KR20070009548A - Handling of Capacity Bottlenecks in Digital Networks - Google Patents

Abstract

The present invention relates to a digital network, in particular a digital home, in which at least two devices or applications having air interfaces that affect internal resource allocation in a system have access to resources of a data transmission medium having a non- constant bandwidth. A method for dealing with capacity bottlenecks in networks. In the case of a resource bottleneck, the data rate of the data stream is reduced and its impact on the associated application is noticed by a minimum of users.

Description

Handling capacity bottlenecks in digital networks

The present invention relates to digital networks, in particular digital home networks, in which at least two appliances or applications with air interfaces that affect internal resource allocation in a system have access to resources of a data transmission medium having an inconsistent bandwidth. To deal with dose bottlenecks.

Information is processed in accordance with the increasing digitization and connection of processing media. As far as home range is concerned, associated networks are referred to as in-home digital networks (IHDN). Televisions, radios, monitors, loudspeakers, cameras, printers, scanners, PCs, telephone services, voice recognition systems, home appliance controls, security devices, and the like can be integrated into this IHDN. In particular, in audio and video media with their high data rates up to 100 mbit / s, conflicts of use often occur when different applications compete for resources (both device and network resources).

This effect is much stronger when the bandwidth of the data transmission medium changes, for example due to external influences. Users of CE appliances expect high quality observed pictures and sounds. The configuration of digital multimedia data at home can be wired or wireless. However, some wirelines, as well as all forms of wireless transmission, guarantee any bandwidth available because they share the transmission medium (cable, air) physically with uncontrolled appliances (vacuum cleaners, microwave ovens, Bluetooth devices). Can not provide. This lack of bandwidth guarantee can cause the available bandwidth to fall below the bandwidth required by the multimedia streams in operation. This may result in less data being sent for some streams than is required for representation in optimal quality. These capacity bottlenecks require prioritization of the data streams based on where the reductions in their required data streams are performed.

In addition to prioritization by brief description of the requirements of the data stream (eg TOS bits in IP), prioritization of the data streams according to the type of information transmitted (e.g., email versus web to video), It has been known for a long time. However, this method is inadequate for prioritization of data streams of the same type (eg, video and audio) that typically occur in digital home networks.

Traffic shape and control algorithms allow prioritization of their queues (eg, CBQ, HTB and PRIO protocols of Linux traffic control mechanisms). This method is particularly suitable for implementing stream priorities, but not suitable for fixing them.

Several methods are proposed for data stream transmission over the Internet using prioritization. In the method of standardized "differentiated services", different service classes are assigned to data packets and get resources consistent with their classes. Different classes may have different priorities. However, this is not defined as certain reference classes are assigned to packets. In addition, there are only a statistically selected number of classes.

All of these systems use priorities or even fixed priority classes predetermined by the applications.

In a real environment, decisions about what elements in the environment should be represented are made based on the virtual location and the virtual direction of the user's perspective. Similarly, in virtual video conferencing systems, decisions are made on the indicated streams of conference participants. This method is used for a continuous rule of stream qualities, which is not suitable for hiding deviations from the maximum quality as much as possible.

Accordingly, it is an object of the present invention to provide a method for enabling substantially high quality adoption of data streams from a user, in which quality adoptions are required due to capacity variation in digital networks with inconsistent bandwidth. According to the invention, this object is that, in the event of a resource bottleneck, the impact on the associated application is minimally observed by the user, and the data rate of the data stream is reduced.

The proposed method provides for digital networks, in particular digital, where at least two appliances or applications with air interfaces affecting internal resource allocation in a system have access to the resources of a data transmission medium having an inconsistent bandwidth. Used to address capacity bottlenecks in home networks. These capacity bottlenecks occur especially when systems without bandwidth guarantees, in particular wireless systems, are affected by external influences, such as domestic appliances, for example. In the scope of the method, the participating applications are given a priority based on their user-subjected quality requirements, and the data stream in which the assigned application has the lowest quality requirements from the user's point of view is the first to be reduced. In this regard, in particular, the following relationships are used. In other words,

Slightly noticed or not noticed streams, such as televisions operating in the kitchen during cooking, can be reduced in quality without the user noticing.

