Applying predictive algorithms to monitor, maintain, and evolve OSS network performance efficiently

Increasingly pervasive Access Networks have become known as Optical Subscriber Services (OSS). Distributing high-bandwidth Internet Protocol-enabled services across all market segments is seen as essential in making the necessary returns on investment over large service areas. Delivering OSS infrastructure and services must occur within a competitive economic environment with levels of sensitivity unheard of in traditional telecommunications practice. Telecom and cable corporations are investing heavily to migrate to an OSS infrastructure for broadband data services, and the subscriber service continues to steadily evolve along with the challenges for both service providers and equipment vendors as new layers of performance and services are built up above basic functions of capacity, scheduling, and supporting interfaces with the user. OSS will support new classes of performance-sensitive multicasting services, in particular, video-on-demand over Multicast Point-to-Point Protocol. Bandwidth fluctuations seen in real-time traffic from media sources must be accommodated by the underlying infrastructure that OSS provides. New enabler architectures must predictively manage OSS bandwidth performance or face regulatory scrutiny relative to new local competition guidelines. Existing technologies that have been used to manage First Generation Broadband-in-Broadcast networks do not scale appropriately to a large number of individual subscribers.