Method for predicting the QoE level of an application intended to run on a wireless user device

Abstract

The present disclosure is directed to an apparatus and method for evaluating a wireless link established between an access point and a user equipment. The apparatus and method include determining a quality of experience level of an application intended to run on the user equipment using a mapping between a parameter representing the QoE of the application under different wireless transmission conditions and a parameter set representing the different transmission conditions of the wireless link.

Description

[0001] The present invention is a divisional application of the following patent application: Application number: 201710214215.7, Application date: April 1, 2017, Invention name: Method for predicting the QoE level of an application intended to run on a wireless user device Technical Field

[0002] The present invention relates to the field of wireless nodes and corresponding devices communicating with each other via wireless communications. Background Art

[0003] Access gateways are widely used to connect devices in the home to the Internet or any other wide area network (WAN). In particular, access gateways used in digital subscriber line (DSL) technology enable high data rate transmission over copper or optical lines. Residential gateways and other devices (such as routers, switches, telephones and set-top boxes) are understood in this context as customer premises equipment (CPE) devices.

[0004] Access gateways, including wireless technologies, play a key role in today's home and professional environments. The mechanism for connecting wireless devices to a local area network (LAN) is called Wi-Fi, which is the brand name of the Wi-Fi Alliance for devices that use the IEEE802.11 family of standards for wireless data transmission. The IEEE 802.11 standard defines two types of wireless nodes, general wireless devices (denoted STA) that can connect to other devices called base stations, and a specific type of STA that controls the network, the access point (denoted AP). A Wi-Fi network, commonly called a WLAN (Wireless Local Area Network), consists of an AP with one or more STAs connected to the AP.

[0005] Due to its flexible and "invisible" nature, many LAN equipment is using Wi-Fi instead of traditional wired Ethernet. However, this widespread use of wireless LANs has exposed a serious disadvantage of using shared media technology: interference. Interference related to Wi-Fi and non-Wi-Fi has led to a degradation of the user experience attributed to the nature of IEEE 802.11. In its most common form, the IEEE 802.11 network applies a medium access method in which conflicts are avoided by sensing that the medium is used (for carrier sense multiple access conflict avoidance, expressed as CSMA-CA). It uses a technique called "idle channel assessment" (CCA). The idle channel assessment determines whether the wireless communication channel is "occupied", for example, "busy" with another wireless communication, and / or has an amount of interference that makes the wireless communication channel unsuitable for communication. In this way, it is determined whether the wireless communication channel can be used for communication or not, for example, occupied or not occupied. The medium access method is also often referred to as "listening before talking", describing the essence of the method. Therefore, interference from any nature may block the medium and force all nodes to remain silent for a certain amount of time.

[0006] Another effect of interference can be packet loss at the receiver side, resulting in a reduction in the physical data rate. In this case, the transmitter’s CCA is not detecting interference, but is reducing the SINR (signal to interference plus noise ratio) of the Wi-Fi packets as seen by the receiver.

[0007] Therefore, in some cases, a Wi-Fi connection may suffer from poor performance or even loss of connection. Some of these situations are obvious and easy to explain to the end user. For example, if the distance between the station and the access point is too large, the signal level is low and the performance will be degraded. Other situations are "invisible" and are not known to the end user, such as hidden nodes. Hidden nodes are invisible to some nodes of the network, resulting in actual failure of the CSMA-CA method, which may cause packet collisions / corruption in the air. In many cases, the end user is not able to diagnose the source of the problem and correct the problem.

[0008] With the recent development of tablets, laptops, and smartphones, the use of Wi-Fi has increased. As a result, home Wi-Fi network connections have become one of the main causes of Internet service provider support costs and help desk calls. In fact, since Wi-Fi connections are susceptible to performance issues attributed to the shared medium, an end user may observe a degradation in the quality of experience (QoE) of the applications he / she is currently running on one of his / her wireless devices, such as an increase in the loading time of a website. The end user may mistakenly believe that there is a problem with the service provided by the Internet Service Provider (ISP).

[0009] As a result, ISPs are looking for ways to better understand the end-user wireless environment, including link quality and performance and its impact on end-user QoE.

