ES2351878T3 - LIMITATION OF TRAFFIC THROUGH ADMISSIBILITY CHECK FOR A NETWORK WITHOUT CONNECTION, ORIENTED TO PACKAGES WITH QOS LEVEL TRANSMISSION. - Google Patents

Abstract

Procedure for limiting traffic in a connectionless, packet-oriented network, for efficient transmission with QoS level of data packets of at least one priority class, in which for a group of data packets to be routed to Admissibility checks are carried out through the network of a priority class, characterized in that - the admissibility checks comprise an admissibility check with respect to the network input and the network output, and - the transmission of the data packet group with the required priority is only admitted, if the admissibility checks have a positive result.

Description

Traffic limitation by checking admissibility for a connectionless, packet-oriented network with QoS level transmission.

The invention relates to a process, to a network, a marginal node and a server for the limitation of traffic on an offline network, packet-oriented, for a Effective transmission with QoS level of data packets from to Less a priority class.

The service integration of telecommunications and data entails a series of requirements regarding the switching technique and networks. To be equally suitable for infotainment and traffic commercial, used networks must have a high capacity, allow Real-time transmission, be reliable and ensure high degree of security Also as an additional condition are Costs as small as possible.

Data services are developed currently largely through IP networks (networks based on the  Internet Protocol), which work offline - at IP level - and package oriented. Progress in the technology of routers have led to the development of IP routers, which with respect to the capacity of switched data traffic and delay times by queues mainly enable telecommunication services and services with wide consumption of band such as video on demand or video conferencing in real time.

Difficulties arise in the case of a degree of high use of the IP network, with which the times of delay increase exponentially or in the case of accumulation in excess data traffic in individual sections, which act then as a bottleneck and limit the rate of transmission.

Due to these difficulties in conventional IP networks, it is not possible, to the extent desired, to guarantee a high quality of service - referred to in the literature in the majority of cases quality of service (QoS, Quality of Service ).

The improvements are aimed at to be able to make better commitments regarding the quality of service, without losing the advantages of reduced complexity and IP network flexibility.

The differentiated services model (Diff-Serv, Differentiated Services ) is based on the observation that the best-effort processing of data packets in the IP network leads to the difficulties mentioned above in quality assurance of service. Conventionally, data packets are transmitted, if possible, quickly and completely, without any guarantees regarding the reliability and security of the transmission. In case of high use or overload of the network there is a worsening of the quality of service due to delays or the discarding of data packets.

The Diff-Serv concept aims at a better quality of service for services with high quality requirements by introducing service classes. In this context, a CoS ( class of service ) model is often also discussed. The Diff-Serv concept is described in the RFCs published by the IETF under numbers 2474 and 2475. RFCs 2638 and 2998 address additional aspects of the concept. Within the framework of the Diff-Serv concept, using a DS field ( Differentiated Services , differentiated services) in the IP header of the data packets by setting the DSCP parameter ( DS codepoint , DS code point) a prioritization is carried out of packet traffic. This prioritization takes place with the help of a " per hop " allocation of resources, that is, packets are processed differently at the nodes according to the class of service set in the DS field by the DSCP parameter. In this context the expression behavior hopping (PBH, per-hop behavior) is used. For example, during a PBH, preference is given to higher service classes with respect to queuing and processing queues at nodes.

The core elements of a network that is based on the Diff-Serv concept are the DS subnets - in the English literature often referred to as the DS domain (DS domain) or single routing domain - and the marginal DS nodes. In many cases a subnet will correspond to the network of a service provider ( service provider domain ). In the case of the marginal nodes DS, the difference between access nodes DS ( ingress nodes , entry nodes) and exit nodes DS ( egress nodes , egress nodes). The data packets arrive through a DS access node to a DS subnet and leave the subnet through a DS output node. A marginal node DS may in this sense bring together the functionality of an access node DS for incoming traffic and an output node DS for outgoing traffic. The functionality of the marginal DS nodes includes the selection of data packets corresponding to the DSCP parameter and the marking of data packets with the help of DSCP parameters. There is also the possibility in the DS subnet, with the help of devices for the management and control of data traffic
( traffic conditioning ) to take control measures such as the measurement of data flows, the distribution of data packets in queues or the discarding of data packets. These control measures are often taken on DS access nodes or DS output nodes. In the context of traffic conditioning, data packets can be classified and compared with a traffic profile provided for the corresponding class of service (eg bandwidth, resources). In case of deviations from the traffic profile, measures can be taken, such as queuing in a queue or discarding data packets.

