DE102018204730B3 - Method for transmitting data - Google Patents

Abstract

The invention relates to a method for transmitting data, wherein a transmission channel to a receiver of several users is shared, wherein users send their data in encoded form asynchronously via the common transmission channel, each user one or more replicas of a data packet to be transmitted to the Transmitting a receiver, wherein the data is decoded on the receiver side using a sliding decoder window, the decoding using a Successive Interference Cancellation (SIC) method in which a certain maximum number of iterations is attempted to decode data packets, the decoding window after the predetermined number of iterations by an offset Δ is moved to a new position, characterized in that the length of this offset Δ and / or the length of the decoding window W depending on the number of transmitted on the transmission channel replicas adjusted asst becomes.

Description

The invention relates to a method for transmitting data, wherein a transmission channel to a receiver is shared by a plurality of users.

If the users access the common channel without any coordination (random access channel), collisions can occur between the individual users on the channel. Due to these collisions, data packets can be lost. It is known to apply so-called Successive Interference Cancellation (SIC) methods to recover damaged packets by eliminating the interference caused by individual packets. For this purpose, several replicas of a data packet are transmitted within a frame. These replicas contain a pointer to all other replicas of the same data package. If one of the replicas is received successfully, the interference caused by you can be removed from the respective total signal for all copies. As a result, if necessary, other packets that were previously subject to interference, can be decoded. The decoding thus takes place by iterative removal of interference.

One method that works as outlined above is the Contention Resolution Diversity Slotted ALOHA Method (CRDSA). Here, a frame is divided into a plurality of time or frequency slots. However, there are also methods in which a frame is not divided into time or frequency slots. Unlike Slotted ALOHA processes, these processes are referred to as Pure ALOHA processes. An example of such a method is Contention Resolution ALOHA (CRA). In such a method, there may be only partial interference in a packet that is still large enough that decoding this interference-prone packet is not possible. In such a situation, a continuation of the decoding process is no longer possible.

In IoT applications or in machine-to-machine data transmission, terminals usually transmit sporadically smaller data packets, which may have, for example, at most a few hundred bits of information. In such applications, random access methods have proven to be particularly efficient.

• [1] E. Sandgren A. Graell i Amat, F. Braennstroem, „On Frame Asynchronous Coded Slotted ALOHA: Asymptotic, Finite Length, and Delay Analysis“, IEEE Transactions on Communications, Vol. 65, No. 2, pp.691-704, Feb. 2017 .
• [2] C. Kissling, „Performance Enhancements for Asynchronous Random Access Protocols over Satellite“, in: Proceedings of the ICC 2011. International Conference on Communication, ICC, 5.-9. Juni 2011 , Kyoto.
• [3] Clazzer, F. and Kissling, C.; „Enhanced Contention Resolution Aloha - ECRA“, 9th International ITG Conference on Systems, Communications and Coding (SCC), Munich, Jan. 2013 .
• [4] DE102012219468 B3 Clazzer, F. and Kissling, C.; „Verfahren zur Übertragung von Daten“, 2012.

Information on the technical background can be found, inter alia, in the following publications:

• [1] E. Sandgren A. Graell i Amat, F. Braennstroem, "On-Frame Asynchronous Coded Slotted ALOHA: Asymptotic, Finite Length, and Delay Analysis", IEEE Transactions on Communications, Vol. 2, pp. 691-704, Feb. 2017 ,
• [2] C. Kissling, "Performance Enhancements for Asynchronous Random Access Protocols over Satellite," in: Proceedings of the ICC 2011. International Conference on Communication, ICC, 5-9. June 2011 , Kyoto.
• [3] Clazzer, F. and Kissling, C .; "Enhanced Contention Resolution Aloha - ECRA", 9th International ITG Conference on Systems, Communications and Coding (SCC), Munich, Jan. 2013 ,
• [4] DE102012219468 B3 Clazzer, F. and Kissling, C .; "Procedure for transferring data", 2012.