[0010] The present invention has been designed in view of the foregoing. Summary of the invention

[0011] According to a first aspect of the present invention, there is provided a computer-implemented method for predicting a Quality of Experience (QoE) level of an application intended to be run on a user equipment, the method comprising:

[0012] - calculating the QoE level of the application using a parameter set representing a transmission condition of a wireless link established between the user equipment and the access point, the parameter set representing the transmission condition of the wireless link being collected by the access point before data is transmitted between the user equipment and the access point.

[0013] This method enables accurate prediction of the expected quality of experience (QoE) of an application intended to run on a user equipment (such as a smartphone) connected to the Internet via a wireless link, such as a Wi-Fi link, established between the user equipment and an access point (such as a residential gateway). This method thus enables the determination of those situations in which the end user's expectations in terms of QoE are not met due to the conditions of the wireless link.

[0014] This is possible due to the knowledge of the parameters representing the transmission conditions of the radio link, ie the CCA statistic and the transmission scheme selected by the transmitter.

[0015] The method further includes, before calculating the QoE level:

[0016] - measuring parameters representative of the QoE of the application under different radio transmission conditions, the different radio transmission conditions being defined by different sets of parameters representative of the transmission conditions of the radio link, called learning sets,

[0017] - Compute a mapping between different learning sets and parameters representing the QoE level of the application.

[0018] Prior to predicting the expected QoE level of an application intended to be run on a user equipment, a learning phase is performed. During the learning phase, parameters representing the QoE of the application are measured a number of times under different wireless transmission conditions in order to establish a correlation between the parameters representing the transmission conditions of the wireless link defining different transmission conditions of the wireless link and the parameters representing the QoE of the application. This is done, for example, by introducing attenuation on the transmission path between the access point and the user equipment, or interference due to different types of wireless communications, e.g. Wi-Fi or non-Wi-Fi communications.

[0019] The method further includes:

[0020] - Determine the parameters of the learning set that have a significant impact on the QoE level of the application.

[0021] Using all parameters of the learning set will not improve the accuracy of the expected QoE level, so using all parameters of the learning set is useless.

[0022] According to another aspect of the method, the application intended to be executed on the user equipment is a web browsing application.

[0023] Web browsing is responsible for the majority of Internet traffic in a local area network, and is therefore an application that end users tend to demand in terms of QoE. Therefore, it is meaningful to predict the expected QoE level for web browsing.

[0024] According to another aspect of the method, the parameter representing the QoE of the application under different wireless transmission conditions is page loading time.

[0025] Page loading time is considered as one of the main indicators when it comes to web browsing experience.

[0026] The method further includes:

[0027] -Calculate the mapping between page load time and mean opinion score.

[0028] Map page load time and mean opinion score to avoid expensive user engagement.

[0029] According to another aspect of the method, the parameters representing the transmission conditions of the wireless link having a significant impact on the QoE level of the application are the average physical layer transmission rate, the frame transfer rate, and the parameters representing the traffic of the transmission medium of the wireless link due to interference from wireless equipment located near the user equipment.

[0030] During the learning phase, it is determined that these four parameters are more relevant for accurate prediction of the QoE level among all the parameters in the learning set.Interference from other wireless equipment is due to different types of wireless communications, such as Wi-Fi or non-Wi-Fi communications.

[0031] According to another aspect of the method, a parameter representing a transmission condition of the wireless link is obtained by monitoring the data transmission condition over the wireless link during normal use of the WLAN.

[0032] According to another aspect of the method, parameters representing transmission conditions of the wireless link are collected by the access point.

[0033] Since the access point is the master node of every wireless LAN, it is the best location to monitor and collect data related to wireless transmissions.

[0034] According to another aspect of the method, calculating the QoE level of the application is based on a mapping of a parameter representing a transmission condition of a wireless link having a significant impact on the QoE level and a mean opinion score.

[0035] The method relies on using the correlation established during a learning phase between parameters representing the transmission conditions of the radio link, which have a significant impact on the QoE level, and the mean opinion score.

[0036] According to another aspect of the method, calculating the QoE level of the application is based on a mapping of a parameter representing a transmission condition of a wireless link and a mean opinion score.

[0037] Another aspect of the present invention is a circuit comprising a processor, a memory and a wireless node, wherein the memory comprises instructions which, when executed by the processor, perform a method according to an embodiment of the present invention.