Normally, in a DS subnet, flows or connections are prioritized. Prioritization takes place in the DS access node for the corresponding DS subnet, if necessary by setting or modifying the DSCP parameter. In the nodes within the network ( core nodes , central nodes) the DSCP parameter is read and assigned to resources correspondingly to the prioritization. The interpretation of the DSCP parameter and the allocation of resources of the individual internal nodes take place independently of each other (jump behavior). In the output node, a modification of the DSCP parameter is eventually reversed in the access node, that is, the DSCP is restored to the original value. In this way, the DSCP can be adapted locally, that is, depending on the subnet and its properties.

The Diff-Serv concept avoids complex path or bandwidth reservation procedures and allows the prioritization of data traffic. In the case of a transmission through several subnets, service classes are established with the help of the so-called Service Level Agreements (SLA) for the entire transmission and are transformed by individual subnets as described above in the traffic conditioning framework. However, in practice, temporary and / or local bottlenecks occur, for example, due to the accumulation of data traffic in individual sections. As a general rule, data packets with the same destination, from the moment they coincide on a node, follow the same adjusted path. In the case of bottlenecks, the Diff-Serv concept ensures that data packets with a low priority first experience delays or discard. The transmission quality for highly prioritized data packets is improved with this, but quality standards cannot be guaranteed, for example, for real-time transmission. Within the framework of the Diff-Serv concept, it would only be possible to guarantee a transmission with QoS level, that is to say a transmission, in which certain quality commitments can be provided and maintained, if the traffic profiles are adjusted with such a reduced use of the subnet, that the load peaks would be absorbed by the reserve bandwidth. As a general rule, this is not done for economic reasons, that is, due to the reduced network utilization resulting from it. For this reason, there is talk in relation to the Diff-Serv concept of a CoS (service class) approach and not of a QoS (service quality) approach.

In packet-based, connection-oriented networks, known primarily as the Asynchronous Transfer Mode (ATM) technique, maintenance of the QoS level is generally caused by limiting the amount of traffic per means of access controls.

H. Hussmann et al. describe in "An edge device for supporting internet integrated services over switched ATM Networks", September 1998 (1998-09), Interworking Conference , Toronto XP002230478, two scenarios in which an admissibility check takes place, which refers to a reservation of resources along a path in a connection-oriented network (ATM), either at the network exit (1st scenario) or at the network entrance (2nd scenario) . The access control, however, is oriented, regardless of where it is performed, to the transmission capacity of the path to be used previously.

WO 01/54448 A discloses a procedure to control access to a communication network, in which an access function assigns an access node a available capacity, which is available for the access node to transmit traffic flows to the communication network. This available capacity, in case of a capacity modification Total network transmission, can be modified so adaptive With this procedure the quantity can be adapted total traffic, which is allowed to access a network, regardless of your destination and limit yourself in this way. With This is intended to achieve a desired quality of service or avoid Discard or loss of data packets. However I don't know Consider the treatment of possible bottlenecks in the area out of a network that works offline through the already described accumulation of traffic flows on the same route or To the same destination. Total allowable transmission capacity in an offline network should therefore be adjusted in a way correspondingly reduced and ineffective.

The invention aims to enable a efficient transmission of data packets with QoS level for networks that work offline, packet switched.

The objective is solved by the objects of claims 1, 13, 18 and 20.

The invention relates to a network with QoS Effective, it works offline and packet switched. He concept for the network comprises the following three considerations:

?

