WO2005115033A1

WO2005115033A1 - Method and radio station for transmitting data via packet-switched transmission links

Info

Publication number: WO2005115033A1
Application number: PCT/EP2005/051993
Authority: WO
Inventors: Markus Kaindl; Günther LIEBL; Florian Reif; Wen Xu
Original assignee: Siemens Aktiengesellschaft
Priority date: 2004-05-18
Filing date: 2005-05-02
Publication date: 2005-12-01
Also published as: DE102004024651A1

Abstract

The invention relates to a method for transmitting data via packet-switched transmission links, according to which the data is coded into data packets by means of a multi-rate encoder. The output processing speed of said multi-rate encoder can be varied. The data of a data packet is divided into data segments and the output processing speed of the multi-rate encoder is dynamically adjusted (21) in such a way that the number of the resultant data segments is minimised (23) and/or that the part of a data segment that is not filled with the data of a data packet is minimised (22).

Description

description

Method and radio station for the transmission of data over packet-switched transmission links

The invention relates to a method for transmitting data via packet-switched transmission links and a corresponding radio station, in particular a mobile station or a base station of a mobile radio network.

An essential goal of future data transmission methods is the provision of flexible packet-switched transmission modes, in particular via the air interface of a mobile radio system.

Flexible multi-rate encoders are known for encoding data, which are variable with regard to their output data rate. It is also known to encode data into data packets using such multi-rate encoders and to divide the data packaged in this way into data segments.

In particular, if the size of the data packets and / or the data segments has rasterized predefined data sizes, the transmission capacity of packet-switched transmission links is not used equally efficiently with every combination of output data rate, data size of the data packets used and data size of the data segments used.

The invention is therefore based on the problem of specifying a technical teaching which enables efficient transmission of data over packet-switched transmission links. The invention is solved by the features of the independent claims. Advantageous and expedient developments of the invention are determined by the features of the dependent claims. Further developments of the device claim, which correspond to the dependent claims of the method claim, are also within the scope of the invention.

The invention is therefore based on the idea of dynamically setting the output data rate of a multirate encoder in such a way that the number of resulting data segments required for the transmission of the data packets obtained is minimized and / or that the part of one which is not occupied by data of a data packet Data segment is minimized.

The knowledge on which the invention is based is that the capacity of packet-switched transmission links can be used particularly efficiently if the data packets are converted into as few data segments as possible and / or if the part of the data segments which is converted after the data packet has been converted into Data segments remain free, is as small as possible.

Complex simulations on which the invention is based showed that these goals can be achieved by dynamically setting the output data rate of the multirate encoder.

The invention can be used particularly advantageously if the data packets and / or the data segments each have a data size which is taken from a large number of predefined data sizes. The data packets and / or data segments can therefore only assume certain predefined rastered data sizes, the data sizes for the Data packets and the data segments in particular are different.

Of course, it is also within the scope of the invention to add a header field to the data packets before it is divided into data segments, which header field can in particular also be subjected to compression.

A preferred implementation for achieving the above-mentioned optimizations, in particular minimizations, is that

Set the output data rate of the multirate encoder in dependence on an error discard rate in such a way that the error discard rate is minimized; The error scrambling rate describes in particular the frequency or frequency with which data packets which are faulty on the input side or are detected as faulty are rejected.

In addition or as an alternative to this, the above-mentioned optimizations, in particular minimizations, can be achieved by setting the output data rate as a function of a delay rejection rate in such a way that the delay rejection rate is minimized. The delay discard rate again describes the frequency and / or frequency of the data packets which are discarded on the receiving end due to an arrival which is delayed beyond a delay threshold value. The delay time can be based, for example, on a time period that lies between the arrival of two successive data packets.