Frame-synchronous [1] or slot and frame-asynchronous [2] random access techniques reduce timing synchronization requirements while providing high throughput and low packet error rate. The receiver performs the decoding within a relocatable decoding window. The received signal is sampled and a series of samples are stored in memory. The number of stored samples depends on the control parameters. After the processing of these samples is finished, the decoding window is advanced by a certain number of samples. Now, the receiver-side decoder will process a number of samples already present in the previous decode window and additionally process new samples that were not present in the previous decode window.

Another method known from the prior art is Enhanced Contention Resolution ALOHA (ECRA) described in Publication [3].

In 1 By way of example, a data transmission is shown in which a plurality of users transmit their data packets to a common transmission channel in the same frequency band (see the upper part of FIG 1 ). In this respect, collisions arise between the individual replicas, which in 1 hatched are shown.

On the receiver side, the replicas located in the size W decoding window are stored on a sample level. The decoder now begins to search for identical replicas. One way to identify such is to use a common preamble on all identical replicas known to the receiver.

Thus, identical replicas can be detected. Subsequently, the data packets are supplied to the channel decoder, which will output a code word. This is then checked by a Cyclic Redundecy Check (CRC). If this test passes, this packet is declared decoded correctly. Subsequently, the Successive Interference Cancellation process can be carried out iteratively in a known manner.

In the in 1 shown decoding all users except User 2, second replica and User 4, second replica can be decoded using the Successive Interference Concellation method. The mentioned 2 replicas of the users 2 and 4 are outside the decoding window, so that the receiver will store the decoded packets for later decoding windows. Since no further replicas are completely within the Dekodierfensters, the decode is shifted by a fixed displacement Δ _W placed. This is in 2 shown. The receiver-side decoder will now do the decoding within the shifted decoding window.

US 2016/0 149 627 A1 describes a method for transmitting data formed according to the preamble of claim 1 of the present application.

The object of the invention is to provide a method in which a plurality of users send data over a common transmission channel to a receiver, which offers improved performance, in particular improved data throughput and / or a reduced packet error rate.

The object is achieved according to the invention by the features of claim 1.

In the method according to the invention, several users share a transmission channel with a receiver. This may, for example, be a satellite connection to a satellite to which several satellite terminals want to transmit data. Here, users send their data in coded form asynchronously via the common transmission channel. Each user submits one or more replicas of a data packet to be transmitted to the recipient. Here, all users can use a common frequency band. The data is decoded on the receiver side using a relocatable decoding window, as already known in the art.
Decoding uses an SIC technique that attempts to decode the packets in the decode window for a given number of iterations. This SIC method is also known from the prior art and is therefore not described in detail.

The decoding window is shifted by a maximum number i of iterations by an offset Δ _W to a new position.

According to the invention, the length of the set Δ _W and / or the length of the decoding window _{W is} adjusted as a function of the number of replicas transmitted on the transmission channel.

The performance of a transmission method depends largely on the length of the decoding window used, the length of the offset and, if appropriate, the number i of the maximum admissible iterations within a decoding window. For example, extending the decode window results in more replicas being stored in the receiver-side memory and processed by the SIC method. Thus, usually more SIC iterations should be performed. On the other hand, if the decoding window is shortened, a lower number of SIC iterations may suffice. The optimum ratio between the length of the decoding window, the length of the offset of the decoding window and the maximum permissible number of SIC iterations is determined and determined in the prior art on the basis of complex experiments.

However, the parameters mentioned also depend on the channel load, that is to say on the number of data packets sent on the transmission channel. According to the invention, it is thus possible to take into account this latter parameter, so that the length of the decoding window and / or the length of the offset can be adjusted during operation. Thus, costly investigations have to previous determination of said parameters are not performed. The method according to the invention thus requires less development work than the methods known from the prior art.

The method according to the invention offers the following advantages: For a low channel load, the receiver-side decoder can increase its operating speed. This can be done by increasing the window length W and / or the offset Δ _W or by reducing the number i of the SIC iterations, as this reduces the number of necessary operations per received sample. For a high channel load, on the other hand, the receiver can increase its performance at the expense of a lower processing speed. Nevertheless, a lower limit for the speed of the decoder can still be set by setting a minimum allowable length of the decoding window Wm, a minimum allowable length of the offset ΔW _m, and a maximum allowable number of SIC iterations. Thus, the receiver is able to handle different channel conditions without the need for elaborate preparations or examinations beforehand.

