Electrical computers and digital processing systems: multicomput – Computer-to-computer protocol implementing
Reexamination Certificate
2000-05-17
2004-01-20
Alam, Hosain T. (Department: 2155)
Electrical computers and digital processing systems: multicomput
Computer-to-computer protocol implementing
C709S204000, C709S241000, C709S217000, C709S219000, C709S226000, C709S229000, C700S002000, C700S005000
Reexamination Certificate
active
06681253
ABSTRACT:
BACKGROUND OF INVENTION
The invention relates to processor systems used in the network elements of telecommunication systems, and particularly to allocation and configuration of resources to digital signal processors (dsp).
With reference to
FIG. 1
, a digital signal processing unit DSPU serving a connection
11
typically comprises, in addition to the actual processor (DSP), an interface I/O (Input/Output) to the connection, and program memory, which usually comprises read-only memory ROM and random-access memory RAM (and other supporting circuits, such as a clock and interrupt circuits, which are not, however, relevant to the invention). The connection
11
is divided into two parts: the part on which actual information is transmitted is indicated by
11
1
and the associated signalling is indicated by
11
2
.
A problem in a system of this kind is that the configuration is inflexible. If there are several processor units DSPU and if different applications are executed by them, the program memory ROM of each processor unit must be sufficiently large to store all the applications. The applications include, for example, different speech coding methods, methods of compressing and decompressing a video signal, echo cancellation, etc. Since in the prior art the applications are installed either permanently or semi-permanently (e.g. in rewritable ROM), the known processor units do not comprise a fast interface for updating the application. Consequently, another problem arises: when an application is being updated, the processor unit is out of use.
In order that not all the applications would be needed in all the processor units, the processor units are typically arranged in pools. The data streams that need a particular application are conducted to the pool serving the application concerned. The problem is that the pools are permanent, so the distribution of different data streams cannot be taken into account.
BRIEF DESCRIPTION OF INVENTION
The object of the invention is to provide a method and an arrangement implementing the method, solving the above problems. The object of the invention is achieved by a method and a system that are characterized by what is stated in the independent claims. The preferred embodiments of the invention are claimed in the dependent claims.
The basic idea of the invention is to use a broadband packet-switched telecommunication network, preferably an ATM network, for transferring applications from a common database of a multiprocessor system to processing units at least at the beginning of each connection and, if necessary, again during the connection. Preferably, the protocol of the ATM network is used, and no conversion to the higher-level protocols of the OSI model is needed.
A first embodiment of the claimed method and arrangement is based on arranging several processing units under a common control unit. The applications are concentrated on a common database. In each connection, the applications (services) required by the connection are identified and subsequently loaded, under control of the control unit, to an available processing unit. The processing unit and the database containing the applications are interconnected using fast connecting technology, preferably a broadband packet network, such as the ATM. Here the fast connecting technology means that an application can be loaded from the database to the processing unit during the connection, without that the user using the connection notices any disturbing delay.
The technology of the invention leads to flexible use of the resources, since in each processing unit memory is needed only to store the largest application (or set of applications). Not all the applications need thus to be stored in the memory. Any processing unit whatsoever can serve any connection whatsoever, since all the applications are available to all the processors. The suitable amount of memory in each processing unit is slightly greater (e.g. 30 to 100%) than the greatest amount of memory needed at a time or the amount of memory required by the largest single application or set of applications, but yet considerably smaller than the combined amount of memory required by the different applications. A set of applications means applications associated with one and the same connection at a time. In the GSM, for example, a set of applications could comprise a full-rate codec and echo cancellation.
The technology of the invention is also reliable in respect of modifications and updating. When an application is updated, the older version of the application can be retained although a newer version is loaded to the database. If the new application does not work properly, the older version can then be immediately put to use. When applications are updated, none of the processing units is out of use. Since the loading of all the applications of all the connections is controlled by the common controller, a further advantage is achieved: the number of times that each application is used can be easily counted, if the operator wants to prepare statistics or if the supplier of the applications wants to charge for the number of times of use, rather than a lump sum.
A second embodiment of the method and arrangement of the invention is a compromise between the prior art (in which the applications are located in permanent memories) and the first embodiment of the invention (in which the applications are loaded separately for each connection). The compromise is achieved, for example, by keeping track of available processing units and of what applications have been loaded in which processing units. At the beginning of each connection (and, where necessary, also during the connection), the applications required by the connection concerned are identified, and an available processing unit to which as large a part as possible of the applications required has been loaded is selected for the connection. The missing applications are then loaded to the processing unit from the database.
The definition ‘as large a part as possible’ can be interpreted in at least two ways: either as few applications as possible need to be loaded to the selected processing unit, or an application (set of applications) that is as small as possible needs to be loaded. If one processing unit can serve several connections, then the ‘available processing unit’ is here a unit that has at least some processing capacity available.
Apart from the above advantages, the second embodiment also has the advantage that if the necessary applications have already been loaded to one processing unit, no applications need be loaded to the processing units. This is particularly useful to the network operator in distributed systems, in which the database and the processing units are a long distance apart, and in which the operator also has to pay for the network capacity reserved for transferring the applications.
Although the maximum advantage is achieved by the invention when there are several processing units, the technology of the invention can also be applied when there is only one processing unit, if the number of the processing units can be expected to rise (e.g. as the traffic load increases). If there is only one processing unit, the selecting of an available processing unit simply means that it is checked whether all the capacity of the processing unit is already in use.
REFERENCES:
patent: 5005122 (1991-04-01), Griffin et al.
patent: 5421009 (1995-05-01), Platt
patent: 5504670 (1996-04-01), Barth et al.
patent: 5546453 (1996-08-01), Hebert
patent: 5600643 (1997-02-01), Robrock, II
patent: 5958010 (1999-09-01), Agarwal et al.
patent: 5970488 (1999-10-01), Crowe et al.
patent: 5999990 (1999-12-01), Sharrit et al.
patent: 6021442 (2000-02-01), Ramanan et al.
patent: 6061442 (2000-05-01), Bhat
patent: 6104721 (2000-08-01), Hsu
patent: 6163535 (2000-12-01), Jordan et al.
patent: 6195678 (2001-02-01), Komuro
patent: 6199093 (2001-03-01), Yokoya
patent: 6208642 (2001-03-01), Balachandran et al.
patent: 6226263 (2001-05-01), Iwase et al.
patent: 6240086 (2001-05-01), Morgan et al.
patent:
Korhonen Jari
Lehtimäki Matti
Suvanen Jyri
Alam Hosain T.
Duong Oanh
Nokia Networks Oy
LandOfFree
Adaptive processor 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 Adaptive processor system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive processor system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3194034