Communication device and method therein for determining contention window size in a communication network

Abstract

The disclosed embodiments relate to a method performed by a device (1200) and also to a device (1200). The method comprises transmitting (1101), to a communication apparatus, a burst containing a subframe; receiving (1102), from the communication apparatus, a HARQ value associated with the subframe; and determining (1103) a random back-off contention window size based on the received HARQ value and also based on previously unused HARQ feedback. The device (1200) is configured to perform the method steps as described above.

Description

Technical Field

[0001] The proposed embodiments relate to channel access in a licensed assisted access system in which spectrum is shared. Specifically, they relate to a method and a first communication device for determining / adapting the size of a random backoff contention window in a communication network. Background Technology

[0002] The 3GPP initiative, known as “Licensed Assisted Access” (LAA), has been proposed to allow Long Term Evolution (LTE) devices, such as User Equipment (UE) or eNB base stations, to operate in the unlicensed 5 GHz radio spectrum. The unlicensed 5 GHz spectrum is used as a supplement to the licensed spectrum. Accordingly, devices use a primary cell (PCell) for connectivity in the licensed spectrum and use carrier aggregation (CA) to utilize one or more secondary cells (SCells) to benefit from additional transmission capacity in the unlicensed spectrum. To reduce the changes required for aggregating licensed and unlicensed spectrum, LTE frame timing in the PCell is simultaneously used in the SCell.

[0003] However, regulations require that transmission in unlicensed spectrum not be permitted without prior channel awareness. Since unlicensed spectrum must be shared with other radios using similar or dissimilar wireless technologies, a so-called Listen-Before-Speak (LBT) approach is applied. Currently, the unlicensed 5 GHz spectrum is primarily used by devices implementing the IEEE 802.11 Wireless Local Area Network (WLAN) standard, also known as "Wi-Fi".

[0004] Regulations can vary by region. For example, in Europe, LBT procedures are governed by the so-called Harmonized European Standards (EN) regulations, also known as EN 301.893, developed by the European Telecommunications Standards Institute (ETSI). For a LAA to operate in the 5 GHz spectrum, the LAA LBT procedure must comply with the requirements and minimum behaviors stated in EN 301.893. However, additional system design and procedures are required to ensure the coexistence of LAAs and Wi-Fi using the EN 301.893 LBT procedure.

[0005] Below, a general description of the technologies involved in LAA is presented, including systems with spectrum licensed therein for LTE and LBT protocols employing, for example, WiFi or WLAN, in order to understand the context of the embodiments herein.

[0006] LTE uses OFDM (Orthogonal Frequency Division Multiplexing) in the downlink and single-carrier FDMA (Frequency Division Multiple Access) in the uplink. Basic LTE downlink physical resources can be considered as follows: Figure 1The time-frequency grid shown is illustrated, where each resource element corresponds to one OFDM subcarrier during one OFDM symbol interval. Uplink subframes have the same subcarrier spacing as the downlink and the same number of (single-carrier) SC-FDMA symbols in the time domain as the OFDM symbols in the downlink. OFDM symbols are also shown as including a cyclic prefix (CP) and a 15 kHz inter-subcarrier spacing. Resource elements are also indicated.

[0007] In the time domain, LTE downlink transmissions are organized into 10 ms radio frames, each consisting of 10 equally sized subframes of length T = 1 ms, such as... Figure 2 As shown in the diagram. For a normal CP, a subframe consists of 14 OFDM symbols. The duration of each symbol is approximately 71.4 μs.

[0008] Furthermore, resource allocation in LTE is described based on resource blocks, where a resource block corresponds to a time slot (0.5 ms) in the time domain and 12 adjacent subcarriers in the frequency domain. A pair of adjacent resource blocks (1.0 ms) in the time direction is called a resource block pair. Resource blocks are numbered starting from 0 at one end of the system bandwidth in the frequency domain.

[0009] Downlink transmissions are dynamically scheduled; that is, in each subframe, the base station or eNB transmits control information informing which terminals (or UEs) the data is being transmitted to in the current downlink subframe and on which radio downlink resource blocks the data is being transmitted. This control signaling is typically transmitted in the first 1, 2, 3, or 4 OFDM symbols of each subframe, and the number n=1, 2, 3, or 4 is called the Control Format Indicator (CFI). Downlink subframes also contain common reference symbols known to the receiver and used for coherent demodulation of, for example, control information. Figure 3 The diagram shows a downlink system with CFI=3 OFDM symbols serving as control (control area).

[0010] Figure 3 The reference symbols shown are called cell-specific reference symbols (CRS) and are used to support multiple functions, including channel estimation for certain transmission modes and good time and frequency synchronization.

[0011] In LTE, there are channels called Physical Dedicated Control Channel (PDCCH) and Enhanced PDCCH (EPDCCH).

