Multiplex communications – Pathfinding or routing – Switching a message which includes an address header
Reexamination Certificate
1998-11-10
2003-11-25
Nguyen, Steven (Department: 2665)
Multiplex communications
Pathfinding or routing
Switching a message which includes an address header
C370S395410, C370S412000
Reexamination Certificate
active
06654374
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to communication networks, and particularly, the present invention relates to providing guaranteed quality of service in a packet switched network.
BACKGROUND OF THE INVENTION
In communications technology, there is an ever-increasing demand for high-performance networks, and in particular, a demand for high-performance Internet access. This increased demand has led to the development of improved networks capable of handling larger volumes of data with smaller delays. Nonetheless, these improved networks each have their own shortcomings.
Communications networks like the Internet are generally formed with a number of transmission links interconnected with switches. A transmission link is any medium through which signals are communicated and can be single or multiple twisted pairs, optical fiber, coaxial cable, radio links, or other mediums. A switch is a device with one or more input ports and one or more output ports. The switch directs bits arriving at an input port to the appropriate output port. Switching in communications is accomplished using one of two methods: circuit switching and packet switching.
Circuit switching has been used since the birth of the modern telephone system. Circuit switching entails establishing an end-to-end connection between two points before any information is communicated. A telephone connection is one example of a circuit-switched connection. In early circuit-switched networks, information could only be communicated over the links if the links were not being used by another party and if an end-to-end connection was already established.
To make the circuit-switched network more economical, Time Division Multiplexing (TDM) was developed. TDM is a method that enables multiple end-to-end connections to share the same link. More specifically, using TDM the respective data streams are transmitted for a designated time period in a round robin fashion over a shared transmission link. For instance, referring to the illustration of
FIG. 1
, to multiplex N data streams using TDM, the first stream is transmitted for T(1) seconds, then the second stream for T(2) seconds, and so on until the N-th stream is transmitted for T(N) seconds. The cycle then repeats starting again with the first stream. Hence, each data stream gets a fraction of the total transmission rate for the shared link, i.e., the stream i (i=1, . . . , N) gets a fraction
T
(
i
)
[
T
(
1
)
+
T
(
2
)
+
…
+
T
(
N
)
]
of the transmission rate of the link. So, using TDM, a link with transmission rate R is capable of carrying data streams with rate R(
1
), . . . , R(N) provided that R(
1
)+ . . . +R(N) is less than R. Therefore, a circuit-switched network with TDM provides a guaranteed bandwidth to a connection from the source of the connection to its destination.
Despite the ability to send multiple data streams through a shared link at a guaranteed rate for each data stream, when a particular data stream does not need its time slot on the link, that fraction of the link capacity is wasted because other data streams are not permitted to use that time slot. Thus, although the circuit-switched network offers the user guaranteed bandwidth available at any time, when the user is not utilizing its allocated bandwidth, the link lays idle for the user's time slot—unavailable for use by any other party, leading to an inefficient utilization of available resources.
Even though there are inefficiencies inherent in circuit switching, such switching has proven satisfactory to carry voice data. Non-voice data, however, tends to be more “bursty” in nature. Hence, when a data burst arrives, having to wait through TDM cycles for transmission will make the transmission of data appear slow—an appearance that might be avoided if all resources were available for use.
To avoid the inefficiencies of circuit switching, the communications industry has made a shift to “packet switching” for non-voice type data transmission. Packet switching does not establish a single end-to-end connection for each data stream. Rather, the data stream is divided into “packets,” each of which carries data as well as various identifying information about each packet. Such identifying information includes routing information, for instance, source and destination addresses for the packet. When a packet arrives at a switch, it is stored in a buffer. The switch looks at the routing information in the packet and routes the packet to the appropriate output port. The packets are then forwarded one “hop” (or switch) at a time, until they reach their destination. Each packet in a particular data stream does not necessarily follow the same path to its destination as the others.
In most networking protocols, the networking system is hierarchically divided into subnetworks that are interconnected. The nodes (e.g., personal computers) in each subnetwork are addressed by a “Layer
2
” address (sometimes called a MAC address). A subnetwork delivers packets between any two of its nodes. These packets are labeled with the Layer
2
addresses of the source and destination nodes of the subnetwork.
Subnetworks can also be interconnected to one another with the use of specialized switches. These switches determine how to handle a packet based either on the Layer
2
addresses in the packet or on some other routing information. For instance, a switch interconnects a number of subnetworks through its several ports. For each of its ports, the switch maintains a list of Layer
2
addresses of the computers of the subnetwork(s) that is(are) attached to the port. When the switch receives a packet, the switch checks the port lists and identifies the port that is connected to the Layer
2
address for the destination of the packet. The switch then transmits the packet on that identified port. Other networking protocols additionally add special routing information in each packet. That routing information may be a connection number or a pair of geographically arranged addresses, as in the case of the Internet Protocol.
Since packets from different data streams are all routed over the same transmission lines, a scheduling method by which to place the packets from the different streams onto the transmission lines is required. One such scheduling method is statistical multiplexing (SM). In contrast with TDM, SM does not allocate fixed periodic fractions of the transmission rate of the link to the different data streams. Instead, the SM schedules packets for transmission depending on the actual needs of the different data streams. In its simplest form, illustrated in
FIG. 2
, statistical multiplexing schedules packets on the transmission line in the order of their arrival at the switch, i.e., on a first come, first served basis. Although such a scheduling method does not waste communication resources, it cannot guarantee a minimum transmission rate for any particular data stream. The packets simply get to their destination when they get there.
A variation of SM called “deficit round robin” (DRR) can provide better transmission rate guarantees and is illustrated in FIG.
3
. Under DRR, to multiplex N data streams, N integers W(
1
), . . . W(N), are first chosen as a weight (W) for each data stream. Then for the first data stream (DS
1
) up to W(
1
) bits are transmitted. If all of the packets of DS
1
are transmitted before transmitting W(
1
) bits or if bit W(
1
) is the last bit of a packet, then a second integer D(
1
) is set to 0. But if bit W(
1
) is not the last bit of a packet, then additional bits are sent until the transmission of the packet is completed, and D(
1
) is set to equal the number of additional bits sent. The procedure is repeated for data streams DS
2
, . . . , DSN with the corresponding numbers W(
2
), D(
2
), . . . W(N), D(N). The cycle then repeats starting with DS
1
, but W(
1
) is replaced with W(
1
)−D(
1
). Hence, W(
1
)−D(
1
) bits are transmitted instead of W(
1
) (in other words, D(
1
) represents a “deficit” of DS
1
). If there
Fawaz Ayman
Walrand Jean
Blakely , Sokoloff, Taylor & Zafman LLP
Extreme Networks
Nguyen Steven
Nguyen Toan
LandOfFree
Method and apparatus to reduce Jitter in packet switched... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus to reduce Jitter in packet switched..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus to reduce Jitter in packet switched... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3116227