High efficiency requires flexibility in the choice of data packet paths or in its distribution. With a dynamic adaptation or modification of roads can be countered for example the formation of necks of bottle of resources.

?

High discharge can be avoided complexity, taking locally, for example in each case by the affected router, decisions (for example on roads, queues, discarding data packets, etc.). Be avoid situations or non-local reserves of roads. The decisions on paths can be taken according to the data package or according to the connection or flow. In this sense the distribution of packages of Individual data is the most flexible.

?

To guarantee QoS, it they require rigorous limits and monitoring regarding the network utilization

This patent application deals with the third appearance.

Through admissibility checks according to the invention it is intended to check if they are available in the network resources corresponding to the requirements in the transmission of a group of data packets of a priority class. According to invention admissibility checks are carried out for data packet groups of at least one priority class that They must be routed through the network. These checks of admissibility include admissibility checks with regarding network input and network output, which also They can be performed at the network input and network output. Alone in the data packet groups, for which all admissibility checks have a positive result, it supports transmission with the priority class of packets of data. On the contrary with the data packet groups, in the that one of the admissibility checks has a result negative, proceed in another way (claim 1).

A group of data packets may come given in this sense by the data packets of a traffic flow or by the data packets accumulated in a physical port, by example in a network access (claim 2). A flow of traffic corresponds for example to a flow, includes packets of data from a connection or comprises data packets with a identical address information, such as all data packets with the same source or the same destination (claim 3).

To guarantee a transmission with level of QoS, an overload of the total network capacity and the emergence of resource bottlenecks. In this sense, in the admissibility check of the data packet groups to be transmitted, parameters such as by example packet transmission rate and / or average data, peak rate, etc., and it can be checked if sufficient capacity is available of transmission for the transmission of traffic flows with the quality of service required. It is also guaranteed by admissibility checks, which on the input side as well as sufficient resources are available on the output side (for example bandwidth, queue capacity, ...) (claim 4). The check can take place depending on the priority class of the data packets or depending on the traffic with the same or higher priority (claim 5). For example without checking the admissibility of resources by the output side could be transmitted from several inputs of the network at the same exit and thus produce at the exit a neck of bottle. Then the quality guarantees could not be fulfilled, but in the best case only - as in the concept Diff-Serv - commitments regarding prioritized processing of traffic flows.

As a criterion for a positive result of the admissibility checks can serve for example a value threshold, which is determined based on the input capacity and network output used in each case, of the total capacity of the network, of the desired quality or priority class etc. For example, for a group of data packets to be transmitted with a priority traffic parameters are recorded such as the rate of transmission of packets and / or average data and the peak rate (claim 7). You can also register the priority class desired. Alternatively, the priority class due is determined to parameters or requirements such as real-time transmission and maximum loss rate

It is also conceivable that for each kind of priority there are several threshold values based on different valuation parameters, which must all be met separately or in corresponding dependencies with each other. In case of a result negative of the admissibility checks can be rejected the transmission of the data packet group (claim 6) or transmission with a lower priority or without prioritize

Due to admissibility checks according to the required quality characteristics can be performed a reserve of resources (eg bandwidth, queues of wait) (claim 8). This reservation will affect, as a rule In general, the affected network access and output as well as the load Total network (eg capacity, queue processing of waits correspondingly to the prioritization).

Compliance with the central traffic parameters, recorded for the respective traffic flows, such as for example the transmission rate, should be monitored if possible, to ensure compliance with the limit values or threshold values for the use of the network ( claim 9). The monitoring function - the English terms traffic enforcement ( policing ) and policing (surveillance) are often found in the bibliography for this purpose - will conveniently compare the traffic parameters indicated in the quality characteristics requirement with the parameters of actual traffic of the corresponding traffic flow. Unregistered data packets can be blocked (claim 10). In the network input, traffic shaping mechanisms can also be used
( traffic shaping ) known as dynamic counter ( leaky bucket ) or token bucket . Possible measures to avoid overload are for example

?

discarding packages data

?

packet marking of data

?

memory storage data packet intermediate

?

switching or blocking of traffic flow

?

the transformation of data packets that breach the agreement or the entire flow of data associated with a lower priority class or processing corresponding to the best effort approach.

