QoS Management of Multi-User EDCA Transmission Mode in 802.11ax Network

Abstract

To overcome the loss of QoS handling resulting from introduction of FDMA transmissions.SOLUTION: A base station supports the construction of a wireless network conforming to IEEE802.11 series of standards, and includes: transmission means for transmitting a trigger frame (TF) conforming to IEEE802.11 series of standards including allocation information of resource units (RU) for OFDMA communication; and first receiving means for receiving data from a terminal station in the RU. In the case where RU allocation information includes information on allocation to the first terminal station, after the first terminal station having received the TF successfully transmits data of a predetermined access category (AC) in RU, MU EDCA mode is enabled for a predetermined duration in transmitting the data of the predetermined AC addressed to the base station, and legacy EDCA mode is enabled in transmitting the data of the predetermined AC addressed to a second terminal station which is different from the base station.SELECTED DRAWING: Figure 10

Description

[Technical field]

[0001] The present invention relates generally to communication networks, and more particularly to a wireless communication method and corresponding devices in a wireless network with an access point (AP) and multiple non-AP stations. The present invention is directed to a communication network that provides channel access to non-AP stations through contention, such as EDCA, to transmit data, and provides secondary access to sub-channels (or resource units) that divide the transmission opportunity TXOP granted to the access point.

[0002] The present invention applies to wireless communication networks, and in particular to 802.11ax networks providing stations with access to 802.11ax composite channels and / or OFDMA resource units enabling them to form 802.11ax composite channels granted, for example, to an access point and to carry out uplink communications. [Background technology]

[0003] The IEEE 802.11 MAC family of standards (a / b / g / n / ac etc.) specifies how Wireless Local Area Networks (WLANs) must operate at the physical and Medium Access Control (MAC) levels. Typically, the 802.11 MAC (Medium Access Control) operating mode implements the well-known Distributed Coordination Function (DCF) which relies on a contention-based mechanism based on the so-called "Carrier Sense Multiple Access with Collision Avoidance" (CSMA / CA) technique.

[0004] The original access DCF method is improved upon the well-known Enhanced Distributed Channel Access (EDCA) method to take into account prioritized data traffic when accessing communication channels of the network.

[0005] EDCA defines traffic categories and four corresponding access categories that allow for different handling of high priority traffic compared to low priority traffic.

[0006] The implementation of EDCA in a station may be done with multiple traffic queues (known as "access categories") serving data traffic at different priorities. Each traffic queue is associated with a respective queue backoff counter. The queue backoff counter is initialized with a backoff value randomly drawn from the respective queue contention parameters, e.g., EDCA parameters, and is used to contend for access to the communication channel to transmit the data stored in the traffic queue.

[0007] The legacy EDCA parameters are the CW for each traffic queue. min , C.W. max and AIFSN. CW min and C.W. max are the lower and upper bounds of the selection range within which the EDCA contention window CW is selected for a given traffic queue. AIFSN stands for Arbitration Inter-Frame Space Number and defines the number of time slots (typically 9 μs) that are added to the DIFS interval (a total number defining the AIFS period) that a station must sense the medium as idle before decrementing the queue backoff value associated with the traffic queue under consideration. This means that, according to EDCA, a station will decrement the queue backoff counter over time as long as the communication channel is continuously sensed as idle for a period longer than the respective arbitration inter-frame space duration.

[0008] Legacy EDCA parameters can be specified in beacon frames transmitted by APs in a network that broadcast network information.

[0009] The contention window CW and the queue backoff value are EDCA variables.

[0010] The conventional EDCA backoff procedure consists of a station randomly selecting a backoff value for its traffic queue backoff counter from its respective contention window CW, and then decrementing the backoff value when it senses the medium as idle after an AIFS period. When the backoff value reaches 0, the station is allowed to access the medium.

[0011] The EDCA queue backoff counter thus serves two purposes for a station: First, it puts the station in a state where it can efficiently access the medium by reducing the risk of collisions. Second, it mirrors the aging of the data contained in the traffic queue (the older the data, the smaller the backoff value) and therefore provides quality of service QoS management by providing different priorities to the traffic queues through different values ​​of EDCA parameters (in particular the AIFSN parameter which delays the start of the decrement of the EDCA queue backoff counter).

[0012] Using the EDCA backoff procedure, stations gain access to the communication network through backoff-based contention.

[0013] Recently, the Institute of Electrical and Electronics Engineers (IEEE) officially approved the 802.11ax task group as the successor to 802.11ac. The primary goal of the 802.11ax task group is to explore ways to improve data rates for wireless communication devices used in high-density deployment scenarios.

[0014] In particular, recent developments in the 802.11ax standard have attempted to optimize the use of communication channels by multiple stations in wireless networks with access points (APs). In fact, typical content includes large amounts of data, e.g., related to high-definition audiovisual real-time / interactive content, which should be transmitted with a required quality of service (QoS).

[0015] Furthermore, it is well known that the performance of the CSMA / CA protocol used in the IEEE 802.11 standard deteriorates rapidly as the number of stations and the amount of traffic increases, i.e. in high density WLAN scenarios.

[0016] In this context, multi-user (MU) transmission has been considered to enable multiple simultaneous transmissions during an AP-granted transmission opportunity in both the downlink (DL) direction from the AP to different users, and the uplink (UL) direction from different users to the AP. In the uplink, multi-user transmission can be used to mitigate collision probability by allowing multiple non-AP stations to transmit simultaneously.

[0017] To make such multi-user transmissions practical, it has been proposed to divide the granted communication channel into sub-channels, also called resource units (RUs), shared by multiple users (non-AP stations) in the frequency domain, for example based on orthogonal frequency division multiple access (OFDMA) techniques. Each RU can be defined by multiple tones, i.e., an 80 MHz channel containing up to 996 usable tones.

[0018] OFDMA is a multi-user variation of OFDM that has emerged as a new key technology to improve efficiency in advanced infrastructure-based wireless networks. OFDMA combines OFDM on the physical layer with Frequency Division Multiple Access (FDMA) on the MAC layer to allow different subcarriers to be assigned to different stations to increase synchronicity. Consecutive subcarriers often experience similar channel conditions and are therefore grouped into subchannels. Thus, an OFDMA subchannel or RU is a set of subcarriers.

[0019] The multi-user property of OFDMA allows the AP to assign different RUs to different non-AP stations to increase contention, which can help reduce contention and collisions within 802.11 networks.

[0020] In OFDMA, different frame transmissions can simultaneously use different subsets of subcarriers in the channel bandwidth. In the downlink direction, the AP can emit parallel transmissions to different receiving non-AP stations. These transmissions are called multi-user downlink transmissions (MU DL). In addition, the AP can provide uplink transmission schedules to non-AP stations, and this kind of transmission scheme is called multi-user uplink (MU UL).

[0021] To support multi-user uplink transmissions, i.e., uplink transmissions to an 802.11ax access point (AP), during a granted TXOP, the 802.11ax AP must provide signaling information for legacy stations (non-802.11ax stations) to configure their NAVs and for 802.11ax client stations to determine the allocation of resource units (RUs) provided by the AP.

[0022] The 802.11ax standard specifies a new control frame, the Trigger Frame (TF), that is sent by the AP to stations to trigger multi-user uplink communication.

[0023] The IEEE 802.11-15 / 0365 document proposes that a "trigger frame" (TF) is transmitted by an AP to solicit the transmission of multi-user uplink (MU UL) OFDMA PPDUs from multiple stations. The TF specifies the resource units provided by the AP to non-AP stations. In response, the stations transmit their MU UL (OFDMA) PPDUs as an immediate response to the trigger frame. All transmitters can transmit data simultaneously but using disjoint sets of RUs (i.e., frequencies in the OFDMA system), resulting in transmissions with less interference.

[0024] A resource unit RU can be reserved for a specific station by the AP, in which case the AP indicates in the TF the station for which the RU is reserved. Such a RU is called a scheduled RU. The indicated station does not need to access the scheduled RU.

[0025] On the other hand, an AP can offer one or more resource units to an 802.11ax station through contention-based access. These RUs are called random RUs and contribute to improving the efficiency of the network with respect to unmanaged traffic to the AP.

[0026] Several communication channels (usually 20 MHz wide communication channels) are pre-empted by the AP, and all control frames, including trigger frames, are duplicated in each of these pre-empted channels so that legacy stations operating in any of these channels can set their NAV.

[0027] The 802.11ax standard allows for several types of trigger frames that trigger various items of information to a station. For example, one trigger frame can be used to retrieve uplink data traffic stored in a station's traffic queue. In another example, a trigger frame can be used to request a Buffer Status Report (BSR) from a station that determines which 802.11ax stations have uplink packets waiting to be transmitted and the associated size of those packets (the amount of data in the station's traffic queue).

[0028] As is immediately evident from the above, the multi-user uplink medium access scheme (or OFDMA access scheme or RU access scheme) makes it possible to reduce the number of collisions caused by simultaneous medium access attempts, and also reduces the overhead caused by medium access, since the medium access cost is shared among several stations. The OFDMA access scheme or RU access scheme therefore appears to be much more efficient (in terms of medium usage) than the traditional EDCA contention-based medium access scheme (for dense 802.11 cells).

[0029] Although the OFDMA or RU access schemes appear to be more efficient, the EDCA access scheme must also survive and therefore must coexist with the OFDMA or RU access schemes.

[0030] This is mainly due to the fact that there are legacy 802.11 stations that still need to have a chance to access the medium, but are not aware of the OFDMA or RU access schemes, and that overall fairness in accessing the medium must be ensured.

[0031] It is also increasingly necessary that 802.11ax stations should have the opportunity to access the medium through traditional EDCA contention-based medium access, for example to transmit data directly to another station (i.e., peer-to-peer (P2P) traffic as distinct from uplink traffic to the AP).

[0032] Therefore, the two medium access methods, EDCA access method and OFDMA / RU access method, must coexist.

[0033] This coexistence has a downside.

[0034] For example, an 802.11ax station and a legacy station have the same medium access probability using the EDCA access scheme, but an 802.11ax station has an additional medium access opportunity using the MU uplink or the OFDMA or RU access scheme.

[0035] As a result, access to the medium is not completely fair between 802.11ax and legacy stations.

[0036] In order to restore some fairness between these stations, solutions have been proposed in which, upon successful transmission of data over an accessed resource unit (i.e., through MU UL OFDMA transmission), the current value of at least one EDCA parameter is changed to a penalized or degraded value to reduce the probability that the station will regain access to the communication channel through EDCA contention. For example, the penalty or degraded value used for the EDCA parameter is more restrictive than the original value (i.e., the legacy value).

[0037] For example, upon successful (MU UL OFDMA) transmission of data in resource units RU reserved for data by the AP, an 802.11ax station is proposed to switch to MU EDCA mode for a predefined duration counted down by a timer (hereinafter denoted HEMUEDCATimer for High Efficiency Multi-User EDCA Timer). In MU EDCA mode, a set of EDCA parameters of the station are modified (specifically penalized) to reduce the probability that the station will again access the communication channel via the EDCA access scheme.

[0038] The penalty or degeneracy values ​​for the set of MU EDCA parameters are provided by the AP in a dedicated information element (usually in a beacon or association frame).

[0039] The technique disclosed in this document only increases the AIFSN value of each traffic queue to be transmitted in the accessed RU, while min and C.W. max It is proposed to keep unchanged the AIFS period. With the corresponding increase in the AIFS period, traffic queues in MU EDCA mode are effectively delayed from decrementing their queue backoff counters when they sense that the medium is not in use. This is especially important in high density environments where the medium is not idle for long periods of time.

[0040] When a station switches to MU EDCA mode, it starts counting down its HEMUEDCATimer. The HEMUEDCATimer is reinitialized every time the station successfully transmits (MU UL OFDMA) data in a newly reserved RU, regardless of the traffic queue from which the data is available for transmission. The initialization value of the HEMUEDCATimer is proposed to be large (e.g., tens of milliseconds) to encompass several new opportunities for MU UL transmission.

[0041] The HEMUEDCATimer mechanism means that the station will stay in the MU EDCA state as long as the AP provides the station with reserved RUs.

[0042] Once the HEMUEDCA Timer has elapsed, the traffic queue in MU EDCA mode is switched back to legacy EDCA mode with legacy EDCA parameters, thereby causing the queue to exit MU EDCA mode.

[0043] Therefore, this mechanism of dual operation modes of legacy EDCA mode and MU EDCA mode promotes the use of MU UL mechanisms by reducing the probability that a station in MU EDCA mode will access the medium using legacy EDCA mechanisms.

[0044] In addition, the document suggests providing a specific value for the AIFSN parameter in the set of shrinkage / penalty parameters provided by the AP, which indicates to stations that the AP should use a very large value for the AIFSN of the traffic queue or queues with which it is associated.

[0045] As stated in the above document, this very large value of AIFSN is proposed to be equal to the HEMUEDCATimer value also provided by the AP, which is typically on the order of tens of milliseconds compared to the worst case AIFS[i] of less than 0.1 milliseconds in legacy EDCA modes.

[0046] "0" has been proposed to be used for a specific value of the AIFSN parameter. This value is not normally allowed for the AIFSN (because AIFS must be at least equal to DIFS), so it is detected directly by stations as the code for setting the AIFSN using the HEMUEDCATimer.

[0047] As a result of this scheme, traffic queues in MU EDCA mode are granted transmission opportunities less frequently through EDCA contention. It therefore seems clear that the use of the code value (until now "0") is aimed at making these traffic queues access EDCA less frequent. On the other hand, this simplifies the process in the AP, as it is no longer necessary to calculate the relevant penalty AIFSN value.

[0048] However, by preventing the backoff counters from changing when a station uses the penalty MU EDCA parameters, this mechanism ensures that the queue backoff counters no longer mirror which traffic queue should have the highest transmission priority in the traditional EDCA sense (e.g., which traffic queue has the oldest data stored). For example, when a station receives a trigger frame with a scheduled RU dedicated to this station, this station can no longer use its backoff counters to handle its QoS and transmit data with the highest priority (not only with respect to its access category, but also with respect to the respective age of the data in the AC[] queues) since their backoff counters have been frozen. Summary of the Invention [Problem to be solved by the invention]

[0049] Therefore, QoS in the network is severely degraded, and there is a need to reintroduce proper QoS operations for traffic prioritization that are compatible with the medium access penalty scheme envisaged in the 802.11ax standard.

