Multiplex communications – Pathfinding or routing – Switching a message which includes an address header
Reexamination Certificate
2000-12-29
2002-12-10
Jung, Min (Department: 2663)
Multiplex communications
Pathfinding or routing
Switching a message which includes an address header
C370S401000, C370S465000, C709S238000
Reexamination Certificate
active
06493341
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to computer network data transmission, and more particularly relates to the cost-efficient use of multiple routers to provide connections with wide area networks, including connections with the global computer network commonly referred to as the Internet.
TECHNICAL BACKGROUND OF THE INVENTION
Many local area networks (“LANs”) are connected to the Internet or another wide area network (“WAN”). LANs may also be connected to one another through the Internet or another WAN. A given LAN, or a given sub-network of a LAN, is connected to the WAN through a device known as a router. For convenience, reference is made hereafter to LANs with the understanding that “LAN” means “LAN or sub-network” unless otherwise stated. Routers use both WAN addresses, such as Internet Protocol (“IP”) addresses, and physical addresses, such as Ethernet addresses. Physical addresses may also be called “data link addresses”.
Each router receives from its LAN all network traffic addressed to a destination outside the LAN, such as data packets addressed to a remote IP address. The router forwards those packets to the next router along a path to the destination. The path often takes the packet through part of the Internet or another WAN. The router likewise receives Internet or other WAN packets from other LANs which are destined for machines within the router's LAN, and re-addresses the packets so they can be delivered using physical addresses which are internal to the LAN. Conversion from an IP address to a data link address such as an Ethernet address may be done using a conventional Address Resolution Protocol (“ARP”). Some known systems use two or more routers with a form of inflexible load balancing, whereby all requests go out over a first router and all responses come back over a second router.
FIG. 1
illustrates a conventional network topology
100
which uses a router to connect a LAN (or sub-network, as noted above) to a WAN. Several nodes
102
are connected by LAN “wires” in a LAN
106
. The nodes
102
may include machines such as desktop computers, laptops, workstations, disconnectable mobile computers, mainframes, information appliances, personal digital assistants, and other handheld and/or embedded processing systems. The “wires”
104
may include twisted pair, coaxial, or optical fiber cables, telephone lines, satellites, microwave relays, modulated AC power lines, and/or other data transmission “wires” known to those of skill in the art. The network
106
may include Novell Netware®, VINES, Microsoft Windows NT or Windows 2000, LAN Manager, or LANtastic network operating system software (NETWARE is a registered trademark of Novell, Inc.; VINES is a trademark of Banyan Systems; WINDOWS NT, WINDOWS 2000, and LAN MANAGER are trademarks of Microsoft Corporation; LANTASTIC is a trademark of Artisoft).
Another “wire”
108
connects a router
110
to the LAN
106
. A wide variety of routers
110
are known in the art. At a minimum, the router
110
maintains a table of routes for different destination addresses. Different routers
110
can handle different physical address types (Ethernet, . . . ). Some routers provide firewall services. Different routers also handle connections that run at different speeds using different line technologies (T1, T3, ADSL, RADSL, . . . ). But in general, some type of high-speed connection
112
connects the router
110
to a WAN
114
.
The Internet or a portion of the Internet may serve as the WAN
114
, or the WAN
114
may be separate from the Internet. “Internet” as used herein includes variations such as a private Internet, a secure Internet, a value-added network, a virtual private network, or a wide area intranet. Another connection
116
connects a server
118
or other destination with the WAN
114
.
Like the illustrated topology
100
, other conventional network topologies utilize one router per LAN (or sub-network). Conventional network topologies do not support the routing of data over multiple routers in any given LAN. For instance, standard TCP/IP stacks are not able to direct data packets from a given LAN to multiple routers when the data needs to be sent to other LANs. Multiple routers may be physically present, but one router is designated as the default gateway for the LAN. This default gateway receives all the traffic for the LAN from outside, and forwards data packets from inside the LAN to the next LAN on their way to their destinations.
The router
110
which serves as the default gateway also maintains a table of routes for different destination addresses. Data transmission generally takes place between two networks over the shortest defined path, where a path is represented as a list of routers which the data has to traverse in order to reach the destination node. For instance, a data packet from a given node
102
addressed with the IP address of the server
118
will be sent from the node
102
over the LAN wires
104
,
108
to the gateway router
110
, will travel from there over the high-speed connection
112
to the WAN
114
(which may forward the packet along a path containing multiple routers), and will finally arrive at the server
118
.
Once a node such as a client PC
102
on the LAN
106
performs the Address Resolution Protocol, the information is stored in an ARP table on the client PC
102
. After this the PC
102
does not send an ARP request until a timeout condition occurs. ARP tables and ARP timeouts are used in conventional systems and they may also be used according to the invention. After an ARP request is sent because of a timeout, or for another reason (e.g., when an ARP table entry is made manually), IP communication starts with SYN packet. SYN packets in and of themselves are known in the art.
Similar steps occur when a packet from the same node
102
is addressed to another node on a distant LAN. In place of the server
118
the path would include another router connected to the distant LAN. In its capacity as gateway for the distant LAN, the distant router would receive the packet from the WAN
114
and deliver it to the distant node.
For clarity of illustration, Internet Service Providers (“ISPs”) have not been shown in FIG.
1
. However, those of skill in the art understand that one or more ISPs will often be located along the path followed by a packet which travels to or from a LAN node
102
over the Internet
114
.
The configuration
100
is widely used but nevertheless has significant limitations. Although the data transmission speed over lines such as the line
112
is relatively high when compared to traditional analog telephone data lines, the available bandwidth may not always be sufficient. For instance, the number of users within the LAN
106
may increase to a point at which the data transmission capacity of the WAN connection
112
reaches its maximum limit. In order to obtain more bandwidth, a company could lease more expensive dedicated data lines
112
which have greater data transmission speeds, such as lines employing T3 or OC3 technologies.
To delay expensive upgrades to line technology and to the corresponding router technology, bandwidth can be used more efficiently. This might be done by compressing data, by combining different types of data to reduce the total number of packets, and by reducing unnecessary access to the WAN
114
through appropriate personnel policies. Tools and techniques for improving router
110
performance are also being developed and made commercially available. In addition, new data transmission technologies like ADSL, RADSL, and others are being proposed and developed. Although these technologies do not have as high a data transmission rate as T3 or OC3, they are several times faster than analog lines.
Moreover, commonly owned copending application Ser. No. 08/859,070 filed May 20, 1997 describes a mux device for assisting the transmission of a user's data between two computer networks. The mux device could be added to a system like that shown in
FIG. 1
to increase the bandwidth of the connection
1
Bhaskar Ragula
Datta Sanchaita
Computer Law++
Jung Min
Ragula Systems
LandOfFree
Combining routers to increase concurrency and redundancy in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Combining routers to increase concurrency and redundancy in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Combining routers to increase concurrency and redundancy in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2939666