Net Neutrality vs. Packet Prioritization

Last article I said that I would detail the concepts of net neutrality and packet prioritization. I have discussed these points at length in several previous articles, and this article is what will be the final philosophical discussion of the issue—I hope. Unfortunately the concept of packet prioritization is being associated with denying certain service and application providers’ unfettered access to the Internet or degrading their service. Packet prioritization is a tool that can improve the user experience of real-time services like VoIP or streaming video. The two concepts can co-exist in harmony to provide a superior Internet experience for users and service providers. What follows is a rather matter-of-fact comparison of the two.

There are two principals for companies that make a business of the Internet should always keep in mind:

1. Internet access has no value without content provided by a variety of sources. This fact is a corollary to Metcalf’s Law.
2. Users expect a certain quality of service from their application providers regardless of the underlying network infrastructure.

So there is a symbiotic relationship between service and access providers. Neither one can stand on their own despite AT&T’s desire to distribute content and Google’s attempt to build a network.

Packet Prioritization	Net Neutrality
Allows time sensitive packets to travel through the Internet first while not restricting or impeding traffic unless a bandwidth limitation exists then traffic is gracefully limited until the condition clears	Does not limit or impede any type of traffic
Reduces impairments to voice and video	Does not attempt to manage the user experience or quality of service
Provides no noticeable impact on other applications like web browsing and e-mail	Some definitions require all packets to be treated equally so browsing and e-mail could arrive before your movie or phone call interrupting the call or movie
Manages bandwidth efficiently to maximize the amount of traffic through the network	Believes throwing more bandwidth will solve the problem.  There will always be more traffic than bandwidth
Can improve the quality of some Internet applications	May restrict access providers from allowing application providers the opportunity to improve the quality of their services
Does not have to be used by all Internet applications	Internet Access providers should not give preference to their content over competing application providers

Net neutrality should be defined not to exclude Internet Access providers from providing packet prioritization. Packetprioritization is a valuable tool that all content providers can use to insure a superior user experience and most efficient use of bandwidth.

Tag: AT&T, Google

Net Neutrality-Why Would We Rely on Congress?

Every time I hear that another bill is introduced in Congress concerning net neutrality I cringe.  Several of the versions introduced recently would cripple the ability to voice and video services reliably over the Internet.  Language in the bills are often vague and open to interpretation that would make implementing any quality of service mechanism in America illegal.  What is ironic is that this move would hurt the companies that support these bills.

Why are Internet application providers supporting legislation that could reduce the quality of their services?  FUD is why.  Google, Yahoo, and others are afraid that their users will not have access or reduced access if they do not pay additional fees to AT&T, Verizon, and Qwest.  They know the ILEC’s history and distrust them.  The ILECs have not been exactly forthcoming in marketing the benefits to Internet application providers.  Recently all three CEO have publicly clarified their intentions.  Richard Notebart, CEO of Qwest, came out with an op-ed piece in the Wall Street Journal a few weeks ago that was very articulate and full of analogies against net socialism, but it was defensive.  This adversarial relationship between access providers and application providers is being perpetuated in the press by companies like CNET.  The industry’s inability to resolve this situation by itself means that the battle will be taken to Congress where the companies with the most influence will determine the outcome.

Why is there so much resistance to prioritizing some services over others?  Is it so complex that only engineers understand the details?  As with relationships, it is all in how you present the issue.  Name changes and big advertising agencies have not changed the nature of how the ILEC communicate with their customers.  Initial statements by the CEO of AT&T and Verizon that Google, Yahoo, MSN, and Vonage are freeloaders were truly argumentative.  I am sure that the middle managers in these companies did not intend for the message to be delivered in that manner;  call it CEO prerogative.  Eventually the rhetoric was toned down and they clarified their intent, but the damage was done.  Application providers and the press were in a defensive posture.

This whole mess could have been avoided by a smart public relations campaign that communicated two simple messages to application providers and the public:

1. All applications are not created equal.  Voice and video services need to move through the network before web browsing and e-mail.


2. All services will operate as they do today without prioritization.  No services will be purposely degraded by implementation of this mechanism, actually all services should operate better.

In reality the ILEC were announcing a new product release so they should have carefully planned its introduction.  Engineers and developers in application providers and the technical press would have been given privy to the details of the implementation.  Given all of the details in a non-confrontational manner, my belief is that they would have supported it and asked when they could sign up for the service.  Now the issue will be resolved in Congress which will give us a non-optimal solution.  

Should packet prioritization be illegal, the United States continue its decline in broadband access and services.  As I have stated in previous articles, we are already behind Asia and some European countries.  This move will facilitate a continued decline.  We can have net neutrality with service prioritization.  They are not mutually exclusive.  Hopefully any legislation passed will account for that fact.

