Optimization of Internet service provider networks

Transporters and service providers nowadays work in a very competitive environment. The customer has multiple choices for access – cable, DSL, satellite, wired wireless IP, hosting and wholesale – and much more when selecting his data center provider. And switching providers is easy. In addition, many airlines have made massive investments in network infrastructure, such as access and backbone networks, and are seeing their revenues decrease due to competition. Service quality is an absolute must, and service providers are looking for new ways to better use their existing infrastructure to differentiate and appropriately charge for this added value.

Triple play packages are a hot topic with new revenue opportunities enabled by fast, cost effective IP network delivery of voice, data and multimedia content. There is a trade-off between density in users, bandwidth and quality of service that carriers must successfully juggle. Qos is key – the services offered such as VoIP, gaming, streaming, etc. In the various application bundles offered, providers can rob their main asset in customer satisfaction. Large amounts of unattended traffic lead to dropped calls, terrible quality calls and unplayable streaming media. Packet loss, latency, and jitter must be effectively managed on request to provide a good end-user experience. There are many threats in the form of viruses, worms, p2p, DDOS attacks and simply a huge amount of traffic that affect performance and lead to customer churn. Providers are always looking for ways to maximize the number of users and traffic on a given link and minimize expiration (multiplexing)

With advanced services and applications such as IPTV, IMS, VoIP and VPLS driving a wider range of advanced requirements, application aware Qos is a must to protect high quality video and voice streams from other applications with minimal configuration.

MPLS-based services have gained widespread acceptance, both from carriers and customers, because they are dynamic and flexible, ideally suited to geographically dispersed environments, and can be delivered over existing network services. Separate MPLS offers only part of the promise of an application-optimized WAN infrastructure. ExtendedMPLS enhances traditional MPLS by providing a unique capability to automatically route and cascade traffic to alternate classes prior to saturation, as well as improved QoS granularity. ExtenedMPLS technology has traditionally been suitable for the end-user market as it requires units on each side of the link, but intuitive providers can bundle the equipment into their managed service offering.

Now an organization can take full advantage of MPLS and implement a complete network infrastructure that recognizes and fulfills the needs of the organization.

“Organizations are increasingly seeing network infrastructure as a commodity – what they’re looking for now is performance guarantees for their critical applications and user groups. Using technology within our service, we can deliver previously unattainable levels of service assurance and network visibility within the most demanding Our goal is to improve the problems of MPLS and maximize bandwidth utilization in any network topology (Ethernet, Fiber, Leased Line, XDSL, Frame Relay, Satellite, etc.) Barry Desmond of former Sirocom.

The best solutions are modular platforms on which a complete portfolio of optional, performance enhancing and security applications can run. (Bandwidth management, firewall, traffic detection, VPN, web caching, content filtering and mail gateway – these can be implemented transparently) Real-time alerts allow the service provider layer 7 to understand traffic patterns and be notified if there is an application on the network that may be outside of a normal usage profile falls. Quotas can also be easily implemented per user. This allows problems to be clearly displayed and network operators to be quickly led to a particular problem. The uniquely flexible platform, which is paramount when dealing with major service providers and the unique feature sets, allows ISPs to offer enhanced services to customers at multiple levels of complexity with real substantial benefits for both the ISP and the customer .

Such technology allows ISPs to simultaneously manage their internet link and links to customers on the same box. With a single unit, one can manage the internet link on the WAN card and all customers as groups on the LAN card (or even both on one card if you wish). One may also have dependencies in this I.e. you can have a burst trigger or threshold on the WAN links that you use to control the LAN groups and you can also group them (usually for service level groups).

The major problem that many face is that they have no control over the speed at which non-TCP / IP traffic gets to them, destined for the WAN. This results in longer queues, more latency and ultimately serious packet loss if the burst doesn’t stop. This has critical implications for latency sensitive applications such as VOIP and video conferencing etc. Devices with tcp capacitance capabilities can control the flow rate of any protocol in any direction. Therefore, it never needs to have queue traffic. The technology revolves around queue avoidance, while competitors have to queue to manage. We allow high priority traffic to flow freely through the network without having to queue packets of any type, completely eliminating packet drops and completely relieving routers and other equipment. The result is a much smoother flowing and stiff network. In fact, high priority traffic has its own lane through traffic that works without delays and it is dynamic and adapts to traffic needs.



Source by Sean Brett