Processing communication traffic

Details Processing communication traffic Processing communication traffic

1999-07-02

2003-06-03

Vincent, David (Department: 2732)

Multiplex communications

Pathfinding or routing

Switching a message which includes an address header

C370S466000

Reexamination Certificate

active

06574224

ABSTRACT:

This invention relates to arrangements and methods for processing communications traffic, and in particular to integrated circuit constructions for performing such processing.
BACKGROUND OF THE INVENTION
Traditionally, two types of legacy telecommunication networks have been developed. The first type is connection oriented and is used for the transport of narrow band voice traffic, typically carried in TDM frames. Such networks comprise for example synchronous or plesiochronous networks. The second type of legacy network is connectionless in nature and is used for the transport of broad band packet or cell-based data traffic. There is currently a drive towards unified networks which provide end to end transport for both voice and data services. However, as there is a well established voice network base, network operators are naturally reluctant to replace such legacy networks. This issue has been addressed by providing broad band (asynchronous) overlay networks which interface with the established TDM networks to provide a voice and data transport function. At the interface between the two networks, an interface function maps TDM frames into packets or ATM cells and vice-versa. ATM is of course just one example of a packet based network.
A goal of development of telecommunications networks is to realise the potential integration of real-time and non-real-time services. The key examples of these two types are voice telephony (real-time) and computer data (non-real-time). Voice telephony is served predominantly by circuit switched connection orientated networks that are arranged to deliver a guaranteed quality of service. Such networks are implemented by transport and switching systems that use a time division multiplex of circuits. Computerdata is served predominantly by the Internet which uses a packet forwarding connectionless mode of operation or Internet protocol (IP), that is a workable paradigm best suited to the burstiness of traffic demand and general non-deterministic nature of this traffic type.
The network supplier industry, predominantly comprising PTTs, is poised to decide the technologies and protocols that will serve the integration of these two different types of service. The two main contenders for universal transport and switching are ATM (asynchronous transfer mode) and IP (Internet protocol), although there are many other legacy systems and nascent technologies that may offer specialised solutions to carrying key services.
ATM (asynchronous transfer mode) has been designed from the outset to adapt to many different types of communications traffic. ATM is a connection orientated network mechanism, allowing dynamic bandwidth configuration and flexibility as a key advantage over circuit switched networks. ATM has adaptation layers for carrying given services over ATM transport and switches. However, the fixed length of ATM cells, while suitable to segmentation of long data packets, thereby simplifying and streamlining switching technologies, is still too large for certain compressed voice services, that suffer a cell assembly or ‘cellification’ delay affecting existing network delay budgets and acceptable voice characteristics.
This issue has been addressed by the definition of Adaptation Layer 2, that is distinct from other ATM adaptation layers, since it de-couples voice packets from ATM cell boundaries, and since mini-packets from several calls can be multiplexed into a single ATM connection. This multiplex is asynchronous to the cell boundary. The multiplex effectively introduces a new switching layer above the ATM layer.
IP was designed for computer communications, although it has been recently demonstrated to be suitable for real-time services if congestion can be controlled to an extent that permits an acceptable quality of service to be achieved. Much activity surrounds investigations and implementation of mechanisms for limiting the degree of congestion and controlling quality of service in IP networks:
The real-time Protocol (RTP) was devised to carry real-time traffic, in particular voice and video, services the existing transport protocols (TCP and UDP) can not cater for. RTP takes into account the lossy behaviour of IP networks when congested, in a manner suitable to real-time services. UDP has no necessary timing or sequence information; TCP does not account for the low-delay and immediacy of real-time service requirements—a late packet should be dropped rather than retransmitted.
Congestion and packet loss can be limited by reserving bandwidth on key routes in an IP network, as implemented in the Reservation Protocol (RSVP). This mechanism in effect makes the connectionless network behave as if connection orientated, since all packets entitled to use the reserved bandwidth must follow an established route.
Other protocols and routing policies have gone further to implement a fully connection orientated paradigm, by establishing routes for traffic in several if not all switching nodes in an IP network. Multiprotocol Label Switching (MPLS) establishes routes for connections in the routing tables of IP nodes, supporting point-to-point and point-to-multi-point connections. Packets bearing a label may be forwarded directly on the pre-configured route, achieving low-latency, and having the benefit of reservation of switching resources and transmission bandwidth. Layer 2 Tunnelling Protocol (L2TP) defines tunnels that behave in a similar manner to labelled routes in MPLS, but offers a further multiplex layer of calls within tunnels.
These IP mechanisms are analogous to ATM QoS and VP and VC partitioning. The particular issue of IP packetisation delay affecting delay sensitive services, has not however been addressed for two reasons:
IP packets may be variable length, and therefore the packetisation delay is arbitrary. Due to present day anomalous pricing of IP service against higher cost PTT telephony, there is no user financial drawback due to the poor bandwidth efficiency of IP header to payload ratio;
IP networks carrying voice have not yet achieved an acceptable QoS in practice, and the packetisation delay is not yet a significant artefact. In time, the delay and efficiency might become issues that detract from the advantages of IP.
Beyond the ability to switch in these adaptation layers and to interwork between them, is the requirement to trunk to legacy TDM and packet based networks. The various demands of IP and ATM adaptation layers, and the likelihood that both IP and ATM technologies will be deployed in the near term for both real-time and non-real-time services, has demonstrated a need for the integration of these technologies.
SUMMARY OF THE INVENTION
An object of the invention is to minimise or to overcome the above disadvantage.
According to a first aspect of the invention there is provided a generic adaptation architecture in which functional partitioning of devices and an architecture that is optimal for variable and fixed packet adaptation layers so as to enable trunking, interworking and switching of the adaptation layers.
The technique provides a functional partitioning of devices and an architecture that is optimal for variable and fixed packet adaptation layers having the greatest possible degree of flexibility in isolation or in combination to serve trunking, interworking and switching of the adaptation layers. The functional partitioning and architecture forms an optimal integration and interworking of AAL based and IP based traffic, and is optimal for a high degree of scalability.
In a further aspect, the invention provides a functional partitioning and architecture that caters for recursive encapsulation of adaptation layers.
In another aspect, the invention provides a functional partitioning and architecture that serves encapsulations of IP in ATM AAL-2 and AAL-5.
The functional partitioning may utilise a fixed length packet switch to switch all traffic in a generic manner, and an input/output buffer paradigm to enable unlimited scaling and interworking.
As the architecture is generic, it is fully extensible to a wide range of

Affiliated with

Also associated with

Rating

Processing communication traffic does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Processing communication traffic, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Processing communication traffic will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3094101