[0012] Both PDCCH and EPDCCH are used to carry downlink control information (DCI), such as scheduling decisions and power control commands. DCI includes:

[0013] - Downlink scheduling assignment includes physical downlink shared channel (PDSCH) resource indication, transmission format, hybrid-ARQ information, and control information related to spatial multiplexing (if applicable). The downlink scheduling assignment also includes commands for power control of the physical uplink control channel (PUCCH) used to transmit hybrid-ARQ acknowledgments in response to the downlink scheduling assignment.

[0014] - Uplink scheduling permission includes information such as Physical Uplink Shared Channel (PUSCH) resource indication, transmission format, and hybrid-ARQ related information. Uplink scheduling permission also includes commands for power control of the PUSCH.

[0015] - Power control commands for a set of UEs, as a supplement to the commands included in the scheduling assignment / grant.

[0016] The PDCCH / EPDCCH carries a DCI message containing one of the groups of information listed above. Since multiple UEs can be scheduled simultaneously, and each UE can be scheduled on both the downlink and uplink simultaneously, multiple scheduling messages may be transmitted within each subframe. Each scheduling message is transmitted on separate PDCCH / EPDCCH resources, and therefore there are typically multiple simultaneous PDCCH / EPDCCH transmissions within each subframe in each cell. Furthermore, to support different radio channel conditions, link adaptation can be used, where the coding rate of the PDCCH / EPDCCH is selected to match the radio channel conditions by adapting resource usage to the PDCCH / EPDCCH.

[0017] Furthermore, in LTE systems, the network uses the PDCCH to notify the UE of downlink data transmission. When the PDCCH is received in subframe n, the UE ( Figure 4 The receiver in the middle is required to decode the corresponding Physical Downlink Shared Channel (PDSCH) and is required to send ACK / NACK feedback in the subsequent subframe n+k. This is in Figure 4 It is shown in the middle.

[0018] ACK / NACK feedback notification from UE to eNodeB or eNB ( Figure 4The transmitter in the UE (eNodeB) determines whether the PDSCH has been correctly decoded. When the eNodeB detects an ACK feedback, it can continue sending new data blocks (new TX) to the UE. When the eNodeB detects a NACK, it will retransmit the coded bits corresponding to the original data block. When retransmission (reTX) is based on a repetition of previously transmitted coded bits, it can be said to be operating under a tracking combination HARQ protocol. When the retransmission contains coded bits that were not used in previous transmission attempts, it can be said to be operating under an incremental redundancy HARQ protocol.

[0019] Depending on whether the UE is simultaneously transmitting the Physical Uplink Shared Channel (PUSCH), the ACK / NACK feedback is sent by the UE using one of two possible schemes:

[0020] If the UE does not transmit PUSCH at the same time, it sends ACK / NACK feedback via the Physical Uplink Control Channel (PUCCH).

[0021] If the UE is transmitting PUSCH at the same time, then send ACK / NACK feedback via PUSCH.

[0022] LTE supports bandwidths greater than 20 MHz. A key requirement for LTE Rel-10 is ensuring backward compatibility with LTE Release 8 (RL-8). This should also include spectrum compatibility. This implies that LTE Rel-10 carriers wider than 20 MHz should behave as multiple LTE carriers to LTE Rel-8 terminals. Each of these carriers can be referred to as a component carrier (CC). Specifically, for early LTE Rel-10 deployments, a smaller number of LTE Rel-10-capable terminals can be expected compared to many legacy LTE terminals.

[0023] Therefore, it is necessary to ensure the effective use of wide-bandwidth carriers for legacy terminals as well; that is, to achieve carriers in which legacy terminals (Rele 8 terminals) can be scheduled across all portions of a wide-bandwidth LTE Rel-10 carrier. A direct way to achieve this is by means of carrier aggregation (CA). CA implies that an LTE Rel-10 terminal can receive multiple carrier clusters (CCs), where each CC has, or at least potentially has, the same structure as a Rel-8 carrier. CA in Figure 5 As shown in the diagram, a UE with CA capability is assigned a primary cell (PCell) that is always active and one or more secondary cells (SCells) that can be dynamically activated or deactivated.

[0024] The number of aggregated component carriers (CCs) and the bandwidth of individual CCs can differ for the uplink and downlink. A symmetric configuration refers to a situation where the number of CCs is the same in both the downlink and uplink, while an asymmetric configuration refers to a situation where the number of CCs differs. It is important to note that the number of CCs configured in a cell can differ from the number of CCs seen by the terminal: even if the cell has the same number of uplink and downlink CCs, the terminal (e.g., the UE) may support, for example, more downlink CCs than uplink CCs.

