Techniques for restoring a telephone connection with satisfactory voice quality for users

Description

[0001] The present invention relates to the technical field of Voice-Over-IP (VoIP) voice calls and in particular to techniques for ensuring or improving their voice quality.

[0002] Even today, voice calls are still sometimes disrupted. This includes traditional telephony, VoIP telephony, and calls made via apps. These disruptions often result in the callers being unable to understand each other, or only very poorly. This usually leads to the call being difficult or even impossible to complete and having to be interrupted.

[0003] Such a call termination is usually inconvenient and frustrating. It can even happen that only one of the participants is experiencing audio problems and therefore "doesn't understand" why the call ended. In such cases of poor audio quality, a back-and-forth exchange typically ensues between the participants, with each repeatedly asking the other, "Hello, can you hear me?"

[0004] Therefore, manually re-establishing the voice call is often the only solution, which can also prove problematic if the other participant is still on the line and hasn't even noticed that their conversation partner has hung up. This leads to unnecessary delays when the voice call needs to be restarted. One problem here is that voice calls made via UDP currently lack any error correction mechanisms.

[0005] Manually restarting a voice call can even increase the risk of an accident if, for example, the sender and / or receiver are driving a car.

[0006] Patent publications US2017 / 339535 and EP3182647 are examples of the prior art.

[0007] The present invention therefore aims to provide techniques that efficiently avoid voice calls with poor voice quality and thus at least partially eliminate the aforementioned disadvantages of the prior art. Another alternative objective of the invention is to initiate an automated re-establishment of a voice call when the current voice call is of poor quality.

[0008] The present invention solves this problem through the features of the independent claims.

[0009] The features of the various aspects of the invention or the various embodiments described below can be combined with each other, unless this is explicitly excluded or is technically impossible.

[0010] The invention provides a method for establishing, restoring, and / or re-establishing a voice connection, in particular a telephone connection. The (new) voice connection preferably has satisfactory voice quality. The method comprises the following steps, the sequence of which may be, in particular, as listed below: Measurement of technical parameters of a VoIP voice call between a sender and a receiver conducted using a communication network, where the technical parameters include call jitter and / or data packet loss; ∘ sender or receiver here refers to the user device with which the respective parties conduct the voice call. This can be a smartphone, tablet, telephone, computer, smartwatch, etc.The communication network can be assigned to a telephony provider and / or an internet provider; the communication network can be a 3G, 4G, 5G, or 6G network – it can be, in particular, a fixed-line or mobile network; the corresponding measurement units, which are set up to measure jitter and / or data packet loss, are optionally assigned to and / or provided for at the receiver, the sender, and / or a server of the communication network; the voice call is carried out, in particular, via UDP and Transport Protocol Layer 4; the technical parameters are passed to a voice quality algorithm, and the voice quality algorithm is set up to determine the voice quality based on these technical parameters.In principle, the voice quality algorithm can be implemented as functional code or as a suitably trained artificial intelligence; calculation of at least one telephony quality score (TQS) that characterizes the voice quality of the voice call using the voice quality algorithm, whereby the calculation is based on the measured values ​​of jitter and / or data packet loss; the TQS can represent a measure of how likely it is that the voice call will potentially drop. The score can be chosen so that a lower value corresponds to poorer voice quality.In principle, however, it is clear to the expert that this is merely a matter of convention and the score could also be chosen so that a higher value corresponds to lower speech quality – accordingly, this finding would be applied to the threshold described below; ∘ both a functional speech quality algorithm and a speech quality algorithm in the form of trained artificial intelligence can, in principle, calculate three different TQS scores. The first – without the order being significant – can be calculated based on jitter "J": TQS(J), the second based on packet loss "PL": TQS(PL), and the third based on a combination of jitter and packet loss: TQS(J, PL). This allows for flexibility in calculating the TQS values ​​and also makes it possible to better adapt these values ​​to possible use cases.In particular, a separate threshold – described below – can be defined for each of these three TQS values. The voice quality algorithm, acting as an artificial intelligence, can be trained as follows: Various datasets of jitter and / or packet loss measurements can be provided, each labeled with corresponding TQS scores. These are introduced to the artificial intelligence during the training phase, enabling it to assign corresponding TQS scores to "unknown" measurements after the training phase. The algorithm then interrupts the voice call between sender and receiver if at least one score falls below a predefined threshold and automatically re-establishes the voice call between sender and receiver.• The interruption affects both the sender and the receiver, so neither blocks the line for a new call; • The interruption of the voice call and the establishment of the new voice call are automated and require no user interaction; • If, as described above, the scores TQS(J, PL), TQS(J) and / or TQS(PL) are calculated, the interruption and automated re-establishment can be triggered if only one, or any combination of, the score exceeds its respective assigned predefined threshold; .