It is preferable that, as the number of replicas transmitted on the transmission channel is reduced, the length of the set Δ _W and / or the length of the decoding window W is increased, and as the number of replicas transmitted on the transmission channel increases, the length of the offset Δ _W and / or the length of the decoding window W is reduced.

Furthermore, it is preferred that, depending on the number of replicas transmitted on the transmission channel, the maximum number i of iterations is adapted for which the SIC method is performed before the decoding window is shifted.

In this regard, it is preferred that as the number of replicas transmitted on the transmission channel is reduced, the maximum number i of SIC iterations is reduced, and as the number of replicas transmitted on the transmission channel increases, the maximum number i of SIC iterations increases becomes.

It is further preferred that the length of the decoding window W and the length of the offset Δ _{W be} set to the minimum allowable length of the decoding window W _m and the minimum allowable offset ΔW _m . By contrast, the maximum number of SIC iterations i is set to the maximum permissible number i _M. Furthermore, it is preferred that after completion of the first decoding operation in a decoding window (that is, when the maximum number i of iterations has been reached) and before the decoding window is shifted for the first time, the position and number of still completely located in this decoding window are not decoded replicas is determined. This can be done for example by physical layer methods using soft correlation in the known preamble of the data packets. If at least one undecoded replica is completely in the decode window, the length of the offset Δ _W and / or the length of the decoder window W and / or the maximum number i of SIC iterations remain unchanged. Actually, it would be useful to reduce the length of the offset Δ _W and / or the length of the decoding window W at this point and / or to increase the maximum number i of SIC iterations. However, this is not possible since the length of the decoding window W and the length of the offset Δ _{W were} previously set to the minimum permissible length W _m and the minimum permissible offset ΔW _m . The maximum number i of SIC iterations has also been set to the maximum allowable value i _M.

A reduction in the length of the decoding window and the offset or an increase in the maximum number i of SIC iterations is thus not yet possible at the present time, so that the values mentioned remain unchanged in this stage of the method.

On the other hand, if the determination of the number and / or position of undecoded replicas still present in this decoding window shows that there is no further undecoded replica completely in the decoding window, the length of the set Δ _W and / or the length of the decoding window W become one fraction α increases their respective length, and / or the maximum number of SIC iterations is reduced by 1.

$$\Delta W_1=\Delta W_0+\alpha \Delta W_0=\left(1+\alpha \right)\Delta W_0$$

$${\text{i}}_1=i_0-1$$

Preferably, the length of the offset Δ _W and / or the decoding window W and / or the maximum number i of SIC iterations can be adapted according to: $${\mathrm{<figure-callout\; id="1"\; label="W"\; filenames="DE102018204730B3\_0028.png,DE102018204730B3\_0030.png"\; state="\{\{state\}\}">W</figure-callout>}}_1=W_0+\alpha W_0=\left(1+\alpha \right)W_0$$

$$\mathrm{<figure-callout\; id="1"\; label="\Delta \; W"\; filenames="DE102018204730B3\_0028.png,DE102018204730B3\_0030.png"\; state="\{\{state\}\}">\Delta \; </figure-callout>}{W}_{}{}_1=\Delta W_0+\alpha \Delta W_0=\left(1+\alpha \right)\Delta W_0$$

$${\text{i}}_1=i_0-1$$

It is further preferred that after completion of the decoding process, the position and number of undecoded replicas still present in this decoding window are determined in a decoding window. If there is no further undecoded replica in the decode window, the length of the offset Δ _W and / or the length of the decoder window W is increased by a portion a of their respective length and / or the maximum number i of SIC iterations is decreased by one as long as the length of the offset Δ _W and / or the decoding window W have not reached their maximum permissible length ΔW _M or W _M and the maximum number i of SIC iterations has not reached its minimum permissible number im.