[0025] Additionally, carrier aggregation is characterized by its ability to perform cross-carrier scheduling. This mechanism allows data transmissions on one CC to be scheduled on another CC via the (E)PDCCH on one CC, using a 3-bit Carrier Indicator Field (CIF) inserted at the beginning of the (E)PDCCH message. For data transmissions on a given CC, the UE expects to receive scheduling messages on the (E)PDCCH on only one CC (either the same CC or different CCs via cross-carrier scheduling); this mapping from (E)PDCCH to PDSCH is also configured semi-statically.

[0026] As mentioned earlier, spectrum sharing is performed in LAA systems, where LTE operates on licensed spectrum and WLAN or WiFi operates on unlicensed spectrum. Below, a brief description of WLAN or WiFi systems is provided, specifically how channels are accessed in WLAN systems.

[0027] In typical WLAN deployments, Carrier-Aware Multiple Access with Collision Avoidance (CSMA / CA) is used for media access. This means that the channel is sensed to perform Clear Channel Assessment (CCA), and transmissions are only initiated when the channel is declared idle. In cases where the channel is declared busy, transmissions are effectively delayed until the channel is deemed idle. When the ranges of several Access Points (APs) using the same frequency overlap, this means that all transmissions associated with one AP may be delayed if transmissions to or from another AP within that range can be detected on the same frequency. In practice, this means that if several APs are within range, they will have to share the channel over time, and the throughput of each AP can be significantly reduced. Figure 6 The diagram shows a general illustration of the Listen Before Talk (LBT) mechanism or procedure.

[0028] After WLAN station A transmits a data frame to station B, station B will send an ACK frame back to station A with a 16 μs delay. This type of ACK frame is transmitted by station B without performing LBT operations. To prevent another station from interfering with the transmission of this type of ACK frame, after observing channel occupancy, the station will delay the transmission for a duration of 34 μs (called DIFS) before reassessing whether the channel is occupied.

[0029] Therefore, a station intending to transmit first performs CCA by sensing the media through a persistent, fixed-duration DIFS. If the media is idle, the station assumes it has ownership of the media and begins a frame exchange sequence. If the media is busy, the station waits for the media to become idle, delays the persistent DIFS, and waits for a further random backoff period.

[0030] To further prevent stations from continuously occupying the channel and thus preventing other stations from accessing the channel, stations that want to transmit again after transmission is completed are required to perform random backoff.

[0031] PIFS is used to gain priority access to the media and has a shorter duration than DIFS. Among other things, it can also be used by stations operating under PCF to transmit priority beacon frames. At the nominal beginning of each contention-free period (CFP), the AP will sense the media. Upon determining that the media is idle for a PIFS period (typically 25 μs), the AP will transmit a beacon frame containing CF parameter set elements and a transport service indication message element.

[0032] It should be mentioned that when the media becomes available, multiple WLAN stations can be ready to transmit, which can lead to collisions. To reduce collisions, the station intending to transmit selects a random backoff counter and postpones the idle time of that number of time slots. The random backoff counter is a random integer drawn from a uniform distribution over an interval of [0, CW]. The default size of the random backoff contention window, CWmin, is set in the IEEE specification. Note that collisions can still occur even with this random backoff protocol when there are many stations accessing the contention channel. Therefore, to reduce persistent collisions, the backoff contention window size CW is doubled whenever a station detects a collision with its transmission, up to a limit Cwmax (which is also set in the IEEE specification). When a station successfully transmits without collisions, it resets its random backoff contention window size back to the default value CWmin.

[0033] It should also be mentioned that for devices that do not utilize the Wi-Fi (WLAN) protocol, EN 301.893 provides the following requirements and minimum behavior for load-based clearance channel assessment.

[0034] 1) Before transmitting or bursting on the operational channel, the device (AP or UE) will perform a CCA check using "Energy Detection". The device will observe the operational channel for a duration of CCA observation time (which will be no less than 20 μs). The CCA observation time used by the device will be stated by the manufacturer. If the energy level in the channel exceeds the threshold corresponding to the power level given in point 5 below, the operational channel will be considered occupied. If the device finds the channel clear, it can transmit immediately (see point 3 below).

[0035] 2) If the device detects that the operational channel is occupied, it will not transmit on that channel. The device will perform an extended CCA check, where the operational channel is observed for a random factor N multiplied by the duration of the CCA observation time. N defines the number of net idle time slots, resulting in the total idle period that needs to be observed before transmission can be initiated. The value of N is randomly selected from the range 1…q each time an extended CCA is requested, and this value is stored in a counter. The value of q is selected by the manufacturer from the range 4…32. This selected value will be declared by the manufacturer. Each time a CCA time slot is considered “unoccupied,” the counter is decremented. When the counter reaches 0, the device can transmit.

