Archive for the ‘Industry’ Category

DPI + Policy Control = True

Monday, January 11th, 2010 by Jon Linden

The rumor is true, DPI and Policy Control are confirmed to be a couple and have been seen hand-in-hand even in public recently. Feelings are mutual and this is bound to last “until death do us part”.

The 3GPP framework rules! At least in the mobile space and it’s also making headway into fixed. That’s good. DPI has become a critical component in a service provider network, which requires DPI to interact with surrounding systems. On top of this complexity grows with tiered services, volume quotas, sponsored sites, and premium gaming and telecommuter services.

This is where the policy server, or PCRF (Policy Charging and Rules Function), comes in.  DPI is an excellent traffic analyzer and policy enforcer. But since DPI in most cases resides inline and manages large volumes of traffic, it can’t afford  “wasting” cycles on polling information from other systems and correlating a lot of conditions.

This is a perfect assignment for the out-of-band PCRF server. PCRF is a decision point – PDP (Policy Decision Point). Once the decision is made it’s sent out to one or more PEPs (Policy Enforcement Point). DPI is a great PEP, but other equipment in the network might occasionally be better suited to host certain policies. The Gx interface in 3GPP enables standardized integration between DPI and PCRF. But the fact is that most installations still use a SOAP interface and a more or less proprietary API, but Gx is starting to make an entrance.

Since DPI and PCRF is part of the same eco system, resolving the same challenges, it’s quite understandable that there is some confusion as to whether or not DPI competes with PCRF vendors like Camiant, Openet, Bridgewater and Volubill. That’s certainly not the case. It’s rather the opposite where everyone’s working with everyone. I dare say it’s flat out promiscuous. Do I also dare to assume there will be consolidation within and between the two segments? Well, let me put it this way: anything else would surprise me.

We, Procera, see that most of our customers and prospects either made a decision or are making a decision on PCRF as they deploy DPI today. This is great. It opens up even more opportunities and options for how to create new services, how to strengthen the operators’ business case, and how to adjust to accommodate the ever-changing reality.

2009 In Retrospect;

2010 A Look In The Crystal Ball

Tuesday, December 29th, 2009 by Jon Linden

Every New Year’s Eve, half an hour before midnight, Swedish television airs a traditional black and white TV episode called the ”The countess and the servant”. The table is set for a party of six, but none of the guests are present. The countess sits at the head of the table and the servant acts the role of each of the guests, toasting five times a dish, quickly getting more and more drunk, stumbling over the head of a bear that decorates the floor.

The financial climate has been one big stumble in a bear market in 2009. Operators and telcos have not been suffering as badly as many other industries – mainly due to two steel baths already endured since the turn of the millennium – and have continued to invest, though the decision has transitioned from the CTO’s to the CFO’s office, adding another layer of involvement and delays to the process. The cutthroat nature of the financial climate has put strain on all vendors, but it strengthens the ones who survive since it requires a compelling offer, obvious value, and a sustainable business model. We all come stronger out of this.

The poor servant on New Year’s Eve doesn’t come out stronger after each course that is followed, in a more slurred voice, by the recurring question “Same procedure as last year, miss Sophie?”, and the countess’ “Same procedure as every year, James”. The DPI industry has been anything but the same in 2009. 2009 was the year DPI grew up, became a commodity term, and got invited to play with the big guys. DPI has become an integrated function in any and all operator networks.

And it will become even more integrated, not only with policy control, but also with charging and billing in the coming 12 months. Intelligence is important for any business decision, and DPI is the intelligent layer in an IP network. The more intelligence, the better the decisions. This is why we’ll see “DPI everywhere”. DPI will be a stand-alone function to accommodate the broad expectations of what DPI should be able to do (throttle, filter, prioritize etc.), but the traffic identification capabilities will also be integrated into other network elements.

I think, or I should say “I hope”, we’ll see an upturn in the economy during the second half of 2010. People will start spending more and operators will start investing for the future to gain market share and keep up with the competition. The competitive landscape is still a bit fuzzy as telcos compete with cable MSOs who compete with mobile operators since they all offer at least three overlapping services. This landscape will transform again in 2010 where the need for speed will narrow everyone’s scope to become best in their field. The belief in the future will drive a fast development, which will benefit LTE that will take off even though 3G and its massive investments are still fresh in mind. More available bandwidth will drive services that consume more bandwidth, which is why video will be the fastest growing application in 2010 as well.

I eagerly look forward to 2010, and as Miss Sophie wraps up dinner, I’m confident it’s the beginning of 365 exciting days until she’s back again and we can conclude another year. Happy New Year!

Happy Holidays!

Tuesday, December 22nd, 2009 by Jon Linden

In the movie The American President, Michael J Fox walks into a room full of eggnog-drinking colleagues and asks, “What’s the occasion?” “It’s Christmas. Didn’t you get the memo?” I hope you got (and read) the memo. The Holidays are here and it’s time for celebration. Personally, I love Christmas and I hope you, just like I, enjoy it with your loved ones. I wouldn’t miss it for the world!

So, let’s get some Holiday spirit! I’ll put on the beard, shout out “ho.ho.ho”, and deliver some Holiday gifts before I shut the lid on my laptop and head home to heat the glühwein.

Gift #1: “From Procera To Julius Genachowsky”
“We admire your guts and wish you glory,
as you and the FCC argue the Net Neutrality story.
Here’s a helmet to protect you from tomatoes and blows,
as both sides put on their hard-hitting shows.
In 2010 we hope for consensus and common sense,
proper guidance would be appreciated your eminence.”

Gift #2: “From Procera To Our Dear DPI Competitors”
“Look in the rearview mirror and you’ll see,
here comes Procera, speeding by raising dust.
A cleaning cloth for your windshield you need,
for you to properly see us swoosh by in a gust.”

