May 30, 2011 by Peter Briggs in the theme A new generation of IP Networking

The Scalable and Adaptive Internet Solutions (SAIL) consortium of leading operators, vendors and research institutions recently summarized some key trends in observed internet traffic.

They report that over the last two years, the internet core has begun to be optimized for delivery of services and content from a decreasing number of dominant sources to reduce cost and improve performance.

Revenue potential is increasingly associated with the delivery of higher value services such as virtual private networks (VPNs) and content delivery networks (CDNs). Old global internet economic models are evolving while new entrants are reshaping both the definition and the value of connectivity.

We know that the drivers for these changes are the worldwide rollout of mobile broadband networks and new, attractive terminals and devices that are being used along with fixed devices to enjoy an ever increasing volume of video content. Network operators are therefore considering how best to monetize the delivery of all these so-called 'cloud' services.

Clearly best-effort networks are unlikely to provide either very profitable operations or the kind of performance that users expect.

The emerging new generation of IP networking is a great balance of smart delivery, simple operation and scalability.

Since overprovisioning won’t maximize operator profit, ecosystem network operators and content and service providers will collaborate to maximize mutual benefit. This needs ‘smart’, service-aware delivery control and associated charging systems.

Operators can be smart about how they use the valuable RAN resources that are really well positioned to maximize value – plentiful enough to drive practical desirability, yet scarce enough to avoid collapse to low-priced commodity.

HetNets will build RAN capacity but require yet more 'smarts' to integrate into a flexible fixed/mobile access network, so users are always ‘best connected’.

Although, on average, IP traffic patterns are becoming simpler and dominated by content delivery, the underlying packet infrastructure still needs the traditional transmission, transport and routing functions – now enhanced with video and content optimization.

Operating this content-optimized network must be highly automated to keep opex in check as the network continually adapts to new traffic delivery patterns, sometimes over very short periods. Maintenance activities, capacity enhancements and inevitable faults must not disrupt valuable services nor require expert-level network re-planning every time. So management systems will allow network operators to deliver ‘more’ but keep control of their operating costs.

Networks will scale with higher bit rates in all domains from wireless and fixed access, microwave links, optical/WDM systems and through the cross-connects, switches and routers of the infrastructure. These higher speeds will cope with the volume of traffic and also reduce latency (especially important in some vertical scientific and enterprise markets).

The need for scale extends to network control plane signaling and the sheer number of devices that need to be addressed and managed.

Packet transport is replacing TDM for packet services. It builds on the traditional transport objectives; it delivers native transport services, provides internal network system interconnectivity, optimizes provisioning, OAM, resilience against faults and provides for planned network maintenance activity. MPLS (-TP) packet transport is also great for carrying emulated services during the transition to all-IP networking.

The new generation of IP networking will deliver marketing promises, user expectations – and an amazing user experience!

May 24, 2011 by Lingaraj Patil in the theme A new generation of IP Networking

The explosive growth in fixed broadband has meant that it has become necessary for many operators to upgrade their access networks.

However the decision is not an easy one, especially because it requires significant capital investment. A detailed up-front financial analysis is the key to ensuring short term as well as long-term success.

For operators with existing copper, it is important to look at bandwidth growth for the next 10 years and the access network evolution that it would entail, as opposed to doing the business case for the next immediate upgrade. How does payback time for copper network enhancements deteriorate over time if the operator is forced to upgrade it every couple of years, as opposed to doing a bigger fiber rollout once?

For operators from low-ARPU countries, recovering capex investment might be difficult. In such cases, it may be helpful to look at open access and Retail Service Provider (RSP) business models, and decide which role they want to play in the value chain. Due to shared resources, open access reduces the ‘cost to connect’ but it also reduces the barriers to entry for new RSPs. Differentiated voice, video and data service offerings become very important. And, more differentiation can be achieved by selecting superior content, a versatile multi-service IP edge and a smart policy manager to provide superior customer experience.

Outdoor fiber deployment constitutes the biggest part of the rollout costs and hence it is extremely important to plan this carefully. Is it better to lay the fiber once and upgrade the active electronics for higher speeds, instead of trying to reduce the splitter ratio to get higher speeds?

It is also important to choose fiber laying technologies that reduce the overall labor hour needed. It may be useful to choose a technology that has a higher proportion of the total cost in the ‘cost to connect’ vs. ‘cost to pass’ since ‘cost to connect’ is incurred only when the customer signs up. This not only spreads out the investment but it also reduces the risk associated with take-up rate uncertainty.

For the opex, care should be taken to reduce the cost of adding a new customer to the network and subsequently servicing the customer. Features like ease of configuration, remote activation and upgrade and reduced truck rolls all become key to increased profitability.