[0050] The present invention seeks to overcome the above problems, in particular the loss of QoS handling brought about by the introduction of MU UL OFDMA transmission.

[0051] With the introduction of 802.1e, data priority is handled via the EDCA backoff mechanism along with the four access category traffic queues. With the introduction of MU UL OFDMA communication, the ability of the EDCA backoff counter to mirror the relative priority of the four AC traffic queues is broken due to the fact that the EDCA backoff counter does not change upon transmission of data over MU UL OFDMA resource units.

[0052] Therefore, the present invention aims to restore EDCA-like behavior for queue backoff counters with the objective of relational mirroring of the relative priorities of AC queues. [Means for solving the problem]

[0053] In this situation, the present invention proposes a communication device operating as a base station for constructing a wireless network compliant with the IEEE802.11 series standard, the communication device having: a first transmitting means for transmitting a signal including information regarding an AIFSN (Arbitration Inter-Frame Space Number) value to another communication device; and a first receiving means for receiving data of a predetermined access category from the other communication device in a resource unit provided by the communication device for communication by OFDMA (Orthogonal Frequency Division Multiple Access), the first receiving means controlling to disable data transmission of the predetermined access category based on EDCA (Enhanced Distributed Channel Access) for a predetermined period of time in response to successful transmission of the data of the predetermined access category when the AIFSN value included in the signal transmitted by the first transmitting means is 0.

[0054] In contrast, a control method for a communication device operating as a base station for constructing a wireless network compliant with the IEEE 802.11 series standard is also proposed, the control method comprising: a transmitting step of transmitting a signal including information regarding an AIFSN (Arbitration Inter-Frame Space Number) value to another communication device; and a receiving step of receiving data of a predetermined access category from the other communication device in a resource unit provided by the communication device for communication by OFDMA (Orthogonal Frequency Division Multiple Access), the control step being such that, when the AIFSN value included in the signal transmitted by the first transmitting means is 0, data transmission of the predetermined access category based on EDCA (Enhanced Distributed Channel Access) is disabled for a predetermined period in response to successful transmission of the data of the predetermined access category.

[0055] Another embodiment of the invention provides a method of communication in a communication network comprising a plurality of stations, at least one of which comprises a plurality of traffic queues serving data traffic at different priorities, each traffic queue associated with a respective queue back-off counter for competing for access to a communication channel to transmit data stored in said traffic queue, the method comprising: decrementing said queue back-off counter over time as long as said communication channel is sensed (usually continuously) as being idle (or unused or available) for a period longer than a respective arbitration interframe space AIFS duration (meaning that each traffic queue detects the medium as being idle for its respective AIFS before decrementing its back-off counter in each new successive time slot in which the network is still sensed as being idle). and upon transmitting (preferably successfully) data stored in any of the traffic queues in an accessed resource unit provided by the access point within a transmission opportunity granted to the access point on the communication channel, switching the traffic queues from a legacy contention mode to a MU contention mode. When one of the queue back-off counters expires (i.e., passes, e.g., reaches zero), determining based on the current (i.e., at the time of expiration) mode of the associated traffic queue whether to access the communication channel to transmit data stored in the associated traffic queue, or to retrieve a new back-off value and reset the expired queue back-off counter without transmitting data from the associated traffic queue on the communication channel.

[0056] These embodiments can reverse the decrement of the backoff counter while still maintaining the penalty scheme envisaged in the 802.11ax standard, thus restoring the aging of data in the traffic queue and therefore restoring QoS.

[0057] This is achieved by the proposed control of medium access upon back-off counter expiry. Contrary to the prior art preventive measures to avoid back-off counter expiry, the proposed scheme is a reactive counter-measure against such back-off counter reaching 0. Thus, a dynamic behavior of the back-off counter is again possible.

[0058] Results of these embodiments include that the backoff counter can again mirror the aging of data in the AC, but also that the restrictive AIFSN value is no longer required.

[0059] Correspondingly, the embodiments also provide a communication station in a communication network comprising a plurality of stations. The communication station includes a plurality of traffic queues serving data traffic at different priorities, each traffic queue associated with a respective queue back-off counter for contending for accessing a communication channel to transmit data stored in the traffic queue; and at least one microprocessor configured to perform the following steps: decrementing the queue back-off counter over time as long as the communication channel is detected as being idle for a period longer than a respective arbitration inter-frame space duration; switching any traffic queue from a legacy contention mode to a MU contention mode upon transmitting data stored in the traffic queue in an accessed resource unit provided by the access point within a transmission opportunity granted to the access point on the communication channel; and determining, upon expiration of one of the queue back-off counters, based on a current mode of the associated traffic queue, whether to access the communication channel to transmit data stored in the associated traffic queue, or to retrieve a new back-off value and reset the expired queue back-off counter without transmitting data from the associated traffic queue on the communication channel. and at least one microprocessor configured to execute the

[0060] This station has the same advantages as the method defined above.

[0061] Optional features of the invention are defined in the appended claims, some of which are described below with respect to a method, but which may be replaced by features of a system dedicated to any communication station according to the invention.

[0062] In an embodiment, if the current mode is the legacy contention mode, the station accesses the communication channel to transmit data stored in the associated traffic queue; On the other hand, if the current mode is the MU contention mode, the new backoff value is retrieved and the expired queue backoff counter is reset without any data from the associated traffic queue being transmitted on the communication channel.

[0063] This ensures to maintain the penalty scheme envisaged in the 802.11ax standard: while the EDCA scheme continues to operate, traffic queues in MU EDCA mode are not granted medium access even if their associated back-off counters with restored dynamicity have expired, and this is the control provided by the present invention.

[0064] In some embodiments, the decision to access the communication channel or to retrieve a new backoff value is further based on data currently stored in the associated traffic queue. This approach allows for tailoring the penalty scheme for some types of data, particularly with the goal of maintaining QoS fairness for data not involved in MU UL transmissions.

[0065] According to a particular feature, if the current mode is the legacy contention mode or if the data stored in the associated traffic queue contains data addressed to another station other than the access point (i.e. it is P2P data), the station accesses the communication channel to transmit the data stored in the associated traffic queue, whereas if the current mode is the MU contention mode and the data stored in the associated traffic queue does not contain data addressed to another station other than the access point, no data from the associated traffic queue is transmitted on the communication channel, the new backoff value is retrieved and the expired queue backoff counter is reset.

[0066] In this configuration, QoS fairness is maintained for P2P traffic since it is not related to MU UL transmissions to the AP. In other words, the present invention allows EDCA access to the medium for P2P data even when stations are in MU EDCA mode of the same AC.

[0067] In this situation, in case of access to a communication channel, only data stored in the associated traffic queue in MU contention mode and addressed to another station different from the access point can be transmitted in the accessed communication channel, which means that only P2P data is allowed in conventional EDCA access while the station (or the corresponding AC) is in MU EDCA mode.

[0068] In some embodiments, the MU contention mode uses the same arbitration interframe space duration as the legacy contention mode. Due to the reactive approach of the present invention, the control over medium access in the penalty scheme no longer depends on the contention parameters, but mainly on an additional check of the current mode upon expiration of any backoff counter. As a result, it is no longer necessary to change the AIFSN, and therefore no longer necessary for the AP to transmit the changed value of the AIFSN. As a result, processing at stations and APs is reduced, as well as the bandwidth usage of transmitting the penalty EDCA parameters.

[0069] In a particular embodiment, each queue backoff counter is reset with an associated backoff value taken from a respective contention window, and the MU contention modes have the same lower bound CW as the legacy contention modes, which together define a selection range within which the size of the contention window is selected. min and / or the same upper boundary CW max is used.

[0070] This configuration simplifies entry and exit into MU contention modes (e.g., MU EDCA mode) by not requiring changes to the contention window, although variations can be envisioned having different boundaries between legacy and MU contention modes.

[0071] In some embodiments, the method further comprises switching traffic queues back to the legacy contention mode upon expiration of an MU mode timer (called HEMUEDCATimer in the standard) at the station, which is initialized when the traffic queues are switched to the MU contention mode. According to a particular feature, the MU mode timer is shared by all the traffic queues and is reinitialized to a predefined duration each time data from any traffic queue is transmitted in an accessed resource unit provided by the access point within any subsequent transmission opportunity granted to the access point on the communication channel. This means that all traffic queues in the MU contention mode exit the MU contention mode when the predefined duration has elapsed without data from the station being transmitted in an RU provided by the AP within a subsequent transmission opportunity.

[0072] In some embodiments, data transmitted in resource units provided by the access point within a transmission opportunity granted to the access point is taken from at least one traffic queue selected based on the current (i.e., upon access to the medium) backoff value of its associated queue backoff counter.

[0073] Thus, fair management of QoS is maintained when implementing the present invention.

[0074] According to a particular feature, the traffic queue selection selects the traffic queue or queues with the smallest current back-off value, thus preserving the EDCA-like behavior of AC queues.

[0075] In an alternative embodiment, data transmitted in resource units provided by the access point within a transmission opportunity granted to the access point is taken from a preferred traffic queue indicated by the access point.

[0076] According to a particular feature, the priority traffic queue indication is included in a trigger frame received from the access point, the trigger frame reserving the transmission opportunity granted to the access point on the communication channel and defining a resource unit RU forming the communication channel including the accessed resource unit.

[0077] This approach allows the AP to drive QoS management.

[0078] In another embodiment, a reset flag is associated with each traffic queue, the reset flag being enabled each time a new backoff value is retrieved and the associated queue backoff counter is reset without data being transmitted from the traffic queue, and being disabled each time data from the traffic queue is transmitted; Data transmitted in resource units provided by the access point within a transmission opportunity granted to the access point is retrieved from at least one traffic queue selected based on the valid or invalid status of the reset flag associated with the traffic queue.

[0079] Preferably, data transmitted in resource units provided by the access point within a transmission opportunity granted to the access point is taken from a traffic queue that has a reset flag enabled.

[0080] The reset flag in these configurations preserves the QoS information for the traffic queue. In fact, due to the retrieval of a new backoff value, the priority of the traffic queue may be lost compared to other traffic queues (because data is not transmitted in MU EDCA mode at the re-retrieval of the new backoff value for this case). Therefore, the reset flag is used to indicate which traffic queue had priority data that should be transmitted urgently when it accessed the RU.

[0081] In some embodiments, the method further includes calculating, at the station, a new back-off value for at least one traffic queue transmitting in the accessed resource unit and resetting the associated queue back-off counter.

[0082] This approach restores fairer QoS management since the same behavior as in conventional EDCA applies here (a new backoff value for a traffic queue every time it transmits).

[0083] In a particular embodiment, new back-off values ​​are calculated only for the transmit traffic queues whose data is transmitted at the beginning of the accessed resource unit. Usually, the first transmit queue will transmit the majority of the data in the accessed resource unit. The other transmit queue or queues transmit only a small amount of data to fill the available bandwidth in the accessed resource unit (given the TXOP). In this respect, retrieving new back-off values ​​for these "secondary" queues is very sensitive to their remaining data, and therefore it has to wait longer than with direct OFDMA access. As a result, the proposed embodiment maintains fairness with respect to these secondary queues by maintaining their same future transmission probability.

[0084] In a variant, a new queue backoff value is calculated for each transmission traffic queue, thus achieving fair QoS management since exactly the same behavior as in conventional EDCA applies.

[0085] In some embodiments, the backoff value used to reset the queue backoff counter is calculated based on a contention parameter set respectively associated with each traffic queue; The method further includes, at the station, resetting the contention parameter sets associated with a traffic queue (preferably each traffic queue) that has remained in MU contention mode for at least a parameter lifetime duration to a default configuration parameter set, For example, the parameter lifetime duration may correspond to at least twice the predetermined duration used to initialize the MU mode timer (HEMUEDCA Timer).

[0086] This contributes to increasing network efficiency. Indeed, if the traffic queues concerned remain in MU contention mode for a long time, their contention parameters, typically EDCA parameters, no longer mirror the actual network conditions. Therefore, by resetting their parameter sets, any old network constraints they may have incorporated are erased. This makes the traffic queues like new traffic queues in the network, with no knowledge of the network conditions.

[0087] In another embodiment, the method further includes, at the station, retrieving a new backoff value and resetting the expired queue backoff counter after data stored in the associated traffic queue is transmitted on the accessed communication channel, thus maintaining traditional EDCA behavior when a station or traffic queue is in legacy EDCA contention mode.

[0088] In yet another embodiment, the method further comprises receiving, at the station, a trigger frame from the access point reserving the transmission opportunity granted to the access point on the communication channel and defining at least one resource unit RU (preferably a plurality of resource units) forming the communication channel including the accessed resource unit, this being in accordance with standard requirements for declaring RUs.

[0089] In yet another embodiment, the transmission traffic queue or queues are switched to the MU contention mode only when the data in the accessed resource unit is successfully transmitted. This configuration ensures fairness. In fact, the basic idea of ​​switching contention modes is that the penalty MU contention mode should only be implemented to compensate for the presence of other transmission opportunities (here through the RU) that mean the data has been successfully transmitted.

[0090] In yet another embodiment, the accessed resource unit over which the data is transmitted is a random resource unit, and the access of the random resource unit is performed through contention using a contention parameter set that is separate from the contention parameter set used to derive the backoff value and reset the queue backoff counter.

[0091] In yet another embodiment, the accessed resource units on which the data is transmitted are scheduled resource units, the scheduled resources being assigned to the station by the access point.

[0092] Of course, some stations may access scheduled RUs while other stations may access random RUs, resulting in various stations being in MU contention mode (for one or more AC queues) at the same time.

[0093] Another aspect of the invention relates to a non-transitory computer readable medium storing a program which, when executed by a microprocessor or computer system in a device, causes said device to perform any of the methods as defined above.

[0094] The non-transitory computer readable medium may have features and advantages similar to those set forth above and below with respect to the method and device.

[0095] At least some parts of the methods according to the invention may be computer-implemented. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software (including firmware, resident software, microcode, etc.) embodiment, or an embodiment combining software and hardware aspects, all of which are collectively referred to herein as "circuits," "modules," or "systems." Furthermore, the present invention may take the form of a computer program product embodied in any tangible medium of expression having computer usable program code embodied therein.

