Electrical computers and digital processing systems: multicomput – Computer-to-computer data addressing
Reexamination Certificate
1999-08-17
2003-09-30
Maung, Zarni (Department: 2154)
Electrical computers and digital processing systems: multicomput
Computer-to-computer data addressing
C709S220000, C709S250000
Reexamination Certificate
active
06629149
ABSTRACT:
FIELD OF THE INVENTION
The field of this invention is networking, and in particular, systems and methods for large scale networking.
BACKGROUND OF THE INVENTION
The architectures presently employed for certain large networks, such as the Internet, often do not scale well. That is, as the size of the network grows, its structure fails to adequately accommodate the growth in demand for services provided on the network, the routing of messages to new nodes, etc. As used herein, the term “network” is a set of computers that are coupled such that they can communicate with each other. Examples of networks include a local area network (LAN) and a wide area network (WAN). As used herein, a WAN can include one or more subnetworks (e.g., LANs) that are coupled together. A subnetwork interfaces to the rest of the WAN through one or more “gateways.” An example of a gateway is a host (a computer with a network address) that is coupled both to a LAN and a WAN, and which handles traffic between devices that comprise the LAN and other devices coupled to the WAN that are not part of the LAN. An example of a LAN is a network that serves a set of users in a single building, and that operates in accordance with the Ethernet protocol that is well-known in the art. An example of a WAN is the Internet. The Internet is a large scale network of subnetworks that communicate using the Internet Protocol version 4 (“IPv4”) described in Douglas E. Comer, Internetworking with TCP/IP, vol.1 , Prentice Hall 1991. As used herein, the term “coupled” means directly or indirectly connected. Thus, if A is directly connected to B, then A is said to be coupled to B. Likewise, if A is directly connected to B, and B is directly connected to C, then A is said to be coupled to C.
The scalablility of a network is often at least partly influenced by its addressing scheme, i.e., the way a node (a device that is coupled to the network) is located on the network. An example of an addressing scheme is IPv4 , which is used on the Internet. IPv4 specifies an address of 32 bits arranged in four octets, i.e., four numbers, each having a value between 0 and 255. An example of an IPv4 address is 132.56.9.234. The number of nodes and other entities on the Internet that use, or could use, a unique address are proliferating at such a rate so as to exhaust the number of addresses available. Furthermore, experience with Internet address administration indicates that practical concerns demand an address space that is substantially larger than the number of devices that require addresses. As used herein, a “device” is any hardware entity capable of receiving and/or sending information over a network. Examples of a device include a host, a client, a handheld appliance that includes a microprocessor, etc.
In response to the problems related to the inadequacies of the IPv4 addressing scheme, the Internet Engineering Task Force (“IETF”, the body that sets the standards for the Internet) has proposed a new standard, Internet Protocol version 6 (“IPv6”), which his described in Stephen A. Thomas, IPing and the TCP/IP Protocols, John Wiley 1996. IPv6 uses an address space comprising sixteen octets, rather than the four used by IPv4. This provides a substantially greater number of addresses for use on the network. However, IPv6 has not been widely adopted for the Internet, largely because its adoption would require a large scale replacement of hardware and software that is presently in place on the Internet. Instead, stopgap solutions have been adopted that can be handled by the existing technology base, but which are temporary and will not be very satisfactory as the Internet develops further.
For example, to overcome the shortage of address space, Internet Service Providers (“ISP”) “lend” an IPv4 addresses to each subscriber that logs in. The subscriber does not have a permanently assigned address, but is rather assigned a temporary address. The temporary address assignment remains effective during the subscriber's session on the network. Another known solution Network Address Translation (“NAT”), which permits the ISP to offer IPv4 addresses to its subscribers even though those addresses are not unique on the Internet, and may be used by subscribers of other ISPs. As packets pass between a subscriber and the WAN (e.g., the Internet) through the ISP, the ISP translates from the address used by the subscriber to a temporarily available address that is unique to the WAN, and vice versa. In the future, an increasing number of users can be expected to be online all the time and will want to publish their addresses so that other can reach them directly. NAT disadvantageously does not allow a user to publish a permanently assigned, globally unique address by which the user can be reached directly from another node on the WAN.
The Federal Communications Commission (“FCC”) has ruled that telephone service subscribers must be permitted to switch from one service provider to another without being forced to change to a new telephone number. That capability, which has been retrofitted at some expense in the telephone network, is disadvantageously not present today in the Internet. A customer that switches from one ISP to another usually is forced to incur the expense and inconvenience of changing her IP address. This is disadvantageous because the IPv4 address embodies information about the internal structure of the network, including the relationship of the ISP to the WAN. This means that in some cases, when an organizational change occurs entirely within the Internet, such as an ISP obtaining a different access provider to the WAN, users are required to change to a new IP address. This can entail considerable expense and inconvenience for users, especially corporate customers who have hundreds or thousands of machines with network addresses. Some large customers have refused to make the change. In this case, it is the service provider that pays a penalty because the ISP address space becomes more fragmented and requires more resources to properly route packets between its users and the WAN.
The present IP and proposed addressing schemes combine device identity information with information about the network structure in the neighborhood of the device, disadvantageously precluding global addressing for mobile devices. Thus, it is not possible to change the location of a device without also changing its address. Also, an IP address presently identifies a device, but not the true endpoint of communication, which can lie within the software of that device. This disadvantageously makes it difficult at best for network service assets to determine the type of traffic that is being carried in order to provide the appropriate quality of service. For example, it may be acceptable for file transfer packets to be delivered with slight delays and with little loss in quality of service. On the other hand, voice or interactive video packets cannot be delivered with substantial delays without significantly degrading the quality of service experienced by the communicating users. Network service providers need to know what type of traffic a given packet or set of packets carries in order to properly route the packet in the appropriate fashion (e.g., with the correct level of priority). Known systems identify types of traffic by having routers and switches snoop into TCP and higher level protocols, which is disadvantageous because the presence of this software in the network prevents subscribers from changing to another protocol when new technology makes such a change advantageous. This disadvantageously reduces the robustness of the network service, and restricts the freedom of the service provider to invent services that are customized and/or customizable for individual data flows and individual applications.
A known method of providing network services is “tunneling.” A packet that is sent to an ultimate destination X is encapsulated with address information that delivers the packet to device Y. Device Y strips the encapsulated address information from the packet, a
Fraser Alexander Gibson
Mapp Glenford
AT&T Corp.
Maung Zarni
LandOfFree
Network system and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Network system and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Network system and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3096330