May 23, 2011 by Ajay Saini in the theme A new generation of IP Networking

Fiber deployment requires good initial investment, but is there a choice not to do it?

Well, alternative operators can try to delay fiber deployment by using their existing infrastructure, by leasing copper from incumbents or by waiting until others deploy fiber and then react.

However, this will not help them become strong long-term leaders or meet their financial goals in today's tough telecom market. It will be difficult for them to offer differentiating services and they will probably hope to make use of any poor reputations incumbents or other competitors may have, which unfortunately from the end-users’ perspective, is not uncommon.

The challenge for incumbent operators is that bandwidth speeds are increasing every year and in spite of various technological efforts, copper or cable bandwidths will not be able to compete with fiber.

Therefore, if an incumbent will not deploy fiber, sooner or later an alternative player will do it. Is this a wise option for incumbent operators?

A follower or second entrant in the FTTH space will struggle to make a similar or better impact, and it will be difficult even for incumbents. Please note that deploying fiber, especially the civil works and network planning aspects take time, which can hinder the ‘go-to- market’ strategy for late entrants.

It is quite clear that fiber will fly and is without any doubt the ‘end game’. Quite often the major driver is fear; however the question is more if one prefers to become a leader in FTTH journey or a follower.

To come out as a strong leader, one has to have a solid flight plan, where the bottom line is to have a higher take-up rate rather than solely focusing on ARPU. Access is all about connecting users and one can only get good revenues from connecting more users. Fiber provides better chance to connect more users.

There have been quite many successful FTTH deployments around the globe and end users are even enjoying speeds of 1 Gbps, which many would not have imagined 10 years back.

Hence, the recommended actions can be to ensure early adoption, come up with efficient time-to-install plan, have an aggressive go-to-market strategy and consider various partnership models to increase as high take-up rates are possible.

This will ensure a smooth take-off rather than being blown away by fiber from the competitors.

May 20, 2011 by Massimo Enrico in the theme A new generation of IP Networking

Returning to the parallels between IT and telecom, there are several immediate attributes required for a new "telecom network OS." It needs to be end-to-end, service-oriented, easy to operate (even by unskilled personnel), scalable and flexible.

In a nutshell, it needs a system that exploits network capability by managing e2e services, really exploiting network convergence, in ways so simple that almost everybody can do it, and so efficiently and flexibly that it can handle the service explosion driven by mobile broadband, video demand and machine-to-machine applications.

This means new network management solutions that tear down existing operational silos; that are systems consolidated and designed to interoperate with each other; that have smart and intuitive user interfaces, automating tasks and are ready to scale up and support new services.

But is this "Windowsification" enough? Is it only an enhancement of existing products? Maybe for the next three to five years, yes, but for a world with more than 50 billion connected devices, with telecoms a fundamental part of life just like energy or water, network management must also sustain a radical shift in business models and take a further step.

Again, the parallels with OS can help, and my thoughts go here to Apple iOS and Android. Their success is not only product-related but is inherently tied to the ecosystems they enable, where users can become developers, where new, previously unforeseen applications and solutions are developed, and where everybody gains from the creative power of the community. Their secret is the clever way they are open and make new business models possible.

The same has to happen in network management. Today it is the exclusive domain of telecom operators; tomorrow it might become the instrument to open the network, under appropriate control, and to allow the end users – first other companies, then even individuals – to define, develop and finally manage the services they need. This will build on the complete, hidden potential of the networks, probably beyond any of our current expectations, and generate new revenues for operators by empowering the market to create and expand itself. That is the way to generate "more" for everybody.

May 19, 2011 by Robert Haim in the theme A new generation of IP Networking

We live in interesting times.

This is especially true for all the players involved in the creation, delivery and consumption of the video ecosystem. There is a plethora of content being created throughout the world; delivery is made by content providers and owners of network infrastructure (service providers); and there are plenty of consumers for these products.

Consumers’ behavior, of course, is the ultimate test of success for the other players, i.e. content, and service providers. The content has to be interesting and its delivery has to be bound within consumers’ acceptance level. Consumers also expect to watch this content anywhere and on any device. Placing control of the consumption in the hands of the consumer is a good idea and, frankly, essential. This brings us to our real topic, as to what is the inter-play among the participants of this ecosystem.

To begin with, content can be delivered in a “structured” way, as in IPTV, or the new nascent non-structured method via Over-The-Top (OTT).

The statistics on the combined expected traffic volume in the network (mobile and fixed) is staggering: 90 percent of the traffic will be video by 2015. Ericsson’s forecast points to a traffic volume of 732 Exabyte (that is 732 followed by 18 zeros) by 2014.

While this data can be perceived as a major challenge for the infrastructure vendors and their customers (i.e. service providers) it must also be viewed as a tremendous opportunity for all the players: more content providers, faster, reliable, and more efficient networks, and satisfied consumers; consequently more revenues for the providers.