[0096] Since the present invention can be implemented in software, the present invention can be embodied as computer readable code provided to a programmable device on any suitable carrier medium. Tangible carrier medium can include storage media such as hard disk drives, magnetic tape devices or solid-state memory devices. Transient carrier medium can include signals such as electrical, electronic, optical, acoustic, magnetic or electromagnetic signals, e.g. microwave or RF signals.

[0097] Further advantages of the present invention will become apparent to those skilled in the art from a study of the drawings and detailed description, in which: Embodiments of the present invention will now be described, by way of example only, with reference to the following drawings, in which: [Brief description of the drawings]

[0098] [Figure 1] 1 illustrates a typical wireless communication system in which embodiments of the present invention may be implemented; [Figure 2a] FIG. 1 illustrates IEEE 802.11e EDCA with access categories. [Figure 2b] FIG. 2 illustrates an example of mapping between eight priorities of traffic classes and four EDCA ACs. [Figure 2c] FIG. 1 illustrates the 802.11e mechanism for counting down the backoff counter. [Figure 2d] FIG. 2 illustrates the structure of a MAC data frame header. [Diagram 3] FIG. 1 illustrates 802.11ac channel allocations supporting composite channel bandwidths of 20 MHz, 40 MHz, 80 MHz or 160 MHz as known in the art. [Figure 4] FIG. 2 illustrates, with a timeline, an example of an 802.11ax uplink OFDMA transmission scheme in which an AP issues a trigger frame to reserve OFDMA resource units on an 80 MHz channel, as known in the art. [Figure 5a] FIG. 2 illustrates an exemplary scenario showing conventional transmission of a trigger frame using an EDCA mechanism, to which a penalty scheme as known in the art can be applied. [Figure 5b] FIG. 2 illustrates an exemplary evolution of a backoff counter and associated data selection as known in the prior art. [Figure 5c] FIG. 2 illustrates the evolution of a backoff counter and associated data selection according to an embodiment of the present invention. [Figure 6] FIG. 1 shows a schematic representation of a communications device or station according to an embodiment of the present invention. [Figure 7] 1 is a block diagram illustrating generally the architecture of a wireless communication device according to an embodiment of the present invention; [Figure 8]FIG. 2 illustrates an exemplary transmission block of a communication station according to an embodiment of the present invention. [Figure 9] 4 is a flow chart showing the main steps performed by the MAC layer of a station when it receives new data to transmit in an embodiment of the invention; [Figure 10] 4 is a flow chart illustrating steps of accessing a medium based on an EDCA medium access scheme according to an embodiment of the present invention. [Figure 11] 1 is a flow chart illustrating the steps of accessing a resource unit based on an RU or OFDMA access scheme upon reception of a trigger frame defining an RU according to an embodiment of the present invention. [Figure 12] 1 is a flow chart illustrating station management switching back (or falling back) from MU contention mode to legacy contention mode in accordance with an embodiment of the present invention. [Figure 13] FIG. 1 is a diagram showing the structure of a trigger frame defined in the 802.11ax standard. [Figure 14a] A diagram showing the structure of standardized information elements used to describe EDCA parameters in beacon frames. [Figure 14b] FIG. 13 illustrates an exemplary structure of a dedicated information element for transmitting degenerate EDCA parameter values ​​and a HEMU EDCA Timer value according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0099] The invention will now be described by way of specific, non-limiting, exemplary embodiments with reference to the figures.

[0100] 1 shows a communication system in which several communication stations (or "nodes") 101-107 exchange data frames over a wireless transmission channel 100 of a wireless local area network (WLAN) under the control of a central station or access point (AP) 110, also seen as a station of the network. The wireless transmission channel 100 is defined by an operating frequency band made up of a single channel or several channels forming a composite channel.

[0101] In the following, the term "station" refers to any kind of station. The phrase "access point station" or shortened "access point" (AP) refers to a station that acts as an access point 110. The terms "non-access point station" or shortened "non-AP station", or client station (STA), refer to the other stations 101-107.

[0102] Access to the shared wireless medium for transmitting data frames is primarily based on the CSMA / CA technique, which employs carrier sensing and collision avoidance by separating simultaneous transmissions in space and time.

[0103] Carrier sensing in CSMA / CA is performed by both physical and virtual mechanisms. Virtual carrier sensing is achieved by transmitting a control frame that reserves the medium before transmitting a data frame.

[0104] Next, a source or transmitting station, including an AP, first attempts to sense, through a physical mechanism, the medium being idle for at least one DIFS (standing for DCF Interframe Spacing) time period before transmitting a data frame.

[0105] On the other hand, if the shared wireless medium is detected to be busy for a DIFS period, the source station continues to wait until the wireless medium becomes idle.

[0106] The wireless communication system of Fig. 1 comprises a physical access point 110 configured to manage a WLAN BSS (Basic Service Set), i.e. a group of non-AP stations that are preregistered with the AP. Such a BSS managed by an AP is called an infrastructure BSS. In the following, the term BSS is used as equivalent to an infrastructure BSS.

[0107] Once a BSS is established, access points can bridge traffic within the BSS or from other networks (e.g., wired networks) into the BSS (or vice versa). Thus, stations in a BSS should only talk to the AP responsible for relaying data frames if those frames are intended for another station in the BSS.

[0108] To access the medium, any station, including the AP, starts counting down a back-off counter designed to expire after a randomly chosen number CW (an integer) of time slots in the so-called contention window [0, CW]. This back-off mechanism or procedure, also called the channel access scheme, is the basis of a collision avoidance mechanism that postpones the transmission time by a random interval, thus reducing the probability of collisions on a shared channel. After the back-off time has expired (i.e., after the back-off counter reaches 0), the source station can transmit a data frame or a control frame if the medium is idle.

[0109] Quality of Service (QoS) management is implemented at station level in such wireless networks through the well-known EDCA mechanism defined in the IEEE 802.11e standard.

[0110] In fact, in the original DCF standard, a communication station only has one transmission queue / buffer, but a delay in the transmission or retransmission of a previous frame prevents the communication from having QoS, since a subsequent data frame cannot be transmitted until the transmission / retransmission of the previous frame is finished.

[0111] FIG. 2a shows the IEEE 802.11e EDCA mechanism with access categories to improve quality of service (QoS) and utilize the wireless medium more efficiently.

[0112] The 802.11e standard relies on a coordination function called the Hybrid Coordination Function (HCF) that has two operating modes: Enhanced Distributed Channel Access (EDCA) and HCF Controlled Channel Access (HCCA).

[0113] EDCA enhances or extends the capabilities of the original DCF access method: EDCA is designed to support prioritized traffic, similar to Differentiated Services (DiffServ), a protocol that designates and controls network traffic by classes so that some specific types of traffic have priority.

[0114] EDCA is the dominant channel access method or mechanism in WLANs because it features a distributed and easy to deploy mechanism, in which stations contend for access to at least one communication channel of a communication network using EDCA contention parameters in order to transmit locally stored data over the accessed communication channel.

[0115] The above deficiency of not being able to have a satisfactory QoS due to frame retransmission delays is addressed by using multiple transmit queues / buffers.

[0116] QoS support in EDCA is achieved by introducing four Access Categories (ACs) and, thereby, four corresponding transmission / traffic queues or buffers 210. Typically, the four ACs are as follows, in descending order of priority: Voice (or "AC_VO"), Video (or "AC_VI"), Best Effort (or "AC_BE"), and Background (or "AC_BG").

[0117] Of course, a different number of traffic queues may be considered.

[0118] Each AC has its own traffic queue / buffer that stores the corresponding data frames to be transmitted on the network. Data frames, i.e., MSDUs, coming from the upper layers of the protocol stack are mapped to one of the four AC queues / buffers and thus input to the mapped AC buffer.

[0119] Each AC also has its own set of queue contention parameters, associated with a priority value and thus defining high or low priority MSDU traffic. Thus, there are multiple traffic queues, each serving data traffic at a different priority. The queue contention parameters are typically set by the CW min Parameter, CW max The parameters include the AIFSN parameter and the TXOP_Limit parameter. min and C.W. maxare the lower and upper bounds of the selection range within which the EDCA contention window CW is selected for a given traffic queue. AIFSN stands for Arbitration Interframe Space Number and defines the number of time slots (typically 9 μs) added to the SIFS interval (a total number defining the AIFS period, see Figure 2b) that a station must sense the medium as idle before decrementing the queue backoff counter associated with the traffic queue under consideration. TXOP_Limit defines the maximum size of a TXOP that a station may request.

[0120] That is, each AC (and corresponding buffer) operates as an independent DCF contention entity with its respective queue backoff engine 211. Each queue backoff engine 211 is thus associated with a respective traffic queue 210 for using queue contention parameters and retrieving a backoff value (from the CW) to initialize a respective queue backoff counter that is used to contend for access to at least one communication channel to transmit data stored in the respective traffic queue over the accessed communication channel.

[0121] The contention window CW and the backoff value are known to be EDCA variables.

[0122] As a result, ACs within the same station compete with each other to access the wireless medium and obtain transmission opportunities, for example, using conventional EDCA access schemes as described above.

[0123] Service differentiation between ACs is achieved by using different CWs. min , C.W. max This is achieved by setting different queue backoff parameters between ACs, such as AIFSN, AIFSN, and / or different transmission opportunity duration limits (TXOP_Limit), which contributes to adjusting the QoS.

[0124] In addition to using a smaller CW on average, the use of different AIFSN values ​​(which postpone the decrement of the backoff counter) gives high priority traffic in EDCA a higher chance of being transmitted than low priority traffic, i.e., a station with high priority traffic will wait, on average, a shorter statistical time before transmitting its packets than a station with low priority traffic.

[0125] Referring to the four AC buffers (210) shown in Figure 2a, buffers AC3 and AC2 are typically reserved for real-time applications (e.g., audio AC_VO or video transmission AC_VI). These buffers have the highest and second highest priority, respectively.

[0126] Buffers AC1 and AC0 are reserved for best effort (AC_BE) and background (AC_BG) traffic. These buffers have the second lowest and lowest priority, respectively.

[0127] Each data unit MSDU arriving at the MAC layer from an upper layer (e.g., link layer) with a certain priority is mapped to an AC according to a mapping rule. Figure 2b shows an example of mapping between eight priorities of traffic classes (user priorities between 0 and 7 according to IEEE 802.1d, i.e., UP) and four ACs. The data frame is then stored in a buffer corresponding to the AC to which it is mapped.

[0128] The effects of different AIFSNs are shown, for example, in Figure 2c.

[0129] Each station must wait a fixed amount of time to ensure that the medium is clear before attempting to transmit. With DCF, the DIFS is constant for all types of traffic, whereas with 802.11e, the fixed amount of time a station must wait depends on the access category and is called the arbitration interframe space (AIFS).

[0130] With AIFS, each traffic queue "i" waiting to transmit must wait until the medium is declared available through a Clear Channel Assessment (CCA) and a Network Allocation Vector (NAV), which are omitted for brevity and will not be discussed here. Once the medium is available, each traffic queue "i" must wait the corresponding AIFS[i] period (including any SIFS period during which it defers access to the medium) before decrementing its associated queue backoff counter.

[0131] Thus, each of the four traffic queues has a defined interframe space value that corresponds to the priority assigned to that queue. For example, the AC_VO queue has the highest priority and therefore the smallest interframe space timer. All AIFS timers (250) assigned by IEEE 802.11e are defined as one Short Interframe Spacing (SIFS) value plus a variable number of slot times (AIFSN) defined by the physical layer coding method in use (CCK, DSSS, OFDM). The value of the EDCA parameter AIFS Number (AIFSN) is administrator configurable, with the default value defined as follows: AC_VO 1SIFS+2*slot time (AIFSN=2) AC_VI 1SIFS+2*slot time (AIFSN=2) AC_BE 1SIFS+3*slot time (AIFSN=3) AC_BG 1SIFS+7*slot time (AIFSN=7)

[0132] The AIFSN value can be provided by the AP in a so-called EDCA parameter set information element (e.g., provided in a beacon frame transmitted by the AP). The AIFSN field in this information element is 4 bits long and has a minimum value of 2 specified in the standard and a maximum value of 15 based on the field length limitations.

[0133] In this manner, arbitration inter-frame spacing allows frames in higher priority traffic queues to gain a statistical advantage by not having to wait too long before decrementing their random backoff counters compared to other queues.

[0134] The figure shows two AIFS[i] corresponding to two different traffic queues. Due to this difference in prioritization, it can be seen that one priority traffic queue starts decrementing its backoff value earlier than the other lower priority traffic queue. This situation is repeated after each new medium access by any station in the network (i.e., when the medium is newly detected as idle).

[0135] To begin transmitting data, a traffic queue at a station first randomly selects a backoff value for its backoff counter. This backoff value must be within the contention window value defined for that traffic queue, as described above. Similar to the AIFS parameter, the difference between the contention windows of various traffic queues helps prioritize traffic in higher priority queues by allowing them to wait shorter time intervals before being allowed to transmit over the air.

[0136] Once the appropriate AIFS[i] period has expired, each traffic queue may begin decrementing its queue backoff counter (251) by one every slot time that passes.

[0137] Then, once the EDCA backoff procedure for a traffic queue (or AC) is completed (at least one backoff counter reaches zero), the MAC controller (reference number 704 in FIG. 7 below) of the transmitting station sends the data frames from this traffic queue to the physical layer for transmission over the wireless communication network.

[0138] Since traffic queues operate simultaneously in accessing the wireless medium, it may happen that two traffic queues of the same station finish their backoff at the same time. In such a situation, the virtual collision handler (212) of the MAC controller makes a selection of the AC with the highest priority (as shown in Fig. 2b) among the colliding ACs and gives up the transmission of data frames from the AC with the lower priority.

[0139] The virtual collision handler then instructs the AC with the lower priority to restart the backoff operation with an increased CW value.