[0038] Another aspect of the present invention is a gateway, which includes a circuit, which includes a processor, a memory and a wireless node, and the memory includes instructions, which execute the method according to an embodiment of the present invention when executed by the processor. Some of the processing implemented by the elements of the present invention can be computer-implemented. Therefore, such elements can be implemented in the form of complete hardware, complete software (including firmware, resident software, microcode, etc.), or the implementation of the combination of software and hardware aspects, which can be generally referred to as "circuit", "module" or "system" here. In addition, such elements can be in the form of a computer program product embodied in any tangible expression medium with a computer-usable program code, and the program code is embodied in the medium.

[0039] Since the elements of the present invention can be implemented in software, the present invention can be implemented as computer readable code for providing to a programmable device on any suitable carrier medium. Tangible carrier media may include storage media such as floppy disks, CD-ROMs, hard disk drives, magnetic tape devices, or solid-state storage devices. Transient carrier media may include signals such as electrical signals, electronic signals, optical signals, acoustic signals, magnetic signals, or electromagnetic signals, for example microwave signals or RF signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Embodiments of the present invention will now be described by way of example only and with reference to the following drawings, in which:

[0041] Figure 1 represents an access point with which the station communicates via wireless communication,

[0042] Figure 2 Indicates based on Figure 1 The data rate of wireless communication,

[0043] Figure 3 A flow chart illustrating the steps of a method for predicting a QoE level of an application intended to run on a station is shown in accordance with one or more embodiments of the present invention. DETAILED DESCRIPTION

[0044] It should be understood that Figure 1 The elements shown can be implemented with various forms of hardware, software or a combination thereof. Preferably, these elements are implemented with a combination of hardware and software on one or more appropriately programmed general-purpose devices, which may include a processor, a memory and an input / output interface. Here, the phrase "coupling" is defined as meaning to be directly connected or indirectly connected through one or more intermediate components. This intermediate component may include both hardware-based components and software-based components.

[0045] This description illustrates the principles of the present disclosure. It will therefore be appreciated that those skilled in the art will be able to conceive of various arrangements that, although not explicitly described or shown herein, embody the principles of the present disclosure and are included within the spirit and scope of the present disclosure.

[0046] All examples and conditional language described herein are intended for teaching purposes to help readers understand the principles of the invention and the concepts contributed by the inventors to advance the art, and should be construed as not being limited to such specifically described examples and conditions.

[0047] In addition, all statements describing the principles, aspects, and embodiments of the present disclosure, and specific examples thereof, are intended to include both structural and functional equivalents thereof. In addition, such equivalents are intended to include both currently known equivalents and equivalents developed in the future, i.e., any element developed that performs the same function, regardless of structure.

[0048] Thus, for example, it will be appreciated by those skilled in the art that the block diagrams presented herein represent conceptual diagrams of illustrative circuitry embodying the principles of the present disclosure. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo-codes, etc. represent various processes that may substantially be represented in a computer-readable medium and thus executed by a computer or processor, whether or not such a computer or processor is explicitly shown.

[0049] The functions of the various elements shown in the figures may be provided using dedicated hardware as well as hardware capable of executing software in conjunction with appropriate software. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of separate processors (some of which may be shared). In addition, the explicit use of the term "processor" or "controller" should not be interpreted as referring exclusively to hardware capable of executing software, but may implicitly include, but is not limited to, digital signal processor ("DSP") hardware, read-only memory ("ROM") for storing software, random access memory ("RAM"), and non-volatile memory.

[0050] Other hardware, conventional and / or custom, may also be included. Similarly, any transformations shown in the figure are merely conceptual. Their functions may be performed through the operation of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or even manually, with the specific technique being selectable by the implementer as more specifically understood from the context.

[0051] In the claims herein, any element represented as a means for performing a particular function is intended to encompass any manner of performing that function, including, for example, a) a combination of circuit elements that perform that function, or b) any form of software, thus including firmware, microcode, etc., combined with appropriate circuitry for executing that software to perform the function. The disclosure as defined by such claims resides in the fact that the functions provided by the various described components are combined and brought together in the manner required by the claims. Therefore, any components that can provide those functions are considered equivalent to those shown herein.

[0052] In the following description, an example method for predicting the QoE level of an application intended to run on a station or user equipment, and an apparatus for performing the method are described. For the purpose of explanation, various specific details are set forth in order to provide a thorough understanding of the preferred embodiments. However, it is apparent to those skilled in the art that the present disclosure may be practiced without these specific details.