[0036] 3) The total time the device uses the operating channel is the maximum channel occupancy time, which will be less than (13 / 32) × qms, where q is as defined in point 2 above. After that, the device will perform the extended CCA described in point 2 above.

[0037] 4) When the device correctly receives a packet intended for it, it can skip the CCA and immediately transmit management and control frames (e.g., ACK and block ACK frames). The consecutive sequence of transmissions performed by the device without executing a new CCA will not exceed the maximum channel occupancy time.

[0038] Note: For multicast purposes, ACK transmissions (associated with the same data packets) from each device are allowed to occur sequentially.

[0039] 5) The energy detection threshold used for CCA will be proportional to the transmitter's maximum transmit power (PH): For a 23 dBme.irp transmitter, the CCA threshold level (TL) will be equal to or less than -73 dBm / MHz at the input to the receiver (assuming a 0 dBi receive antenna). For other transmit power levels, the CCA threshold level TL will be calculated using the formula TL = -73 dBm / MHz + 23 - PH (assuming a 0 dBi receive antenna and PH specified in dBm eirp).

[0040] exist Figure 7An example of the LBT mechanism in EN 301.893 is described.

[0041] Regarding LAA systems, until now, the spectrum used by LTE has been dedicated to LTE. The advantage of this is that the LTE system does not need to worry about coexistence with other non-3GPP radio access technologies in the same spectrum, and spectral efficiency can be maximized. However, the spectrum allocated to LTE is limited and cannot meet the increasing demands for higher throughput from applications / services. Therefore, a new research project has been initiated within 3GPP to extend LTE to develop and utilize unlicensed spectrum in addition to licensed spectrum.

[0042] Using LAAs for unlicensed spectrum, such as Figure 8 As shown, the UE connects to a PCell operating in the licensed spectrum and one or more SCells operating in the unlicensed spectrum. In this application, we denote the SCell in the unlicensed spectrum as an LAA secondary cell (LAA SCell). The LAA SCell can operate in DL-only mode or with both UL and DL services. Furthermore, in future scenarios, LTE nodes can operate in standalone mode on unlicensed channels without assistance from licensed cells. The unlicensed spectrum can be used simultaneously by multiple different technologies as defined. Therefore, the LAA described above can coexist with other systems such as IEEE 802.11 (Wi-Fi or WLAN).

[0043] To ensure fair coexistence with Wi-Fi (WLAN) systems, transmissions on SCell will conform to the LBT protocol to avoid collisions and severe interference with ongoing transmissions. This includes two aspects: performing LBT before initiating a transmission, and limiting the maximum duration of a single transmission burst. A single transmission burst refers to a transmission performed by a node after successful channel contention. The maximum transmission burst duration is country-specific and / or region-specific. For example, the maximum burst duration is 4 ms in Japan and 13 ms in Europe according to EN 301.893. Figure 9 Examples are shown in the context of LAA using carrier aggregation and LBT, with different examples of the duration of bursts transmitted on the LAA SCell constrained by, for example, a maximum allowed transmission duration of 4 ms.

[0044] A basic LAA coexistence protocol with a fixed random backoff contention window size (such as the random backoff contention window size specified in ETSI EN 301.893) can handle networks with a small or moderate number of nodes accessing the channel with contention. Additional measures may be needed to handle situations where a large number of nodes are operating on the same channel.

[0045] Existing random backoff contention window protocols are based on receiving a single ARQ feedback value (ACK / NACK) after the transmission of a burst of data. In the case of LTE, a Hybrid ARQ (HARQ) protocol is followed instead of a simple ARQ protocol. Therefore, multiple retransmissions based on HARQ feedback may be required before a single ARQ feedback value at a higher layer is generated.

[0046] Furthermore, multiple UEs can communicate with the eNB in ​​a single subframe. Additionally, a single LAA transmission can consist of multiple subframes. Moreover, transmissions to or from a single UE can have multiple HARQ feedback values. This is the case, for example, when the transmission is a multi-codeword transmission. Therefore, there are various ways in which multiple feedback values ​​can be received corresponding to a single transmission burst after successful channel contention. Existing random backoff contention window protocols are not suitable for handling HARQ feedback.

[0047] It should also be mentioned that LTE is characterized by HARQ feedback being available only after a fixed, predetermined time delay (e.g., 4 ms, corresponding to multiple subframes), whereas in other systems, feedback is assumed to be available after a very short time interval following the end of transmission, and this very short time interval can be shorter than the delay defined in LTE. These systems do not effectively handle systems similar to LTE, where the feedback delay is much larger. Summary of the Invention

[0048] The purpose of the embodiments described herein is to provide a method and arrangement in a first communication device or a first communication device that allows determining the size of a random backoff contention window for the next channel contention based on one or more HARQ feedback values. Therefore, adaptation / variation of the random backoff contention window size is achieved, thereby allowing fair coexistence operation between the cooperative channel LAA and WiFi, even when a large number of devices or equipment are accessing the contention channel.