The technical challenges include: 1. delivery and operation of a highly efficient and reliable network, 2. infrastructure must allow the consumer to be in control of the content in terms of time and place.

There are also business challenges: how do the content providers and service providers work together? Clearly, the service providers do not want to become simple conduits for the content providers’ data (see data given in the previous paragraph) and the content providers want to ensure that their content is delivered in the most efficient way.

Finally, what kinds of business models work with consumers? In North America, consumers – author included – have been, for lack of a better word, spoiled with flat rate, “all-you-can-use” pricing models, with no cap on monthly usage. This model is not necessarily going to be relevant given the expected video traffic in the network.

The success of OTT video business depends on new more sophisticated pricing and business models for profitable delivery of these services. For example:

1. Collaborate with content providers to deliver superior video quality and content than that delivered via Best Effort pipes

2. Offer value-added network services such as bandwidth on-demand to enhance OTT

3. Allow users to control their services using advanced policy management – this provides a new stream of revenue and helps reduce churn by more effectively addressing users’ needs

Sophisticated pricing and business models can be implemented via intelligent and flexible networks, through “Smart Networking” – broadband policy management, tools such as Deep Packet Inspection (DPI), and application and service layer traffic management.

Other tools that can complete the above solutions include Content Delivery Network (CDN) and Transparent Internet Caching (TIC) to ensure bandwidth optimization and enhanced user experience. And Asset or Content Management to manage different metadata format, content type, and workflow processes are also necessary.

May 16, 2011 by Alan Repech in the theme A new generation of IP Networking

"Projections show that the network will need to support Tb/sec transmission rates by 2015-2016."

Is this yet another prognostication by some futurist pundit disconnected from reality? Nope.

It’s an excerpt from a grant notice from the United States Department of Energy posted April 7, 2011.

The grant is titled, "Terabit networking for Extreme-Scale Science." The grant description goes on to say, "While industry will supply the underlying optical technologies to build the network, DOE will need to develop and deploy the tools and services that will allow scientists to exploit this advanced infrastructure." The grant application window is planned for closure on May 23, 2011, so there’s not much time left to submit your proposal.

Extreme-scale science includes projects such single-cloud-granularity climate modeling "accurate enough to inform critical policy decisions (IEEE Computer, 2009)," and as stated above assumes, "industry will supply the underlying optical technologies."

So while "industry" is just now rolling out 100Gbps optical links and 1Gbps microwave systems, the demand for 10 times that is already upon us. And extreme-scale is not confined to the esoteric niches of high end computer science.

Mobile broadband growth has been well-documented, streaming video is commonplace, and business Ethernet services are growing fast. The number of end users and end devices is approaching extreme scale ("50B by 2020"). And it seems our "smart" access devices are actually getting "dumber." We’re running task-specific apps on phones and tablets as opposed to bloated, general purpose software applications on laptops. Cloud services and servers are critical components in the new model. Users demand quick, instant-on access and bandwidth headroom to accommodate upload/download bursts. Consider the number of Facebook users (very, very many) compared to the number of Facebook servers (very, very few).

The extreme-scale number of end users and devices needing always-on access to a very small number of extreme-scale servers are driving the need for bigger, faster, extreme-scale transport.

May 13, 2011 by Massimo Enrico in the theme A new generation of IP Networking

When I talk with colleagues and customers about the new generation of IP networking, there is one word that is used very frequently: "more."

When "more" relates to bandwidth, services, capacity, scale and quality of experience, everybody wants to hear it. Everybody likes the great business and social opportunities that we can foresee in the years ahead. But there is less enthusiasm when "more" is associated with complexity, cost, time for provisioning the network, and increases in overall operational cost driven by the type of services, number of connections and size of the network.

The key to handling the trade-off between positive and negative here is network management, the software that operates the network.

After working for many years in the telecom world, I know that network management is the big challenge for transforming network expectations and promises into reality. We are going to face the same experience we had several years ago in computers: where we once used to speak of CPU speed and RAM capacity, we gradually shifted the discussion to operating systems and finally tools and services. The scope, the service, is becoming king. This concept is now so important that we have entered the cloud era where the means – whether hardware or software – simply disappear.

Similarly, in the telecom world, the network – converged, with a lot of capacity and growing quickly – is becoming a given, just like the x86 platform in the computing world. The question is how to exploit this potential efficiently. To achieve that, we need to rethink network management, similarly to the shift that happened in the move from DOS to Windows and finally to iOS and Android. We will look more closely at that in my next entry.

May 11, 2011 by Dwight Witherspoon in the theme Living in a Networked Society

ICT is making life better for you, me and billions of other people around the world. Just increasing broadband penetration alone has been shown to increase GDP, boost social cohesion and improve efficiency in service delivery.