Furthermore, it is preferred that, if at least one undecoded replica is located completely in the decoding window, the length of the offset Δ _W and / or the length of the decoding window W are reduced by a portion γ of their respective length and / or the maximum number i SIC iterations is increased as long as the length of the offset Δ _W and / or the length of the decoding window W have not reached their minimum allowable length ΔW _m or W _m and the number i of SIC iterations does not reach its maximum permissible number i _M Has. At this point it is assumed that it is possible to reduce the length of the decoding window or the offset and to increase the maximum number of SIC iterations, since these values do not change as before their first shifting of the decoding window. Maximum values are located.

$$\Delta W_{j+1}=\text{min}\left(\left(1+\alpha \right)\Delta W_j,\Delta W_M\right)$$

$${\text{i}}_{\text{j}+1}=\text{max}\left(i_j-\mathrm{1,}{i}_m\right)$$

It is particularly preferred that if there is no further undecoded replica in the decoding window, the length of the offset Δ _W and / or the length of the decoding window W and the maximum number i of SIC iterations are adjusted as follows: $$W_{j+1}=\text{min}\left(\left(1+\alpha \right)W_j.W_M\right)$$

$$\Delta W_{j+1}=\text{min}\left(\left(1+\alpha \right)\Delta W_j.\Delta W_M\right)$$

$${\text{i}}_{\text{j}+1}=\text{Max}\left(i_j-\mathrm{1,}{i}_m\right)$$

$$\Delta W_{j+1}=\text{max}\left(\gamma \Delta W_j,\Delta W_m\right)$$

$${\text{i}}_{\text{j}+1}=\text{min}\left(i_j+\mathrm{1,}{i}_M\right),$$

wobei γ<=1.When at least one undecoded replica is completely in the decode window, it is preferred that the length of the offset Δ _W and / or the length of the decoder window W and the maximum number i of SIC iterations be adjusted as follows: $$W_{j+1}=\text{Max}\left(\gamma W_j.W_m\right)$$

$$\Delta W_{j+1}=\text{Max}\left(\gamma \Delta W_j.\Delta W_m\right)$$

$${\text{i}}_{\text{j}+1}=\text{min}\left(i_j+\mathrm{1,}{i}_M\right).$$

where γ <= 1.

In the following, preferred embodiments of the invention will be explained.

• 1 und 2 die Durchführung eines SIC-Verfahrens mit verschiebbarem Dekodierfenster, wie es aus dem Stand der Technik bekannt ist,
• 3 ein Flow Chart einer Ausführungsform des erfindungsgemäßen Verfahrens.

Show it:

• 1 and 2 the implementation of an SIC method with a displaceable decoding window, as known from the prior art,
• 3 a flow chart of an embodiment of the method according to the invention.

1 and 2 have already been explained in connection with the prior art.

The following is a description of the flow chart of the 3 , In step 1, the length of the decoding window W ₀ at time 0, that is, before the first shifting of the decoding window to W _{m is} set. The offset of the decoding window at the same time ΔW ₀ is set to ΔW _m . Both values are thus set to their minimum. The maximum number i _{0 of} the admissible iterations at time 0 is set to i _M , that is to say to its maximum.

Subsequently, the SIC procedure is performed as known in the art. This is repeated until the maximum number of iterations (in this case i _M ). Upon completion of these iterations, the number and position of the undecoded replicas still in this decode window is determined.

If at least one non-decoded replica is located completely in the decoding window, step 3b) takes place, that is to say the length of the decoding window W and the offset Δ _W and the number of maximum permissible iterations i remain unchanged. If, on the other hand, no further undecoded replicas are completely in the decoding window, step 3a) takes place, ie the length of the decoding window is calculated according to: $${\mathrm{<figure-callout\; id="1"\; label="W"\; filenames="DE102018204730B3\_0028.png,DE102018204730B3\_0030.png"\; state="\{\{state\}\}">W</figure-callout>}}_1=\left(1+\alpha \right)W_0.$$

The length of the offset is calculated according to: $$\mathrm{<figure-callout\; id="1"\; label="\Delta \; W"\; filenames="DE102018204730B3\_0028.png,DE102018204730B3\_0030.png"\; state="\{\{state\}\}">\Delta \; </figure-callout>}{W}_{}{}_1=\left(1+\alpha \right)\Delta W_0.$$

By contrast, the maximum number of permitted iterations is reduced by 1 according to i ₁ = i ₀ - 1

After the decoding window has been shifted by the determined offset ΔW ₁ , the SIC procedure is performed again. Upon completion of the maximum allowable iterations, it is rechecked if any undecoded replicas are in the decode window.