[0049] Therefore, according to one aspect of an exemplary embodiment, a method or arrangement in a first communication device is provided for adapting and / or determining a random backoff contention window size in a licensed assisted access system including a primary cell and one or more secondary cells. The method includes: transmitting at least one burst comprising one or more subframes to one or more second communication devices; wherein prior to the transmission of the at least one burst, a successful Listen-After-Talk (LBT) procedure is performed, i.e., the channel is determined by the first communication device to be free and capable of transmitting the at least one burst. The method further includes: receiving corresponding HARQ feedback values ​​for the at least one subframe in the burst. The HARQ feedback is received from the at least one second communication device. The method further includes: determining a random backoff contention window size based on each received HARQ feedback value, and also based on previously unused HARQ feedback values ​​available to the first communication device when performing the LBT procedure to access the channel.

[0050] According to another aspect of an exemplary embodiment, a first communication device or arrangement within a first communication device is provided for adapting and / or determining a random backoff contention window size in a licensed assisted access system including a primary cell and one or more secondary cells. The first communication device includes: a transmitter module or transmitter circuit configured to transmit at least one burst comprising one or more subframes to one or more second communication devices; wherein a successful Listen-After-Talk (LBT) procedure is performed prior to the transmission of the at least one burst, i.e., the channel is determined by the first communication device to be free and capable of transmitting the at least one burst. The first communication device further includes a receiver module or receiver circuit configured to receive corresponding HARQ feedback values ​​for at least one subframe in the burst. HARQ feedback is received from the at least one or more second communication devices. The first communication device further includes a processing module, processing circuit, or processor configured to determine the random backoff contention window size based on each received HARQ feedback value, and also based on previously unused HARQ feedback values ​​available to the first communication device when performing LBT procedures to access the channel.

[0051] An advantage associated with the proposed embodiments is that it allows for fair coexistence between the cooperative channel LAA and WiFi. Attached Figure Description

[0052] Figure 1 This is a diagram showing the known physical resources of the LTE downlink.

[0053] Figure 2 This is a diagram illustrating the known LTE time domain structure.

[0054] Figure 3 A simplified diagram of a downlink subframe in LTE is shown.

[0055] Figure 4 An example of HARQ operation in LTE is shown.

[0056] Figure 5 This shows an example of carrier aggregation in LTE.

[0057] Figure 6 This illustrates the LBT protocol in a WiFi system.

[0058] Figure 7 The LBT is shown in EN 301.893.

[0059] Figure 8 This shows a UE with carrier aggregation capability configured with one PCell and one LAA SCell.

[0060] Figure 9 This demonstrates the use of carrier aggregation and LBT for LAA on unlicensed spectrum.

[0061] Figure 10 The document describes the use of multiple HARQ feedback values ​​to determine the random backoff contention window size during LBT procedures, according to exemplary embodiments described herein.

[0062] Figure 11 This is a flowchart illustrating the main method steps performed by the first communication device according to the proposed embodiment.

[0063] Figure 12 This is a block diagram illustrating a first communication device according to the proposed embodiment. Detailed Implementation

[0064] Briefly, exemplary embodiments are provided of a first communication device and a method thereof for adapting and / or determining the size of a random backoff contention window in a licensed auxiliary access system including a primary cell and one or more secondary cells. The first communication device may be a network node, such as a base station serving a secondary cell (SCell) and a PCell, or the first communication device may be a user equipment (UE) configured with one primary cell and at least one SCell.

[0065] The embodiments in this document teach how to determine a variable random backoff contention window size for a LAA SCell using parameters, metrics, signals, and procedures that are compatible with the LTE specification and have minimal impact on it. The solutions described below are embodiments in which the contention window size can be varied based on one or more HARQ feedback values. The embodiments also describe how the contention window size can be varied in this way, while taking into account the HARQ feedback delay, which can potentially be larger than the length of a subframe or even a transmission burst.

[0066] The proposed random backoff contention window variation technique for the LBT protocol will now be described. This is generally applicable to both DL and UL transports, and to both FDD and TDD systems. Below, the contention window from which the random backoff counter for a new LBT attempt is extracted is denoted as CW, such that the extracted counter falls within [0, CW]. The default random backoff contention window size is indicated by CWmin.

