Distributed type switching system

Details Distributed type switching system Distributed type switching system

1999-08-13

2002-01-01

Olms, Douglas (Department: 2661)

Multiplex communications

Pathfinding or routing

Switching a message which includes an address header

C370S400000, C370S412000, C370S498000

Reexamination Certificate

active

06335934

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a switching system adapted to perform switching processings of voice, data, image information and the like and more particularly to a time-division channel arrangement suitable for performing switching of information, in a unit of a block accommodating communication control information, between switching nodes or modules in a star-type configuration.
2. Description of the Related Art
In the field of switching systems, studies have concentrated on the realization of high-speed and wide band performance, increase in the capacity of the channel and advanced improvements in reliability. Conceivably, a specified expedient to this end resides in the decentralization of a channel system, especially, a multiplexer stage. This expedient takes advantage of the decentralization of load which contributes to improvement in throughput increase in the capacity of the overall exchange and risk-decentralization owing to the decentralization of the channel system.
In the past, the exchange has played a major role in switching low-speed telephone voice, but it is expected that high-speed data communications as applicable to image data will be in great demand in future. Under the circumstances, the exchange is required to be drastically improved in call throughput but throughput of a processor can not be extended unlimitedly. To promote the throughput of the processor, the distribution of function and the distribution of load based on a multi-processor scheme may conceivably be adopted. This countermeasure, however, invites complexity of software and besides suffers from a bottleneck caused by communications between processors and peripherals such as magnetic disks, printers, etc., thus failing to eliminate the limitation imposed on the throughput.
Incidentally, from the standpoint of the enlargement of the system and the decentralization of risks such as system break down, call congestion or intended or unintended destruction of a switching system, it is desirable that individual function units of the exchange be realized with individual modules and these modules be physically distributed. In such an instance, unless an independent distribution configuration is adopted wherein processors per se are distributed, the load will be concentrated on a central processor and inconveniently the number of lines for controlling each distributed module will be increased.
In a distributed type switching system in which the speech path system is distributed and the control system is concentrated at a location, throughput of the control system is limited and the merits of the distribution of load can not fully mature to advantage. Under the circumstances, a foreseen switching system is desired to be of an independent distributed configuration in which not only the speech path system but also the control system is distributed. The use of independent modules is however problematic in that when a module communicates with another module, the sending module has to know whether a channel to the partner module is idle and whether an outgoing circuit from the partner module is idle. In other words, resource management is needed. Even in the ordinary distributed system, resource management is often concentrated at a location and all modules interrogate a common managing unit. Consequently, a bottleneck of processing is caused by the managing unit, especially, in a large-scale system. On the other hand, the system of full independent distributed modules is forced to take either a way to enable one module to constantly know the status of all of the other modules or a way to permit one module to confirm the status of the partner module each time a call is set up. In the former way, one module, when its status changes, is required to inform all remaining modules of the change or all of the modules must mutually confirm their status periodically. Even with the above procedure completed, when a module has only one idle circuit, there is a possibility that the remaining modules will transmit communication requests to that module at the same time. In the latter way, the above problems are not encountered but the necessity of mutual communications among all of the modules is by itself problematic. As a countermeasure, it is conceivable to establish communication lines in a mesh configuration among the modules. However, this becomes costly due to the complicated physical geometry and due to the fact that additional management of communication is needed.
Incidentally, in recent years, the trend of integratedly communicating various kinds of information including voice and data has become active. This trend originates from a desire for efficiency and economization but realization of the integrated communications needs integrated switching processings of various kinds of information. In particular, it is desired that switching of circuit switching information such as voice of which real-time base processings are required and switching of storable data or storage switching information which has hitherto been handled by a packet exchange can both be effected integratedly through the same channel.
A time-division channel arrangement for integratedly handling voice and data, that is, a so-called integrated channel arrangement is known as disclosed in JP-A-61-60044 entitled “Block switching system” and JP-A-60-127844 entitled “Circuit/packet integrated switching system”. In the former literature, also as disclosed in Proceedings of International Switching Symposium 1987 SESSION B. 7.1 “ELASTIC BASKET SWITCHING—A NEW INTEGRATED SWITCHING SYSTEM FOR VOICE AND HIGH-SPEED BURST DATA—” by S. Morita et al, a boundary identifier indicative of the boundary between communication channels is dynamically set in accordance with the amount of information in a set-up call in order that either of information requiring real-time base processing and information requiring burst base processing can be switched one-dimensionally. In the latter literature, also as disclosed in Proceedings of International Switching Symposium 1984 SESSION 42B-3 “SYNCHRONOUS COMPOSITE PACKET SWITCHING FOR ISDN SWITCHING SYSTEM ARCHITECTURE” by T. Takeuchi et al, a plurality of distributed communication nodes are connected in loop, the communication node being a switching module having a subscriber interface or a trunk interface, and each switching module packets circuit switching information such as voice and packet switching information such as data in blocks of fixed length in accordance with destination switching modules and adds a destination node number to each block, so that the information may be switched through the loop channel.
In the aforementioned “Block switching system”, the time switch function of the channel can efficiently be realized but because of each block being of a variable length, the space switch function of the channel for mutually rearranging blocks on a highway is difficult to achieve. Accordingly, this system is unsuitable for a large-scale exchange.
In the aforementioned “Circuit/packet integrated switching system”, on the other hand, all of the switching modules access the loop and in order to prevent the overall throughput from being decreased, the loop must be operated at so high a speed that the total of amounts of throughput of the individual modules can be handled by the loop. Accordingly, the larger the scale of the exchange, the more the high-speed device will be used in the interface between each switching module, and the loop becomes expensive.
In order to solve these problems, it is effective to use a block of fixed length which accommodates connection information and to construct a self-routing network which does not use a loop. However, while the use of the loop is convenient in that all the blocks carried on the loop can be processed sequentially with ease, a plurality of blocks used in, for example, the star-type configuration must be rearranged to prevent them from colliding with each other at a time within the central node. To this end, avoidance of

Affiliated with

Also associated with

Rating

Distributed type switching system 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 type switching system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Distributed type switching system will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2833364