A transmission with QoS level of flows of traffic with corresponding priority class or traffic class presupposes a corresponding processing of the priority class. It is convenient to load only part of the total capacity with prioritized traffic. The other part of the network capacity is loaded then with non-prioritized traffic. This non-prioritized traffic can then be processed correspondingly to the principle of best effort (claim 12). The checks of admissibility can then be limited to prioritized traffic (claim 11). Through this limitation of prioritized traffic it is guaranteed that the capacity of the network can be used completely, without loading spikes having a negative effect on prioritized traffic. The level of quality, with which transmits non-prioritized traffic, then acts almost as memory intermediate for prioritized traffic. A possible way to proceed, to set limits for use with prioritized traffic, is to indicate for each priority class a percentage utilization fixed maximum for traffic with the same priority class or a higher. For example, in a network with two priority classes for traffic with the highest traffic class could be set on 30% use limit and set the limit for the traffic with the lower or higher priority class at 60%. For non-prioritized traffic, then a capacity would remain 40% minimum.

In the transmission of data packets they can thus guarantee quality characteristics in a specific way priority classes, which enable a transmission with level of QoS. The concept according to the invention starts in this sense from the observation that an adequate quality of service (QoS) must Orient with respect to the respective service. For example organs Human sensory can process incomplete information in a certain measure, without a subjective loss of quality. For interactive machine control (for example remote control of robots) the requirements are clearly higher in certain circumstances. Then criteria should apply correspondingly more rigorous. Depending on the service they can defined therefore criteria or limits, which guarantee a QoS level transmission. In the case of oriented transmission to packages these criteria refer among others to

?

the type and extent of possible loss of information

?

fixed delays and / or variables

?

temporal consistency (sequence) of the information.

According to the invention, the network is equipped with means for conducting an admissibility check in network input and network output for the transmission of a group of data packets to be routed through the network of a priority class (claim 13). Can be sized in the sense of a transmission of prioritized data packets with QoS level. Sizing can take place in this regard according to limit values for factors or criteria that determine the QoS (claim 14). Depending on the service, values may be indicated maximum for statistical probability, with which they can exceeding the limit values for the factors that determine the QoS, without significant loss of quality. These maximum values for statistical probability can influence the sizing of the network (claim 15).

Complying with the limits of use for the network - possibly within the framework of a statistical probability service-specific default - and limitation of prioritized traffic as well as through a good distribution of traffic and traffic limitation at the entrances and exits of the network or physical ports may indicate statistical values for quality loss factors. With the help of these values Statistical and its variance can be guaranteed QoS services.

In many cases it is convenient that the network corresponds to a single routing domain or network of a service provider (service provider domain) (claim 16). This way they can be integrated in a way easier the functions of admissibility checks and of Network monitoring. It is not necessary to provide the External confidential network performance data. If of transmission across several domains of unique Routing or service provider networks is guaranteed through the guarantee through the network management in the respective network and checking network accesses and outputs, which is possible transmission of QoS level across several networks according to the invention connected to each other and if necessary in a central computer connected to a network according to the invention. So Internet with QoS level transmission can be built from of networks operating according to the invention, comprising example in each case a single routing domain. In this meaning can be used, as in the Internet framework conventional, IP-based networks (claim 17), which by example transmit non-prioritized traffic so corresponding to the principle of best effort.

An important role corresponds to him in this meaning to the marginal nodes of the networks, through which data traffic arrives at the respective network or leaves again the same. Through traffic eligibility checks prioritized with respect to the resources in the network input and in the network output can be guaranteed that in the network transition Maintain the quality of service. For checking admissibility marginal nodes may be equipped with means for carrying out the admissibility checks in the network input or network output (claim 18). Nodes marginals can also present means for the functions of monitoring (claim 19). It can also be arranged, alternatively or in addition to the proportion of the functions in marginal nodes, a server with means for checks of admissibility (claim 20) and optionally for monitoring (claim 21) in the network.