The invention finds a particularly preferred use in a

AMR (Adaptive Multrate) voice transmission in real time via packet-switched transmission channels, in particular a mobile radio network. In this case, the AMR is selected Mode (codec mode), that is to say the setting of the output data rate of the multi-rate encoder, preferably in such a way that the data size of the data segments selected by the mobile radio network for the current transmission channel is optimally used. The aim here is also in particular to minimize the number of data segments required and the segmentation overhead of the last data segment for an AMR data packet, that is to say to minimize the part of a data segment which is not occupied by data from an AR data packet.

However, the invention is in no way limited to AMR voice transmission, but can also be used in general for packet transmissions of other multi-rates or variable-rate signals (voice, image, multimedia signals).

Exemplary embodiments of the invention are explained in more detail below, for the illustration of which the figures given below are used:

FIG. 1 flow diagram of an exemplary embodiment of the invention;

FIG. 2 shows a schematic diagram of the output rate optimization;

Figure 3 Block diagram of a radio station.

The invention is explained below by way of example on the basis of the implementation of low-rate multimedia applications, such as voice or IP, via GERAN (GSM / EDGE Radio Access Network), via packet-switched transmission links, which are based in particular on the GPRS (General Packe Radio Service) protocol. Speech signals are first subjected to flexible multi-rate coding and converted into data packets. An adaptive multirate encoder (AMR) is used as the multirate encoder, as described in "ETSI, GSM 06.90, version 7.2.1, Release 1998". The selected AMR codec mode determines the output data rate and the data size of the data packets output by the multi-rate encoder, the data size corresponding to one of several predefined discrete data sizes.

These data packets generated in the network layer are converted into data segments in another protocol layer, in particular the RLC (Radio Link Control) / MAC (Medium Access Control) layer, which are finally transmitted via a radio channel after any further processing steps , The data size of the data segments also corresponds to one of several predefined discrete data sizes and is set by a radio network controller depending on the channel conditions.

The invention optimizes the transmission quality in that the AMR codec mode is selected in such a way that the number of data segments required for the transmission of the resulting data packets is minimized and / or that the part of a data segment not occupied by data of a data packet is minimized.

For this purpose, in a first step 1, as shown in FIG. 1, the delay period between the received data packets is determined at the receiving end. If the time interval between two data packets is above a predetermined delay threshold value, the codec mode is reduced by one step in step 2. So the multi-rate Encoder data packets with a smaller data size output. If there are repeated delays that are above the delay threshold, the codec mode is reduced step by step by one level until codec mode 0 or AMR475 is reached, at which the multi-rate encoder data at a rate of 4.75 kilobytes per second outputs.

If, on the other hand, the determined delay time does not exceed the predefined delay threshold value, the error discard rate is used to set the codec mode in step 3, which describes the frequency with which data packets received incorrectly at the receiving end are discarded. If the packet loss rate is currently increasing, the codec mode is reduced in step 4. If, on the other hand, the packet loss rate does not increase, it is checked in step 5 whether the packet loss rate is decreasing. If the packet loss rate decreases, the codec mode is increased in step 6. If the packet loss rate does not decrease, a segment optimization described below is carried out in step 7.

If both the delay discard rate and the error discard rate are negligible, an attempt is made to make optimum use of the segmentation used by the data link layer by changing the current codec mode. For this purpose, it is first checked whether a data segment can be dispensed with by reducing the codec mode. If it is not possible to do without a data segment even when using the lowest-rate codec mode, the data segment is optimally filled. In this context, optimal means that the codec mode is increased as long as no additional data segment ment must be used. In the best case scenario, the data segment is completely filled.

Figure 2 shows an example in the upper part of three data segments ds, on which the data of a data packet are divided. The first two data segments ds (left and middle data segment) are completely filled with data of the data packet. The third data segment ds (right data segment), on the other hand, is only partially filled with data of the data packet. The data segments ds are obviously not used optimally, since the last data segment is not completely filled. In this case, it is checked in step 21 whether it is possible to dispense with a data segment ds. If it is not possible to dispense with a data segment, the number of data segments is retained and all data segments are filled in as best as possible in step 22. If, on the other hand, it is possible to dispense with a data segment, a data segment is dispensed with in step 23 and the remaining data segments are optimally filled.