[0053] The CPE device includes, but is not limited to, for example, a controller (e.g., a microprocessor), a memory in which an operating system for operating the CPE device is stored, a wireless node for wireless communication, and a circuit for a broadband connection (e.g., an xDSL connection). The wireless node includes, but is not limited to, a software driver, a physical layer with a data buffer, and an antenna. Such a CPE device is, for example, an access gateway.

[0054] The wireless nodes are controlled by software drivers that perform multiple background tasks during the operation of the wireless nodes, such as dynamic rate adaptation, packet aggregation, channel quality monitoring, etc. On top of the signal manipulation, the wireless driver also embeds the IEEE802.11 protocol stack with associated IEEE-defined management and control messaging. The wireless driver will therefore inject multiple management and control packets in the data stream, making it difficult to analyze the link by only transparently viewing the data frame exchanges.

[0055] Figure 1 Schematically depicted is an arrangement illustrating wireless communication: an access point 1 communicates with a station 2 or user equipment 2 via a wireless link 3. The station 2 may be, for example, a smartphone, a tablet, a laptop, etc. The access point 1 comprises a circuit comprising a microprocessor 10, a memory 11, a wireless node 12 for a wireless link, and a monitoring application 13. The station 2 comprises a second circuit comprising a microprocessor 20, a memory 21, and a wireless node 22 for a wireless link. The wireless node 12 comprises a physical layer 14 and a link layer 15, and the wireless node 22 comprises a physical layer 24 and a link layer 25. The access point 1 is in particular a CPE device, such as a residential gateway that establishes a home network of an end user with the station 2. A monitoring application 13 is included for analyzing and evaluating the wireless link 3, and in particular for retrieving parameters representing the transmission conditions of the wireless link 3.

[0056] The monitoring application 13 comprises instructions for the microprocessor 10 and the monitoring application 23 comprises instructions for the microprocessor 20, which are included to diagnose the wireless link 3 and collect a set of information about the wireless link 3. The set of information includes, among other things, the actual data rate, the physical layer data rate, the number of spatial streams, the channel bandwidth, the medium availability and the received signal strength indicator (RSSI). The monitoring data is collected in a passive mode, where data transmission is monitored between the access point 1 and the station 2 and vice versa.

[0057] Figure 2The possibilities that must be considered when diagnosing the Wi-Fi performance between an access point 1 and a station 2 are illustrated. A unidirectional link 3' from the access point 1 to the station 2 is examined. The theoretical maximum data rate 30 for this link, referred to here as MaxNegotiatedPhyRate or MaxPhyRate, is given by the capabilities of the access point 1 and the station 2, which is for example 130 MB / s in the case where the IEEE 802.11n standard with a 20 MHz channel bandwidth and two spatial streams is chosen for the transmission between the access point 1 and the station 2. This is therefore the maximum achievable link speed, 100%, which is only a theoretical value, since in most cases physical limitations come into play: for example the received signal strength RSSI on the station side is reduced due to the distance between the access point 1 and the station 2 and due to the path loss caused by any walls or other obstacles and reflections. The number of spatial streams must also be determined. Therefore, the actually achievable data rate 31 (referred to here as PhysLimitsPhyRate) is less than the data rate 30.

[0058] Additional performance may be lost due to interference close to station 2, which is not visible by access point 1, here called far-end interference FEIF: it could be any microwave source, like an RF baby phone, a microwave oven or a hidden Wi-Fi node, and results in an additional reduced data rate, here called TrainedPhyRate 32. Similar interference may occur at access point 1, here called near-end interference NEIF: this will reduce the available data rate 32 to data rate 33, MediumBusyOtherWiFi. Additional performance may be lost by sharing the medium with other Wi-Fi traffic, which may be caused by WLAN traffic in the home network, but also by Wi-Fi traffic from neighboring networks.

[0059] In order to monitor the data traffic at the physical layer, layer 1 of the OSI (Open Systems Interconnection) model, the transmission conditions of the traffic transmitted and received by the Wi-Fi nodes of the residential gateway, the residential gateway includes a monitoring application that receives all received and transmitted packets. The monitoring application can access the following blocks:

[0060] Transmit (TX) packet queue, TX packet

[0061] Receive (RX) packet queue, RX packet

[0062] Transmit / Receive Signal Indicator (RSSI)

[0063] While there are solutions describing the link between network transmission conditions, such as packet loss, latency and bandwidth, and QoE, none of them teach the link between parameters representing the transmission conditions of a wireless link, such as Wi-Fi Link, Phy Rate; Wi-Fi intermediate services; etc., and QoE.

