Electrical computers and digital processing systems: multicomput – Network computer configuring – Reconfiguring
Reexamination Certificate
1997-05-02
2001-04-17
Burgess, Glenton B. (Department: 2757)
Electrical computers and digital processing systems: multicomput
Network computer configuring
Reconfiguring
C709S220000, C709S222000, C710S010000, C710S009000, C710S003000
Reexamination Certificate
active
06219697
ABSTRACT:
MICROFICHE APPENDIX
This application contains a microfiche appendix, which consists of three sheets of microfiche and a total of 196 frames.
NOTICE REGARDING COPYRIGHT
A portion of the disclosure of this patent document contains matter subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent disclosure document as it appears in the Patent and Trademark Office files and records but otherwise retains all copyrights whatsoever.
BACKGROUND OF THE INVENTION
A. Field of the Invention
This invention relates to the operation of a high speed data network which interconnects different application modules, and more particularly relates to a method and apparatus for operating IP protocol over a high-speed bus such as an IEEE 1394 high-speed bus.
B. Description of Related Art
When communication is necessary among heterogeneous systems (i.e. different vendors and standards), the software to communicate between the systems can be extremely difficult to develop. The potential exists for different vendors to use different data formats and data exchange conventions so that the differing machines are unable to “talk” directly to each other. To combat this potential problem, standardizing organizations have established computer architectures and structures to facilitate the communication between heterogeneous systems. One accepted structuring technique is layering. The communications functions are partitioned into a vertical set of layers whereby each layer performs a related subset of functions required to communicate with another system. Each layer then relies on the next lower layer to perform more primitive functions and to conceal the details of those functions as well as to provide services to the next higher layer. In this manner, the larger problem of communicating between differing systems can be structured into a discrete subset of layered subproblems.
One model of layering is the Open Systems Interconnection (OSI). The Open Systems Interconnection uses the layering framework with the goal to create an open systems networking environment where any vendor's computer system, connected to any network, can freely share data with any other computer system on that network or linked network. Many computer systems have a structure based on the OSI model, which organizes the communication proceeds into seven different categories and places these categories in a layered sequence based on their relation to the user.
The three lowest layers in the OSI model (layers 1 through 3) are the physical, link and network layers, each layer dealing with network access. The physical layer is concerned with transmission of unstructured bit stream over physical media, such as cables, and is considered to be the lowest layer. The link layer provides for the reliable transfer of information across the physical layer, sending blocks of data with the necessary synchronization, error control, and flow control. The network layer determines how data is transferred between computers and addresses or routing within and between individual networks.
One example of a physical and link layer medium is the IEEE 1394 high speed serial bias. IEEE 1394 is a data transport bus that supports up to 63 nodes per bus and up to 1023 buses. The bus can be a tree or daisy-chained (devices connected in series) or a combination of both and can support both asynchronous and isochronous data. Further, the Internet Protocol (IP) is a connectionless protocol (i.e. mode of operation in which a packet header is encoded with sufficient information to permit independent delivery of the packet) that operates at the network layer. The IP protocol is a standard describing software that monitors the internetwork addresses for different nodes, routes outgoing messages, and recognizes incoming messages. IP, which works in conjunction with the Transmission Control Protocol (TCP), (and identified as TCP/IP) provides communication across interconnected networks, between computers with diverse hardware architectures and various operating systems. TCP and IP are two of the more widely used protocols in the family of Internet protocols.
However, there are several problems when integrating the IEEE 1394 protocols with the IP protocols. First, there are differences in the protocols between the sender and the receiver of data. The IEEE 1394 is designed as a memory read/write bus where the sender of data must have information on the receiver's memory structure. In particular, the sender or initiating node reads from or writes to the memory location of the receiver or target note. The sender node is required to know the memory architecture of the receiving node in order to perform memory read/writes. In contrast, the IP protocol transfers data without the sender knowing about the receivers memory architecture. As such, the data transfers of IP cannot be directly accomplished on the IEEE 1394 architecture. Further, the IEEE 1394 is essentially a connectionless acknowledged protocol. Each IEEE 1394 packet of data, other than a packet that is broadcast to all nodes, generates an acknowledgment from the receiver. If the acknowledgment indicates a failure in the reception of the packet, the sender retransmits the packet. Second, the packet formats of the IEEE 1394 are incompatible with the IP packet formats. A packet header is the portion of the message sent that contains information which guides the message to the correct destination.
Third, there are addressing problems when using the IP protocol as the transport protocol for data transfers over the IEEE 1394 bus. The IP protocol uses logical addresses to identify each node in the network with the IP addresses as 32-bit values depicted in dotted-decimal fashion (e.g., 149.112.234.1 for IP host and 149.112.234.2 for another IP host). An application on one IP host that needs to transfer data to another application on a different IP host will use the IP address to direct the data transfer (e.g., use the IP address 149.112.234.2 to transfer data to the second application). However, the IEEE 1394 protocol requires the use of a physical address of the destination node. Thus, a conversion between the IP address to the IEEE address must be accomplished before the data can be handed to the physical layer for transmission.
Prior methods of obtaining the physical address, such as through the use of the Address Resolution Protocol (ARP), either cannot or should not be used with the IEEE 1394 network.
Ordinarily, when an IP host wishes to obtain the address of a destination node, it broadcasts an ARP query with the IP address of the destination node. All IP hosts on the local network receive the ARP broadcast and check to see if the IP address contained in the ARP query is the same as the local IP interface address. The node whose address is equal to the address contained in the ARP query responds with an ARP response that contains the physical address for that node. However, most ARP implementations cannot support the dynamic changing of the hardware address and therefore cannot be used on an IEEE 1394 bus. The IEEE 1394 physical addressi is generated dynamically by the IEEE 1394 and can change upon a bus reset. The physical address of an IP interface is set at the time of interface initialization and cannot be changed without bringing the IP interface down and then reinitialized with a new physical address. If this is done, all of the IP traffic on the interface is stopped and is only restarted after the IP interface is up again. Any TCP connections using that interface will be torn down.
Further, the requirement of broadcasting an ARP inquiry and the address resolution latency are undesirable. All nodes on the network are forced to process the ARP request to determine if they have to respond to it. The determination of the physical address must be done quickly enough so that no data is lost on the network. During this determination, the data is buffered. However, the data can only be buffered for a finite period of time due to physical constraints
Lawande Sachin S.
Ling Salim
3Com Corporation
Burgess Glenton B.
McDonnell & Boehnen Hulbert & Berghoff
Salad Abdullah E.
LandOfFree
Method and apparatus for operating the internet protocol... 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 for operating the internet protocol..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for operating the internet protocol... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2516142