 | Issue: | EMEA 2013 |
| Article no.: | 10 | |
| Topic: | Diameter signaling: the good, the bad, and the future | |
| Author: | Doug Suriano | |
| Title: | CTO | |
| Organisation: | Tekelec | |
| PDF size: | 301KB |
About author
Doug Suriano is the CTO of Tekelec where he leads product development for intelligent mobile broadband solutions. His R&D team is responsible for architecting session, policy and subscriber data management products for software-defined networks. Mr Suriano joined Tekelec from Dynamicsoft, Inc. where he was VP, Engineering. Prior to Dynamicsoft, Inc., Mr Suriano served as Chief Information Officer for QAD Inc., an enterprise resource planning software developer, responsible for the company’s global information technology infrastructure, business systems and development support. Mr Suriano also managed IT for the U.S. Marine Corps.
Doug Suriano earned a Master’s of Science degree in information technology at the U.S. Naval Postgraduate School and a Bachelor’s of Science Degree at the U.S. Naval Academy.
Article abstract
Diameter is an authentication, authorization and accounting protocol for telecommunications networks that will generate a significant amount of signaling, as more sophisticated policy and charging systems require more messaging and communication. The dynamics of chatty apps and LTE-driven services will drive operators to lay a foundation for self-organizing, self- optimizing, software-defined networks that autonomously respond to surges in signaling and other unpredictable events. As such, more robust systems are needed for monitoring traffic, load balancing and congestion control, as well as enhancing the customer experience and potential for revenue generation.
Full Article
Diameter signaling [‘Diameter’is an authentication, authorization and accounting protocol for telecommunications networks] surges are to be expected as more sophisticated policy and charging use cases trigger more signaling messages. But the increased pinging and communication should be embraced as a harbinger of more revenue opportunities for operators.
The multi-tasking, mobile-social nature of subscribers, and the multi-session nature of new devices are making the concept of ‘busy hour’ obsolete. The majority of devices today are continually connected and perpetually busy thanks to ‘chatty’ apps like Facebook, Skype and WhatsApp. Constant signaling is also driven by the dynamics of Voice over LTE (VoLTE), Rich Communication Suite (RCS), and new machine-to-machine (M2M)-, over-the-top (OTT)- and advertising-driven business models.
For operators in Europe, the Middle East and Africa, market pressures, regulatory changes and spectrum auctions are expected to drive more LTE roll outs.
TeliaSonera, Telekom Austria, Deutsche Telekom and O2 Germany have been among the first, and Telefónica Spain, Vodafone and others have also announced more LTE spend for this coming summer.
Further, by the end of 2015, about 7.5 million Middle East subscribers and 11 million African subscribers will be on LTE, according to figures issued by Signals and Systems Telecom.
As LTE momentum builds across EMEA, there will be more sophisticated roaming arrangements, concurrent data sessions, video streaming, quality of service (QoS) guarantees and behavioral changes via social networking. All of this will cause increases in Diameter signaling. In fact, Tekelec’s LTE Diameter Signaling Index® predicts that LTE Diameter signaling traffic in EMEA will see a 320 per cent compound annual growth rate (CAGR) from 2011 to 2016.
That CAGR will be driven by increased Diameter-based communication among gateways, policy servers, charging systems, subscriber profile databases, and mobility and session management functions.
Some of the biggest increases in signaling traffic will be due to online charging and policy functions.
Charging alone can generate multiple Diameter messages per session – depending on the type of session, billing arrangements, and other factors pertinent to the service. VoLTE and RCS and anything requiring dynamic charging, quality of service, and shared usage plans will further drive up Diameter signaling.
Also, increased mobility will require that policy be active for longer durations, maintaining session ‘state’ and tracking subscriber usage for hours or days as subscribers move around networks.
For example, if you think about policy in the context of authentication, messages have to be invoked every time a subscriber powers on an LTE device, or activates a new service. An event like a new iPhone launch – where millions of devices get activated within a condensed window – can suddenly generate a surge of Diameter messages. Similarly, an event like the Olympics can trigger millions of messages when smartphones and other devices become a conduit through which people share their experiences.
The good news is that the swell in Diameter traffic will correlate to the sophistication of services and devices. More sophistication and more personalization means more revenue-generating opportunities.
The bad news, however, is that a wait-and-see approach to addressing Diameter traffic is no longer acceptable. As operators accommodate popular applications like Apple’s FaceTime over cellular or QoS-driven VoLTE and video, or tailor Mobile Ads with subscriber data, Diameter signaling will need to be addressed. If not, it will become increasingly difficult to personalize mobile data services. Operators need to apply advanced policy rules, and they will have to manage the consequent increase in Diameter messages among policy and charging systems, and subscriber databases.
Why a centralized diameter approach?
‘Signaling storms’ have already hurt network performance and compromised the customer experience. In the shift from unlimited to usage-based data plans, for example, signaling surges were experienced by several operators.
Operators that are now in the planning stages have learned from others to incorporate Diameter signaling strategies into their network architectures. The lesson: a separate, intelligent New Diameter Network is needed from the outset of LTE networks, and cannot wait until ten million or 20 million subscribers join the network. A meshed architecture—the direct connections between each network element—simply cannot scale sufficiently, even at low subscriber numbers.