[0064] Figure 3 A flow chart illustrating the steps of a method for predicting the QoE level of an application intended to be run on a station according to one or more embodiments of the present invention is shown. The application is for example web browsing, video streaming, Voice over IP (VoIP), etc. The application is run on a station 2 or a user equipment 2 such as a smartphone or a computer connected to the Internet via a Wi-Fi link. In the following example, the application considered is web browsing.

[0065] In order to determine the impact of Wi-Fi transmission conditions on the QoE of a given application, multiple transmission conditions of the Wi-Fi link are defined, and for each of these transmission conditions, parameters representing the transmission conditions are passively collected by the access point during a learning phase. These collected parameters are then processed by the monitoring application 13, 23. At the same time, i.e. during the learning phase, the station 2 performs a series of web browsing tasks under different transmission conditions of the Wi-Fi link and measures parameters representing the QoE each time.

[0066] Then, a mapping is performed between the parameters representing the transmission conditions of the Wi-Fi link processed by the monitoring application 13, 23 and the measurement parameters representing the QoE.

[0067] The learning phase comprises steps E1 to E6 of the method according to an embodiment of the invention.

[0068] This mapping is then used to predict the QoE level of applications intended to run on the station. To this end, parameters representing the transmission conditions of the Wi-Fi link are passively collected by the access point. Then, using the mapping obtained during the learning phase, the expected QoE level of the application is obtained.

[0069] Therefore, during step E1 , different transmission conditions of the Wi-Fi link are determined. To this end, the wireless transmission conditions are modified, for example along the axes: PhyRate and medium availability.

[0070] The change of PhyRate is achieved by introducing attenuation on the transmission path established between the access point and the station. For example, attenuation of 6, 12, 15, 18, 19 and 20 dB is introduced, thereby reducing the PhyRate.

[0071] The medium may be unusable due to interference from both Wi-Fi and non-Wi-Fi communications. To simulate non-Wi-Fi communications, a narrowband sine wave signal is generated to block the access point CCA. To simulate Wi-Fi communications, a second pair of access points / stations generates competing Wi-Fi traffic that blocks the medium.

[0072] Different transmission conditions of the Wi-Fi link correspond to different combinations of PhyRate and medium availability.

[0073] During a step E2 , for each transmission condition defined by the monitoring application 13 , 23 during step E1 , the access point passively collects a set of parameters representative of the transmission conditions of the Wi-Fi.

[0074] Meanwhile, during step E3, station 2 performs a series of web browsing tasks under different transmission conditions of the Wi-Fi link determined during step E1 and measures each time a parameter representative of QoE. In the case of web browsing, such a parameter representative of QoE is, for example, page loading time.

[0075] Thus, during step E3 , station 2 accesses 10 times a web page, such as google.com, twitter.com, amazon.fr, etc., and measures each time the page loading time.

[0076] During step E4, the page loading time of the web page is mapped with the mean opinion score (MOS) defined in ITU-T Recommendation G.1030. MOS is an integer between 1 and 5, 1 corresponding to the lowest QoE score and 5 corresponding to the highest QoE score. Therefore, the longer the page loading time, the lower the MOS. For example, a MOS of 1 corresponds to a page loading time of 10 seconds, because users usually lose focus after 10 seconds of waiting; and a MOS of 5 corresponds to an average page loading time per web page.

[0077] During a step E5 , a mapping is calculated between a set of parameters representing the transmission conditions of the Wi-Fi link and parameters representing the QoE level of the application measured for each web page access.

[0078] For example, a vector representing both transmission conditions and MOS is generated for each web access. RRSI and average Tx / RxPHY rates are strongly correlated with changes in the SNR of the Wi-Fi link because they indirectly measure link quality. Information representing medium unavailability due to Wi-Fi and no-Wi-Fi traffic is also reported. These vectors then feed a predictor such as a support vector regressor (SVR).

[0079] The mapping obtained during step E5 enables the expected QoE level to be estimated taking into account only the Wi-Fi effects.

[0080] In step E6 , a selection is performed to determine, from the set of parameters collected by the access points, which parameter representative of the transmission conditions of the Wi-Fi is most relevant for predicting the QoE level of a given application.