Below are two variants of the object of the invention within the framework of the exemplary embodiment. Show

Figure 1: a system with data transmission through a network according to the invention

Figure 2: a network according to the invention

Figure 3: a schematic representation of different paths for routing two flows in a network according to the invention

Figure 4: a schematic representation of different paths for routing two flows with the same destination in a network according to the invention.

For reasons of clarity it should be assumed that the invention will be applied within the framework of a telephone call to over an IP network, IPN - voice over IP is spoken in this case (VoIP) In the case of this IP network, IPN, it can be for example of a single Internet routing domain. The calls Phone numbers are subject to real-time requirements. Therefore it prioritize the associated data traffic. Naturally the object of the invention can be applied analogously to all others services where a traffic prioritization of data. Examples of such services are video on demand, web conferences, multimedia applications, etc.

A system with VoIP transmission is schematically represented in figure 1. Through the AN-A and AN-B access networks (AN: for Access Network ), the TLN-A and TLN-B telecommunications terminals are connected to a public network, which includes the IP network, IPN . Within the framework of the two variants of the embodiment, it is assumed that a connection to the TLN-B terminal is established from the TLN-A terminal for the purpose of a telephone call. In this sense, a distinction is made between service level SL (SL: for service level ) and network level (NL: for network level ), which is marked in the figure by a dashed line. At the SL service level, the SIG (VA, DS) signaling (SIG (VA, DS) signaling for: connection establishment signaling and service control) of the connection establishment and the service control takes place. To this end, the CCP-A and CCP-B control devices (CCP: for Call Control Point , call control point) for example media gateway controllers or switching facilities are connected to the networks of access AN-A and AN-B of terminals TLN-A and TLN-B. The transmission of useful data takes place at the NL network level and is at least partly routed through the IPN network according to the invention (IPN: for IP Net , IP network). The IPN network works offline and packet oriented. Useful data packets, which can be transmitted within the framework of the telephone call from the TNL-A terminal to the TLN-B terminal, arrive through the marginal node IgNd (IgNd: for ingress node ) to the network IPN and leave it again through the marginal node EgNd (EgNd: for egress node , egress node).

In the course of connection establishment at service level SL is signaled from the terminal TLN-A through the access network AN-A to CCP-A control device The desire for connection. The TLN-A terminal is authenticate, for example by name or address. Then the terminal is identified and located that receives the TLN-B call or the device from CCP-B associated control. It is usually transmitted a connection establishment message from the device CCP-A control to the control device CCP-B. In the control device CCP-B relevant information is extracted and in the network of AC-B access checks the availability of the TLN-B terminal as well as information relevant. The CCP-B control device gives finally know the connection establishment message to CCP-A control device and the necessary information for the connection such as address information of the TLN-B terminal. TO connection establishment can then be made at level of service SL. In the case of a connection establishment satisfactory, then useful data can be exchanged at the level NL network

In the chosen example it is exchanged as data Useful real-time voice information, that is with level of QoS For the transmission with QoS level they are made according to the invention admissibility checks. Within the framework of these admissibility checks a signaling takes place, by example in the transmission of the desired quality requirements, for the transmission of the result of the checks of admissibility, etc. Hereinafter this signaling will be called QoS signaling. Two variants are distinguished, as the QoS signaling in the context of admissibility checks at the SL service level or at the NL network level.

In the case of a QoS signaling at the level of SL service the procedure according to the invention can proceed as follows: through level control devices service, CCP-Ig and CCP-Eg, se identify the marginal nodes IgNd and EgNd, through which Useful data is transmitted as useful data packets. The CCP-Ig control devices and CCP-Eg represented in Figure 2 may, but not they have to be identical to the control devices CCP-A and CC-P associated with networks of access AN-A and AN-B.