[0140] Fig. 2d shows the structure of a MAC data frame and a QoS control field 200 contained in the header of an IEEE 802.11e MAC frame. The MAC data frame also contains, among other things, a Frame Control Header 201 and a Frame Body 202. As shown in the figure, the QoS control field 200 consists of 2 bytes and contains the following information items: - bits B0 to B3 are used to store a traffic identifier (TID) 204 identifying the traffic stream. The traffic identifier takes the value of the transmission priority value corresponding to the data carried by the data frame (user priority UP values ​​from 0 to 7, see Fig. 2b) or, for other data streams, the traffic stream identifier TSID value from 8 to 15; Bit B4 is used by non-AP stations to distinguish the meaning of bits B8 to B15, as detailed below; - bits B5 and B6 define an ACK Policy subfield that specifies the acknowledgement policy associated with the data frame. This subfield is used to determine how the receiving station must acknowledge the data frame, i.e., with a Normal ACK, a No ACK or a Block ACK; - bit B7 is reserved, meaning that it is not used by the current 802.11 standard; and - If bit B4 is set to 1, bits B8-B15 represent the "Queue Size" subfield 203, which indicates the amount of buffered traffic for the given TID at the non-AP station sending this frame. This queue size value is the total size of all packets buffered for the specified TID, rounded up to the nearest multiple of 256 octets, expressed in units of 256 octets. The access point receiving the frame can use this information to determine the next TXOP duration to grant to the station. A queue size of 0 indicates that there is no buffered traffic for that TID. A queue size of 255 indicates that the size of that TID 204 is unspecified or unknown. - If instead of using "Queue Size", bit B4 is set to 0, bits B8 to B15 represent the "TXOP Duration Requested" subfield, which indicates the duration in units of 32 μs required by the transmitting station for its next TXOP for the specified TID. Of course, "TXOP Duration Requested" provides an equivalent request to "Queue Size", since they both take into account all packets buffered for the specified TID.

[0141] The 802.11e MAC frame format, and more specifically the QoS control field 200, is being maintained for up and comer versions of the standard as described herein.

[0142] To meet the ever-increasing demand for faster wireless networks that support bandwidth-intensive applications, 802.11ac targets larger bandwidth transmissions through multi-channel operation. Figure 3 shows the 802.11ac channel allocations that support composite channel bandwidths of 20 MHz, 40 MHz, 80 MHz, or 160 MHz.

[0143] IEEE 802.11ac introduces support for a limited number of predefined subsets of 20 MHz channels that form a unique predefined composite channel configuration that is available for reservation by any 802.11ac station on the wireless network transmitting data.

[0144] These predefined subsets are shown in the figure and correspond to channel bandwidths of 20 MHz, 40 MHz, 80 MHz, and 160 MHz, as compared to only 20 MHz and 40 MHz supported by 802.11n. In effect, the 20 MHz component channels 300-1 through 300-8 are concatenated to form a wider composite communication channel.

[0145] In the 802.11ac standard, the channels of each predefined 40 MHz, 80 MHz or 160 MHz subset are contiguous within the operating frequency band, i.e., no holes (missing channels) are allowed in the ordered composite channels within the operating frequency band.

[0146] A 160 MHz channel bandwidth consists of two 80 MHz channels that may or may not be contiguous in frequency. An 80 MHz channel and a 40 MHz channel consist of two adjacent or contiguous 40 MHz and 20 MHz channels, respectively. However, the present invention may have embodiments with any composition of channel bandwidth, i.e., a composition that includes only contiguous channels within the operating band or a composition formed of non-contiguous channels.

[0147] Stations are granted TXOP through the Enhanced Distributed Channel Access (EDCA) mechanism on the "primary channel" (300-3). Indeed, for each composite channel with a certain bandwidth, 802.11ac designates one channel as "primary", meaning that it is used to compete for access to the composite channel. The primary 20 MHz channel is common to all client stations (STAs) that belong to the same basic set, i.e., all client stations managed by or registered to the same local access point (AP).

[0148] Meanwhile, in order to ensure that other legacy stations (i.e. stations not belonging to the same set) do not use the secondary channel, it is provided that the control frames reserving the composite channel (e.g. RTS / CTS frames) are duplicated over each 20 MHz channel of such composite channel.

[0149] As covered above, the IEEE 802.11ac standard allows for the bonding of up to four and even eight 20 MHz channels. Due to the limited number of channels (19 in the 5 GHz band in Europe), channel saturation becomes an issue. In fact, in densely populated areas, the 5 GHz band tends to become saturated even with 20 MHz or 40 MHz bandwidth per WLAN cell.

[0150] Developments in the 802.11ax standard aim to improve the efficiency and utilization of wireless channels in dense environments.

[0151] From this perspective, one can consider the multi-user (MU) transmission property, which allows multiple simultaneous transmissions in both the downlink (DL) and uplink (UL) directions between different users and the access point. In the uplink, multi-user transmission can be used to mitigate the collision probability by allowing multiple non-AP stations to transmit simultaneously to the AP.

[0152] To make such multi-user transmission practical, it is proposed to divide the licensed 20 MHz channel (400-1 to 400-4) into at least one subchannel, also called subcarriers or resource units (RUs) or "traffic channels", and preferably into multiple subchannels 410 (basic subchannels), which are shared in the frequency domain by multiple users, for example based on Orthogonal Frequency Division Multiple Access (OFDMA) techniques.

[0153] This will be explained with reference to FIG.

[0154] In this example, each 20 MHz channel (400-1, 400-2, 400-3 or 400-4) is subdivided in the frequency domain into four OFDMA subchannels or RUs 410 of size 5 MHz. Of course, the number of RUs into which a 20 MHz channel is divided may be different from four. For example, between two and nine RUs may be provided (each thus having a size between 10 MHz and about 2 MHz). It is also possible to have RU widths larger than 20 MHz when included within a wider composite channel (e.g. 80 MHz).

[0155] Contrary to MU downlink OFDMA (supported by a specific indication inside the PLCP header), which allows an AP to transmit multiple data to multiple stations directly, a triggering mechanism is employed whereby the AP triggers MU uplink communications from various non-AP stations.

[0156] To support MU uplink transmissions (during TXOPs pre-acquired by the AP), the 802.11ax AP must provide signaling information for both legacy stations (i.e., non-802.11ax stations) to set their NAV and for 802.11ax client stations to determine resource unit allocations.

[0157] In the following description, the term legacy refers to non-802.11ax stations, meaning previous technology 802.11 stations that do not support OFDMA communication.

[0158] As shown in the example of FIG. 4, the AP transmits a trigger frame (TF) 430 to the targeted 802.11ax stations. The bandwidth or width of the targeted composite channel is signaled in the TF frame, which means that a 20 MHz value, a 40 MHz value, an 80 MHz value or a 160 MHz value is signaled. The TF frame is transmitted over the primary 20 MHz channel and is duplicated (copied) on each other 20 MHz channel forming the targeted composite channel. Due to the duplication of the control frame, any consecutive legacy station that receives the TF frame (or its copy) on its primary channel is expected to then set its NAV to the value specified in the TF frame. This prevents these legacy stations from accessing the channel of the targeted composite channel during the TXOP.

[0159] Based on the AP's decision, the trigger frame TF can define multiple resource units (RUs) 410. The multi-user nature of OFDMA allows the AP to assign different RUs to different client stations to promote contention, which can help reduce contention and collisions within the 802.11 network.

[0160] As already discussed, RUs having widths greater than 20 MHz can be defined. As an example, a 996 tone RU can be provided by an AP to encompass an 80 MHz communication channel and thus be equivalent in capacity to a single user 80 MHz communication. It is thus understood that such 80 MHz communication is still a MU UL communication within the RU in the sense that it is triggered by a trigger frame issued by the AP.

[0161] The trigger frame 430 may specify "scheduled" RUs that may be reserved by the AP for some specific stations, in which case access contention for such RUs is not required for these stations. Such RUs and their corresponding scheduled stations are indicated in the trigger frame. For example, a station identifier, such as an association ID (AID) assigned to each station at registration, is added in association with each scheduled RU to indicate the stations that are enabled to use each scheduled RU. Such a transmission mode coexists with the conventional EDCA mechanism, and uplink data transmitted to the AP is taken from the EDCA queue 210.

[0162] The trigger frame TF may also specify "random" RUs in addition to or instead of "scheduled" RUs. Random RUs may be accessed randomly by stations of the BSS. In other words, the random RUs specified or assigned by the AP in the TF may serve as a base element for contention between stations intending to access the communication medium to transmit data. A collision occurs when two or more stations attempt to transmit simultaneously via the same RU. An AID equal to 0 may be used to identify a random RU.

[0163] For the 802.11ax standard, a random assignment procedure based on an additional backoff counter (OFDMA backoff counter, or OBO counter or RU counter) for RU contention by 802.11ax non-AP stations can be considered. That is, a random assignment procedure can be considered to enable non-AP stations to perform access contention among themselves and transmit data via random RUs. The RU backoff counter is completely different from the EDCA backoff counter 211. However, it is assumed that data transmitted in the accessed OFDMA RU 410 is sourced from the same EDCA traffic queue 210.

[0164] The RU random allocation procedure includes a first step of determining from the received trigger frame a subchannel or RU of the communication medium available for contention (so-called "random RU") for one of the multiple 802.11ax stations with a positive RU backoff value (initially captured within the RU contention window range), a second step of checking whether the value of the RU backoff value local to the station under consideration is less than or equal to the number of random RUs detected as available, and a third step of randomly selecting an RU to transmit data thereafter from among the RUs detected as available if the checking is successful. If the second step does not check that the value is less than or equal to the number of random RUs detected as available, a fourth step (instead of the third step) is performed to decrement the RU backoff counter by the number of random RUs detected as available.

[0165] As can be seen above, a station is not guaranteed to perform an OFDMA transmission via a random RU every time a TF is received, because at least the RU backoff counter is decremented by the number of proposed random RUs upon each reception of a trigger frame, which makes the data transmission different for subsequent trigger frames (depending on the current value of the RU backoff number and also the number of random RUs provided by each received TF).

[0166] Returning to Figure 4, due to the various possible accesses to the RUs, some of the RUs are unused (410u) because no station with a RU backoff value smaller than the number of available random RUs randomly selected one of these random RUs, while some other RUs are in collision (410c, for example) because at least two of these stations randomly selected the same random RU. This shows that due to the random determination of which random RU to access, collisions may occur across some RUs while other RUs remain unused.

[0167] When a station transmits data to the AP using scheduled and / or random RUs, the AP responds with a multi-user acknowledgment (not shown) to acknowledge the data in each RU.

[0168] The MU Uplink (UL) medium access scheme, including both scheduled and random RUs, has been shown to be highly efficient compared to traditional EDCA access schemes, especially in high-density environments such as those envisioned by the 802.11ax standard, because it reduces both the number of collisions caused by simultaneous medium access attempts and the overhead caused by medium access.

[0169] Meanwhile, the EDCA access scheme and the MU UL OFDMA / RU access scheme must coexist, especially to allow legacy 802.11 stations to access the medium and also allow 802.11ax stations to initiate communication with other non-AP stations.

[0170] While the EDCA access scheme alone provides fair access to the medium across all stations, when this scheme is combined with the MU UL OFDMA / RU access scheme, a drift in fairness is introduced because, compared to legacy stations, 802.11ax stations have additional opportunities to transmit data through the resource units provided in the AP-granted transmission opportunities.

[0171] Solutions have been proposed to restore some fairness between stations.

[0172] For example, in co-pending UK Application No. 1612151.9, filed July 13, 2016, upon successful transmission of data over an accessed resource unit (i.e., through an UL OFDMA transmission), the current value of at least one EDCA parameter is changed to a different value (an MU EDCA parameter) in order to reduce the probability that a station will access the communication channel through (traditional EDCA) contention.

[0173] In this framework, a mechanism is proposed to reduce the probability that a station will perform an EDCA-based transmission (i.e., using the EDCA medium access method) as soon as it has successfully transmitted its data using the MU-UL mechanism. This reduction is achieved by using the well-known EDCA parameter set (AIFSN, CW min and C.W. max This is done by modifying the

[0174] The proposed mechanism configures each transmit traffic queue (or MU node) to an MU EDCA mode (or MU node) in response to successful transmission of data in the accessed MU UL OFDMA resource unit for a predefined duration known as the HEMUEDCATimer, in which the respective EDCA parameter set is changed to a MU parameter set that is different from the legacy EDCA parameter set used in the different legacy EDCA modes.

[0175] To switch from the legacy EDCA contention access mode to the MU EDCA mode, a station must update its EDCA parameter sets (AIFSN, CWSN, IF ... min , and / or C.W. max) can be changed. Switching back to legacy EDCA mode can be done upon expiration of the HEMUEDCATimer. Note that this timer is reset to its initial value every time the station transmits new data again (from any AC) during a newly accessed resource unit provided by the AP. The initialization value of HEMUEDCATimer is proposed to be large (e.g., tens of milliseconds) to encompass several new opportunities for MU UL transmission.

[0176] The modified values ​​of the EDCA parameter sets (ie, the MU parameter sets for the four traffic queues) may be transmitted by the access point in dedicated information elements that are normally transmitted within beacon frames that broadcast network information to stations.

[0177] This specification provides a specific configuration for transmit traffic queues in MU EDCA mode that tend to make less frequent EDCA accesses to the medium. The AP specifies this specific mode of operation by indicating a specific value (usually 0) of the AIFSN parameter in the set of MU EDCA parameters. Such a specific value means for a station that it uses a very large value of its AIFSN in MU EDCA mode. This value shall be equal to the HEMUEDCATimer transmitted by the AP (recall that the value should be large, on the order of tens of milliseconds, compared to the lowest AIFS[i] in legacy EDCA mode, which is less than 0.1 milliseconds).

[0178] Unfortunately, the known mechanisms for controlling fairness fluctuations by significantly modifying EDCA parameters, especially the AIFSN value, reduce the chance that the queue backoff counters of each traffic queue in MU EDCA mode are changed (decremented), thereby reducing their effective utilization in determining the relative priority of the queues. Thus, in effect, the queue backoff values ​​no longer mirror which traffic queue should have the highest transmission priority in the EDCA sense (e.g., the traffic queue with the oldest data stored in it).

[0179] It is therefore no longer possible for stations to adhere to the QoS principles as set out in the 802.11e standard.