Scalability problems abound in the SS7 world, too. But where SS7 did handle congestion, traffic overload and traffic throttling, Diameter does not!
Operators assumed that IP networks would handle those responsibilities on a ‘best-effort’ basis, but now it has become obvious that the Diameter protocol has to be deterministic to address issues at the IP layer.
Currently, Diameter is at the end of year two of a five-year development cycle and the IETF is leading standardization to incorporate new capabilities into the protocol. In the meantime, operators should increase their focus on the signaling core.
Building dynamic and resilient networks with a diameter signaling router
As operators and vendors learn more about the impact of applications on the Diameter network, operators are focusing more on the signaling network and the role it plays in supporting innovative and sophisticated applications.
Traffic engineers that understand how a signaling network works are looking more to Diameter Signaling Routers (DSRs) as a means to prevent further outages and optimize what they can do to enable or participate in new mobile-data business models.
The DSR enables an architecture that reduces the cost and complexity of the core network, and it helps IP networks to grow incrementally in support of increasing service and traffic demands. Additionally, the DSR facilitates network monitoring by providing a centralized vantage point in the signaling network.
When evaluating solutions, operators should consider how robust and proven a solution is in 3G mobile data and LTE networks, as well as its roadmap for the future.
Important for a solution today is the ability to handle hundreds of thousands of messages; tens of millions of concurrent sessions; and millions of subscribers and devices. Also important are extensive scalability and congestion control features. But for the future, there has to be a path to virtualization, as operators are going to want to move the DSR and related elements like policy and subscriber data management into the cloud. As more architectural agility becomes necessary to handle signaling and data traffic surges, the DSR’s operations, administration and maintenance (OAM) functions will have to be virtualized.
Already, operators in the Asia-Pacific region and IP Exchange (IPX) providers are deploying DSRs with virtualized elements.
As the DSR becomes increasingly critical, it will become a key component within an independent control layer – one that should be comprised of signaling, policy and subscriber data management. Tekelec calls that independent layer the ‘New Diameter Network,’ a Diameter signaling layer in the network core that handles routing, traffic management, load balancing and protocol interworking. In that context, the DSR integrates with the Policy Server (PCRF) in order to define business rules for new service plans. It also integrates with subscriber data management (SDM) to personalize services and evolve them according to consumer circumstances and preferences.
As a part of the NDN, the DSR, Policy Server and SDM manage the constant pinging among essential LTE network elements, including:
• Offline Charging
• Policy Servers (PCRF)
• Online Charging (OCS)
• User Data Repositories
• Mobility Management Entities (MMEs)
• Policy Control Enforcement Points (PCEFs)
• Session Management, like a Call Session Control Function (CSCF)
In essence, the DSR becomes the network’s central nervous system, monitoring all Diameter traffic and making decisions regarding load balancing and congestion control based on its knowledge of the entire network, as opposed to a specific segment of the network.
These capabilities lay the ground work for software-defined ThinkingNetworks™ – networks capable of self-organizing, self-optimizing, and self-determining responses to unprecedented and unpredictable events. In essence, networks that have to think for themselves, yet remain malleable enough that network architects can adjust network resources as needed.
While operational efficiencies and cost savings are some of the benefits of ThinkingNetworks, it will ultimately be innovation and personalization that are the crowning benefits.
To get to that point, there are four phases to undergo:
Phase 1, the creation of the New Diameter Network – a centralized, intelligent Diameter control layer governing all equipment that uses the Diameter protocol.
Phase 2 comes when the NDN moves to the cloud, creating a fundamental shift in the way operators improve network scalability and flexibility in terms of sessions, transactions, and throughput. Tekelec calls this phase Cloud XG™, as the next-generation of cloud will depend on network function virtualization and software-defined networks (SDNs). Here, operators will look to dynamically add and remove compute resources to available hardware, and to dynamically manipulate traffic flows entering or leaving the cloud.
Phase 3 is MobileSocial™, where subscribers are intimately known by their providers as individuals and as members of greater social circles – personal and professional. This is where the rubber meets the road, as operators become digital-lifestyle providers, customizing customer experience according to real-time data coming from networks and Big-Data sources.
As operators unleash the valuable contextual data they hold, they build opt-in, mobile advertising, and other personalized offers to OTT application and content providers.
Phase 4 culminates with ThinkingNetworks, which embody all critical components of the previous phases. Founded on policy-directed SDN control and network event listeners (such as event records, analytics, alarms and probes), ThinkingNetworks give an analytical view of the relationship between all of the services supported – the characteristics of each as well as the impact of one service on another.
The end goal is to significantly reduce network costs by giving service providers the ability to dynamically assign compute and storage resources where and when needed in the network, as well as to enhance service, application, device- and user-awareness in order to further personalize services.
The ThinkingNetworks vision provides a roadmap to turning those goals into reality with carefully planned steps that get operators from where they are today to where they want to be tomorrow as Digital Lifestyle Providers.