[0011] The method according to the invention thus makes it possible, for the first time, to automatically and efficiently re-establish a voice connection when the voice quality is too poor or when there is a risk that the voice call might soon be terminated. This enables users to communicate with each other as efficiently as possible, which reduces waiting times and – as explained above – can minimize traffic risks.

[0012] In a preferred embodiment of the invention, the method is implemented on a server of a telephony provider or communications network provider, on which the sender and / or the receiver is implemented.

[0013] The transmitter and / or receiver communicate with each other in a way that allows them to determine the relevant measurements. The transmitter and / or receiver also have access to other measurement data, such as the strength of the radio signal, which is not readily available to the communication network provider. The transmitter and / or receiver can be configured, or set to be configured, to determine the relevant measurements. For example, if the transmitter and / or receiver detects that a threshold has been breached, it can send a signal to the other device to interrupt the voice call, preventing the other device from blocking the line and allowing it to independently re-establish the call.

[0014] On the other hand, the server-side solution can be implemented independently of the sender and / or receiver, which also saves resources on the part of the sender and / or receiver, as they do not have to constantly perform the measurements that are already available to the server. Furthermore, in the event of a connection failure, the sender or receiver can no longer communicate that the voice call should be re-established – the server, however, can still communicate with at least one of the devices.

[0015] To measure packet loss in VoIP telephony, specialized network monitoring devices or software designed to monitor VoIP call quality can be used. A commonly used device is a VoIP quality monitor or network analyzer. These devices can monitor network traffic and provide statistical information about packet loss, latency, jitter, and other important metrics.

[0016] There are also specialized VoIP test devices that can be used to test the performance of VoIP connections by measuring packet loss and other metrics. Such devices are often used by network engineers and IT professionals to monitor the quality of VoIP networks and diagnose problems.

[0017] Furthermore, software solutions such as Wireshark can be used to analyze network traffic and identify packet loss in VoIP telephony. Wireshark is an open-source network protocol analyzer that enables detailed analysis of network traffic and is available for various operating systems.

[0018] In particular, jitter measurements can be obtained as follows: continuous measurement of jitter during a voice call conducted over UDP and the Layer 4 transport protocol. If the jitter exceeds certain thresholds in either direction (upstream or downstream), this affects the TQS(J, PL) or the TQS(J). A calculation of the jitter can be performed based on the definition in IETF RFC 1889, finalized on April 30, 2024, for the RTP protocol. For this purpose, the document https: / / www.ietf.org / rfc / rfc1889.txt is incorporated in its entirety into the disclosure of the invention.

[0019] In a preferred embodiment, an audio message is played by the sender and / or the receiver indicating the termination of the current voice call and / or the re-establishment of the voice call. The audio message can be transmitted, in particular, using the "Media Resource Function" service.

[0020] This has the advantage that users of the devices are not "surprised" when the voice call suddenly ends and they know what will happen next, namely that a reconnection will take place.

[0021] Preferably, the audio message is stored in the sender and / or the receiver, or sent from a server to the sender and / or the receiver.

[0022] If the audio message is stored in the sender and / or receiver, it can still be played even if the communication connection fails. Storing it on a server has the advantage that the devices do not need to be specially configured, and the method can also be applied to conventional devices already on the market.

[0023] Preferably, confirmation is requested from the sender and / or the receiver to end and / or re-establish the voice call. This confirmation can be given by the user on their device or specified in the settings.

[0024] This offers the technical advantage that a restart of the conversation does not necessarily occur if, for example, it is not desired at all or a conversation has already been virtually ended, thus conserving communication resources.