[0180] This will now be explained with reference to Figure 5a, which shows an example scenario of an 802.11ax network implementing the MU EDCA mode as described in the above-mentioned document. In this scenario, a first trigger frame that is entirely random (meaning that this frame specifies only random RUs) is sent by the AP to poll stations with uplink data, followed by a second trigger frame that is entirely scheduled (meaning that this frame specifies only scheduled RUs).

[0181] Since the receiver access point performs contention on behalf of non-AP stations in uplink OFDMA, the access point should know an indication of both which non-AP stations have uplink packets and what the size of their buffer 210 is. In fact, if a non-AP station that does not have an uplink packet is polled for uplink OFDMA transmission, the resource units allocated for MU UL OFDMA transmission will be wasted, thus resulting in under-utilization of the wireless medium.

[0182] The standard proposes that a buffer status report from an 802.11ax station can be utilized to support efficient MU-UL operation by an access point. To do so, upon receiving a trigger frame 430-BSR containing a request indication for a buffer status report, the 802.11ax station responds with a frame containing a queue size subfield 203 in its QoS control field 200. The indication for the buffer status report can be, for example, a "trigger type" provided in the trigger frame, with a particular value indicating such a buffer status request. The trigger frame 430-BSR is seen by the station as a trigger frame for a buffer status report (BSR).

[0183] Preferably, the trigger frame 430-BSR is broadcast by the AP to reach all stations in the BSS, and almost all of its resource units are random type resource units, which allows all stations to have a random chance to provide a queue size report. In addition, the maximum number of resource units is provided by the trigger frame 430-BSR to reach the maximum number of stations, i.e., the widest communication channel with the narrowest resource unit size is required.

[0184] To minimize the duration of the TXOP 490 to obtain a buffer status report, frames transmitted within the resource unit 410-BSR should be constrained and of the same size to avoid inefficient padding. For example, a QoS_Null frame seems appropriate to provide such a constraint. This particular QoS data frame contains a QoS control field with queue size information but no data payload.

[0185] Current versions of IEEE 802.11ax extend the use of queue size information 203 in a new QoS Control field, i.e., HE Control, potentially replacing the QoS Control field in the 802.11ax frame, to signal several, and preferably all, traffic queues 210 of a station, rather than just one as proposed in the 802.11e standard.

[0186] Once the access point has obtained the buffer reports for a set of stations in its BSS, it can specifically poll those stations through scheduled resource unit allocations, which are transmitted using a trigger frame 430-D for data transmission. Stations with allocated resource units can then send their buffered data in a longer TXOP_TF data The emitting stations emit in their assigned resource units 410-D during 491. Since MU UL / DL OFDMA transmissions in all resource units of a composite channel should be time aligned, stations can provide padding payload 411-D if there is no more data that can be transmitted in the assigned resource units. This can occur, for example, if there is no more data buffered for transmission or if the emitting station does not want to fragment any remaining data frames.

[0187] The access point can manage the resource unit size according to the needs of the reported stations. The access point can schedule a resource unit(s) during the TXOP to any of the stations that sent a report.

[0188] When a station transmits data to the access point using resource units 410-D, the access point responds with a multi-user acknowledgment 440 to acknowledge the data in each resource unit, which ends the granted TXOP period.

[0189] Due to polling by the AP of a given traffic queue and subsequent OFDMA transmission of data from that traffic queue, this traffic queue may switch to MU EDCA mode with the AIFSN set to the HEMUEDCATimer. Only other (non-transmitting) traffic queues may continue to access the communication channel using the EDCA method.

[0190] If the AIFSN value provided by the AP (e.g., in the EDCA parameter set information of the beacon frame) is 0, data of the access category for which EDCA access to the wireless network has been gained is not transmitted over the wireless network, thereby preventing traffic queues of stations in MU EDCA mode from accessing the medium.

[0191] In this example shown in the figure, station STA4 is polled by the AP for transmission of data from traffic queue AC_VI (slot 410-D4). Due to the switch to MU EDCA mode, traffic queue AC_VI can no longer request EDCA access unless the HEMUEDCA timer expires. However, station STA4 can still request EDCA access for other traffic queues (until the other traffic queues are served in the next MU UL transmission, if one occurs). This is because traffic queue AC_VO, which is in legacy EDCA mode, can actually access the communication channel using EDCA access and then empty itself via a single user transmission 495-SU (which is TXOP 492).

[0192] FIG. 5b provides further detail on the penalty mechanism described above.

[0193] In this figure, four values ​​530 represent the queue backoff counters BC[AC] (and their backoff values) associated with the four traffic queues 210. Graphical codes are used to distinguish the different states that the queue backoff counters can be in. The graphical codes are provided directly in the figure as a legend.

[0194] In the first phase shown (each phase corresponds to a period from when the network becomes available to the end of a granted TXOP), station 502 accesses the medium through EDCA when BC[VI] (the counter in the dashed box) has reached 0, but the best queue backoff value of AP 501 has only reached 4. The black portion 540 before decrementing the queue backoff value corresponds to AIFS[AC] (its different sizes are not shown).

[0195] Then, video data from AC[2] (i.e., traffic queue of AC_VI) is transmitted during the granted TXOP 550. The new queue backoff value of AC_VI is retrieved (white number in black box).

[0196] In the second phase, the AP 501 first accesses the network after counting down its AIFS 540 and its backoff value 710, and then transmits a trigger frame 1300. The TF 1300 serves at least one scheduled RU of the station 502.

[0197] The TF 1300 does not indicate a prioritized AC in the corresponding field (the prioritized AC level field 1330 shown in FIG. 13 set to 0), so the station 502 needs to determine which AC queue has the highest priority to select the corresponding data for MU UL OFDMA transmission.

[0198] To do this, station 502 selects the AC queue with the smallest current backoff counter value. In this example, the first AC queue (corresponding to the VO queue) is selected because its associated queue backoff value is equal to 1 (compared to 4, 6, and 12 for the other AC queues).

[0199] Thus, data from the AC_VO can be transmitted in the accessed resource unit (560).

[0200] After a successful MU UL OFDMA transmission 560 of data, the corresponding traffic queue enters MU EDCA mode, in which the EDCA parameter set for this transmission traffic queue, here AC_VO, is modified with the set of MU values. Traffic queues in MU EDCA mode are illustrated using bold boxes.

[0201] During this second phase, the queue backoff counter 531 associated with the transmit traffic queue AC_VO is frozen, which means that this queue backoff counter is not updated and maintains its previous value, here "1."

[0202] Thus, the third phase begins with station 502 entering a deferred transmission state by waiting 565 for the expiration of the AIFS[i] timer before decrementing queue backoff value BC[i] 530 .

[0203] For traffic queues in MU EDCA mode with more restrictive EDCA parameters, such as a very large AIFSN, the modified value of AIFS 565 causes the corresponding queue backoff value BC[AC] to be decremented less frequently. Therefore, when the AP 501 transmits a new trigger frame 1300-2 based on EDCA-based access to the medium, the station 502 to which the new scheduled RU is served determines which traffic queue has the smallest current queue backoff value.

[0204] This, in turn, is BC[3] with a backoff value of 1. This means that station 502 again transmits 570 data from the AC_VO in the accessed RU.

[0205] After some time, if some traffic queues enter MU EDCA mode, especially if all traffic queues enter MU EDCA mode, the associated queue backoff values ​​are blocked, so that MU UL OFDMA transmission priority is always given to the same traffic queue, and the backoff-based QoS requirements are no longer met.

[0206] The lack of dynamicity of the backoff counter due to its freezing in the case of MU UL OFDMA transmission should be restored so that the backoff counter still effectively mirrors the relative priority of the AC queues. Advantageously, this restoration should maintain the penalty scheme that reduces the EDCA-based transmission probability of AC queues in MU EDCA mode, while also maintaining the principle of gradual change of the backoff counter.

[0207] It is included in this framework that the present invention proposes, inter alia, not to penalize the AIFSN of the AC queue, while keeping the queue back-off counter decremented by introducing a reactive measure upon back-off counter expiry.

[0208] Recall that the queue backoff counter is decremented over time as long as the communication channel is sensed as idle by the stations for longer than the respective arbitration interframe space duration, and that any traffic queue is switched from the legacy contention mode to the MU contention mode upon transmitting data stored in that traffic queue in an accessed resource unit provided by the access point within a transmission opportunity granted to the access point on the communication channel.

[0209] Therefore, in order to efficiently restore the dynamics of the back-off counters while maintaining the penalty scheme, the present invention proposes that, when one of the queue back-off counters expires, a station decides, based on the current mode of the associated traffic queue, whether to access the communication channel to transmit data stored in the associated traffic queue, or to retrieve a new back-off value and reset the expired queue back-off counter without transmitting data from the associated traffic queue on the communication channel.

[0210] In effect, if the current mode is the legacy contention mode, the station accesses the communication channel to transmit data stored in the associated traffic queue, which is the legacy EDCA method. Otherwise, i.e., if the current mode is the MU contention mode, no data from the associated traffic queue is transmitted on the communication channel, a new backoff value is retrieved, and the expired queue backoff counter is reset.

[0211] Slight variations can be envisaged, such as making the decision to access the communication channel or to fetch a new back-off value also based on the data currently stored in the associated traffic queue, e.g., only if no data intended for non-AP stations is stored in the associated AC queue under MU EDCA mode, in order to apply the penalty of not transmitting data even if the back-off counter has expired.

[0212] By reverting the decrement of the backoff counters of non-transmitting AC queues in MU mode, the aging characteristics of the EDCA backoff counters are ensured to be restored, and therefore the QoS is ensured to be restored. Thus, the dual function of EDCA backoff is restored, allowing the station to apply relative QoS priorities among AC queues.

[0213] In addition, a reactive measure to reset an expired backoff counter without transmitting data in the case of MU mode ensures that penalties continue to be applied to AC queues in such MU mode, reducing EDCA accesses.

[0214] Thus, the present invention works inversely to known techniques that are essentially preventative measures to avoid allowing the backoff counter to reach zero in order to ensure reduced EDCA access.

[0215] As will become further apparent, the method of the present invention is easier to implement in a standard environment, especially in the transmit state machine of an 802.11 device.

[0216] The results of one embodiment of the present invention are now described with reference to Fig. 5c, which uses the same sequence as Fig. 5b to show the restoration of QoS through relative EDCA-based prioritization between ACs.

[0217] The first phase remains the same.

[0218] During the second phase, station 502 receives TF 1300 from AP 501. The AC queue selection algorithm at the station determines that traffic queue AC_VO has the highest priority due to the smallest queue backoff value (as in the example of FIG. 5b). MU UL OFDMA transmission is performed using data from AC_VO (AC[3]).

[0219] After successful transmission 560 of data in the accessed RU, the AC_VO enters MU EDCA mode (backoff value in bold box), and an embodiment of the present invention contemplates selecting a new backoff value 531 (here white number in black box with value "15") for the transmit traffic queue AC_VO from the current, unchanged associated contention window.

[0220] Also, AIFS[AC_VO] is not changed when entering MU EDCA mode, which means that the penalty due to MU mode is not implemented through the degradation of EDCA parameters.

[0221] Then, in a third phase, the backoff counters are decremented given their respective AIFS. Here, the backoff counter 531 of the AC_VO is decremented by one for each time slot in which the medium is sensed to be idle, even if the AC_VO is in MU mode. This is obtained for an unmodified (or in some embodiments, very slightly penalized) value AIFS[AC_VO].

[0222] Again, station 502 receives TF 1300-2 from AP 501. The AC queue selection algorithm at the station determines that traffic queue AC_VI has the highest priority due to the smallest queue backoff value (since BC[VO] now has a value of 14). Therefore, MU UL OFDMA transmission in the accessed RU is performed using data from AC_VO.

[0223] Comparing with Fig. 5b, it can be seen that another traffic queue is requested for MU UL OFDMA transmission in the third phase thanks to the new backoff value 532 retrieved to reset the AC_VO backoff counter, and thus the relative priority among the traffic queues is restored for the OFDMA transmission traffic queue.

[0224] After successful transmission 570 of data in the accessed RU, AC_VI also enters MU EDCA mode (backoff value in bold box; whereas AC_VO is already in MU EDCA mode), and a new backoff value 532 (here white number in black box with value "9") is taken out of the contention window for transmit traffic queue AC_VI.

[0225] Then, in the fourth phase, the medium is sensed as idle and the backoff counters are decremented given their respective AIFS. Here, even though two AC queues, AC_VO and AC_VI, are in MU mode, the backoff counter 531 of AC_VO goes from 14 to 7, while the backoff counter 532 of AC_VI goes from 9 to 2.

[0226] At the same time, the backoff counter of AC_BE expires, causing transmission of data from AC_BE 580. A new backoff value of "15" is retrieved for the expired backoff counter 533.

[0227] Then, in the fourth phase, the medium is sensed as idle and the backoff counters are decremented given their respective AIFS. Here, during the first two time slots after the SIFS period, AC_VO backoff counter 531 goes from 7 to 5, while AC_VI backoff counter 532 expires.

[0228] Since the traffic queue AC_VO is now in MU mode, a new backoff value ("8" in this example) is retrieved and the expired queue backoff counter 532 is reset without data from the associated traffic queue being transmitted on the communication channel. As a result, the backoff counters continue to decrement (i.e., do not stop decrementing), particularly the backoff counters of the APs that also expire. Thus, the station 502 can receive the new TF 1300-3 from the AP 501 and execute the AC queue selection algorithm at the station.

[0229] For example, the station determines that traffic queue AC_VO has the highest priority due to the smallest queue backoff value (since BC[VI] has been reset in the meantime), and therefore, a MU UL OFDMA transmission 590 in the accessed RU is performed using data from AC_VO.

[0230] However, in an embodiment, the reset of any AC queues that are in MU mode can be stored and this information can be used to select data to be transmitted. For example, a reset flag (not shown) can be associated with each traffic queue. This reset flag is enabled (i.e., set to TRUE) each time a new backoff value is retrieved and the associated queue backoff counter is reset without data from the traffic queue being transmitted (i.e., when the AC queue is in MU mode). This is the case, for example, for traffic queue AC_VI during the fifth phase (this traffic queue has been reset with a new backoff value of "8").