In my next blog article, I will compare principles of net neutrality with packet prioritization.

Tags: net neutrality AT&T Verizon Qwest

What Comes Around Goes Around

John Chambers, CEO of Cisco Systems, stated in his keynote speech at Interop 2006 that that the network will become more intelligent to ease application development.  Simply stated, Cisco will build systems for carriers that contain more intelligence taking control from the network end-points.  This philosophy reminds me of the Bell System days where we built intelligent networks for dumb appliances to hang off the end.  It seems that we have gone a full circle back to the Bell System philosophy.  Everything old is new again.

Cisco was founded on empowering businesses with the ability to design their corporate networks to meet their own business needs.  Their philosophy was that the network was dumb and the end-points intelligent.  AT&T and its Baby Bells were professing the opposite that having intelligence in the network enabled more powerful services at less cost.  Cisco and other companies determined to create a new communications paradigm (you don’t hear that word much any more) based on their success at proving that the old Bell mentality was obsolete.

They were successful at unseating the Bells.  IP won out as the protocol of choice over a fairly transparent optical network.  TDM and ATM are giving way to Ethernet and IP/MPLS over DWDM.  Innovation is coming not from the big iron vendors, but from companies developing equipment and applications based on intelligent network end-points.  VoIP, Skype, BitTorrent, Wi-Fi, DVR, TiVo, camera phones, iPods and countless other devices are products of this pendulum swing.  This millennium we have seen an unprecedented period of innovation enabled by thousands of entrepreneurs that do not require billions of dollars to bring a new application or service to market.

The paradox of success is that the larger a company becomes the less they innovate and the slower they grow.  Cisco is experiencing this paradox.  They have grown to become the 800 pound gorilla.  To continue a reasonable growth trajectory, they have to come up with new communications paradigms that require carriers and enterprises to purchase new hardware, software, and services.  Without major shifts in philosophies, their growth will level out and possibly decrease.  It is a perfectly natural response, and you have to acknowledge Chambers’ drive at trying to appear as the great innovator.

Personally I do not believe that the industry is ready to shift innovation back to large corporate labs like Bell Labs, PARC, IBM, and RCA Labs, and control to their corporate sponsors.  Those companies are relegated to our technology history books. We will witness the same happen to Cisco Systems, Intel, and Microsoft over time.  These are all great companies, but they cycle of innovation will swallow them as well.  The Internet and intelligent network end-points has spawned the Cheap Revolution that empowers any smart individual or small group with entrepreneurial spirit the ability to be the next Skype.  

Tag: Cisco, Chambers, Interop

Comments on Net Neutrality

I wrote the following article as a response to the frequent comments on CNET’s Buzz Out Loud podcast misrepresenting the RBOCs intentions to prioritize traffic on their networks.  The hosts commented on my article but they did not seem to comprehend what I was saying.  The press is misrepresenting the concept of net neutrality.  Packet prioritization and net neutrality are not mutually exclusive.

Once again Congress may screw up a valuable resource.  Senator Ron Wyden's bill will cripple Internet access services to millions of Americans and stifle the creation of new and innovative services.  There is no question that Internet access providers should not give preference to their services or content that they distribute.  Likewise, they should not negatively impact any potential service competitor's traffic that goes over their networks.

The press, Congress, and many others are confusing bandwidth and speed with packet prioritization. All packets travel through the network at the same speed.  The speed at which a packet travels is somewhat less than the speed of light dictated by the medium they travel through.  Along the way they are slowed down by transport equipment, routers, and switches.  So saying that there is a "fast lane" is technically incorrect.

Traffic becomes limited because interconnections and equipment are limited in the amount of packets that they can process.  We call the rate at which information flows through equipment or interconnections: bandwidth. Bandwidth is always a limiting factor. Internet carriers attempt to fully utilize their bandwidth to reduce cost and improve margins.  There are many places in the network where bandwidth can become constrained, but the most common location is the last mile to the consumer.

When there is more traffic than bandwidth, then some packets have to wait their turn until bandwidth is freed to pass them on to the next point in the network.  Currently all packets are treated equal so e-mail is just as important as web browsing as VoIP calls.  Here's the rub.  If there are several web browsing, p2p, and e-mail packets waiting to go through the pipe, VoIP packets will have to wait a long time to reach their destination. As long as they only wait a couple of hundred milliseconds, users will not hear the delay.  If the wait is longer then users will hear garbled speech, echo, and in extreme cases the call will be disconnected.