[0025] In a preferred embodiment, the automated rebuild is delayed until the transmitter and / or receiver have a sufficiently strong radio signal. A sufficiently strong radio signal might be present at two out of five bars, where five bars represent the best possible signal. A sufficiently strong radio signal might also be present at approximately 20% of the maximum signal strength.

[0026] If the server handles the re-establishment and termination of the voice call, the sender and / or receiver can communicate their radio signal strength to the server. The server only triggers the re-establishment of the voice call if the signal strength is sufficiently good. Since the process is initiated by the sender and / or receiver, the respective end devices are already aware of the signal strength and can use this information to determine whether a voice call can be re-established. This has the technical advantage of increasing the likelihood that the re-establishment will result in a sufficiently high-quality voice call.

[0027] In one embodiment, the voice quality algorithm is based on a trained artificial intelligence. Such an AI, particularly one trained for pattern recognition, is especially well-suited to predicting whether a voice call might be likely to fail in the near future due to poor quality. The voice quality algorithm can compare TQS values ​​at different times and, in particular, consider any downward trend. For example, a sharp drop in TQS value can also be a criterion for interrupting and restarting the voice call, even if the current TQS value is still above the threshold.

[0028] The measurement of technical parameters can be performed at predefined time intervals. These intervals are chosen to ensure that poor call quality is detected in a timely manner while simultaneously conserving resources, as excessively frequent measurements increase computing power and energy consumption. Shorter predefined intervals are possible in server-side implementations compared to those implemented on the sender or receiver, as devices like smartphones need to manage their energy efficiently. Suitable predefined intervals are 10 seconds, preferably 5 seconds.

[0029] The time intervals are dynamically shortened if the voice quality of the voice call deteriorates and / or the time intervals are dynamically lengthened if the voice quality of the voice call improves.

[0030] This has the advantage that the time intervals adapt to the specific "situation" of the voice call, thus ensuring efficient resource utilization. For example, if the voice call quality deteriorates, this could be because the user is moving out of range. In this case, it is relatively likely that their signal will weaken further soon, making more frequent measurements advisable. Conversely, if the voice call is consistently of good quality—for example, because the user is connected at home via their Wi-Fi network—the quality is highly likely to remain good, allowing the time interval to be extended. It is also possible that the sender, receiver, and / or server use different time intervals.

[0031] In one embodiment, parameters of the voice call are reselected, thus improving the quality of the voice call.

[0032] The poor quality of the voice call may be due to technical parameters of the voice call that can be changed. For example, to re-establish the connection, the sender and / or receiver may be configured to dial into a new base station, be assigned a different network slice, have app settings changed, increase the radio power of the base station, sender and / or receiver, and / or route communication within the complication network differently.

[0033] According to a second aspect of the invention, a server of a telephony provider is specified, wherein the telephony provider provides at least partially a communication network for a voice call between the sender and the receiver, wherein the server is configured to perform the steps of the method described above.

[0034] The advantages are essentially analogous to those described in connection with the procedure.

[0035] According to a third aspect of the invention, a transmitter is specified, wherein the transmitter is configured to carry out the steps of the method described above.

[0036] The advantages are essentially analogous to those described in connection with the procedure.

[0037] According to a third aspect of the invention, a receiver is specified, wherein the receiver is configured to carry out the steps of the above-described method.

[0038] The advantages are essentially analogous to those described in connection with the procedure.

[0039] According to a fourth aspect of the invention, a computer program product is specified, wherein the computer program product is implementable on a computing unit, the computing unit performing the steps of the method described above by means of the computer program product.

[0040] Further advantageous design features of the present invention are defined in the patent claims.

[0041] Preferred embodiments of the present invention are explained below with reference to the accompanying figures: Fig. 1: shows a communication system for providing a voice call with satisfactory voice quality;

[0042] Numerous features of the present invention are explained in detail below with reference to preferred embodiments. The present disclosure is not limited to the specific combinations of features mentioned. Rather, the features mentioned here can be combined arbitrarily to form embodiments according to the invention, unless expressly excluded below.

[0043] Fig. 1 shows a communication system 100 for providing a voice call with satisfactory voice quality.

[0044] The communication system 100 comprises a transmitter 110, a communication network 120, and a receiver 130. The transmitter 110 and the receiver 130 are, for example, smartphones 110 and 130, respectively, which are operated by their respective users to make a voice call using VoIP. The communication network 120 includes the typical physical resources, such as base stations, etc.