[0231] In that case, when the station receives TF 1300-3, it determines that traffic queue AC_VI has a reset flag set to TRUE. Thus, the MU UL OFDMA transmission in the accessed RU is performed using data from AC_VI (AC[2] as shown in the example of the figure). This means that the data to be transmitted in the resource units provided by the access point within the transmission opportunity granted to the access point is taken from at least one traffic queue selected based on the enabled or disabled status of the reset flag associated with the traffic queue. In particular, the data to be transmitted in the resource units 590 provided by the access point within the transmission opportunity granted to the access point is taken from the traffic queue with a enabled reset flag.

[0232] On the other hand, the reset flag is disabled (i.e., set to FALSE) every time data from the traffic queue is transmitted. In this example, the reset flag of AC_VI is reset after successful transmission 590.

[0233] According to the illustrated embodiment, a new backoff value (here a white number in a black box having a value "11") for the backoff counter 531 associated with the transmit traffic queue AC_VI is retrieved.

[0234] This exemplary scenario clearly shows that the full functional behavior of EDCA backoff, and therefore QoS, is restored; in particular, the dynamic relative EDCA-based priorities between traffic queues are restored.

[0235] 6 shows a schematic diagram of a communication device 600 of the wireless network 100 configured to implement at least one embodiment of the present invention. The communication device 600 may preferably be a device such as a microcomputer, a workstation or a lightweight portable device. The communication device 600 comprises a communication bus 613. To this communication bus, preferably, the following are connected: · A central processing unit 611, denoted by a CPU such as a microprocessor; · a read-only memory 607, denoted ROM, storing a computer program implementing the invention; a random access memory 612, denoted RAM, storing the executable code of the method according to an embodiment of the invention and registers adapted to record variables and parameters necessary for implementing the method according to an embodiment of the invention; and At least one communication interface 602 connected to a wireless communication network 100, for example a wireless communication network according to the 802.11ax protocol, through which digital data packets or digital data frames or control frames are transmitted. Frames are written from a FIFO transmit memory in RAM 612 to the transmit network interface, or read from the receive network interface and written into a FIFO receive memory in RAM 612, under the control of a software application running in CPU 611.

[0236] Optionally, the communication device 600 may also include the following components: · data storage means 604, such as a hard disk, storing a computer program for implementing the method according to one or more embodiments of the present invention; a disk drive 605 for a disk 606, adapted to read data from or write data onto the disk 606; A screen 609 that displays the decoded data and / or acts as a graphical interface for the user via a keyboard 610 or any other pointing means.

[0237] Communications device 600 may optionally be connected to various peripherals, such as, for example, a digital camera 608. Each peripheral is connected to an input / output card (not shown) to provide data to communications device 600.

[0238] Preferably, a communications bus provides communication and interoperability between the various elements included in or connected to communications device 600. The term bus is not limiting, in particular a central processing unit is operable to communicate instructions to any element of communications device 600 directly or by way of another element of communications device 600.

[0239] The disk 606 may optionally be replaced by any information medium, such as, for example, a rewritable or non-rewritable compact disk (CD-ROM), a ZIP disk, a USB key or a memory card, and generally by an information storage means that can be read by a microcomputer or microprocessor, that may or may not be integrated in the device, possibly removable, and that is adapted to store one or more programs that, when executed, make it possible to carry out the method according to an embodiment of the invention.

[0240] The executable code may optionally be stored either in a read-only memory 607, in the hard disk 604 or in a removable digital medium such as, for example, the disk 606 as previously described. According to an optional variant, the executable code of the program may be received via the interface 602 by means of the communications network 603 for storage in one of the storage means of the communications device 600, such as the hard disk 604, before being executed.

[0241] The central processing unit 611 is preferably adapted to control and direct the execution of instructions or parts of the software code of the program or programs according to the invention, which instructions are stored in one of the aforementioned storage means. On power-up, the program or programs stored in a non-volatile memory, for example the hard disk 604 or the read-only memory 607, are transferred into the random access memory 612, which then contains the executable code of the program or programs and registers for storing variables and parameters necessary to implement the invention.

[0242] In a preferred embodiment, the apparatus is a programmable apparatus that implements the invention using software, although the invention may alternatively be implemented in hardware (e.g. in the form of an application specific integrated circuit or ASIC).

[0243] 7 is a block diagram illustrating generally the architecture of a communications device or node 600 adapted to at least partially carry out the present invention, in particular one of the stations 100 to 107. As shown, the station 600 comprises a physical (PHY) layer block 703, a MAC layer block 702, and an application layer block 701.

[0244] The PHY layer block 703 (e.g., the standardized PHY layer of 802.11) has the task of formatting any 20 MHz channel or composite channel, modulating onto or demodulating from these channels, and therefore transmitting and receiving frames such as 802.11 frames over the used wireless medium 100. These frames are, for example, control frames (RTS / CTS / ACK / trigger frames) based on a 20 MHz width interacting with legacy 802.11 stations or 802.11ax stations in legacy mode, single user frames such as MAC data frames and management frames (e.g., for trigger frames), as well as OFDMA type MAC data frames, preferably with a width smaller than the 20 MHz legacy (typically 2 MHz or 5 MHz) to and from the wireless medium.

[0245] The MAC layer block or controller 702 preferably comprises a MAC 802.11 layer 704 which performs conventional 802.11ax MAC operations, and one additional block 705 which at least partially performs an embodiment of the present invention. The MAC layer block 702 may optionally be implemented in software, which is loaded into RAM 612 and executed by CPU 611.

[0246] Preferably, an additional block called EDCA medium access module 705 implements the portions of the invention related to station 600, namely, remembering that "MU mode" has been established for a given AC once it has been provided with a successful transmission in a resource unit, continuing to decrement the respective backoff counters of the AC(s) regardless of the contention mode of the associated traffic queues (and in particular regardless of whether their associated traffic queues are in MU mode), optionally remembering when the backoff counters in MU mode have been reset, and also preventing EDCA medium access when an AC backoff in MU mode is reduced to 0.

[0247] The MAC 802.11 layer 704 and the EDCA medium access module 705 interact with each other to provide management of dual modes (single-user EDCA and multi-user UL OFDMA) as described below.

[0248] At the top of the diagram, an application layer block 701 executes applications that generate and receive data packets, e.g. data packets of a video stream. The application layer block 701 represents all stack layers above the MAC layer according to ISO standardization.

[0249] Next, the embodiments of the present invention will be described using various exemplary embodiments. The proposed example uses a trigger frame 430 (see FIG. 4) transmitted by an AP for multi-user uplink transmission, but an equivalent mechanism can be used in a centralized or ad-hoc environment (i.e., without an AP). This means that the operations described below with respect to an AP can be performed by any station in an ad-hoc environment.

[0250] These embodiments are primarily described in the context of IEEE 802.11ax by considering OFDMA resource units, however the application of the present invention is not limited to that in conjunction with IEEE 802.11ax.

[0251] Also, the present invention does not necessarily rely on the use of the MU access scheme as described in 802.11ax, but any other RU access scheme that defines an alternative medium access scheme that allows simultaneous access by stations to the same medium may be used.

[0252] FIG. 8 illustrates an exemplary transmission block of a communication station 600 according to an embodiment of the present invention.

[0253] As mentioned above, the station comprises a channel access module and possibly a RU access module, both of which are implemented in the MAC layer block 702. The channel access module comprises: a plurality of traffic queues 210 serving data traffic of different priorities; A plurality of queue backoff engines 211, each associated with a respective traffic queue that uses a set of EDCA parameters, and in particular, that calculates a respective queue backoff value that is used by an associated backoff counter to contend for access to at least one communication channel for transmitting data stored in the respective traffic queue, which is an EDCA access scheme.

[0254] The RU access module includes an RU backoff engine 800 that is separate from the queue backoff engine. This RU backoff engine 800 uses the RU contention parameters to calculate an RU backoff value that is used by the RU backoff counter to contend for access to an OFDMA random resource unit specified in a received TF (e.g., transmitted by an AP) for transmitting data stored in any traffic queue in the OFDMA RU. The RU backoff engine 800 is associated with a transmission module called OFDMA muxer 801. For example, the OFDMA muxer 801 is responsible for selecting data to be transmitted from the AC queue 210 when the RU backoff value, described below, reaches 0.

[0255] The conventional AC queue backoff register 211 drives medium access requests along the EDCA protocol (channel contention access scheme), while in parallel the RU backoff engine 800 drives medium access requests over the OFDMA multi-user protocol (RU contention access scheme).

[0256] Since these two contention access schemes coexist, the stations implement a medium access mechanism with collision avoidance based on the calculation of the following backoff values: - A queue backoff counter value corresponding to the number of time slots a station waits (in addition to the AIFS period) after the communications medium is detected as idle before accessing the medium. This is EDCA whether in a degenerate or non-degenerate state. - RU backoff counter value corresponding to the number of idle random RUs the station detects after a TXOP is granted to the AP or any other station for the composite channel formed from the RU and before accessing the medium. This is OFDMA. One variation to the RU backoff counter countdown based on the number of idle random RUs can be based on a time-based countdown.

[0257] Figure 9 illustrates, by means of a flow chart, the main steps performed by the MAC layer 702 of a station 600 when it receives new data to send (an MSDU packet from a higher layer, e.g. the application layer 701). The figure illustrates the traditional FIFO feeding associated with 802.11.

[0258] Initially, at the start, none of the traffic queues 210 stores any data to transmit. As a result, no queue backoff value 211 is calculated. The corresponding queue backoff engine or the corresponding AC (Access Category) is said to be inactive. As soon as data is stored in a traffic queue, a queue backoff value is calculated (from the corresponding queue backoff parameter) and the associated queue backoff engine or AC is said to be active.

[0259] When a station has data ready to be transmitted on the medium, this data will be stored in one of the AC queues 210 and the associated backoff 211 will be updated.

[0260] More details are provided below.

[0261] In step 901, new data is received from an application running locally on the device (e.g., application layer 601), from another network interface, or from any other data source. This new data is prepared for transmission by the station.

[0262] In step 902, the station determines which AC queue 210 the data should be stored in. This is typically done by looking up the TID (traffic identifier) ​​value attached to the data (according to the lookup shown in FIG. 2b).

[0263] Next, step 903 stores the data in the determined AC queue, which means that the data is stored in an AC queue that has the same data type as the data.

[0264] In step 904, a conventional 802.11AC backoff calculation is performed by a queue backoff engine associated with the determined AC queue.

[0265] If the determined AC queue was empty immediately before the storage in step 903 (ie, the AC was initially inactive), a new queue backoff value for the corresponding backoff counter needs to be calculated.

[0266] Thus, the station calculates the queue backoff value to be equal to a random value selected in the range [0,CW], where CW is the current value of CW for the considered access category (as specified in the 802.11 standard and updated in accordance with some embodiments of the present invention, e.g., as described in step 1080 below). Recall that the queue backoff value is added to the AIFS to enforce the relative priority of the different access categories. CW is the random value selected in the range [CW min ,C.W. max] is the congestion window value selected from the boundary CW min and C.W. max depends on the access category under consideration.

[0267] As a result, the AC is activated.

[0268] The above parameters CW, CW min , C.W. max The AIFSN, AIFSN, and backoff values ​​form the EDCA contention parameters and variables associated with each AC. They are used to set the relative priority of accessing the medium for different data categories.

[0269] EDCA parameters (e.g., CW min , C.W. max , and AIFSN) usually have fixed values, while the EDCA variables (CW and backoff values) vary with time and medium availability. As can be easily seen from the above, thanks to the present invention, it is no longer necessary to degenerate the EDCA parameters when applying penalizing the MU mode. Of course, the EDCA parameters may still be changed, but there is no need to do so.

[0270] In addition, the station may support a UL RU OFDMA random access procedure (as described above). In that case, step 904 may include calculating an RU backoff value, if necessary. The RU backoff value needs to be calculated if the RU backoff engine 800 was inactive (e.g., because there was no data in the traffic queue until the previous step 903) and if new data destined for the AP is received.

[0271] The RU backoff value is determined in a similar manner to the EDCA backoff value, i.e., by the dedicated contention window [0,CWO] and the selection range [CWO min ,CWO max ] can be calculated using dedicated RU competition parameters such as

[0272] It should be noted that some embodiments may provide a distinction between data that can be transmitted over a resource unit (i.e., compatible with MU UL OFDMA transmission) and data that cannot be transmitted. Such a determination may be made during step 902, and a corresponding marking item may be added to the stored data.

[0273] In such a case, the RU backoff value is calculated only if the newly stored data is marked as compatible with a MU UL OFDMA transmission (either scheduled or random).

[0274] After step 904, the process of FIG.

[0275] Once the data is stored in the AC queue, the station can directly access the medium through the EDCA access scheme, as described below with reference to FIG. 10, or through resource units provided by the AP by one or more trigger frames, as described below with reference to FIG. 11.

[0276] 10 illustrates, by way of a flow chart, steps for accessing a medium based on a conventional EDCA medium access scheme for the purpose of handling both legacy and MU contention modes according to an embodiment of the present invention. In particular, the reactive countermeasure of the present invention is implemented in step 999 by a non-AP station 600 in this example.

[0277] For clarity, the steps of Figure 9 are also referenced since data stored in the AC[] queue is a prerequisite for temporary EDCA medium access.

[0278] Steps 1000-1020 show the conventional waiting introduced in the EDCA mechanism to reduce collisions on the shared wireless medium based on a backoff countdown. In step 1000, station 600 senses the medium and waits for it to become available (i.e., the detected energy falls below a given threshold on the primary channel).

[0279] Once the medium has been idle for an AIFS[i] period (at least a DIFS period, see FIG. 2d) plus one slot time, step 1010 is executed. In this step, station 600 begins decrementing all active (non-zero) AC[] queue backoff counters 211 by one. In other words, the station decrements the queue backoff value for each basic unit of time that the communication channel is detected as idle.

[0280] Next, in step 1020, station 600 determines whether at least one of the AC backoff counters has reached zero.

[0281] If the AC queue backoff has not reached 0, station 600 waits for another backoff time slot (typically 9 μs) and thus loops back to step 1000 to sense the medium again during the next backoff time slot. This allows the AC backoff counter to be decremented in each new backoff time slot in which the medium is sensed as idle as soon as their respective AIFS[i] expires.