But as we combine the power of ICT with the dramatic growth of cities globally, we are seeing the emergence of a new phenomenon within the Networked Society: the connected, empowered city.

Ericsson and consultancy firm Arthur D. Little have developed the Networked Society City Index, which ranks 25 cities according to ICT maturity and development. The aim is to inspire and assist networked societies around the world by highlighting the ways top-performing cities successfully use ICT to grow.

I see ICT becoming even more important for cities. The index report points out that more than half of the world’s population already lives in urban areas, a figure set to top 60 percent by 2030. People are flocking to the cities, because that’s where the opportunities are, the jobs are, and where the money is. All of this is putting more pressure on our growing cities.

The best performers in the index – topped by Seoul, Singapore and Stockholm – use their extensive investments in ICT to tackle issues such as environmental management, infrastructure, public security, healthcare quality and education.

In lower-ranked cities, which have underinvested in ICT, governments are encouraged to provide digital access and ICT training for the underprivileged to reduce the digital divide.

As well as all its social benefits, ICT is a vital way to reduce environmental impact, making cities more sustainable while their populations continue to grow.

I see the Networked Society City Index as a way of stimulating discussion and generating ideas about how the cities of the world can harness the power of ICT to meet their challenges. What are the challenges your city is facing? And how do you see ICT helping?

May 10, 2011 by Elisa Bellagamba in the theme A new generation of IP Networking

Is MPLS-TP a new technology, similar to MPLS?

Absolutely not!

MPLS-Transport Profile (TP), as the name indicates, is a ‘profile’ of MPLS, hence it is MPLS!

If I were a mathematician, I would have formulated it like this:

MPLS-TP = transport subset_of_(MPLS + OAM_tools + NMS_or_Control_Plane)

By ‘subset_of_MPLS’ I mean restricting the scope to only those MPLS mechanisms that operate in a ‘connection-oriented’ way, in that they don’t lose or delete precious information along the traffic path that is useful to identify the traffic stream.

This ‘deletion’ is not necessarily a bad thing. It is in fact used in the core network for scaling purposes, but it is something that would change the nature of transport networks.

‘OAM (Operation and Maintenance) tools’ are at the heart of MPLS-TP. They are extensions to existing tools augmented with some new ones designed with transport networks in mind, but capable of running in any kind of MPLS networks in general and allowing traffic to be monitored in a deterministic way.

The new point here is that they can run in-band together with the traffic and are independent of the need for a full-IP network.

Protection switching, then, can be performed within 50 ms, which is a very strict requirement coming from transport networks.

Unlike Ethernet-based OAM, the IETF BFD based OAM is consistent with the MPLS architecture and existing MPLS deployments that can be enhanced for interoperability.

‘NMS_or_Control_Plane’ refers to the possibility of controlling and configuring MPLS-TP networks either with a GMPLS (Generalize Multi-Protocol Label Switching) control plane or with a centralized NMS (Network Management System). The latter closely resembles the SDH way of managing networks.

And this is in fact one of the most remarkable aspects of MPLS-TP: it introduces no real difference to the way operators are used to running their TDM networks. Think about how OPEX is kept low in this way and you’ll realize the importance of this point even more.

So…when will the standardization be ready?

This is the most frequently asked question. Speaking in IETF language, the working group will have finished its work in May/June and the last RFCs will be published during October/November.

But speaking in terms of deployments, it is already feasible right now because the standards are already stable enough for implementation.

Unified MPLS…what is it? Yet another flavor of MPLS?

No, this is simply a name we are using to emphasize the point that MPLS-TP is MPLS as much as traditional MPLS.

For the very purpose of having a single, transparent and optimized e2e network architecture, we are reflecting on the nuances of customizing and tailoring MPLS to meet operators’ needs. But all these ways are simply MPLS, unified MPLS.

What do you think about it?

May 9, 2011 by Elisa Bellagamba in the theme A new generation of IP Networking

Let’s zoom in on the cost-reduction aspect following up from the previous posts on the wider IP networking scope.

I think everybody will agree when I say that, to achieve a more optimized and cost-efficient network, we need to simplify the multitude of network layers currently existing in many deployments. Together with that, we see that metro and aggregation networks are transitioning from the TDM to the packet model due to the progressively higher bandwidth demands of applications, and the gains that the burst tolerance of large pipes, shared protection and statistical multiplexing can offer.

This is where MPLS comes into the game.

Multi-Protocol Label Switching is a mature technology that has been playing a dominant role in core networks for 15 years, delivering a diverse and rich set of services and applications. Just that? Well, no. The biggest strength of MPLS is the fact that it can run over a multitude of different layer 2 technologies and, at the same time, it allows multiple services to be converged over a single network.

