Packet network

Details Packet network Packet network

1998-11-02

2003-03-25

Ton, Dang (Department: 2661)

Multiplex communications

Data flow congestion prevention or control

C370S411000

Reexamination Certificate

active

06538989

ABSTRACT:

The present invention relates to a packet network and to elements and hosts therein.
Prior to sending any information across a traditional connection-oriented network, a user is allocated a circuit, either by provision or by on-demand signalling. During the allocation phase it can be arranged for the circuit to meet user specified performance criteria and this is typically controlled by the connection admission rules. Consequently, once an appropriate circuit has been established information can be sent between host machines to a specified Quality of Service (QoS).
In contrast a traditional connectionless network has no requirement for circuit allocation and such networks do not incorporate connection admission control. This has the result that, during periods of network congestion, the meeting of performance criteria, and thus the satisfaction of Quality of Service requirements (such as end to end delay and delay variation) cannot be guaranteed.
It is now becoming increasingly clear that future networks will need to support services akin to those provided by traditional connection-oriented networks and also services akin to those provided by traditional connectionless networks. Furthermore, it will be essential for those services to be supported with minimal complexity and within acceptable performance trade-offs.
For the past decade, the vision for future broadband multimedia networks has been that they would be based on Asynchronous Transfer Mode (ATM) technology and the associated networking standards. However, during the early stages of standardisation, it was decided that ATM networks would be connection-oriented and not support a native connectionless mode of operation. Also ATM standards have tended to concentrate on service class optimisations, rather than taking a broader network view. For instance optimising bearer utilisation does not necessarily require the bandwidth utilisation to be optimised for each individual service class supported. More broadly it seems likely that applications are likely to evolve at a more rapid pace than is feasible for networks to track.
Potentially undesirable aspects and complexities of traditional ATM service for future networks include:
Statistical multiplexing within a service class requires complex connection admission control (CAC) based on leaky bucket source traffic descriptors.
Buffer size needed to achieve zero cell loss is indeterminate when using statistical multiplexing.
The fixed ATM cell size is unlikely to suit all services.
In a switched network a signalling phase is required irrespective of the traffic type and potentially this could create a performance bottleneck for some types of connectionless traffic.
Cell header size is a further bandwidth overhead in addition to that already imposed by any higher layer protocols (for AAL
5
adaptation the cell header wastes ten percent of available bandwidth, for other AAL's it is higher).
Traditional IP networks have evolved from the concept of connectionless transport methods which to date have offered users only a “best effort” service. However, a new service model, a so called integrated services (IS) Internet, is now being proposed and addressed by the IETF (Internet Engineering Task Force) Integrated Services Working Group.
The IS Internet will support a number of service classes with particular QoS provisions.
The principal service classes being proposed are:
Guaranteed Service (GS), which supports a guaranteed bandwidth and well defined delay bound;
Controlled Load (CL) which supports a more loose guarantee of bandwidth; and the traditional Best Effort (BE).
The term flow is used to denote a stream of one or more packets produced by a given application and transmitted with a particular QoS requirement.
To support the provision of different service classes, in contrast to the present day TCP/IP protocol suite, the IS Internet will require flow state information in all the network routers.
So far formal analysis has concentrated on the Guaranteed Service class for which it is proposed that the guaranteed delay bound will be met by using token bucket traffic descriptors in the CAC algorithm and scheduling schemes like Weighted Fair Queuing (WFQ). Although absolute delay bounds are guaranteed by this approach, it appears that they could be excessively pessimistic and under certain circumstances result in unnecessarily complex processing (i.e. under some circumstances the potential benefits may be outweighed by the additional complexity).
Should this prove to be the case, alternative and more simple solutions giving more realistic delay bounds will be desirable.
Potentially undesirable aspects and complexities of Guaranteed Service for future networks include:
The delay control offered by WFQ may become negligible when GS is operating under conditions of time contention rather than bandwidth contention (i.e. no statistical multiplexing).
Although WFQ gives both bandwidth sharing and strict flow isolation, the need for this may diminish as the buffer backlog bound decreases and/or an appropriate upper limit is imposed on the maximum datagram size.
Statistical multiplexing within the GS class requires complex CAC based on token bucket traffic descriptors.
Resource Reservation Set-Up Protocol (RSVP) signalling is required to establish whether or not the requested end to end delay bound can be supported.
At present WFQ is applied on a per flow basis so it is computationally intensive and this may lead to a potential performance bottleneck as network speeds increase and/or datagram sizes decrease (less time to compute datagram scheduling).
Furthermore, traditionally telecommunications networks have been designed on a basis that network usage patterns and traffic statistics are well understood and relatively stable parameters but it is now becoming obvious from the growth of the Internet and the way in which it is being used that these parameters are becoming increasingly uncertain. It is also anticipated that this trend is set to continue well into the future.
Consequently, future network designs must be robust enough to cope with these uncertainties but they should not be at the expense of utilising unnecessarily complex network control techniques. Present indications suggest however that this may well happen in the area of QoS performance guarantees.
According to the invention, there is provided a packet network element comprising at least one input for receiving flow based packets; at least one output of predetermined bandwidth; wherein a received packet is associable with a first or second class of service; means for directing each received packet on the basis of its class to a first or a second corresponding packet buffer, said first packet buffer being allocated a predetermined portion of the output bandwidth; said second packet buffer being allocated the remaining portion of the output bandwidth; bandwidth requirement determination means for determining a bandwidth requirement associated with at least said first class flows; means for allowing admission of the first class flow packets to the first packet buffer if said bandwidth requirement can be met; and means for directing packets from the first and second packet buffers to an output.
In this way a comparatively simple network element architecture advantageously offers a service similar to that associated with a traditional connection-oriented network to a first class of packets but offers a service similar to that associated with traditional connectionless networks to a second class of packets.
Preferably, said means for allowing admission is operable to apply a peak rate test, allowing admission to the first buffer if the currently unused portion of said predetermined portion of the output bandwidth is able to meet said peak rate bandwidth requirement.
In some circumstances, this allows the formulation of a fixed delay bound whilst avoiding the potential processing bottlenecks that may limit the throughput of more complex schemes—first class flow packets admitted to the first packet buffer will be provided the g

Affiliated with

Also associated with

Rating

Packet network does not yet have a rating. At this time, there are no reviews or comments for this patent.

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

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3019423