Multiple processors in a row for protocol acceleration

Details Multiple processors in a row for protocol acceleration Multiple processors in a row for protocol acceleration

1997-08-21

2002-12-03

Dinh, Dung C. (Department: 2153)

Electrical computers and digital processing systems: multicomput

Network-to-computer interfacing

C709S230000, C709S241000

Reexamination Certificate

active

06490631

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to the use of multiple processors in a sequential configuration for accelerating the encoding and decoding of a protocol for data transfer. More particularly, the invention relates to the use of multiple processors sequentially arranged to perform a respective protocol conversion application to the data as it passes from one processor to another processor in the sequential chain.
2. Description of the Related Art
Protocols are the standards that are used to specify how data is represented when it is transferred from one machine to another machine. Protocols specify how the transfer occurs, how errors are detected, and how acknowledgements used in a data transfer procedure are passed from machine to machine.
In order to simplify protocol design and implementation, communication tasks are typically segregated into separate subtasks that can be solved independently of each other. Each subtask is typically assigned a unique protocol, or protocol layer.
Layered protocols provide a conceptual framework for protocol design. In a layered protocol, each layer handles one part of the communications task, which typically corresponds to one protocol. For layered protocols, the ith software implementing layer on the destination machine receives exactly what the ith software implementing layer on the source machine sends. That is, the ith layer at the destination machine receives exactly the same object sent by ith layer at the source machine.
The layered protocol concept allows the protocol designer to focus attention on one layer at a time, without being concerned about how the other layers perform. For example, in a file transfer application, the protocol designer need only be concerned with two copies of the application program executing on two machines and what needs to be done in order to perform the file transfer exchange between the machines. The protocol designer makes the assumption that the application at the destination machine receives exactly what the application at the source machine sends.
FIG. 1
shows an example of a layered protocol, in which four layers are utilized to send data between Host A and Host B. The highest layer is the Application layer
110
, the second highest layer is the Transport layer
120
, the third highest layer is the Internet layer
130
, and the lowest layer is the Network interface layer
140
. As can be seen from
FIG. 1
, the Application layer
110
of the destination machine, say Host B, receives exactly the same message sent by the Application layer
110
of the source machine (Host A). The Transport layer
120
of the destination machine receives exactly the same packet sent by the Transport layer
120
of the source machine. The Internet layer
130
of the destination machine receives exactly the same datagram sent by the Internet layer
130
of the source machine. Lastly, the Network Interface layer
140
of the destination machine receives exactly the same frame sent by the Network Interface layer
140
of the source machine.
A frame of data passes from one machine to the other machine over the physical network
150
, as seen from FIG.
1
.
FIG. 1
shows a simple transfer over a single network
150
, and
FIG. 2
shows a more complex transfer over multiple networks, using a router R between the first network
160
and the second network
170
. As can be seen from
FIG. 2
, message delivery uses two separate network frames, one frame
180
for the transmission from Host A to router R, and another frame
185
from router R to host B. The frame
180
delivered to router R is exactly the same frame sent from host A, but it differs from the frame
185
sent between router R and host B. The same is true of the datagram
190
sent between host A and router R, and the datagram
195
sent between router R and host B.
The application layer
110
and the transport layer
120
deal with end-to-end issues, and are designed so that the software at the source (i.e., host A) communicates with its respective equivalent at the ultimate destination (i.e., host B). Thus, the packet
197
received by the transport layer
120
at host B is identical to the packet
197
sent by the transport layer
120
at host A. Further, the message
199
received by the application layer
110
is identical to the message
199
sent by the application layer
110
at host A.
The higher level layers deal primarily with end-to-end transfers, and the lower level layers deal primarily with single machine transfers. Thus, the ultimate destination (i.e., host B) may not receive the identical datagram
190
sent by the ultimate source (i.e., host A). For example, the header field of the datagram
190
is changed as it passes through the router R, for example.
FIG. 3
shows how the different layers are used to send data from a source machine to a destination machine over multiple networks. A sender on the source machine
310
transmits a message, which the IP layer
320
places in a datagram and sends across the first network
330
via the interface
340
. The intermediate machine
350
receives the message on the first network
330
via its interface
340
, passes the message up to the IP layer
320
of the intermediate machine
350
, and routes the message onto a second network
360
via its interface
340
. The intermediate machine
370
receives the message on the second network
360
via its interface
340
, passes the message up to the IP layer
320
of the intermediate machine
370
, and routes the message onto a third network
380
via its interface
340
. The destination machine
390
receives the message on the third network
380
via its interface
340
, the IP layer
320
of the destination machine
390
extracts the message, and the message is passed up to the higher layers
395
of protocol software, to be eventually received at the receiver
397
. Note that the message was not passed up through the higher levels of protocol software by each of the intermediate machines
350
,
370
, since they had no need to extract the message, but only to pass it on to the desired destination machine.
Conventional layered protocols include TCP/IP, X.25 and ISO (also known as OSI). The TCP/IP protocol is a four-layered protocol, and the ISO protocol is a seven-layered protocol. The seven layers of the ISO protocol are: application layer (layer
7
), presentation layer (layer
6
), session layer (layer
5
), transport layer (layer
4
), network layer (layer
3
), data link layer (layer
2
), and physical hardware connection layer (layer
1
).
The X.25 network consists of packet switches that contain the logic needed to route packets through the network. Hosts attach to one of the packet switches using a serial communication line, and hosts must follow a predetermined procedure in order to transfer packets onto the network and retrieve packets from the network.
At the physical layer, X.25 specifies a standard for the physical interconnection between host computers and network packet switches, as well as the procedures used to transfer packets from one machine to another.
The data link layer specifies how data travels between a host and the packet switch to which it connects. The data link layer defines the format of frames and specifies how the machines are to recognize frame boundaries, and well as providing error detection.
The network layer specifies the functionality for completing the interaction between the host and the network, and it defines the basic unit of transfer across the network. The network layer includes the concepts of destination addressing and routing. The network might allow packets defined by network layer protocols to be larger than the size of frames that can be transferred at the data link layer. The network layer software assembles a packet in the form the network expects and uses the data link layer to transfer it (presumably in multiple packets) to the packet switches. The network layer also responds to congestion problems on the network.
The transport layer provides end-to-end relia

Affiliated with

Also associated with

Rating

Multiple processors in a row for protocol acceleration does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Multiple processors in a row for protocol acceleration, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multiple processors in a row for protocol acceleration will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2990174