Recorded streams have the potential to be watched multiple times. During their recording, they have greater importance as streams watched live, while they have less importance during reproduction.

Streams watched by multiple people are of greater importance, especially when the rest are guests than those watched by only one person.

There are broadcasts and programs that are less important to the user, or some of which are considered less important, for example, video information provided at a music station.

At different times, streams of different users, for example, children's streams in the afternoon and adult streams in the evening, may have priority.

Different sources or end appliances raise different user predictions about the degree of interference. Obstructions about transmissions from the Internet or mobile appliance will be tolerated earlier than those from televisions in the living room.

In the case of a shortage of resources that occur, streams with the lowest priority are the first to degrade or be completely interrupted in quality. As a result, the user experiences as a phenomenon of not recognizing or less disturbing the requested degradation. This leads to a higher quality level subjectively.

In another embodiment of the invention, in order to select a data stream to be reduced, the priority for each stream is determined regularly by the users from the information about the context and from the contents of all the streams, and different applications / Operating system components are informed about this determined priority for each stream. This ensures that the applications can respond without delay in case of resource bottlenecks.

According to the invention, the applications in the system transmit their quality requirements in combination with their connection requirements. This ensures that the minimum requirements of the appliances or the applications are taken into account when calculating priorities. In conclusion, retransmission factors such as burning failures on CD due to data stream break off are avoided, for example.

In an advantageous embodiment of the invention, the user is able to reject the quality limitation of the system, which determines the next highest priority and notifies the applications about it. This allows the user to do so that the subjective quality the user experiences as the quality of the rejection can be offset.

Advantageously, user rejection is stored by the system and taken into account when determining future priority calculations. As a result, the system acquires learning capabilities to enhance its practical usefulness. The efficiency of the system is continually improved by the permanent adoption of the calculations on the quality experienced by the user as the actual quality.

Other embodiments of the invention are defined in the dependent claims.

These and other features of the present invention will be apparent from and elucidated with reference to the embodiments described below.

1 is a diagram of a method according to the invention.

As shown by the example in FIG. 1, in a digital home network (IHDN), different applications 1a, 1b, 1c contend for resources. The example above relates to a wireless network in accordance with the IEEE 802.11b standard. The network includes, for example, a television set 1a (in the kitchen) and a computer 1b in addition to the hi-fi equipment 1c. The applications 1a, 1b, 1c receive their data streams from a home archive 2. A network 5 supporting quality of service (QoS), in particular priority-based QoS, exists between the home archive 2 and the applications 1a, 1b, 1c. In the case of a wireless network according to IEEE 802.11b, this is substantially the support of the priority classes of IEEE 802.11e. However, similarly, IEEE 802.15.3 or home plug 1 can be used. When the network does not provide QoS, an exceptional case arises that the applications must implement QoS.

The applications 1a, 1b, 1c send their QoS requirements together to request their essential data streams from the QoS prioritizer 3. When legacy applications having direct access to the network 5 without considering the prioritizer 3 are involved, the corresponding accesses are made by the network 5 by the prioritizer 3. Detected by the prioritizer itself when communicating to or monitoring the network traffic. In these cases the missing requirements are replaced with default values, possibly depending on the type of applications or stream detected.

The prioritizer 3 checks on the basis of the application requirements 4 whether it results in a structural overload of the requested stream available network resources. In this case, it rejects the requirement. In the opposite case, the prioritizer applies the requirement 4 with information about the content (content metadata) 9 of the stream, the user's current desires and their current environment, and their actions 8. Strengthen. He receives this information from additional components such as the home archive 2, the user database 6 and the context detector 7 which are parts of the system. In principle, however, external sources may also be relevant. The requirement data enhanced in this new system is combined with the enhanced requirement data of already existing streams and checked for the presence of specific situations (such as the situations described above). Based on this, the priority is determined for the new stream and the priorities 10 of the existing streams are possibly changed. The new and changed priorities 10 are then communicated to the network 5 or the applications 1a, 1b, 1c in a corresponding format, in the example as priority classes according to IEEE 802.1d. When they are communicated to applications that do not implement QoS itself, these applications communicate the priorities 10 to the network 5.