[0282] If at least one AC queue backoff reaches 0 (which occurs three times in the scenario of FIG. 5c), the conventional step 1030 of initiating the EDCA transmission phase is not automatically reached.

[0283] In accordance with the present invention, the station will decide based on the current mode of the traffic queue associated with the expired queue backoff counter whether to access the communication channel (through the conventional EDCA process beginning at step 1030) to transmit the data stored in the associated traffic queue, or to retrieve a new backoff value and reset the expired queue backoff counter without transmitting data from the associated traffic queue on the communication channel (this is one improvement of the present invention that restores the dynamics of the backoff counter in MU mode).

[0284] This decision step is implemented in the proposed embodiment through step 1099, whereby the normal behavior of the EDCA state machine is modified for the purpose of applying an EDCA medium access penalty for non-AP stations 600 in MU mode.

[0285] Thus, a new test 1099 determines whether the traffic queue associated with the expired queue backoff counter is in penalty MU mode. The traffic queue can be flagged with "MU mode" set to 1 (performed by step 1140, described further below), so that the station 600 can easily know that the traffic queue is in penalty mode of EDCA medium access.

[0286] One main advantage of this approach is that it is still possible to reuse the hardware / state machine of the standard backoff decrement mechanism, specifically the basic mechanism that ultimately allows to initiate a medium access request once the backoff counter reaches 0. With the present invention, this request is simply conditional on the mode of use (MU mode on or off). The new test 1099 can be easily implemented in hardware.

[0287] If the determination at test 1099 is positive (the current mode is in MU mode), the station postpones any EDCA medium attempts, without transmitting any data over the EDCA access, and loops back to step 904 to retrieve a new backoff value by resetting the expired backoff counter. This step thus reintroduces EDCA backoff management regardless of "MU mode".

[0288] In case of a reset, a reset flag (introduced in the above description of the scenario of Fig. 5c) associated with the concerned traffic queue may be enabled to remember that this traffic queue had the highest QoS before the reset, and this flag is used to select the most priority data for OFDMA transmission in step 1030.

[0289] When check 1099 results in the re-obtaining of the backoff value without an EDCA transmission, the EDCA parameter / variable CW min , C.W. max Note that there is no need to change CW, , or Current CW, because nothing new is known about the state of the EDCA medium since no EDCA transmissions are allowed (essentially those values ​​change according to the correct behavior of EDCA transmissions), so any update of those values ​​is pointless and wasteful.

[0290] Optionally, these EDCA parameters / variables may be updated only when a significant amount of time (e.g., at least several times the HEMUEDCATimer value 1425) has been spent by the corresponding traffic queue in "MU mode". This is because these frozen values ​​established so far no longer mirror the actual network conditions. Thus, the EDCA mode may start from standardized initial values ​​as if this were an initial (first) transmission using EDCA. In other words, the station may reset the contention parameter set associated with a traffic queue (preferably each traffic queue) that has remained in MU contention mode for at least the parameter lifetime duration to a default parameter set. For example, the parameter lifetime duration may correspond to at least twice (or more) the predefined duration (i.e., the value of the HEMUEDCATimer) used to initialize the MU mode timer.

[0291] On the other hand, if the determination in step 1099 is negative (the current mode is in the legacy EDCA mode), step 1030 is executed, in which station 600 (more precisely, virtual collision handler 212) selects the active AC queue with the highest priority that has a queue backoff counter of 0. This corresponds to the first and fourth phases of FIG. 5c.

[0292] In step 1040, an appropriate amount of data is selected from the selected AC for transmission.

[0293] Next, in step 1050, station 600 begins EDCA transmission, e.g., if an EDCA TXOP is granted upon successful execution of an RTS / CTS exchange. Station 600 then transmits selected data over the medium during the granted EDCA TXOP.

[0294] Next, in step 1060, station 600 determines whether the EDCA transmission is finished, and if so, step 1070 is executed.

[0295] In step 1070, station 600 updates the EDCA contention window CW of the selected traffic queue based on the status of the transmission (positive or negative ack, or no ack received). Typically, station 600 updates the EDCA contention window CW of the selected traffic queue based on the status of the transmission (positive or negative ack, or no ack received). Typically, station 600 updates the EDCA contention window CW by a maximum value that depends on the AC type of the data. max On the other hand, if an EDCA transmission is successful, the contention window CW is doubled until it reaches a minimum value CW, which also depends on the AC type of the data. min is set to.

[0296] Once again, thanks to this invention, CW min and C.W. max Recall that the CW can be the same in both legacy and MU contention modes. min and C.W. max is not degraded during step 1170 described below. Of course, the present invention does not prohibit these values ​​from being degraded when switching to MU mode (step 1170), for example based on the "MU EDCA parameter set" element 1420 received in a management frame (usually a beacon frame) issued by the AP. In that case, step 1070 determines whether the degraded CW min and C.W. max is carried out with respect to

[0297] Note that if the data to be transmitted comes from a traffic queue whose reset flag is enabled, then this flag is disabled, and in effect the new backoff value retrieved mirrors the new relative priority.

[0298] Then, if the selected traffic queue is not empty after transmitting the EDCA data, a new associated queue backoff counter is randomly selected from [0, CW] through a loop back to step 904. This means that after the data stored in the associated traffic queue is transmitted on the accessed communication channel, the station retrieves a new backoff value and resets the expired queue backoff counter.

[0299] This completes the process of FIG.

[0300] In a slight variation of the presented check 1099 already introduced above, the check 1099 deciding whether to access the communication channel or to fetch a new back-off value can also be based on the data currently stored in the traffic queue associated with the expired traffic queue, in order to adjust the penalty scheme for some types of data, especially with the intention of maintaining QoS fairness for data not related to the MU UL transmission (which triggered the penalty scheme).

[0301] We focus on peer-to-peer (P2P, ie, station-to-station) data transmission between non-AP stations.

[0302] The Direct Link Setup (DLS) protocol defined in the 802.11e standard allows direct point-to-point frame transfer within a basic service set.

[0303] The 802.11z standard subsequently introduced Tunneled Direct Link Setup (TDLS), which allows devices to perform more efficient direct station-to-station frame forwarding without support from an access point. The Wi-Fi Alliance added a certification program for TDLS in 2012, describing this feature as a technology that allows stations to link directly with each other when connecting to a traditional infrastructure network.

[0304] More generally, transmissions to an Independent BSS (IBSS) station (ie, the destination is not registered with any BSS) can be considered as P2P communication.

[0305] Both DLS and TDLS require stations to be associated with the same access point. As a result, communications within a P2P group can be viewed as occurring in parallel with communications in the infrastructure network (which includes access point 110). That is, stations simultaneously involved in P2P communications and the BSS network have their transmit queue(s) 210 serviced with data from both traffic modes.

[0306] As a result, the penalty scheme as envisaged by the 802.11ax standard penalizes uplink traffic to the AP but blocks any legacy EDCA access for P2P communication.

[0307] To address this situation, the test 1099 may be slightly adapted as suggested above.

[0308] In particular, according to the adapted check 1099, if the current mode is a legacy contention mode or if the data stored in the associated traffic queue contains data addressing another station other than the access point (i.e., the data is P2P data), the process proceeds to step 1030, through which the station accesses the communication channel and transmits the data stored in the associated traffic queue.

[0309] Conversely, if the current mode is MU contention mode and the data stored in the associated traffic queue does not contain data addressing another station other than the access point (i.e., the data is P2P data), the process loops back to step 904. This means that no data from the associated traffic queue is transmitted on the communication channel, a new backoff value is retrieved, and the expired queue backoff counter is reset.

[0310] In other words, test 1099 is amended as follows: - The test result is positive (yes) if (MU mode is active) AND (AC queues with expired backoffs do not hold P2P traffic); Otherwise, the test result is false and the process proceeds to step 930.

[0311] If the result of the check 1099 is false due to the AC queue being in MU mode but P2P data being present (leading to accessing the communication channel and transmitting data), it seems appropriate to give priority to the transmission of such P2P data as opposed to data intended for the AP that can be transmitted in the next accessed RU. This means that during the data selection step 1040, the P2P criterion is implemented, i.e. in case of access to the communication channel, only data stored in the associated traffic queue in MU contention mode and addressing another station different from the access point is transmitted in the accessed communication channel.

[0312] A station can implement various methods to distinguish between P2P data and UL data intended for the AP. For example, the distinction between the two types of data can be made using the destination address (receiver address or "Address 1" field of the MAC data frame for Fig. 2d). That is, a destination address set to the AP identifies UL data, and any other destination address identifies P2P data. Alternatively, pending MAC frames transmitted in the direct path to a (T)DLS peer station or a station in the IBSS can be assigned an UPLINK_FLAG parameter set to 0. When this UPLINK_FLAG parameter is set to 1, it indicates that the data is in the uplink direction towards the AP only.

[0313] Thanks to this variant of the invention, P2P traffic can still be emitted by the stations.

[0314] Figure 11 illustrates, by means of a flow chart, the steps of accessing resource units based on the RU access method (random RU or scheduled RU) upon receiving a trigger frame 430 defining the RU, for example, this illustrates the behavior of station 502 in phase 2 or phase 3 or phase 5 of Figure 5b.

[0315] In step 1110, the station determines whether a trigger frame 430 has been received from an access point in the communication network, which reserves transmission opportunities granted to the access point on a communication channel and defines resource units RU forming the communication channel. If so, the station analyzes the contents of the received trigger frame.

[0316] In step 1120, the station determines whether it can transmit data via one of the RUs specified in the received trigger frame. This determination may involve one or both of two conditions, particularly regarding the type of RU.

[0317] By analyzing the contents of the received TF, the station determines whether the specified RU is a scheduled resource unit assigned to the station by the access point. This can be done by looking for its own AID in the received TF, which will be associated with the particular scheduled RU used for transmission.

[0318] Also, by analyzing the contents of the received TF, the station determines whether one or more random RUs, i.e., RUs whose access is through contention with dedicated RU contention parameters (including the above-mentioned RU backoff value 800), are defined in the TF. If so, the station also determines whether it is possible to select one random RU with its current RU backoff value (especially if the RU backoff value 800 is smaller than the number of available random RUs in the current TF).

[0319] If one scheduled RU is assigned to the station, i.e., the station is allowed to access one random RU, then the station determines the size of the random / scheduled RU or RUs to be used and step 1130 is executed. Otherwise, the station decrements the RU backoff value 800 based on the number of random resource units specified in the received trigger frame, and the process ends since the station cannot access the RU (scheduled or random) specified by the received TF.

[0320] In step 1130, the station selects at least one of the traffic queues 210 from which data will be selected for transmission and adds data from the selected queue or queues to a transmission buffer until the amount of data reaches the size of the selected resource unit to be used.

[0321] By restoring the countdown of the EDCA backoff counter according to the present invention, and thus restoring the dynamics of the respective backoff values, the backoff values ​​can be used as a means to apply priority selection. For example, this can be done by selecting the traffic queue(s) 210 with the smallest associated queue backoff value. The selection of the traffic queue thus depends on the value of the EDCA backoff 211 (this method ensures that the station adheres to the EDCA principles, and also ensures the correct QoS for its data).

[0322] In one variant, the AP may indicate a preferred AC in the parameters of the trigger frame. Thus, the station may instead select a non-empty traffic queue associated with a data type that matches the data type associated with the resource unit in which the selected data is to be transmitted. Such a specified data type may be, for example, a traffic queue indicated by the AP in the trigger frame using the preferred AC field 1340 of FIG. 13 when the AC priority level field is set to 1.

[0323] In this variant, data to be transmitted in resource units provided by the access point within a transmission opportunity granted to the access point is taken from a priority traffic queue indicated by the access point, and an indication of this priority traffic queue is included in a trigger frame received from the access point, which reserves the transmission opportunity granted to the access point on a communication channel and defines a resource unit RU forming the communication channel including the accessed resource unit.

[0324] In another variation based on the reset flag introduced above (when describing the scenario of FIG. 5c or when looping back to step 904 from a new test 1099), the station may preferably select data from the traffic queue that has been reset in the meantime (i.e., there has been no transmission from the traffic queue since the last OFDMA transmission) in order to give priority to these data that have the highest priority at a given time due to their back-off values ​​before their relative priorities are slightly degraded by the back-off counter reset.

[0325] When data is selected from a traffic queue that has the reset flag enabled, the reset flag will be disabled such that the relative priority of that traffic queue is given only by the associated backoff value.

[0326] Also note that if P2P traffic is being handled, step 1130 may be limited to selecting UL data, ie, data intended for the AP.

[0327] The station may then store the selected traffic queue as the transmit traffic queue in step 1140. For example, the first traffic queue selected (i.e., the first time step 1130 is performed) may be stored as the primary traffic queue and any other selected traffic queues may be stored as secondary traffic queues.

[0328] In step 1150, the station determines whether the amount of data stored in the transmit buffer is sufficient to fill the selected resource unit.

[0329] If not, there is still room in the resource unit for additional data to serve another AC queue. A second AC queue (called a secondary AC) can be determined using the same criteria as defined above for the remaining traffic queues. The process then loops back to step 1130, during which another traffic queue is selected. In such a way, the transmission buffer is filled progressively until it reaches the size of the selected resource unit.

[0330] Thus, it can be seen that multiple transmit traffic queues of the same station can be involved during a MU UL OFDMA transmission, thereby causing multiple queues to enter MU EDCA mode.

[0331] In one variation that avoids mixing of data from two or more traffic queues (i.e., data for a selected RU is selected from a single traffic queue), padding data can be added to completely fill the selected RU to ensure that the entire duration of the RU has energy that can be detected by legacy stations.

[0332] According to another variation for implementing a particular data aggregation rule, if the initially selected traffic queue does not have enough data to completely fill the accessed resource unit, data from a higher priority traffic queue may be selected.

[0333] Once the transmit buffer of the selected RU is filled, step 1160 initiates MU UL OFDMA transmission of the data stored in the transmit buffer to the AP based on the OFDMA subchannel and modulation specified in the received trigger frame, specifically the RU definition.

[0334] The transmission is then performed and, preferably, if the transmission is successful (i.e., an acknowledgement is received from the AP), the transmit queue identified in step 1140 enters MU mode. One or more transmit queues may already be in MU mode.