[0045] Sufficient or satisfactory voice quality of the voice call can be achieved by the inventive method 140 described below, wherein the communication system 120 is configured to execute the inventive method 140.

[0046] Method 140 includes at least the following steps: Step 141: Measurement of technical parameters of a VolP voice call between a sender and a receiver made via a communications network, wherein the technical parameters include call jitter and / or packet loss;

[0047] Step 142: Passing the technical parameters to a speech quality algorithm;

[0048] Step 143: Calculation of at least one telephony quality score (TQS) that characterizes the voice quality of the voice call using the voice quality algorithm based on the measured values ​​of jitter and / or data packet loss;

[0049] Step 144: Interruption of the voice call between the sender and the receiver if at least one score exceeds a predefined threshold and automated re-establishment of a voice call between the sender and the receiver;

[0050] Procedure 140 thus enables the parties to be provided with a voice call of sufficient quality, or it allows the parties to avoid having to manually initiate a new call. While the procedure does not necessarily guarantee that the new voice call will be of better quality than the previous one in every case, experience shows that poor voice call quality is improved in many cases (approximately 9 out of 10 cases) by re-establishing the call.

[0051] The procedure 140 can be implemented on the transmitter 110, on a server 150 of the communication network 120, and / or on the receiver 130. If the procedure is implemented on the transmitter 110 and / or the receiver 130, a data connection 160 can be routed directly between the transmitter 110 and the receiver 130. If the procedure is implemented on the server 150, a data connection 161 passes through this server, enabling it to perform the relevant measurements and also to interrupt and re-establish the voice connection.

[0052] Here's another possible implementation: When implementing the procedure on the sender's end, a second call (conference) can be established to the same recipient number (instead of to a different participant as usual) to use this new line instead of the old one. This is essentially a "re-dial" while the active but "bad" call is still in progress. The recipient then hears the call waiting notification.

Claims

1. Method for providing a voice-over-IP VoIP voice connection having satisfactory voice quality, comprising the following steps: • measuring technical parameters of the VoIP voice connection, which is made using a communication network, between a transmitter and a receiver, the technical parameters including a jitter of the call and a data packet loss, and the measurement being carried out in predefined time intervals; • transferring the technical parameters to a voice quality algorithm, • calculating at least one telephony quality score TQS, which characterizes the voice quality of the voice connection, by means of the voice quality algorithm on the basis of the measured values of the jitter and / or the data packet loss; • interrupting the voice connection between the transmitter and the receiver if the at least one score falls below at least one predefined threshold value, the interruption affecting both the transmitter and the receiver, and reestablishing the voice connection between the transmitter and the receiver by automated means, • the time intervals for the measurement being dynamically shortened if the voice quality of the voice connection deteriorates and / or the time intervals being dynamically extended if the voice quality of the voice connection improves.

2. Method according to one of the preceding claims, wherein an audio message that indicates the termination of the current voice connection and / or the reestablishment of the voice connection is played by the transmitter and / or the receiver.

3. Method according to Claim 2, wherein the audio message is stored in the transmitter and / or in the receiver or is transmitted to the transmitter and / or the receiver by a server.

4. Method according to one of the preceding claims, wherein an acknowledgment is requested from the transmitter and / or from the receiver to terminate and / or reestablish the voice connection.

5. Method according to one of the preceding claims, wherein automated reestablishment is delayed until the transmitter and / or the receiver have a sufficiently good radio signal available.

6. Method according to one of the preceding claims, wherein the voice quality algorithm is based on trained artificial intelligence.

7. Method according to one of the preceding claims, wherein parameters of the voice connection are reselected so that the quality of the voice connection is improved.

8. Server of a telephony provider, at least portions of which provide a communication network for a voice connection between a transmitter and a receiver, the server being configured to perform the steps of the method according to one of Claims 1 - 7.

9. Transmitter for making a voice call, the transmitter being configured to perform the steps of the method according to one of Claims 1 - 7.

10. Receiver for making a voice call, the receiver being configured to perform the steps of the method according to one of Claims 1 - 7.

11. Computer program product implementable on a computer unit, the computer unit using the computer program product to perform the steps of the method according to one of Claims 1 - 7.