For example, the prioritizer 3 assumes that the user will primarily observe the acoustically corresponding broadcast and rarely watch it, so that the lower priority 10 is taken from the kitchen to the television set 1a. We will assign it to the video data stream. In contrast, if the CD recorder integrated in the hi-fi equipment 1c has requested a stream from the archive 2 or from the PC 1b, the prioritizer 3 can determine that the interference of the stream has been recorded. High priority (10) will be assigned to this stream, as this may result in uselessness.

When the newly requested stream is a stream with scalable bandwidth, the prioritizer 3 does not determine the priority for the stream, but rather the priority for each of its transmission bands.

The current data flow 51 is actually transmitted under the responsibility of the network 5 alone. When resource bottlenecks occur, the QoS-enabled network 5 automatically reduces the bandwidth occupied by less important streams by only partially transmitting or delaying packets of these streams or not. When the applications 1a, 1b, 1c implement QoS, this reduction applies to them. When the user observes a decrease in quality contrary to the system's prediction and experiences it as an obstruction, he communicates it to the system via the user interface 11 of the television set 1a. This information of user rejection is passed to the QoS prioritizer 3 which calculates an alternative priority 10 and communicates it to the applications 1a, 1b, 1c and thus adopts the data streams. The user denial is stored by the QoS prioritizer 3 for the calculation of future priorities.

The prioritizer 3 is informed of changes in the prioritization, so that the data 52 of the user's current situation or the content of the stream, as well as the longer lasting bandwidth limit of the network 5, can be sent. Effect, possibly changing the priorities 10 of already existing streams and notifying the network 5 or the applications 1a, 1b, 1c.

In the above example, when the context detector 7 determines that a "kitchen work" condition applies to a user in the kitchen (for example, all kitchen products are powered off or the user is in the direction of the television set) As he continues to see), he notifies the prioritizer 3 of this change in such a condition while increasing the priority of the stream to the television set 1a.

Claims (11)

At least two appliances or applications 1a, 1b, 1c having air interfaces that affect internal resource allocation in the system may have digital networks 5 having access to the resources of a data transmission medium having a non- constant bandwidth. ), In particular, a method for handling capacity bottlenecks in digital home networks, In the case of a resource bottleneck, the data rate of the data stream is reduced and its impact on the associated application is minimally perceived by the user. The method of claim 1, In order to select the data stream to be reduced, the priority for each stream is determined regularly from information about the context of users and the content of all streams. Way. The method of claim 2, The transport network (5) is characterized in that it is informed about the determined priority (10) for each stream. The method of claim 2, The digital home, characterized in that the different applications 1a, 1b, 1c are informed about the determined priority 10 for each stream that is directly processed by themselves or delivered to the network 5. Capacity bottlenecks in networks. The method according to claim 3 or 4, One priority for each transmission band is assigned to scalable data streams. The method according to any one of claims 2 to 5, In the event of a resource bottleneck occurring, the data streams with the lowest priority are the first priority to be degraded or completely interrupted in quality. The method according to any one of claims 1 to 6, And said applications (1a, 1b, 1c) in said system transmit their quality requirements in combination with their connection requirements. The method according to any one of claims 1 to 7, Quality requirements of said applications (1a, 1b, 1c) are defined by said system from default entries. The method according to any one of claims 1 to 8, And the user can reject quality restrictions of the system. The method of claim 9, After denial of quality restriction, the system determines the next highest priority 10 and notifies the network 5 or the applications 1a, 1b, 1c therefor. Of capacity bottlenecks in the field. The method of claim 9 or 10, User denials are stored by the system and taken into account when determining a future priority (10) calculation.

Images