Gift #3: The nicest gift we always save for our customers;
“’Accuracy and control redefined’ we say,
small words, great values, that we already provide today.
In a world that constantly changes, you’re asking for facts,
insight, intelligence, quotas and network stats.
We promise you much and we stick to our pledge,
here’s new PacketLogic functionality – once again cutting edge!”

Before I rip off my Santa beard and switch back to my normal tone of voice, I would also like to hand out “a Happy Holidays” to all Procera people around the world. You’re a marvelous group of individuals who take pride in over-delivering – time and time again! I’m so proud of you, and I’m privileged to work with you! Keep up the good work in 2010!

2009 is coming to an end, and I’m confident 2010 will be a great year. Maybe the best ever?!? Time will tell, but right now I’m out of time.

From all of us, to all of you – Happy, Happy Holidays!

Best Regards
Procera Networks

Mobile Internet – Just A Bit Pipe?

Wednesday, December 2nd, 2009 by Jon Linden

I’ve argued that mobile Internet is “just another Internet access”. This used to be a controversial and provocative statement when talking to mobile operators who defended their premium network investments they made,  in order to deliver pretty fast Internet connectivity.

Imagine my surprise when I recently attended a traffic management conference in London where all mobile operators in chorus surrendered to being just a bit-pipe competing with wireline broadband… What happened?

I assume they’re realizing that they don’t have the recipe for the all-resolving secret sauce, which is why it currently feels a bit hopeless. Walled gardens didn’t work, a price war has driven prices down, customers don’t pay for added-value services, and another network upgrade, this time to the fourth generation, is just around the corner.

The good thing is that this puts the spotlight on the fundamentals – production cost. You consider how you can limit OPEX by minimizing helpdesk calls, you look at how you can maximize the utilization of your network, you want to automate up-sales, and you implement cost thresholds like volume quotas and international roaming control. Oh by the way, these are all things you can do using DPI.

Once we’re past this realization we must get ourselves out of this sorrow and start looking forward. Mobile Internet certainly has added value over wireline. The growth in mobile Internet is probably the best validation of this. But speed, i.e. bandwidth, is not one of them. Maybe it’s time to take the lead on selling non-bandwidth-centric services? How can you leverage mobility? For what customers is mobile Internet the only, or the best, option? How will built-in 3G modems in laptops impact the market and the ability to sell pre-paid services?

This is putting a lot of demand on the business developement and product management people at mobile operators. It’s time to get your moneys worth guys! Every case is certainly unique. Do you provide mobile and fixed line, do you go after specific customer segments, or are you the low price option? This will impact your strategy more than ever.

So I guess I must convert to the other side and argue “Mobile Internet is more than just a bit-pipe”. It doesn’t feel as controversial, but a man’s got to do what a man’s got to do.

Do’s and Don’ts in Bandwidth Control

Wednesday, December 2nd, 2009 by Jon Linden

Suddenly it all happens at once. We went from zero to two events focused on DPI in one week – Light Reading’s virtual tradeshow Policy Control & Deep Packet Inspection, and Informa’s Broadband Traffic Management event IRL in London early last week. These events are good validation that DPI has grown up to get proper attention.

I was on-site at the event in London – and I apologies for the hotel lobby background noise on my line in the LR panel discussion. But I also had the pleasure to be on a panel in London together with Benny Lim from SingTel and David Hodgers from O2 Ireland, moderated by Mark Newman from Informa. The topic was “Lessons Learnt: The Do’s and Don’ts of Bandwidth Control”.

So what are the lessons learnt on bandwidth control? Let me quickly go through my conclusions and what I told the audience. When looking at do’s and don’ts I see two categories: what works and what ‘s ethical. The latter is more challenging with today’s fast moving social media where bad news travels fast and subscribers vote with their feet.

How do you manage this? Well, don’t expect to “get away with it”. Internet users are savvy and operators in London confirmed that non-transparency came back to bite them. You must be able to defend what you’re doing, and a good test is if you’d accept it yourself. Greed is not an acceptable reason, though a viable business case normally is.

Do’s and don’ts also depend on what you’re trying to achieve. DPI and bandwidth control spans over a broad range of applications, from network protection and congestion control to quality assurance and service differentiation. We tend to be very introvert and technology-driven in this industry, but you must start from business benefits and objectives in order to define intentions and make it understandable to your end-users.

The key to what works is knowing what’s going on. It’s not an option to guess what your customers do and how your network is doing. And conditions change – all the time. What used to be 80% filesharing a year and a half ago is today replaced with 40% streaming video as the largest bandwidth consumer. A policy rule-set that is not up-to-date will cause some strange consequences.

I wish I could provide a blueprint for what works, but every situation is unique. Your product mix, fixed and/or mobile, target groups, price positioning, and on top of this there are geographical differences like what applications are used and what bandwidth control practices are accepted.

This means that the number one thing to do is try. You will have to try it out and assess the outcome in your specific situation and your environment. I know this sound scary, but be adaptive, be quick, and make sure you make qualified decisions based on facts rather than guesstimates and assumptions.

Let me also summarize my don’ts: Don’t lie – be transparent. Don’t insert packets – honor integrity. Don’t get introverted and super techie – understand your business and your customers. And don’t do nothing – this is the most expensive decision of them all…

Let me wrap up with one more suggestion for what you should do: Keep it simple! Complexity kills this cat. If it’s too complicated you won’t be able to communicate to your subscribers, it’ll be impossible to update and change, and you won’t be able to measure if you successfully achieved what you intended to. This is hard to make simple – which is why we, Procera, are here to help ;-)