You can see it as a layer 2.5 technology which acts as a very efficient "common" denominator across layer 1 and layer 2 technologies.

This is where its big success is coming from and why it has been seen as the perfect candidate to harmonize the metro and aggregation networks, maybe even extending all the way out to the access nodes.

Of course, the metro and aggregation networks have very different characteristics and requirements than core networks. For example, a much higher number of nodes (which means individual nodes need to be cheaper), simple point-to-point and ring topologies (compared with the mesh topologies in the core) and, last but not least, the need to monitor the traffic in a deterministic way in order to be able to meet the agreed SLA (service level agreement) and switchover within 50ms for the traffic in case some kind of failure occurs.

This is where MPLS-TP (Transport Profile) comes into the game. I’ll elaborate on that more tomorrow…

May 6, 2011 by Tomas Hasselrot in the theme A new generation of IP Networking

Are you addressing all subscriber segments? Is the service performance consistent? Have you explored business opportunities with content providers? Policy control is an essential tool for broadband operators to grow their subscriber bases, minimize churn and launch new services.

The key to success in any business is customer satisfaction. Policy control has mainly been discussed from the operators’ perspective, even though the true benefits are experienced by the subscribers. Combining clear terms and conditions with policy control helps you manage subscriber expectations, secures the user experience and gives subscribers the possibility to handle their subscriptions at self-service portals. All this gives subscribers a sense of control. The effect? Well, more loyal subscribers and increased revenues over time.

Naturally, there are also important benefits that have a direct impact on the bottom line. Not all subscribers have the same needs and requirements. By introducing truly differentiated offerings, new segments can be addressed and additional subscribers can be acquired. Typically, this would be a tiered service plan, comprising a range of bit rates and coupled with fair usage policies. Again, the cornerstone and enabling factor is policy control, and the result is a larger subscriber base.

Advanced policy-control solutions go beyond these basic capabilities. They create business opportunities by supporting service-level agreements and revenue-sharing mechanisms. This means that you form partnerships with third-party content providers and offer over-the-top services to your subscribers.

Having introduced tiered service plans and over-the-top services, it is critical to have effective tools for traffic management. While providing good quality of service, you want to avoid overdimensioning the network. The solution once again is policy control. Identifying and prioritizing traffic from premium subscribers and traffic associated with delay-sensitive applications will protect your margins as well as brand image.

The question is not if you should implement a policy-control solution but what characteristics to look for when choosing one. Two aspects are more important than others: end-to-end support and convergence. An effective solution must support enforcement of policies close to the subscribers to optimize the utilization of the critical radio and transport network resources. To be future-proof, it should also support multiple radio-access technologies (such as GSM, WCDMA and LTE) as well as fixed-mobile convergence with common subscriber profiles across the two domains. This enables bundling of services in new, creative ways and gives a competitive edge.

Policy control is arguably one of the hottest topics in the telecom industry, and for good reason. Broadband operators and subscribers clearly benefit from this type of solution.

May 4, 2011 by Bala Thekkedath in the theme A new generation of IP Networking

As time-to-market for new applications becomes a critical factor for business success, communications service providers are no longer looking for boxes that can provide a set of features.

What draws them to a vendor is a solution or an application that can be deployed and monetized at the earliest. The composition of the solution (an all-Ericsson solution versus a cherry-picked solution) is probably less significant today than the mere existence of a proven solution. A fully pretested solution significantly reduces the cost of new technology introduction for service providers and starts generating revenue much faster than a traditional network rollout that involves significant interoperability testing and vendor management costs.

Add in a highly competent professional services organization from equipment vendors and the entire technology migration cycle becomes as painless as it could be.

With the burden of keeping up with rapidly evolving communications technology outsourced, service providers now have the opportunity to channel their resources to areas of their core competence, which for most operators are formulating innovative service offerings and efficient service marketing. New service offerings reinforce the evolution in modes and types of communication and further the Networked Society and the connected lifestyle. And so the cycle goes.

What is critical in this whole process is the move from traditional product lifecycle management to solution lifecycle management. Solutions undergo a process similar to a traditional product lifecycle where every solution is designed, verified, documented and supported. Tactical product roadmap decisions driven by a broader solutions strategy are the key to a leaner, targeted and efficient portfolio.

Ericsson’s mobile transport solution is a classic example of how this approach benefits operators and end users. Products that comprise the mobile transport solution are designed keeping the integrated solution in mind. Individual product features are designed to drive synergies in the larger solution. For example, eight priority queues on the microwave transport segment allow a common quality of service (QoS) profile to be enforced from radio access to the mobile core. The ability to standardize QoS markings and policies, in conjunction with a service-aware policy control mechanism, helps service providers deliver revenue-generating, differentiated services.

