As service providers deploy larger and larger SIP access networks, scalability problems are presenting unique challenges, particularly from an operational standpoint. Deployments that scale beyond the number of users serviceable by a single Oracle Communications Session Border Controller (SBC) – as well as deployments that use a geographically redundant SBC for catastrophic fail over purposes – encounter edge reachability problems. In general there are two coarse techniques that carriers use today to support end-point populations that exceed one Oracle Communications Session Border Controlled capacity: they either use a DNS-based distribution mechanism, or they will pre-provision endpoint to point to specific SBCs (manually load balancing them). Each of these solutions has its drawbacks. End users – many of them familiar with load balancing equipment deployed to scale protocols such as HTTP or SMTP – have expressed interest in a device that will perform dedicated load balancing for their SIP endpoint.

The Subscriber-Aware Load Balancer addresses the need for scaling a network edge to millions of endpoint. Designed as a standalone system, the network architect can deploy a physical or virtual Subscriber-Aware Load Balancer capable of supporting an extremely large number of endpoints, where an endpoint is defined as a unique source and destination IP address. The Subscriber-Aware Load Balancer aggregates signaling from large endpoint populations to reduce the edge reachability problems.

The network architect reduces this problem by deploying clusters of SBCs supported by the Subscriber-Aware Load Balancer. These SBCs can be operating as either Physical Network Functions (PNFs) and Virtual Network Functions (VNFs). The Subscriber-Aware Load Balancer supports clusters of homogenous or heterogenous groups of PNFs and/or VNFs.