As a general rule, a plurality of control devices will be involved in signaling at the service level inter-city telephone communications SL. These control devices then have direct access only to a section or a partial network of the entire transmission segment for useful data. The control devices CCP-Ig (CCP-Ig: for call control point at ingress node , call control point at the entry node) and CCP-Eg (CCP-Eg: for call control point at egress node , point of call control at the point of egress) are characterized in that they communicate with the marginal nodes IgNd and EgNd. The two control devices CCP-Ig and CCP-Eg in this regard can also coincide.

Figure 2 shows additional marginal nodes. The internal CoNd nodes (CoNd: for core node , central node) are not represented individually. An NCS control server (NCS: for network control server ) is also shown, through which network management tasks are performed.

The admissibility checks according to the invention NAC (NAC: for network admission control , network admission control ) with respect to the network input and the network output take place for example in the marginal nodes IgNd or EgNd. The requirements for the VoIP telephone call between terminals TLN-A and TLN-B are transmitted to the marginal nodes IgNd and EgNd via the control devices CCP-Ig and CCP-Eg of the SL service level. The requirements may include, in addition to the relevant traffic parameters, such as bandwidth and QoS requirements, additional parameters regarding reliability, security, etc. The result of the admissibility checks in the network limits is signaled to the control devices CCP-A and CCP-B. Depending on the results of the admissibility checks, on the side of A, the transmission of useful data is released or blocked and an alternative route for the transmission of useful data is optionally sought (in the English bibliography also known as Bearer Redirection (bearer redirection) Admissibility checks can take place after or during the connection establishment at the service level SL In the admissibility checks during the connection establishment, in case of a negative result, the establishment may eventually be interrupted of connection.

In QoS signaling at the NL network level it is not possible to begin admissibility checks until after connection establishment at service level SL. After a successful connection establishment at service level SL is releases the QoS signaling at the NL network level and transmits the relevant information, such as information from direction from the side of B. This can take place through a corresponding message from the control device CCP-A to a device located in the access network AC-A, for example media gateway. For the QoS signaling can also be made available from the level SL service for example from the control device CCP-A a program structure, for example a Java Applet of the device in the access network. With the help of the address information consulted within the establishment from service to service level and if necessary with the help of downloaded program structure is sent at the network level a signaling message to the terminal on the B side TLN-B or the access network on the B side AN-B Through this and optionally messages from additional signaling are located the marginal nodes IgNd and EgNd and the checks on the marginal nodes begin admissibility. The results of the checks of admissibility are signaled below by level messages network NL to the access network on the side of A AN-A or to the participant on the side of A TLN-A.

In addition to the admissibility checks with respect to the network input and output, a admissibility check relative to the total capacity of the net. This admissibility check can take place, by example, also in one of the marginal nodes IgNd or EgNd, of distributed form on both marginal nodes IgNd and EgNd or in a server planned for it on the network.

Figure 3 shows a representation Schematic of different paths for routing two flows in a network according to the invention. The network limit is represented by a dotted line. The network nodes are illustrated by circles, cutting the circles, which represent marginal nodes, the dotted line. With the help of arrows, They represent possible paths for flows. Dashed arrows they refer to possible paths for a flow, which enters the network in the marginal node C and transmitted to the marginal node D, where the Flow data packets leave the network again. Through Continuous arrows represent possible paths for a flow, which is routed from marginal node A to marginal node B. In most internal nodes there is more than one alternative or ramifications for the flow path. Then for a determined flow the different possible road sections alternatives from an internal node to the next node, is say for the "next jump", they are called compartments of branch of the flow in the corresponding internal node. In the internal nodes, which are in possible paths of both flows, indicates whether the branching compartments of flows are identical i, partially disjoint t or disjoint d.

Possible paths are determined by the network parameters, such as topology, capacity of the individual road sections, delay times, etc. The decisions on the branch of the road compartment, through which a packet or group of packets continue to be transmitted, they are taken locally based on traffic parameters in force at that time. In this way a use is achieved relatively uniform network and bottlenecks are avoided load.