As more and more of the network operations become part of the vendor’s portfolio, efficient networks and faster time-to-money are no longer an issue for the "customer" or "service provider." It becomes more personal. How does this shift in operator demand and business models change ways of working at equipment vendors? Will this solution focus foster cooperation or enhance competition within the vendor community?

May 2, 2011 by Christian Vogt in the theme A new generation of IP Networking

With the recent depletion of the Internet Assigned Numbers Authority’s inventory of unallocated global IP version 4 addresses, the discussion on the deployment of IPv6 has gained new momentum.

Certainly, the depletion as such is no immediate hazard to the functioning of the internet. Translation of IPv4 addresses, as a means to share addresses across multiple devices, has been a good-enough workaround and will continue for some time. But the depletion is resounding evidence that the IPv4 internet has reached its limits. And whereas the discussion on IPv6 transition previously mostly involved the engineering community, more recently top executives of major networking companies are taking a position.

Widespread IPv6 deployment has for a long time been impeded by an unfortunate incentives structure. Because IP is an end-to-end protocol, and because the technically recommended deployment model, Dual Stack, calls for IPv4 and IPv6 to be used side by side, local adoption of IPv6 has been associated with high costs, high risk and a lack of immediate benefits. As a consequence, IPv6 deployment stalled soon after the economies of scale of large software companies had facilitated support in operating systems and mainstream applications.

Universal IPv6 deployment will therefore significantly depend on “divide-and-conquer” techniques that simplify local adoption and provide immediate benefits. Even though Dual Stack continues to be the general recommendation, the following three divide-and-conquer techniques will play a key role as well:

• Enabling network operators to offer IPv6-based internet connectivity to end users without requiring a complete network upgrade

• Making content and applications available via IPv6 without requiring the originating provider to upgrade

• Retaining reachability of IPv4-only content and applications while enabling the IPv6 internet to jettison IPv4 support.

At Ericsson, we have played a major role making these divide-and-conquer techniques possible. The all-IP architecture of mobile networks decouples end-user traffic from network control functions, and thus allows operators to roll out IPv6 to end users without major network upgrades.

As part of the Third Generation Partnership Project, we are making key contributions to the development of this architecture. Content-delivery networks are essential in making content and applications available via IPv6 without requiring the originating provider to upgrade.

We are now collaborating with the world’s largest content-delivery network operator, Akamai, to help optimize content delivery in mobile networks.

And in the long term, the Internet Engineering Task Force’s emerging standard for IPv6-IPv4 translation will enable parts of the internet to become IPv6-only without losing connectivity to legacy content and applications. We have been supporting the development of IPv6-IPv4 translation, by prototyping, testing, and eventually productizing the technology.

The tools for IPv6 transition exist and hopefully will enable a faster deployment going forward. And as the number of IPv6 adopters increases, so will the motivation for others to follow. There is little doubt that, at some point, the IPv6 internet will reach the critical mass that it needs to live on its own – without backwards compatibility to IPv4. Opinions just differ on when this will happen. What do you think?

World IPv6 Day - event page

Apr 28, 2011 by Matthew Smith in the theme A new generation of IP Networking

Due to the way that we use the network in our daily lives, the IP network architecture is changing.

It’s changing in ways that we never could have imagined when spending on infrastructure was through the roof in the days of the internet boom. The mantra then was ‘build it and they will come.’ Well, many built, and then experienced a pretty bad time around a decade ago. But now ‘they’ have come, and come in a massive way.

And the networks (although able to adapt themselves massively to date) were not designed with the forces of mobility, cloud, smart devices, apps, user-to-user, network and machine-to-machine in mind. So a new generation of IP networking arrives, one that can act as a strategic asset and differentiator for an operator, helping them diversify revenue streams, achieve lower total cost of ownership, and – most importantly – keep customers happy and coming back for more.

Network convergence has meant many things over the years. And as services are converging on IP, so is the network itself. However, it’s not just about IP for the sake of IP. Not many can afford to roll out expensive and power-hungry IP engines everywhere in the network. A network must be architected to be simpler – the majority of traffic today is actually headed to very few locations and this has meant that packet and transport have finally come together, helping to flatten, de-layer the network with new mixes of lower-layer technologies evolving to complement IP, such as unified MPLS and WDM/OTN.

The network must also become smarter: injecting intelligence, with features like policy control at the optimum point: the edge and mobile core. The network needs to be aware of different users, their devices, content and locations like never before.

Finally, you need to make the network more scalable. And whilst adding higher capacity (in access – fiber or radio, or in transport) is critical, it is not the only scale concern. You also need to deal effectively with control plane scale due to all of those diverse IP sessions, services, subscriptions, deal with that signaling challenge, and manage the third dimension of the huge proliferation of devices and applications.