In order to enable the data segments to be filled optimally, the following difference matrix is introduced.

ϋttfft rtna

This has the size 8 x 8 and includes the change in the size of the user data when changing from a codec mode to the their codec mode. For example, line 5 of the matrix contains all differences in the size of the user data when changing to mode 0 to 7.

The use of this difference matrix is briefly explained below using an example:

With each data packet that the multirate decoder receives and decodes on the receiving side, information about the currently used codec mode and the currently used coding scheme is also received. However, the number of the current coding scheme is not explicitly transmitted, only the size of the data segments currently used and the degree of their occupancy. These two values are used to determine how many bytes are unused within a data segment and are therefore still available. An attempt is made to find a codec mode which either allows this data segment which is not completely filled to be dispensed with or, if this is not possible, to find a different codec mode with which all data segments are optimally filled. The difference matrix described above is used. If the data packet just received has been encoded with codec mode 5, for example, and the data segments are not optimally filled, line 5 of the difference matrix searches for a better codec mode, i.e. searched for a codec mode with which the data segments can be filled optimally. This codec mode is then used on the transmission side for the next data packet. For this purpose, appropriate signaling can take place from the receiving side to the transmitting side.

A jump from codec mode 5 to codec mode 1 leads, for example, to a reduction in the size of the useful data of the data packets by 6 bytes. A jump from codec mode 5 to codec mode 8, on the other hand, leads to an increase in the size of the useful data by 12 bytes. This information can easily be found in the above-mentioned difference matrix: the line number corresponds to the number of the current codec mode (from 1 to 8), the column number is the number of the codec mode that would have to be requested next from the remote station. The respective entries indicate how many bytes are used more or less when changing the codec mode.

FIG. 3 shows a radio station FS, in particular a mobile station or a base station, which is set up to transmit data over packet-switched transmission links.

The radio station has a processor device PE, which is set up, for example in terms of programming, in such a way that the data are encoded into data packets by means of a multirate encoder, the multirate encoder being variable with regard to its output data rate. The data of a data packet are divided into data segments, and the output data rate of the multirate encoder is set dynamically in such a way that the number of resulting data segments is minimized and / or that the part of a data segment which is not occupied by data of a data packet is minimized.

Claims

claims

1. Method for the transmission of data via packet-switched transmission links, - in which the data are encoded into data packets by means of a multi-rate encoder, the output data rate of the multi-rate encoder being variable, - in which the data of a data packet are divided into resulting data segments, - in which the Output data rate of the multirate encoder is set dynamically such that the number of resulting data segments is minimized and / or that the part of a resulting data segment not occupied by data of a data packet is minimized.

2. The method according to claim 1, wherein the data packets have a data size which is taken from a plurality of predefined data sizes.

3. The method according to any one of the preceding claims, wherein the data segments have a data size that is taken from a plurality of predefined data sizes.

4. The method according to any one of the preceding claims, in which a header field is added to the data packets.

5. The method according to any one of the preceding claims, in which faulty data packets are discarded at the receiving end, and in which the output data rate is set as a function of the error rejection rate in such a way that the error rejection rate is minimized.

6. The method as claimed in one of the preceding claims, in which incoming data packets arriving after a delay beyond a delay threshold value are discarded, and in which the output data rate is set as a function of the delay discard rate in such a way that the delay delay discard rate is minimized.

7. radio station (FS) for the transmission of data via packet-switched transmission links, with a processor device (PE) which is set up in this way,

that the data are encoded into data packets by means of a multirate encoder, the multirate encoder being variable with respect to its output data rate,

- in which the data of a data packet are divided into resulting data segments,

- in which the output data rate of the multirate encoder is set dynamically in such a way that the number of resulting data segments is minimized and / or that the part of a resulting data segment not occupied by data of a data packet is minimized.