physical layer all the way up to the transport layer), architectures, sessions and services are built to be secure. This means that only extremely specialised DPI technologies can possibly decrypt and decipher the information transmitted. Yet none of these technologies ultimately mean much compared to a user willingly giving up information to another party within the Internet ecosystem. That party, is usually an OTT; such as Google, Facebook, LinkenIn, Amazon, Skype and the like. In essence, even though the ISP can ‘theoretically’ see all traffic the user is surfing, the OTT does in fact have the unadulterated private info from the user. One of these scenarios (i.e. the Telco that could ‘see’ everything via DPI) is a remote future possibility, while the other is already a reality (i.e. the OTT’s DO already see everything!). Another element that is ignored is that Telco’s, due to their geographical nature, ultimately give the users more ‘freedom to choose’ than many of these ‘global scale’ OTT providers.
To put this another way: the ISP/Telco ‘might’ use the user’s private information for advanced services and possibly for law enforcement (=lawful interception). The OTT, meanwhile, is already using the user private content for advertising and related services. For all intent and purposes, most successful OTTs of today thrive due to access to, and the questionable use of, the internet users’ private information.
As far as privacy issues are concerned, the OTT is clearly more guilty and has little argument to offer against the potential abuse of user’s private information by the ISP/Telco.
But as indicated above, this is the easiest rebuttal to offer…
The real Rebuttal — the Internet User
The second consideration with regards to matters of internet privacy in the Net neutrality discussion is the internet user himself. The user is already too exposed. There are too many OTTs gasping to get as much access to his personal and private information as possible. Without Net neutrality, the user will apparently now also have the ISP/Telco to keep in mind when he goes online.
But even if this is a valid concern (it actually is), then the discussion has to be much broader. It has to include the regulation of technologies, services and platforms that are used on the Internet. This is to ensure that the user’s privacy is always protected. In other words, this is not a Net neutrality problem. It is an online information safety, security and usage problem.
Real rebuttal. Done!
The Internet Cost Problem
The argument that the Net neutrality advocates are making is that the differentiation of data traffic will involve the use of expensive NGN technologies. The cost of most of these technologies will be passed down to internet users. The non-Net-neutrality regime will thus result in higher data prices. The ISPs/Telcos will be further incited to charge users for access to traffic-heavy sites such as Netflix and Youtube.
This, however, is clearly a fear of the unknown. The reality is: the non-Net-neutrality environment will not be the same as today. In many ways it will result, overall, in a better user experience of the internet. This also will apply to users who may opt for the cheapest packages which might result in slower internet speeds. These users will undoubtedly still be able to access many services (even if at a higher fee) at a much better QoE than they would have ever imagined in our current Net neutrality environment.
Ultimately, someone has to absorb the costs of our ever-expanding and increasingly struggling internet infrastructure. The current ecosystem only allows one entity, the ISP/Telco, to absorb those costs. In the future non-Net-neutrality environments, there will be an opportunity for many more players to play a role in reducing these costs and improving the infrastructure:
The OTT will have increased incentives to subsidize the user data packages, so that more users can use the OTT service. Governments will have increased incentives to subsidise citizens’ access to the internet (for access to e-Government), so that more users can access e-Government services/informationTelcos/ISPs will have increased incentives to continue improving the infrastructure at a much faster pace, as they will have more at stake in terms of competing services, technologies and packages from their competitors in both mobile and fixed Telco environments.
In the end, all these parties need to work together; for the future of the Internet, depends on them.
You Are Probably Confused, Don’t Be
Just know you’re being lied to.
Just know Telco’s are the only ones who actually spend money on maintaining your ‘internet speeds’.
Just know Amazon and Google are not your darlings.
What Should Happen
We need new companies that work closely with Telcos to destroy the Google/Facebook and co. monopoly on content delivery.
The most obvious solution is: we need more Telco’s! Competition is still sorely lacking at the ‘infrastructure level’ of Telecommunications. Governments heavily regulate many new network companies out of existence all in the name of ‘spectrum policy’, so-called ‘fair’ internet pricing, and monopoly strategies sponsored and protected by politicians.
The Net neutrality advocacy is only a plot to give power to the same politicians, similar monopolies (like Google, Facebook and Amazon), to the ultimate ‘unfairness’ of the same users!