A new generation of IP networking – based on the idea of being architected to be simple, smart and scalable – can help respond to the challenge of adapting better to new business models.

Apr 27, 2011 by Mikael Eriksson Björling in the theme Living in a Networked Society

We live in a visual world. In fact the world has never before produced as many images, moving pictures and commercials as our eyes meet every day.

The increased individualism and the digitalization of media and communication make it more important than ever for people to express who they are using visual digital expressions. Digital appearance is today as important as physical appearance. Expressing identity, individuality and belonging is a major function of social networking services like Facebook.

Digital appearance is expressed through signs. With pictures, links, film clips, text, sound, frequency of status updates, likes, connections and comments, people paint a picture of their lives.

Semiotic studies is the discipline of studying signs and sign systems. Simply described, a sign is anything we can interpret. Smoke, for example, is (often) a sign for fire.

Understanding semiotic codes – what a certain image, link or status message express – becomes a key to understanding the social online world. All these signs are chosen carefully to represent who I am in the digital arena. Going forward this is an area that we all will experience as more and more important as we continue to represent ourselves on the net.

When I meet young people today they often say they don’t want Mum and Dad as friends on Facebook. The main argument for that is that their parents would not understand the signs and the codes they are using in communication with their friends – smoke doesn’t always mean fire (you can take it easy Mum!).

Young people are developing intricate language and rules for online communication and behavior (I will come back to these rules in a coming blog post).

Five to ten years from now your online story, your digital presence and your appearance will be of major importance when applying for a job. Without a personal online brand and identity, you won’t qualify for an interview.

What strategy do you have regarding your online presence and appearance?

Apr 26, 2011 by Håkan Eriksson in the theme A new generation of IP Networking

In the transformation of mobile networks into full IP broadband networks, advanced service-control functions are required at the edge of the network, in the so-called edge routers.

Differentiated management of Quality of Service (QoS) for different users, services and devices, provides "smart pipes" controlled by the policy control server. At the same time the edge router must handle the dramatically increasing data volumes. In a similar way, the fixed broadband edge routers increasingly require service awareness, driven by more bandwidth and QoS-sensitive applications such as HDTV.

Network traffic is becoming dominated by video downloads (streaming), so-called Over-the-Top (OTT) application data from clouds, and high-data-rate visual communication. Providing the best quality by optimizing QoS across radio access and edge routers is becoming a key differentiator for operators.

As a consequence of the dramatic data increase in the cellular networks, there is an increasing focus on small cell solutions to increase the capacity, complement the cellular network, or provide a combination of both. This takes the form of (1) WiFi complement, (2) Femto base stations or (3) Pico base stations.

The first – WiFi complement – is the most straightforward since the infrastructure is, in many cases, already there and the heavy data users are often stationary and indoors.

Femto base stations are in reality very similar to a WiFi modem, in that they also need to be connected to your fixed broadband, but with the important difference that they use the cellular air interface, instead of WiFi. Because most, if not all, data-heavy devices already have WiFi, the Femto does not really solve any capacity problem, but can be useful to improve indoor coverage for voice.

Pico base stations, unlike Femto, are small base stations that are connected directly to the operator’s radio access network, and are also controlled by the operator in terms of who is using them and so on. Pico adds substantial capacity and will be a necessary component of our mobile broadband offering. While realizing the difficult business case for the Femto to compete with WiFi in the home, the Femto industry is gradually moving towards the Pico solution.

Given the above multi-access scenario, it is important that the smart pipe logic ensures that a user is always best connected, be that over a cellular Macro cell, cellular Pico cell or WiFi access point.

Apr 21, 2011 by Rachel Ooi in the theme Mobile advertising: a market on the move

In my previous post , I talked about how particular consumer segments are very relevant for location-based coupons to work. The ‘work-hard, play-hard’ consumer segment was targeted by Sensis for their location-based mobile ad campaign.

This group has a few interesting traits. They are in their mid-20s to mid-30s, not married, professional, and have a fairly high disposable income level. They fancy night life, enjoy cool cars and hot dates, and love to party on Thursday-Saturday nights.

With the consumer’s name remaining anonymous, it is not difficult to model his or her lifestyle and behavior, with a proxy set of data and parameters. It is possible to predict their interests, purchasing behavior, as well as their propensity to accept campaigns and offers.

For example, take ‘Leon’ who is searching for cinemas while downtown at about 5 pm on a Friday. Leon will probably go for a glass of wine or have dinner as well. Offers such as a free preview of a mobile via a mobile trailer, together with click-to-buy tickets and wine and dine discounts will probably drive a high response. The same principle applies to how mega supermarkets apply data-mining models to predict cross-selling behavior of beer and diapers to fathers with young kids.

