Aktuelles
25. Mai 2007
TELES Informs: The Carrier-grade Super Softswitch
What does the switch of the future look like? What do carriers expect from a “super softswitch” today? What does the “carrier-grade” term from the TDM world really mean and why is it becoming ever more necessary? The heart of every NGN is the softswitch – the software-based telephone switch for IP networks. As the “brain” or “intelligence” of the network, the softswitch controls the media gateways in the NGN environment and takes over the complete signaling.
With a flexible architecture – as foreseen in NGN – the softswitch must make efficient voice routing possible; simultaneously making open interfaces for new applications and services available to the carrier.
In addition, the typical step-by-step migration from a TDM-based network to a Next Generation Network must be taken into consideration. The softswitch should be able to integrate so-called "legacy equipment" into the new network. The existing TDM network still represents in most cases the revenue generator for the carrier and it must be ensured that the softswitch provides all possibilities for smooth integration/migration.
One aspect which is often overlooked in the softswitch and VoIP discussion is that the softswitch should bring well-known carrier-grade attributes from the TDM world. The term "carrier-grade" is often to be found in the glossy brochures of NGN equipment manufacturers. But what is exactly is behind this? What is "carrier-grade" exactly?
The reliability rate of 99.999% (“five nines”) is often given with the “carrier-grade” term, whereby the question “which reliability rate?” must also be asked: that of the individual softswitches or of the whole network? As well as the reliability rate, carrier-grade also means that the softswitch is scalable, can be simply maintained and allows easy integration with third party components (“interoperability”).
Network Concept for 99.999% Reliability
The “five nines” percentage is the golden standard when speaking about network availability and uptime. In order to make this value clear: five nines means that a network is available all year round with an allowable downtime of five minutes in that year, with all maintenance work including, for example, software upgrades already included!
A softswitch should not only be delivered on service platforms with 99.999% availability, it is also necessary that the software itself is fault tolerant designed in order to properly treat faults and, accordingly, ensure the highest level of availability possible. It is not enough that the software alone displays five nines availability. End-to-end availability must be provided.
All hardware components must be included for a statement of network availability. However, because the softswitch is the central component in the network, its robustness and rate of reliability must correspondingly be the highest in the network. In order to reach 99.999% network availability, a rate higher than the “golden standard” should be aimed for in the softswitch. Despite the fact that a highly available platform doesn't guarantee a highly available softswitch, the subject should still be touched upon here as the hardware platform directly influences the softswitch availability. These days, commercial servers and blade-based servers are typically deployed as softswitch hardware architecture.
Commercial Servers
In the case of server-based softswitches, commercial NEBS (Network Equipment Building System – American standard) compliant computers are typically deployed. This standard sets the dimensioning, safety and environmental conditions for telecommunications equipment. NEBS guidelines are one reason, amongst others, why telecommunications networks still work during and after earthquakes and storms.
The disadvantage of these servers is that the scalability is not optimal. Higher scalability – if at all possible – usually requires the deployment of additional servers which leads to additionally required space and cabling and resulting higher operational costs.
The distribution of the signaling traffic on multiple computers should also be taken into consideration. One further disadvantage can also be that other network components like, for example, SS7 signaling gateways or media gateways are only available in a different form and must accordingly be mounted in a different rack.
Blade-based Platforms
Blade-based systems allow so-called COTS (Commercial Off-the-Shelf) processor blades to be deployed in a unified NEBS compliant chassis. Simply deploying additional blades in the chassis allows the network to be extended - without requiring extra cabling or space. Thanks to the more compact architecture, the high power density as well as the simpler cabling, these systems allow the service and maintenance costs to be reduced in direct comparison with conventional commercial servers.
In the case of blade-based systems, there are mostly third party products (e.g., media gateway blades, SS7 blades, IP router blades et al.) which can be seamlessly integrated with the blade chassis with minimum space and simple cabling requirements.
The only disadvantage is that the initial investment costs are higher in the blade chassis and the blade server than is the case with comparable commercial servers. However, this picture changes as soon as network extensions become necessary.
Redundancy
When we talk about redundancy in a super softswitch, we don’t only mean redundant hard and software. Geographic redundancy is also an important key component for robustness and should be supported by the super softswitch.
Once a softswitch is able to manage 100,000 ports in an NGN scenario, a single site no longer becomes sustainable due to reasons of reliability and availability. Even in cases where there is a failure at one site, in an NGN softswitch functionality should always be at full capacity.
By today’s standards, it would be an untenable state of affairs if the telephone was not available as a means of communication in the face of a terrorist attack or other disaster. In order to avoid such instances, a super softswitch and its redundant components should be distributed across different locations. Only in this way can a higher degree of availability and robustness corresponding to the 99.999% be reached.
Interoperability
The structure of an NGN creates the possibility to construct a network based on open standards which is also made up of the “best-in-class” components of a number of different manufacturers. Given the diversity of applications and manufacturers currently in the VoIP/NGN environment, it is no surprise that there are also a number of different protocols.
A Super Softswitch should therefore support as many signaling protocols as possible in order to guarantee “multi-vendor” interoperability. Because protocols develop with the times, it is vital that the softswitch manufacturer offers an “Interoperability Partner Program” and is able to adapt the protocols to its products accordingly. Only this way is it possible to build a future proof NGN!
As well as the signaling protocols, the softswitch must be equipped with the appropriate OSS/BSS interfaces. Diverse fulfilment, assurance and billing interfaces are indispensable in a state-of-the-art softswitch. Interoperability, based on open standards, allows the carrier to provide the best end-to-end solution while reducing system integration costs at the same time.
Summary
Today’s telecommunications world is changing constantly and every carrier is compelled to migrate his network to an NGN architecture or build anew in order to take advantage of the opportunities which arise from new IP-based applications.
A carrier-grade super softswitch is the heart of the VoIP/NGN and therefore must fulfil the requirements mentioned at the beginning of this article. For a successful VoIP/NGN start, the balancing act between provision of innovative services and the well-known robustness and performance of the TDM world must be mastered, as the carrier-grade attributes form the basis for the accustomed to availability and quality of a telephone service.