[0067] The embodiment proposes LBT for data (burst) transmission, for example, carried on PDSCH or PUSCH. The receiver transmitting the data is configured to provide HARQ feedback to the transmitter to indicate whether the data was successfully received (ACK) or unsuccessfully received (NACK) according to the LTE specification. The random backoff contention window size CW is modified by the transmitter (first communication device) based on the HARQ feedback. The modification is based on all previously unused HARQ feedback received that is available when performing LBT operation to access the channel. The receiver means one or more second communication devices that provide HARQ feedback to the transmitter device (i.e., the first communication device transmitting the burst).

[0068] As described above, a transmission burst refers to a transmission performed by a first communication device (UE or eNB) after successful channel contention. A transmission burst may contain one or more subframes, each containing a transmission to one or more second communication devices (UE or eNB). The transmission of a subframe contains one or more codewords, and based on these codewords, one or more HARQ feedbacks are transmitted to the first communication device. Each transmission burst is preceded by a successful LBT procedure, in which the first communication device determines that the channel is free and transmission is possible.

[0069] Reference Figure 10 The figure illustrates a scenario where multiple HARQ feedback values ​​are used to determine the random backoff contention window size during LBT procedures.

[0070] If a particular subframe occurs after at least (x+1) subframes, it is assumed that the HARQ feedback value used for transmission in that subframe is available to the first communication device for use in the LBT procedure. This includes the time x ms for the feedback to be available, plus a gap for handling delays, assumed to be less than 1 ms. For example, in LTE, x could be equal to 4 ms, and therefore in this case, if a particular subframe occurs after at least five subframes, it is assumed that the HARQ feedback value used for transmission in that subframe is available to the first communication device for use in the LBT procedure.

[0071] Figure 10The contention window (dashed line) for an LBT procedure occurring in a transmission burst starting in subframe 6, using HARQ feedback values ​​from subframe 0, is shown. Within subframe 0, a single HARQ feedback is received for user 1 (U1), while multiple HARQ feedback values ​​are received for multiple codewords transmitted to user 2 (U2). HARQ feedback values ​​for subframe 1 are not available in this example and are therefore not used even though they belong to the same transmission burst. As shown, the HARQ feedback values ​​are used as input to box 1001, which is configured to determine the random backoff contention window size (exit 1) for the next LBT procedure.

[0072] The diagram also illustrates the LBT procedure (dotted line) for the next transmission burst, in which, in this example, the next transmission burst uses previously unused HARQ feedback values ​​available for previous burst transmissions and HARQ feedback values ​​for subsequent burst transmissions. Here, the feedback values ​​span multiple codewords, users, subframes, and transmission bursts. The previously unused HARQ feedback values ​​for previous transmissions and the HARQ feedback values ​​for subsequent burst transmissions are used as inputs to box 1002 to determine the random backoff contention window size (exit 2) for the next LBT procedure.

[0073] Below, based on some exemplary embodiments described herein, the combination of multiple HARQ values ​​that can be executed by the first communication device is explained.

[0074] According to one exemplary embodiment, multiple HARQ values ​​can be combined into a single feedback value. For example, if all received HARQ feedback values ​​are NACK, then the (valid) HARQ values ​​are combined to form a single (valid) NACK. If all feedback values ​​turn into ACK, then the combination of ACKs generates a single ACK value.

[0075] In another exemplary embodiment, if a single HARQ value in the middle of the received HARQ values ​​is NACK, then the combination produces a single NACK value.

[0076] In another exemplary embodiment, if a portion of the received HARQ feedback value for a NACK exceeds a certain threshold, multiple HARQ feedback values ​​or valid HARQ feedback values ​​are combined to form a single (valid) NACK. This threshold is a design parameter.

[0077] In another exemplary embodiment, if a portion of the received HARQ feedback value for an ACK exceeds a certain threshold, multiple HARQ feedback values ​​or valid HARQ feedback values ​​are combined to form a single valid ACK. This threshold is also a design parameter.

[0078] It should be mentioned that multiple HARQ feedback values ​​or valid HARQ feedback values ​​can be sorted into a list before being used to determine the contention window size. For example, HARQ feedback values ​​corresponding to multiple codewords for a user (UE) and multiple users (UEs) in a subframe can be sorted in descending order of the transmission rate determined by the modulation and coding scheme used for transmissions performed by the first communication device. HARQ feedback values ​​for different subframes and for different transmission bursts can be sorted in chronological order.

[0079] In another exemplary embodiment, the HARQ feedback values ​​of multiple codewords corresponding to a user (UE) and multiple users (UEs) in a subframe are sorted in ascending order of the transmission rate as determined by the modulation and coding scheme used for transmissions made by the first communication device. Signal-to-noise ratio (SNR) values, if available at the transmitter (first communication device), can also be used to sort the HARQ values ​​in the list.

