Resource Control Point
- Details
- Category: Products
The RCP provides a software solution for real-time, network state-aware admission control, along with multi-service integration with voice and video applications or service delivery platforms (SDPs). The RCP dramatically improves the quality of end-user experience for real-time IP services; optimizes network utilization, which minimizes the need for over-provisioning and over-investment in the access and aggregation networks; and enables automatic application-driven real-time control.
IPTV, VoIP or other demanding and potentially high-value applications require network resources to be carefully controlled to meet and exceed expectations on performance and Quality of Experience (QoE). The Resource Control Point (RCP) provides policy- based Resource and Admission Control in IP networks to ensure QoE and resource availability to mission critical applications.
The RCP is a software solution that logically resides between the transport network layer and the applications. From its control plane position the RCP dynamically detects network topology changes and controls the use of network resources for applications to provide QoS guarantees. When an end-user or end-device attaches to the network, the subscriber profile is acquired by the RCP and applicable subscriber policies are activated in the policy enforcement function. When application specific sessions/flows are initiated, the admission control decisions can be made to accept or deny the session with respect to subscriber profiles and network resources. The RCP may activate per session/flow policy enforcement as a result of admission decisions.
The RCP supports managed services by interfacing to application servers and policy enforcement devices (RCEF/PCEF). It supports over the top IP services by interfacing to devices performing packet inspection and policy enforcement functions (DPI/PCEF). The above illustration shows how the RCP operates in a multi-service environment where application servers are part of service control.
Product Features
The architecture of RCP consists of a collection of distributable software components. This allows the RCP to scale from very small deployments where all components are located on the same hardware to very large deployments where the different components are distributed and scaled over multiple servers/blades.
Resource and Policy-based Admission Control
The RCP can be deployed in a multi-service environment where the network resources are shared between the services. Multiple services and applications can be aggregated and controlled into a few classes of service (CoS) at the network level. The mapping of application services to network CoS is controlled by policies.
The RCP implements advanced policy functions that provide admission control. The network topology, resource allocation and reservation state across it are dynamically represented in the resource map together with subscriber and access line policy data. This data is dynamically attached to the topology using the mapping of physical network location and subscriber assigned IP address. The resource map may be distributed over multiple platforms for scalability reasons and it accepts changes in real-time.
Subscriber profiles and topology resources
The RCP can be provisioned manually with policies and resources through the DSM Graphical User interface and by uploading policy files to the internal repository (resource map) of the RCP. However, a key strength of the RCP is that large parts of the information can be auto-provisioned through various interfaces to the surrounding elements in the NOC and in the network. Auto provisioning saves management overhead and effectively avoids duplication of data management tasks which may otherwise lead to data inconsistencies between different systems due to the human factor.
Systems Management
Operax delivers a Distributed Systems Manager (DSM) component that provides system administrators with functionality for centralized deployment and management of large scale Operax RCP deployments. It also provides interfaces for FCAPS management.
Interfaces
The RCP is compliant with the ETSI TISPAN RACS and is partly compliant to the PCRF of the 3GPP PCC architecture. In addition there are WebServices interfaces and APIs for vendor specific interface integration.
- Resource and Policy based Admission Control (RaPC) and Policy Enforcement
- Rx / Diameter; RFC-3359, RFC-3588, 3GPP TS 29.214
- Gq’ / Diameter; RFC-3359, RFC-3588, ETSI TS 183 017
- Rq / Diameter; RFC-3359, RFC-3588, ETSI ES 283 026
- Operax WebServices RaPC; This interface is friendly to IPTV and web- services applications and comes with a full WDSL description.
- Gx / Diameter (on roadmap); RFC-3359, RFC-3588, 3GPP TS 29.212
- Pkt-mm2 / COPS; PKT-SP-MM-I05
- Subscriber
- e4 / Diameter; RFC-3359, RFC-3588, ETSI ES 283 034
- RADIUS; RFC 2866
- Pkt-mm4 / RADIUS; PKT-SP-MM-I05
- Topology resources
- BGP, OSPF, IS-IS, SNMP
- Operax Provisioning Tool for file-based provisioning
- Operax IRc XML/SOAP Topology and Resource Provisioning API; This is an open XML API for topology and resource provisioning of aggregation networks.
Deployment
The RCP can be deployed in a few central NOC locations or in a larger number of regional POP locations. The RCP in defense battle fields is typically deployed on standard servers in military vehicles in the field (e.g., Command Posts).
In NOC deployments the RCP meets telecom carrier grade high-availability requirements. It can be scaled to any size of network by adding more servers, so that large size networks can be covered through divide and conquer where different servers have responsibility for different parts of the network. In a carrier grade configuration the RCP servers are deployed with 1+1 resilience (active + hot standby) where dynamic state is maintained across the pairs to allow sub-second fail-over with fully maintained dynamic state.
In battle field deployments of the RCP, geographical resilience (N-way active) is used. The objective is to share state between all instances of the RCP in the field deployment so that the system is fully functional as long as any single instance is alive. The solution is designed to synchronize state between the instances with small overhead in order to handle network split and rejoin scenarios with minimal bursts of control traffic.
Platforms
The platform in this context is the combination of hardware and operating system. The RCP is currently supported on CentOS, RedHat and SUSE Linux operating systems using x86_32 or x86_64 hardware. The RCP is a software product that includes all application level software and middleware required to run it on the above mentioned platforms.
The RCP is distributable on standard hardware nodes, e.g. a rack-mount server or a CPU blade (e.g., ATCA) in chassis. Depending on performance requirements, the RCP can also be deployed on virtualization platforms such as VMware, Citrix/Xen and VirtualBox.
The product can be ported to other platforms depending on agreements.
Updated: Friday, 26 August 2011 10:13
