Telephonic communications – Plural exchange network or interconnection – With interexchange network routing
Reexamination Certificate
2001-12-22
2004-11-30
Tieu, Benny (Department: 2642)
Telephonic communications
Plural exchange network or interconnection
With interexchange network routing
C379S112010, C379S112040
Reexamination Certificate
active
06826273
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to a method for preventing, or mitigating the effects of, an overload condition in a circuit switched telecommunication network. In some embodiments, such a network includes one or more arrangements where a group of client terminals are statistically multiplexed onto one or more transmission paths, the signals of active ones of the terminals are brought to a head end terminal (HET), and those signals are then communicated from the head end terminal to a central office switch of the network (henceforth, near-end office) over a shared transmission path. By statistically multiplexing of terminals onto the transmission paths it is meant that the transmission paths do not have sufficient capacity to service all terminals simultaneously. One example of such an arrangement is a network that employs a digital (subscriber) loop carrier system. Another example is a network that provides circuit switched telephony over hybrid-fiber coax (HFC). Still another example is a network where HET
20
is a PBX. In this example, however, terminals that are connected to the PBX are not employing a shared transmission resource in order to reach the PBX. Closely related to the above are arrangements that employ central office switch concentrators. In the context of this disclosure, therefore, a concentrator at the ingress of a near-end switch is considered to be a head end terminal.
In all such arrangements there exists an element that concentrates traffic from numerous client terminals onto a transmission path, but that path is capable of providing simultaneous service to only a fraction of the client terminals. Normally this is acceptable because not all client terminals are likely to need service at the same time, and the shared transmission path is engineered to have sufficient capacity to insure that the probability of a client terminal being denied service because the transmission path is fill is below a chosen, specific, level.
One aspect of such prior art arrangements is that the near-end office provides the basic functions needed to establish connections, such as dial tone, digit detection, etc. Employing the dial tone generation and digit detection circuitry of the near-end office simplifies the architecture and reduces cost by avoiding duplication of features both at the HET and the near-end office. The disadvantage of this arrangement, however, is that when circuits in the shared transmission path are congested, new callers cannot connect to the near-end office and, therefore, do not even get a dial tone. Consequently, dialing digits are not transmitted by client terminals (or, if transmitted, are not received) and that leads to difficulty in implementing intelligent admission controls. This condition, unfortunately, has been observed more and more in practice because, with the increased popularity of the Internet, more computers utilize the network, and computers connected to the Internet often have holding times (the time intervals between off-hook and on-hook) that are much longer than those that were expected when the shared transmission path were originally engineered.
The above describes one source by which congestion can occur at, or prior to, the near-end office. However, even in network arrangements that do not have a shared transmission path between client terminals and the switch at the near-end office (near-end switch, for short), congestion can occur, simply because of an extraordinary use of the near-end switch in response to some event, such a radio “call-in” contest. That can occurs of course, when components of that near-end switch that are shared, such as a concentrator or a digit collector, become overloaded. It might also be caused by a large number of computer users whose computers automatically keep redialing a busy Internet Service Provider (ISP) modem bank in an effort to obtain a modem the moment one becomes free.
The central network (i.e., the network core excluding the near-end offices and their connections to users) is often protected from some of these overload conditions by means of a “choke” network. This is typically implemented with a limited number of trunks that are set up in a trunk group specifically for a “call-in” contest telephone number. Once these trunks are busy, the switch returns a re-order tone (fast busy) to any additional user attempting to access the “call-in” number. This protects the central switches from dealing with a large number of calls to a number that is almost certainly busy, and allows normal use of the core network. The end offices themselves are, however, not protected from rapid retries, except in the sense that a user is likely to give up after a few busy signals. Alas, the likelihood of giving up quickly has shank in recent years because many telephones have redial buttons (and computers have modem software that has redial options) which permit redialing with a very small effort on the part of the user. Moreover, the re-tries have become more closely spaced in time.
SUMMARY
Congestion problems in prior art networks are alleviated, and an advance in the art is achieved with a method that works toward insuring that unused capacity will always exist in elements of a network that are resources which are shard by a plurality of users and which, consequently, can be overloaded. In response to each a request to establish a connection over a path in the network, pursuant to a predetermined algorithm the method either services the request, declines to service the request, or services the request after dropping an established connection. In one embodiment, when unused capacity on the path is above a preselected level, all requests are serviced. When unused capacity falls below a preselected threshold, a probabilistic approach is taken as to whether to service the request or not, and as whether to drop an existing call in order to service the request. The probability that the request is satisfied without dropping an existing call reduces as the load on the path increases.
REFERENCES:
patent: 6606380 (2003-08-01), Petersen et al.
patent: 6731739 (2004-05-01), Riihinen et al.
Kaplan Alan Edward
Ramaswami Vaidyanathan
AT&T Corp.
Brendzel Henry T.
Le Karen
Tieu Benny
LandOfFree
Method for preventing overload condition in a circuit... 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 for preventing overload condition in a circuit..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for preventing overload condition in a circuit... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3360944