[0080] In another exemplary embodiment, based on any of the aforementioned criteria applied only to the first M subframes in the sorted list of multiple HARQ feedbacks or valid HARQ feedback values, the multiple HARQ feedback values ​​or valid HARQ feedback values ​​sorted according to the above example are combined to form a single valid NACK. This includes cases where M=1, such that the combined valid HARQ feedback value is the same as the value of the first HARQ feedback or valid feedback value in the list.

[0081] According to one exemplary embodiment, multiple HARQ feedback values ​​can be used in different ways for different groups of packets to determine how the contention window size changes. For example, the HARQ feedback values ​​corresponding to codewords in multiple-input multiple-output (MIMO) transmissions, different users in subframes, users in different subframes, and users in different transport bursts can all be used in different ways to determine how the contention window size changes. Some exemplary implementations of this embodiment are described below.

[0082] (1) As described above, HARQ feedback values ​​for different codewords transmitted from the first communication device to a single user (UE) are combined to form a single (valid) HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value per user. HARQ feedback or valid HARQ feedback values ​​for all users within a subframe are further combined to generate a single valid HARQ feedback value for the subframe. HARQ feedback values ​​for all subframes across all transmission bursts are sorted in a list in chronological order as described above and are used as input to the random backoff contention window size determination algorithm, as will be described.

[0083] (2) In another example, HARQ feedback values ​​for different codewords used by a user are first combined, and then, as described in the embodiment above, the user's HARQ or valid HARQ feedback values ​​in a subframe are combined to generate a single valid HARQ feedback value per subframe. Subsequently, as described above, the single values ​​in each subframe are further combined to generate a single valid HARQ feedback value per transmission burst. The listed HARQ feedback values ​​or valid values ​​per transmission burst are then sorted in the list according to their chronological order and used as input to the random backoff contention window size determination algorithm.

[0084] (3) In a variation of the above exemplary embodiment, the HARQ feedback values ​​in the list of HARQ feedback values ​​or valid values ​​for each transmission burst can be combined to generate a single valid HARQ feedback value for determining the size of the random backoff contention window.

[0085] (4) According to another example, as described above, HARQ feedback values ​​for different codewords transmitted to a single user (UE) can be combined to form a single valid HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value per user. HARQ feedback values ​​from all users (UEs) across all subframes and all transmission bursts are sorted in a list without any further combination and are used as input to the algorithm for determining the size of the random backoff contention window.

[0086] (5) In another example, HARQ feedback values ​​for all codewords belonging to all users (UEs) in a subframe are combined to form a single valid HARQ feedback value as described, such that there is one HARQ feedback or valid HARQ feedback value per subframe. The HARQ feedback values ​​for all subframes across all transport bursts are sorted in a list in chronological order and used as input to the algorithm for determining the random backoff contention window size.

[0087] (6) According to another example, as described above, the HARQ feedback values ​​for all codewords belonging to all users (UEs) in a subframe and for all subframes in a transport burst are combined to form a single valid HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value per transport burst. The HARQ feedback values ​​for all transport bursts are sorted in a list in chronological order and are used as input to the algorithm for determining the random backoff contention window size.

[0088] (7) In another example, all codewords belonging to all users in a subframe, HARQ feedback values ​​in a transport burst and across multiple transport bursts in all subframes are combined to form a single valid HARQ feedback value as described, such that there exists a generated HARQ feedback or valid HARQ feedback value. This single HARQ feedback or valid HARQ feedback value is used as input to the random backoff contention window size determination algorithm.

[0089] (8) In another example, all available HARQ values ​​can be used directly as input to the algorithm for determining the random backoff contention window size without any combination or sorting.

[0090] As mentioned above, the HARQ feedback value is used as input to the algorithm for determining the random contention window size of the first communication device. An example of such an algorithm is described below.

[0091] According to an exemplary embodiment, as described above, the random backoff contention window size when performing LBT operation at the beginning of a transport burst containing one or more subframes is determined as a function of a sorted list of past HARQ feedback values. The window size, denoted as CW, can therefore be considered as the sorted list of HARQ feedback values ​​HARQ2, HARQ2, ..., HARQ2 according to the following equation. K Functions:

[0092] .

[0093] Where k is the number of the last HARQ feedback value in the list.

[0094] In the first non-restrictive random backoff contention window size adaptation embodiment, whenever a NACK feedback is received, the random backoff contention window size CW is multiplied by a factor of two until it reaches a maximum value Cwmax; and whenever an ACK feedback is received, it is reset to CWmin. The size of the contention window used to perform LBT can be expressed in this case as:

[0095]

[0096] The multiplication factor a is 2, and n indicates the number of NACKs received after the last received ACK.

[0097] According to the example, n is set as follows:

[0098] n = floor((number of NACKs received after the last received ACK) / N)

[0099] The function floor() returns an integer value no greater than the input n. In other words, on average, for every N NACK feedback values, the window size CW is increased by a multiplication factor a, where N is a parameter that can be used to control the aggressiveness of the LBT algorithm.