If this is not the case, the length of the next decoding window is determined as follows: $$W_{j+1}=\text{min}\left(\left(1+\alpha \right)W_j.W_M\right).$$

The length of the offset of the decoding window is determined as follows: $$W_{j+1}=\text{min}\left(\left(1+\alpha \right)\Delta W_j.\Delta W_M\right).$$

The number of maximum allowed iterations is reduced by 1 and calculated according to: $$i_{j+1}=\text{Max}\left(i_j-\mathrm{1,}{i}_m\right)$$

If, on the other hand, there is at least one undecoded replica in the decoding window, the length of the next decoding window is determined according to: $$W_{j+1}=\text{Max}\left(\gamma W_j.W_m\right).$$

The length of the next offset of the decoding window is determined according to: $$W_{j+1}=\text{Max}\left(\gamma \Delta W_j.\Delta W_m\right).$$

By contrast, the maximum permissible number of SIC iterations is increased by 1 until the maximum i _{M is} reached. This is done according to the equation: $$i_{j+1}=\text{min}\left(i_j+\mathrm{1,}{i}_M\right)$$

Subsequently, the decoding window is shifted by the determined offset and the SIC method is performed again. This is followed by a renewed query for remaining undecoded replicas in the decoding window and a repetition of the method steps 4a) or 4b). In the following, the performance of the method according to the invention compared to the prior art will be described. In the prior art comparison method, fixed values were used for the length of the decoding window, the length of the offset and the maximum number of permissible SIC iterations. A contention resolution ALOHA (CRA) transmission was assumed, and the results presented are also transferable to other asynchronous or frame-asynchronous random access methods. The method according to the invention can be used for all asynchronous or frame-asynchronous transmission methods. While the method may be used for synchronous methods such as CRSDA, it is only possible to adjust the number of SIC iterations, whereas the length of the decoder window and the length of the offset can not be adjusted.

The simulations carried out used two replicas per transmission and a packet length of 1000 bits of information encoded at a rate (in bits / symbol) of 1. The virtual frame for each user has been set to 100 packet lengths. The received SNR value was set to 6 dB and the Shannon limit was used as the decoding limit. The parameters mentioned were used both in the method according to the invention and in the method according to the prior art.

The following values for the respective parameters were also used for the method according to the invention: parameter value W _m 3 * VF W _M 20 * VF ΔW _m 0.01 * VF ΔW _M VF i _m 2 i _m 20 α 0.2 γ 0.5

For the method according to the prior art, however, were determined: $$W=3*VF.\Delta W=\frac{1}{3}*VF\text{and I}=10$$

The values were not changed during the entire simulation.

The following is a comparison of the Packet Loss Rate (PLR), which is the probability that the packet of any user was not received correctly. This was considered for different channel loads: Existing channel load [pk / pk duration] PLR according to the method of the invention PLR according to the prior art 0.2 0.00008 0.00020 1.0 0.00056 0.00087 1.5 0.01126 0.03174

It can be seen that PLR enhancement occurs at all channel loads. At a higher channel load of 1.5, there is an improvement by a factor of 3. In addition, there is the advantage that the method according to the invention automatically adapts to the conditions of the channel and a prior adaptation of the parameters with respect to the window length of the offset and the maximum permissible number SIC iterations is no longer necessary.

The method according to the invention can be used in all random access transmission methods, in particular in frame-asynchronous or frame and slot asynchronous random access methods. Typical applications could be for example a machine-to-machine data transmission in I-oT applications, but also data transmissions from or to satellites or via terrestrial connections.