Figure 4 differs from Figure 3 because now the two flows leave the network at the marginal node B. In the internal nodes, which are in possible paths of both flows, the distribution compartments of both flows are identical, which is illustrated with an i in the nodes. The distribution compartments, within the framework of traffic control, or traffic shaping, they can also differ from each other. Without however for flows with the same destination they will be at least for the most part identical part. Through the admissibility checks according to the invention is achieved that the marginal node B and the nodes internal, which are topologically adjacent to the marginal node, not are overloaded by the data packets that arrive from both flows. When both streams cannot be transmitted with the Registered parameters, for example one is not supported. In the case of the limitation of the admissibility checks at the entrance of network could occur in the constellation of figure 4 a neck bottle in the marginal node B, which would go against the QoS warranty commitments.

Claims (21)

1. Procedure for limiting traffic on a connectionless, packet-oriented network for efficient transmission with QoS level of data packets of at least one priority class, in which for a group of data packets that must routing through the network of a priority class admissibility checks are carried out, characterized in that - admissibility checks include an admissibility check regarding the network input and network output, and - data packet group transmission with the required priority is only supported, if the checks of Admissibility have a positive result. 2. Method according to claim 1, characterized in that - the group of data packets is given by the data packets of a traffic flow or by the packets of priority class data accumulated on a port. 3. Method according to claim 2, characterized in that - the traffic flow is given by a flow, is associated with a connection or comprises data packets with a common address information. 4. Method according to one of the preceding claims, characterized in that - in the framework of the checks of admissibility the degree of utilization relative to total network capacity and available bandwidth of network access on the input side and on the output side that they must be used in the transmission of the data packet group. 5. Method according to one of the preceding claims, characterized in that - in the framework of the checks of admissibility the degree of network utilization is taken into account and of the accesses to the network on the input side and on the side of output to be used in the transmission of the packet group of data, the traffic of the traffic flow priority class that has to be transmitted and / or the traffic of a priority class, which is greater than or equal to the priority class of the package group of data. 6. Method according to one of the preceding claims, characterized in that - in case of a negative result of the admissibility checks the packet group is rejected data. 7. Method according to one of the preceding claims, characterized in that - For the group of data packets to be routed through the network, traffic parameters are recorded in the network, which include information regarding the transmission resources needed.
rivers. 8. Method according to claim 7, characterized in that - according to the traffic parameters registered, Make a reservation of transmission resources. 9. Method according to claim 7, characterized in that - during the transmission of the traffic flow Check that the registered traffic parameters are met. 10. Method according to claim 7, characterized in that - unregistered data packets are block / discard 11. Method according to one of the preceding claims, characterized in that - only checks are made of admissibility for prioritized traffic. 12. Method according to one of the preceding claims, characterized in that - Unprioritized data packets are transmitted with best-effort quality. 13. Offline, packet-oriented network, for the transmission of data with means for the realization of a admissibility check at the network input and output of network for the transmission of a group of data packets of a priority class, to be routed through the network. 14. Network according to claim 13, characterized in that - the network is sized according to limit values for factors that determine the QoS. 15. Network according to claim 14, characterized in that - the network is sized according to maximum values for the statistical probability, with which the values are exceeded limit of the factors that determine the QoS. 16. Network according to one of claims 13 to 15, characterized in that - the network corresponds to a single domain routing 17. Network according to one of claims 13 to 16, characterized in that - a data transmission can be made to over the network with the help of the IP protocol. 18. Marginal node for use in a connectionless, packet-oriented network according to claim 13, characterized in that - the marginal node is equipped with means to performing an admissibility check at the entrance network or network output for the transmission of a stream of traffic of a priority class, to be routed through of the network. 19. Marginal node according to claim 18, - with means for monitoring Traffic parameters registered in the network. 20. Server for use in a connectionless, packet-oriented network according to claim 13, characterized in that - the server is equipped with means for performing an admissibility check at the entrance of network and network output for the transmission of a group of data packets of a priority class, to be routed via network. 21. Server according to claim 20, - with means for monitoring Traffic parameters registered in the network.