Multiplex communications – Channel assignment techniques – Messages addressed to multiple destinations
Reexamination Certificate
1998-06-01
2002-04-09
Olms, Douglas (Department: 2732)
Multiplex communications
Channel assignment techniques
Messages addressed to multiple destinations
C370S235000, C370S397000, C370S399000, C370S401000, C370S409000
Reexamination Certificate
active
06370152
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to digital data communication systems, and is particularly directed to a new and improved management agent, that may be readily employed for simple network management protocol (SNMP), being distributed among individual channel units of a frame relay switching system, rather than in a proxy device interposed between the network and the channel units, with the address of an individual device being encoded in the community string portion of the SNMP packet.
BACKGROUND OF THE INVENTION
Frame relay (FR) switching networks enjoy widespread success for supplying digital data services—principally in support of local area network (LAN) integration across wide area facilities. Frame relay derives its benefit by multiplexing ‘frames’ of data, or packetized data traffic, from each of a plurality of physical circuits into a common or shared backbone network, so as to reduce overall transport requirements, by creating more efficient use of the available transport bandwidth.
Prior to the advent of frame relay switching systems, LAN integration required the purchase of point-to-point digital data services delivered via individual physical circuits. Among the advantages frame relay provides over point-to-point services are enhanced network reliability and troubleshooting, through redundant and highly manageable switching networks, multiple end-to-end connections via permanent virtual circuits (PVCS) over single physical interfaces, and lower deployment costs that often result in lower consumer fees.
In a typical implementation, rather than being dispersed throughout the network, the switching backbone is confined to a very limited geographic area, requiring significant back-haul support. In addition, in stark contrast to the widespread availability of central office voice switches, the number of available data switches is relatively small. Also, due to the nature of physical interface design, a frame relay switch is often not necessarily employed in a manner that makes full use of its frame processing functionality. By accessing the frame relay switch through point-to-point facilities, both data transport time and idle time are presented to the switch interface. The frame relay switch then becomes limited at the point of its physical port access, often mandating the purchase and installation of additional frame relay switches and line cards.
In order to better utilize frame relay switch capability, as well as increase the utilization of back-haul transport bandwidth, it is desirable to efficiently extend frame multiplexing to central offices, remote terminals (RTs) and customer premises equipment (CPE) environments in a manner that address current deployment schemes. For this purpose, it has been proposed to implement frame relay switching networks in an architecture of the type diagrammatically shown in FIG.
1
.
In this architecture, a plurality of frame relay channel units (or frameports)
11
terminate a large number of access lines
13
from associated CPEs
15
, and are coupled to an aggregate data link
21
from the network
23
. The aggregate data link customarily has a data transport capacity that is considerably less than the cumulative data capacity of all of the access lines
13
. Included within this architecture is an outgoing data transport access mechanism that statistically multiplexes outgoing data from the various CPEs
15
into a single data stream for transport over the aggregate data link
21
. The use of statistically multiplexing serves to reduce the back-hauling of circuits to the data switch, thereby more efficiently utilizing aggregate data transport resources, since the data is multiplexed at the point of access (the respective frame relay channel units
11
), rather than at the switch.
In order to provide an orderly and efficient allocation of the limited bandwidth of the aggregate link
21
to each access line
13
requesting transmission service, yet without allowing any individual access line to tie up the aggregate data link
21
during periods of inactivity or congestion, the statistical multiplexing of data from the access lines
13
onto the aggregate data link
21
is based upon a prescribed opportunity-to-transmit arbitration mechanism, that is distributed among all users of the system, so that the state of the system is effectively shared with each access line.
In the above-referenced co-pending '461 application, we describe a new and improved arbitration scheme, which calculates an arbitration value for each data line based upon a combination of parameters, including queuing delay and information customized for the particular configuration and traffic rate of that line. The calculated arbitration value is combined with a request to transmit or start bit and a physical location address code to produce a composite arbitration code, the respective bits of which are clocked onto a wire-ORed arbitration bus. As each channel unit's arbitration code is clocked onto the arbitration bus, a respective channel unit will perform a bit-by-bit comparison with the state of the bus as driven by other channel units, to determine whether that channel unit is to be given priority to transmit. Whenever a channel unit sees a difference between its bit value and that on the bus, the channel unit knows it has a lower priority and terminates its participation in the arbitration cycle. As a result, at the end of the arbitration cycle, only the channel unit having the largest priority will not have dropped out, allowing it to take ownership of the aggregate data link.
In the incoming or receive direction (from the network to the channel units), the system typically employs some form of supervisory management entity
25
, such as a Hewlett-Packard “Open View” platform as a non-limiting example, which is connected to the network and manages the routing of data transmissions from the network to the individual channel units serving the end customers equipment, using an embedded in-band management scheme, such as industry standard SNMP, referenced above. Where the management scheme assigns an Internet protocol (IP) address per device, the overall management process becomes prohibitively complex and costly, as there can be literally thousands of IP addresses to allocate, assign and maintain.
An alternative approach, diagrammatically illustrated in
FIG. 2
, involves using a single, intermediary proxy device
30
as the management agent between the network manager and the data communication devices to be managed. This approach has the advantage that it allows a single supervisory device to manage a large number of data communication devices using a single IP address over a single access line. In addition, if the respective communication links
32
between the proxy agent device
30
and the managed devices employ a proprietary communication protocol, the managed devices can be made less expensive than would be the case if these devices were required to completely support SNMP protocol. However, the use of a proxy device has its shortcomings.
First of all, since the proxy device forms a junction of all incoming data packets from the network, a failure in the proxy device will result in a complete loss of management for all of the managed devices connected to it. Secondly, if a proprietary protocol is used between the proxy agent and the managed devices, the system becomes vendor-specific, rather than conforming with an industry standard, such as SNMP. Further, as new data communication devices are developed, the software within the proxy agent may have to be upgraded to allow it to be a proxy for the new devices, thereby adding to system complexity and cost.
SUMMARY OF THE INVENTION
In accordance with the present invention, the above-discussed shortcomings of conventional data routing management schemes are obviated by distributing the management agent among individual channel units of the system, without using any intermediate proxy device. Distributing the management agent is implemented by means of an address protocol emp
Schofield Wade S.
Venters W. Stuart
Williams Philip David
Adtran Inc.
Allen, Dyer, Doppelt Mlibrath & Gilchrist, P.A.
Hom Shick
Olms Douglas
LandOfFree
Distributed SNMP agent for frame relay switching network does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Distributed SNMP agent for frame relay switching network, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Distributed SNMP agent for frame relay switching network will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2828293