Claims (9)

A method of transmitting data, wherein a transmission channel to a receiver is shared by a plurality of users, wherein users send their data in encoded form asynchronously over the common transmission channel, each user transmitting one or more replicas of a data packet to be transmitted to the receiver the data is decoded on the receiver side using a relocatable decoding window, wherein the decoding uses a Successive Interference Cancellation (SIC) method in which for a given maximum number of iterations it is attempted to decode data packets, the decoding window being after the determined number Iterations is offset by an offset Δ _W to a new position, characterized in that the length of this offset Δ _W and / or the length of the decoding window W is adjusted depending on the number of replicas transmitted on the transmission channel. Method according to Claim 1 characterized in that, as the number of replicas transmitted on the transmission channel increases, the length of the offset Δ _W and / or the length of the decoding window W is reduced, and as the number of replicas transmitted on the transmission channel decreases, the length of the offset Δ _W and / or the length of the decoding window W is increased. Method according to Claim 1 or 2 characterized in that, depending on the number of replicas transmitted on the transmission channel, the maximum number i of iterations is adapted for which the Successive Interference Cancellation method is performed before the decoding window is shifted. Method according to Claim 3 characterized in that, as the number of replicas detected on the transmission channel is reduced, the maximum number i of SIC iterations is reduced, and as the number of replicas transmitted on the transmission channel increases, the maximum number i of SIC iterations is increased. Method according to Claim 1 to 4 characterized in that after completion of the first decoding operation, that is prior to the first shifting of the decoding window, the position and number of undecoded replicas still remaining in that decoding window is determined, wherein if at least one undecoded replica is completely in the decoding window, the length of the set Δ _W and / or the length of the decoding window W and / or the maximum number i of SIC iterations remain unchanged, if no further undecoded replicas are completely in the decoding window, the length of the set Δ _W and / or Length of the decoding window W are increased by a portion α of their respective length and / or the maximum number i is reduced by 1 at SIC iterations. Method according to Claim 5 , characterized in that the length of the offset Δ _W and / or the decoding window W and / or the maximum number i is adapted to SIC iterations, according to $$W_1=W_0+\alpha W_0=\left(1+\alpha \right)W_0$$

$$\Delta W_1=\Delta W_0+\alpha \Delta W_0=\left(1+\alpha \right)\Delta W_0$$

$${\text{i}}_1=i_0-1$$

Method according to Claim 1 to 4 , characterized in that after completion of the decoding process in a decoding window, the position and number of undecoded replicas still in this decoding window is determined, wherein, if there are no further undecoded replicas completely in the decoding window, the length of the set Δ _W and / or the length of the decoding window W is increased by a portion α of its respective length and / or the maximum number i of SIC iterations is reduced by 1, as long as the length of the offset Δ _W and the decoding window W their maximum allowable length ΔW _m or W _M and the maximum number i of SIC iterations have not reached their minimum allowed number. Method according to Claim 7 , characterized in that when at least one undecoded replica is completely in the decoding window, the length of the set Δ _W and / or the length of the decoding window _W are reduced by a portion γ of their respective length and / or the maximum number i of SIC It increases iterations by 1 as long as the length of the offset Δ _W and / or the length of the decoding window W have not reached their minimum allowable length ΔW _m or W _m and the maximum number i of SIC iterations does not exceed their maximum allowed number i _{M has} reached. Method according to Claim 7 or 8th , characterized in that, if there is no further undecoded replica completely in the decoding window, the length of the offset Δ _W and / or the length of the decoding window W and the maximum number i of SIC iterations are adjusted as follows: $$W_{j+1}=\text{min}\left(\left(1+\alpha \right)W_j.W_M\right)$$

$$\Delta W_{j+1}=\text{min}\left(\left(1+\alpha \right)\Delta W_j.\Delta W_M\right)$$

$${\text{i}}_{\text{j + 1}}=\text{Max}\left(i_j-\mathrm{1,}{i}_m\right)$$

if at least one undecoded replica is located completely in the decoding window, the length of the offset Δ _W and / or the length of the decoding window W and the maximum number i of SIC iterations are adjusted as follows: $$W_{j+1}=\text{Max}\left(\gamma W_j.W_m\right)$$

$$\Delta W_{j+1}=\text{Max}\left(\gamma \Delta W_j.\Delta W_m\right)$$

$${\text{i}}_{\text{j + 1}}=\text{min}\left(i_j+\mathrm{1,}{i}_M\right)$$

where γ <= 1.

Images