[0081] These selected parameters, called parameters representing the transmission conditions of the radio link having a significant impact on the QoE level of a given application, are obtained by feeding the predictor with different combinations of parameters representing the transmission conditions of the radio link and checking the accuracy of the predicted QoE with a MOS corresponding to the set of parameters representing the transmission conditions of the radio link from which the combination of parameters representing the transmission conditions of the radio link was extracted.

[0082] For example, for web browsing, parameters representing the transmission conditions of the wireless link that have a significant impact on the QoE level are Average Tx PHY Rate, Frame Transfer Rate, MediumBusy, and MediumBusyOtherWi-Fi.

[0083] In step E7 , parameters representing the transmission conditions of the Wi-Fi link are periodically collected passively by the access point.

[0084] In step E8, the QoE level is calculated for a given application (e.g. web browsing) by using the mapping obtained in step 5. The parameters used to calculate the QoE level are a set of parameters representing the transmission conditions of the Wi-Fi link collected by the access point, or parameters representing the transmission conditions of the wireless link that have a significant impact on the QoE level, as selected during step E6, depending on the application intended to run on the station.

[0085] Although the present invention has been described above with reference to specific embodiments, the present invention is not limited to the specific embodiments, and modifications will be apparent to those skilled in the art, which are within the scope of the present invention.

[0086] With reference to the foregoing illustrative embodiments, many additional modifications and variations will suggest themselves to those skilled in the art, which are given by way of example only and are not intended to limit the scope of the invention, which is determined solely by the appended claims. In particular, different features from different embodiments may be interchanged where appropriate.

Claims

1. A computer-implemented method for predicting a quality of experience (QoE) level of an application intended to run on a user equipment, the method comprising: - passively measuring, by an access point, a set of parameters representing transmission conditions of a wireless link established between the user equipment and the access point, the set of parameters including Phy rate and medium availability associated with Wi-Fi communications and non-Wi-Fi communications; - calculating the QoE level of the application using the parameter set representing the transmission conditions of the wireless link established between the user equipment and the access point, the parameter set representing the transmission conditions of the wireless link being measured by the access point before data is transmitted between the user equipment and the access point.

2. The method according to claim 1, comprising: Before calculating the QoE level: - measuring a parameter representing the QoE of an application under different combinations of parameters of the transmission conditions in the parameter set representing the transmission conditions of the radio link, - calculating a mapping between different combinations of parameters of the transmission conditions in the parameter set representing the transmission conditions of the radio link and parameters representing the QoE level of the application.

3. The method according to claim 2, further comprising: - determining one or more parameters of said parameter set representing transmission conditions of the radio link having a significant impact on the QoE level of said application.

4. The method according to any one of claims 1-3, wherein the application intended to be run on the user equipment is a web browsing application.

5. The method according to claim 4, wherein the parameter representing the QoE of the application under different wireless transmission conditions is page loading time.

6. The method according to claim 5, further comprising: -Calculate the mapping between page load time and mean opinion score.

7. The method of claim 4, wherein the parameters representing the transmission conditions of the wireless link having a significant impact on the QoE level of the application are an average physical layer transmission rate, a frame transfer rate, and a parameter representing the traffic of the transmission medium of the wireless link due to interference from wireless equipment located near the user equipment.

8. The method according to any of the preceding claims 1-3, wherein the parameter representing the transmission condition of the wireless link is obtained by monitoring data transmission over the wireless link.

9. The method according to any one of claims 1 to 3, wherein: Calculating the QoE level of the application is based on a mapping of a parameter representing a transmission condition of a wireless link having a significant impact on the QoE level and a mean opinion score.

10. The method according to any one of claims 1 to 2, wherein: Calculating the QoE level of the application is based on a mapping of a parameter representing a transmission condition of a wireless link and a mean opinion score.

11. A circuit comprising a processor, a memory and a wireless node, the memory comprising instructions which, when executed by the processor, perform the method according to any one of the preceding claims 1-10.

12. A gateway comprising circuitry comprising a processor, a memory and a wireless node, the memory comprising instructions which, when executed by the processor, perform the method according to any of the preceding claims 1-10.

13. A processor-readable medium having instructions stored therein for causing a processor to perform a computer-implemented method for predicting a Quality of Experience (QoE) level of an application intended to run on a user equipment according to any one of claims 1 to 10.

Images