This is the reason why we need to prioritize certain traffic. Packets from services like VoIP, IPTV and interactive gaming need to be moved to the head of the queue to provide an excellent user experience; otherwise, consumers will not find them acceptable over today’s alternatives. Each packet should be labeled depending on the class of service it requires enabling bandwidth to be used more efficiently and better quality of service for all users.

Throwing more bandwidth at the problem is not the solution because bandwidth is always consumed as soon as it is provided by new services and greater usage of existing services.  The bandwidth answer is a never ending spiral.  The only true solution is creating tiered services for all service providers to use.

Notice that I have not mentioned who is providing these services.  The point is that all service providers whether RBOC, cable company, or Internet service provider benefits from implementing quality of service (i.e. packet prioritization). Actually tiered services provide a level playing field for Vonage, Netflix, or any other IP service providers as long as they can purchase these service guarantees for the same price that the access providers’ charge their internal content distributors.  In other words, the RBOC or cable companies do not give preference to their own services. The way services are delivered today, the RBOC or cable company can control the quality of their VoIP or IPTV services; thereby, providing a higher quality user experience than service providers that use their networks. The current situation is not fair. Google, Yahoo!, Netflix, Vonage, and all of the other IP service providers should welcome this new service from the RBOC. It offers entrepreneurs that do not have billions of dollars to build a network the ability to create new and innovative services.

What alarms me is that the industry cannot work out this issue on their own and are relying on 535 old men and women that have their staff print their e-mail to decide the future of the Internet in the United States. These people do not have a clue of what to do about this issue. Instead they depend on industry lobbyists to help draft this legislation. Wyden’s bill is too restrictive and will stifle competition. Frankly it scares the hell out of me.  Barton’s bill just gives more power to the FCC over the Internet. Neither one is good for our industry. Unfortunately there are too few people in the industry that understand the issue and the ones that do are less than eloquent in explaining it. The bottom line is that we should let the free market sort it out. There are already sufficient laws in place if the RBOC or cable companies give preference to their own services. Let the market work!

Tag: net neutrality

Postage Due

A recent article in the New York Times detailed AOL’s and Yahoo’s plans to charge a quarter cent for companies sending their subscribers’ e-mail. Their objective is to dramatically cut down on the spam and phishing that ends up in our mailboxes. The scheme will allow those that pay postage for their message to receive priority and bypass black-lists and spam filters. In addition, addresses in user’s address books will not be subject to paying postage and receive the same service.

There are many reasons why this scheme is not a solution to the spam problem. First, spammers will still send messages. Some will pay the fee. It is still cheaper than direct mail. Second, the ingenious spammers will use zombie machines to send messages to recipients in user’s address books. Third, it will impede senders of legitimate e-mail. Under this scheme, if I met someone in a business meeting and I wanted to send them a follow up message, I would have to ask them to add me to their address book or pay the quarter cent. Not much, but it is an additional expense I thought came with my Internet service. Corporate users will just consider this the cost of doing business, but personal users will find paying postage burdensome. I suppose that it will put and end to those pesky chain letters that people are always sending me. Proponents claim postage is just the solution that we need to reduce spam to a trickle.

What is puzzling is that these same proponents are opponents to the RBOC charging for tiered service. According to the NYT article, AOL and Yahoo are positioning the postage as a superior class of service just like the RBOC. Where is the opposition to this scheme? Tom Evslin, ex-executive of AT&T’s Worldnet service, is against it for one. His argument that it will adversely impact subscription services and legitimate e-commerce is extremely compelling and lucid. I have no problem if service providers want to charge extra for a service like certified mail that validates and guarantees the identity of the user as long as it does not impact how I send mail today. Will it stop spam? No way.

The widespread use of digital signatures is another effective way to prevent spam, but the cost of a digital signature has to come down to under $5 per year before we see widespread use. You can buy a domain name for $1.99, why not a digital signature? What I would like to know is how AOL and Yahoo will make the service any different than using a digital signature? It seems to me that they work about the same way. Why not credit senders a quarter cent until they pay for the digital signature. That way there would be a cap on how much user’s are charged. Once everyone has a digital signature, the e-mail client would verify the message header against the digital signature. If they did not match, then it would be filed as junk mail. Other solutions like Sender ID or DomainKeys are effective at preventing spam and they do not require user intervention.

Spamming is an arms race. Legislating it away did not solve the problem. Spammers either ignore the laws or moved off shore. Charging will not solve it. They will hijack machines or continue to use best effort and hope they get through. There are even ways around Sender ID or DomainKeys, but these solutions are more effective than blocking messages that do not pay a toll. So go ahead and charge for a higher class of e-mail. If users want to only accept mail from certified senders and people in their address book then that is their prerogative. Anything less is creating a walled garden and against the spirit of the Internet.

Tags: AOL , Yahoo, e-mail, Tom Evslin