With my past in analytical marketing and specialized consulting expertise on CRM analytics, I know that modeling data-mining tools for the purpose of customer retention/attrition, churn management, cross-selling/up-selling, credit card frauds and bad debts management is not something new – and it is a common practice.

The same proxy modeling approach and intelligent algorithms can be applied to mobile advertising as well. Therefore, it is possible to predict a consumer’s propensity to accept an offer, their propensity to purchase an item, and probability to cross-sell services based on a series of search or browsing activities or interactions with few applications used. With basic consumer segmentation by clusters as foundation, proxy behavior can be modeled across distinct segments based on a defined set of variables and parameters, tested logics and algorithms.

With consumers’ revealing their needs through usage and interactions, marketers now have the opportunity to pursue precise profiled targeting to fulfilling consumers’ these desires and needs.

This is not all intrusive but delivers a promising consumer and user experience. This is the missing link between advertising sellers and consumers in the value chain today.

Do you agree? Is this the missing link?

Apr 20, 2011 by Don McCullough in the theme A new generation of IP Networking

The conversation has changed. Revenue is the new discussion topic for mobile operators and network suppliers.

Yes, box-on-box technical comparisons and total cost of ownership reductions are still important, but the network is now part of the business model discussion. It has to be since operators cannot save their way to victory in a MBB world that is moving toward a Networked Society with 50 billion connected devices. That reality has changed the conversation about the network.

In 2011, the IP network has to be a business advantage. Specifically, packet core, deep packet inspection, policy control are the way operators are providing differentiated services.

Ericsson is working with operators who are offering access to Facebook, Twitter or YouTube as add-ons paid services to the basic flat rate data service. DPI identifies these special service flows and packet core alerts a subscriber management system to exempt these bits from the limits on that basic service.

The backhaul and packet backbone systems provide priority for these service flows within the standard best efforts traffic flows. The IP network makes it possible to implement the operator's marketing insights and earn more revenue while offering customers what they want most.

Then take it another step, where the evolved packet core can optimize the use of spectrum in the radio network. Spectrum is the most limited resource in the MBB network. Policy control that does not take each cell spectrum usage situation in real-time into account when applying policy control is wasting precious resources. And it has to work with backhaul, packet core, backbone and subscriber management. We'll come back to EPC and spectrum optimization later.

Today, let's open up the revenue conversation. When I say "revenue opportunity" do you think "network"? And if the conversation has changed, is it translating into revenue yet?

How do we break through the IP, packet core and policy control hype to make more money? Or is it just hype? What do you think?

Apr 19, 2011 by Mikael Eriksson Björling in the theme Living in a Networked Society

Gone are the days when the TV schedule was an important part of our lives.

Today there is an overflow of content, ranging from traditional media to all the content created and published by users around the world.

According to the Pingdom website, in 2010, there were about two billion internet users generating two billion tweets per year, writing 152 million blogs and uploading 20 million videos to Facebook every month, 35 hours of video to YouTube and 3000+ photos to Flickr every minute.

To this we can add billions of comments and status updates every day.

The content visible in our feeds and timelines provided by our nearest and dearest is almost infinite.

What article, blog post or YouTube video should I read or watch? We learn to judge a recommendation in a split second, mostly by judging who the sender is.

Much of the media we consume is communicated and recommended by friends who act as filters and we quickly learn who has the same tastes and interests as we do.

Media and communication has become totally mixed. We express ourselves by linking to film clips, music or articles we like. These links represent us and our digital appearance.

Today there are smart applications on the market such as Zite and Flipboard that render our feeds, media and communication sources into a nice readable magazine.

Communication and media becomes one.

Going forward, we will see applications that learn our patterns and taste, that are sensitive and smart and that take location, mode, presence, and many other available data into consideration, providing us with our very personal media and communication experience.

What do you think about this kind of future?

Apr 18, 2011 by Matthew Smith in the theme A new generation of IP Networking

As we move into the networked society, where anything that will benefit from being connected will be connected, operators face both an opportunity and threat from the combined force of the popularity and proliferation of a wide variety of smart devices, video/OTT and the rush to the cloud.

Through progressive generations of broadband access technology to LTE and GPON, the backhaul, transport and edge/core network has adapted to respond to the differing challenges: be they capacity, application or device specific.

Here on Thinking Ahead, we are launching this new theme to find out what you think about how the whole IP network can be refreshed to become the strategic differentiator that enables network operators to cope with the changed business climate.

Covering a broad range of converged network technology trends, this theme aims to generate a discussion around the 4th generation IP network, which will enable operators to deliver differentiated, personalized services from a simpler, smarter and more scalable IP infrastructure.

What will this infrastructure mean for you?