[0335] The MU mode timer can be initialized with the HEMUEDCATimer, and the MU mode timer runs continuously as time progresses. Note that if the MU mode timer has already run out (meaning the station was already in MU EDCA mode) when the new transmission 1160 is successfully completed, the MU mode timer is reinitialized to the HEMUEDCATimer again to keep the station in MU EDCA mode for the next HEMUEDCATimer period.

[0336] Thus, where appropriate, step 1170 may be performed to determine new value(s) for modifying the value(s) applied to one or more EDCA parameters of one or more traffic queues to one or more penalty values.

[0337] As mentioned above, the penalty scheme is now implemented entirely through new step 1099 in the EDCA state machine in this embodiment, so step 1170 is now optional (compared to the prior art document).

[0338] Not changing the EDCA parameters means that the MU contention mode uses the same arbitration interframe space duration as the legacy contention mode, and the MU contention mode uses the same lower bound CW min and / or the same upper boundary CW max Both of these bounds define a range of choices within which the size of the contention window is chosen.

[0339] By using the same value, the bandwidth used by the AP to transmit the parameters in the beacon frame is advantageously reduced. In fact, the "MU EDCA Parameter Set" element 1420 can be significantly reduced to only report the HEMU EDCA Timer value. In other words, fields 1421-1424 can be removed from 1420.

[0340] In addition, since the EDCA backoff values ​​are now changing over time again, periodic polling of some traffic queues is performed, and thus a single MU mode timer is sufficient to efficiently drive the station out of MU mode.

[0341] Next, optional step 1180 is performed in which the QoS characteristics supported by the backoff counter are also restored, so that the backoff values ​​are recalculated more frequently.

[0342] In this optional step, a new EDCA backoff value may be retrieved for each transmit traffic queue even if the current backoff value of the transmit traffic queue has not been decreased to 0, in order to maintain relative priority among the traffic queues (the retrieved value is likely to be a larger value with respect to other traffic queues, such as pending traffic queues, which are therefore prioritized for the next transmission).

[0343] For example, the station calculates or retrieves new backoff values ​​for at least one traffic queue transmitting in the accessed resource unit and resets the associated queue backoff counters, which is the case for backoff counters 531 and 532 at the end of the second and third phases of the scenario of FIG.

[0344] In a first embodiment, the new queue backoff value is calculated only for the transmit traffic queues for which data is being transmitted at the beginning of the accessed resource unit. Preferably, only the primary traffic queues identified in step 1140 are involved in the reset with the new queue backoff value.

[0345] In a second embodiment, a new queue backoff value is calculated for each transmission traffic queue identified in step 1140 .

[0346] It can be seen that even if some backoff values ​​are re-fetched (and thus the corresponding backoff counters are reset), the situation continues where these values ​​are again decremented, thus providing a new opportunity to reach 0, in which case the test 1099 is performed again to operate according to the current contention mode.

[0347] Although the above embodiment of the present invention provides a penalty scheme in the case of using both scheduled and random access in UL MU resource units, it can be envisaged to only apply the penalty scheme in response to successful transmissions in scheduled RUs. This is motivated in the sense that penalizing MU EDCA parameters (1150) is applied with respect to the AP's behavior (if the AP actually decides to grant (schedule) UL access to a given station). In the case of randomly accessed RUs, the AP does not specify a particular station (i.e., does not explicitly indicate that it is penalizing one station in particular), so the penalty is not considered to be applied to the EDCA access mode.

[0348] Figure 12 shows, by means of a flow chart, the station management for switching to and from the legacy EDCA mode in the above example. This management is based on the HEMUEDCA timer described above. In fact, the station will stay in the MU EDCA mode as long as this MU mode timer has not elapsed.

[0349] Thus, in step 1210 it is checked whether the HEMUEDCATimer has elapsed / expired, i.e. whether it has reached the value 0.

[0350] If so, the station switches back to EDCA mode in step 1220, for example, by setting the "MU mode" flag to 0 for all traffic queues.

[0351] In this embodiment, all degenerate traffic queues share the same predefined degenerate duration HEMUEDCATimer to exit the degenerate MU EDCA mode at the same time, which may mean restoring legacy EDCA parameters if the MU mode involved degenerate EDCA parameters.

[0352] Note that due to the reinitialization of the MU mode timer upon each new successful MU UL OFDMA transmission by the station, expiration of the MU mode timer occurs only when no data is transmitted from the station in the RU provided by the AP within a subsequent TXOP granted to the AP for a predefined degenerate duration HEMUEDCATimer.

[0353] The process then ends in step 1230.

[0354] Figure 13 shows the structure of the trigger frame as specified in the 802.11ax draft standard.

[0355] The trigger frame 1300 consists of a dedicated field called the "User Info" field 1310. This field contains a "Trigger dependent Common info" field 1320, which contains an "AC Preference Level" field 1330 and a "Preferred AC" field 1340.

[0356] The Priority AC field 1340 is a 2-bit field (values ​​0-3) that indicates the AC queue in which data should be transmitted by a station in the RUs assigned to that station in the trigger frame.

[0357] The AC priority level field 1330 is a bit that indicates whether the value of the priority AC field 1340 is meaningful. If the field 1330 is set to 1, the station should consider the priority AC field 1340 when selecting data in step 1130. If the field 1330 is set to 0, the station is allowed to transmit data from any AC queue, regardless of the value of the priority AC field 1340.

[0358] Other fields of the trigger frame are specified in the 802.11ax standard.

[0359] The AP can also be responsible for broadcasting EDCA parameters for both EDCA and MU modes (if they have degenerate parameter values). The AP preferably performs the broadcast using a well-known beacon frame dedicated to configuring all stations in the 802.11 cell. Note that if the AP is unable to broadcast the EDCA parameters, the stations are configured to fall back to default settings as specified in the 802.11ax standard.

[0360] FIG. 14a shows the structure of a standardized information element 1410 used to describe legacy EDCA parameters in a beacon frame.

[0361] Fields 1411, 1412, 1413, 1414 describe the parameters associated with each traffic queue 210. For each traffic queue, subfield 1415 describes the EDCA parameters, namely, AIFSN as the delay before starting to decrease the associated backoff value, and minimum CW min and maximum CW max It includes ECWmin and ECWmax as the contention window values, and finally the TXOP limit as the maximum transmission data time for an 802.11 device.

[0362] All other fields of the information element are as described in the 802.11 standard.

[0363] This standardized information element 1410 is used by a station to configure itself into legacy EDCA mode.

[0364] 14b shows an example structure of a dedicated information element 1420 for transmitting parameter values ​​for MU mode according to the present invention, including possible degenerate EDCA parameters (if present) and a HEMU EDCA timer value (always present). The dedicated information element 1420 can be included in a beacon frame transmitted by an AP.

[0365] A dedicated information element 1420 contains, for each AC queue, the reduced EDCA parameters (1421, 1422, 1423, 1424) used by stations in MU mode. This dedicated information element also contains a subfield 1425 that specifies the value of the HEMUEDCATimer.

[0366] Each subfield 1421, 1422, 1423, 1424 contains the reduced AIFSN value (if present) and the reduced ECWmin and ECWmax values ​​(which may be the same as the legacy EDCA values) for the corresponding traffic queue. A value of 0 in the AIFSN field indicates that the AIFS is equal to the HEMU EDCA Timer value set in the MU EDCA Timer subfield 1425.

[0367] Of course, if the MU mode uses legacy EDCA parameters, fields 1421, 1422, 1423, and 1424 can be omitted, thereby reducing bandwidth usage.

[0368] The MU EDCA timer subfield 1425 indicates the HEMU EDCA timer value in units of 8 TU (a time unit is a length of time equal to 1024 μs).

[0369] In this example, the set of non-degenerate values ​​and the set of degenerate values ​​(if any) as well as the HEMUEDCATimer value are transmitted in a beacon frame, which is transmitted periodically by the access point to broadcast network information about the communication network to multiple stations. In a variant, they can be included in a Probe Response frame or a (re)association Response frame.

[0370] Although the present invention has been described above with reference to specific embodiments, the present invention is not limited to these specific embodiments, and modifications that are within the scope of the present invention will be apparent to those skilled in the art.

[0371] With reference to the above exemplary embodiments, many further modifications and variations will occur to those skilled in the art. It should be noted that the above exemplary embodiments are given by way of example only and are not intended to limit the scope of the present invention, which is defined only by the appended claims. In particular, different features from different embodiments can be interchanged as necessary.

[0372] In the claims, the term "comprises" does not exclude other elements or steps and an unspecified number includes the singular and the plural. The mere fact that different features are recited in mutually different dependent claims does not indicate that a combination of these features cannot be used to advantage.

Claims

1. A base station for constructing a wireless network conforming to the IEEE 802.11 series standard, A transmitting means for transmitting a trigger frame conforming to the IEEE 802.11 series standard, the trigger frame including resource unit allocation information for performing communication by OFDMA (orthogonal frequency division multiple access); a first receiving means for receiving data from a terminal station in the resource unit; having When the resource unit allocation information includes allocation information for a first terminal station, the first terminal station that has received the trigger frame is controlled so that, in response to successful transmission of data of a predetermined access category in the resource unit, an MU EDCA mode is enabled for a predetermined period in transmission of data of the predetermined access category addressed to the base station, and a legacy EDCA mode is enabled in transmission of data of the predetermined access category addressed to a second terminal station different from the base station. A base station comprising:

2. 2. The base station of claim 1, wherein data transmissions of the predetermined access category destined for the second terminal station different from the base station are peer-to-peer traffic different from uplink traffic destined for the base station.

3. 2. The base station according to claim 1, wherein a peer-to-peer group for data transmission of the predetermined access category destined for the second terminal station different from the base station is maintained in parallel with a network established by the base station.

4. 2. The base station of claim 1, wherein the first terminal station and the second terminal station are associated with the same base station.

5. The base station according to claim 1 , wherein the predetermined period is specified by a HEMUEDCATimer.

6. The base station according to claim 1, characterized in that, when the predetermined period has elapsed, the first terminal station is controlled to enable transmission of data of the predetermined access category addressed to the base station in the legacy EDCA mode.

7. 2. The base station according to claim 1, further comprising: a second receiving means for receiving, in the legacy EDCA mode, data of a predetermined access category transmitted using a contention window determined based on a lower limit value and an upper limit value of the contention window.

8. 2. The base station according to claim 1, wherein even when the first terminal station is controlled to operate in MU EDCA mode for the predetermined access category, the first terminal station is not controlled to operate in MU EDCA mode for data transmission of an access category different from the predetermined access category.

9. 2. The base station according to claim 1, wherein the predetermined access category is any one of voice (AC_VO), video (AC_VI), best effort (AC_BE), and background (AC_BG).

10. A base station for constructing a wireless network conforming to the IEEE 802.11 series standard, A transmitting means for transmitting a signal including information regarding an Arbitration Inter-Frame Space Number (AIFSN) value; A first receiving means for receiving data of a predetermined access category from a first terminal station in a resource unit allocated by the base station for communication by OFDMA (Orthogonal Frequency Division Multiple Access); having When the AIFSN value included in the signal transmitted by the transmitting means is 0, a first terminal station that has received the signal transmitted by the transmitting means is controlled so that, in response to successful transmission of the data of the predetermined access category in the resource unit, EDCA transmission of the data of the predetermined access category addressed to the base station is disabled for a predetermined period of time, and EDCA transmission of the data of the predetermined access category addressed to a second terminal station different from the base station is enabled. A base station comprising:

11. 11. The base station of claim 10, wherein the EDCA transmission of the data destined for the second terminal station is peer-to-peer traffic distinct from uplink traffic destined for the base station.

12. 11. The base station of claim 10, wherein the EDCA transmission of the data destined for the second terminal station is performed without support of the base station.

13. 11. The base station of claim 10, wherein a peer-to-peer group for EDCA transmission of the data destined for the second terminal station is maintained in parallel with a network established by the base station.

14. 11. The base station of claim 10, wherein the first terminal station and the second terminal station are associated with the same base station.

15. The base station according to claim 10, wherein when the AIFSN value is 0, the first terminal station is controlled not to perform EDCA transmission for a period specified by a HEMUEDCATimer value.

16. The base station according to claim 15, characterized in that, when the first terminal station is switched to the MU contention mode, the first terminal station is controlled to return data transmission of the access category to the legacy contention mode in response to an MU mode timer having elapsed.

17. 11. The base station according to claim 10, wherein the predetermined access category is any one of voice (AC_VO), video (AC_VI), best effort (AC_BE), and background (AC_BG).

18. A communication method for a base station that constructs a wireless network conforming to the IEEE 802.11 series standard, comprising: A transmission step of transmitting a trigger frame conforming to the IEEE 802.11 series standard, the trigger frame including resource unit allocation information for performing communication by OFDMA (orthogonal frequency division multiple access); a first receiving step of receiving data from a terminal station in the resource unit; having When the resource unit allocation information includes allocation information for a first terminal station, the first terminal station that has received the trigger frame is controlled so that, in response to successful transmission of data of a predetermined access category in the resource unit, an MU EDCA mode is enabled for a predetermined period in transmission of data of the predetermined access category addressed to the base station, and a legacy EDCA mode is enabled in transmission of data of the predetermined access category addressed to a second terminal station different from the base station. A communication method comprising:

19. A program for causing a computer to execute the communication method according to claim 18.

20. A communication method for a base station that constructs a wireless network conforming to the IEEE 802.11 series standard, comprising: a transmitting step of transmitting a signal including information regarding an Arbitration Inter-Frame Space Number (AIFSN) value; A first receiving step of receiving data of a predetermined access category from a first terminal station in a resource unit allocated by the base station for communication by OFDMA (Orthogonal Frequency Division Multiple Access); having When the AIFSN value included in the signal transmitted by the transmitting step is 0, a first terminal station that receives the signal transmitted by the transmitting step is controlled so that, in response to successful transmission of the data of the predetermined access category in the resource unit, EDCA transmission of the data of the predetermined access category addressed to the base station is disabled for a predetermined period of time, and EDCA transmission of the data of the predetermined access category addressed to a second terminal station different from the base station is enabled. A communication method comprising:

21. A program for causing a computer to execute the communication method according to claim 20.

Images