[0100] In another example, the window size CW can be a non-exponential function, for example:

[0101] CW = CWmin × g(n)

[0102] in, g (n) can be an m-order polynomial:

[0103] g (n) = c m n m + c m-1 n m-1 + … +_c1 n + c0

[0104] In another example, neither sorting nor combining is performed on the received HARQ feedback values. For example, HARQ feedback values ​​are treated exactly the same in the contention window size determination / change algorithm. That is, there is no difference in how HARQ values ​​are processed depending on whether the values ​​belong to the same user, multiple users, multiple subframes, etc. For example, the contention window CW(i) = f(HARQ ACK / NACK) for the i-th LBT operation can be defined as follows:

[0105] If NACK_ratio < T0

[0106] If NACK_ratio ≥ T0

[0107] Where the multiplication factor a is 2, NACK_ratio = (number of HARQ NACKs) / (total number of available unused HARQ feedback values), T0 is a threshold ranging from 0 to 1, and x is a function of NACK_ratio. A sample implementation for x is given below.

[0108] If x=1, and NACK_ratio ≥ T0

[0109] The following is another example implementation for x.

[0110] If x=0, and T0 ≤ NACK_ratio < T1

[0111] If x=1, and T1 ≤ NACK_ratio < T2

[0112] …

[0113] If x=m, and Tm ≤ NACK_ratio ≤ 1

[0114] It should be mentioned that different random backoff window sizes can be maintained and adapted for data transmission and for the transmission of management and control information. Non-limiting examples of the transmission of management and control information are discovery reference signal (DRS) transmission; master control information block (MIB) and / or system information block (SIB) signals (transmitted by the first communication device).

[0115] According to one exemplary embodiment, the size of the random backoff window used for managing and controlling information transmission can be fixed by the first communication device, while the size of the window used for data transmission can be determined / adapted by the first communication device based on any of the above embodiments.

[0116] In another non-limiting embodiment, the random backoff window size for managing and controlling information transmission can be determined / adapted using a lower growth rate than that used for data transmission. As a non-limiting example, the multiplication factor used for managing and controlling information transmission is set to a smaller value than the multiplication factor used for data transmission. As a second non-limiting example, the random backoff window size for managing and controlling information transmission is adapted using a polynomial function (as described above), while the random backoff window size for data transmission is adapted using an exponential function (as described above).

[0117] Figure 11 This is a flowchart illustrating the main method steps performed by the first communication device according to the embodiments described above.

[0118] As shown, the methods include:

[0119] (1101) Transmit at least one burst comprising at least one subframe to one or more second communication devices; wherein the transmission of the at least one burst is preceded by a successful Listen-Before-Speak (LBT) procedure, i.e., the channel is determined by the first communication device to be free and capable of transmitting the at least one burst.

[0120] (1102) Receive the corresponding HARQ feedback value for at least one subframe of the burst. The HARQ feedback is received from the at least one second communication device; and

[0121] (1103) The random backoff contention window size is determined based on each received HARQ feedback and also based on the previously unused HARQ feedback value available to the first communication device when performing the LBT procedure to access the channel.

[0122] As described above, the method performed by the first communication device further includes combining multiple HARQ values. For example, the method includes: if all received HARQ feedback values ​​are NACK, then combining (valid) HARQ values ​​to form a single (valid) NACK. If all feedback values ​​turn into ACK, then the method includes combining ACKs to generate a single ACK value.

[0123] In another exemplary embodiment, the method includes combining the results in a single NACK value if a single HARQ value in the middle of the received HARQ values ​​is NACK.

[0124] In another exemplary embodiment, the method includes combining multiple HARQ feedback values ​​or valid HARQ feedback values ​​to form a single valid NACK if a portion of the received HARQ feedback value for a NACK exceeds a certain threshold. This threshold is a design parameter.

[0125] In another exemplary embodiment, the method includes combining multiple HARQ feedback values ​​or valid HARQ feedback values ​​to form a single ACK value if a portion of the received HARQ feedback value for an ACK exceeds a certain threshold. This threshold is also a design parameter.

[0126] It should be mentioned that multiple HARQ feedback values ​​or valid HARQ feedback values ​​may be sorted into a list by the first communication device before being used to determine the contention window size. For example, the method includes sorting them in descending order of the transmission rate as determined by the modulation and coding scheme used for transmitting the HARQ feedback values, wherein the HARQ feedback values ​​correspond to multiple codewords for users (UEs) and multiple users (UEs) in subframes. The first communication device may then sort the HARQ feedback values ​​for different subframes and for different transmission bursts in chronological order.

[0127] In another exemplary embodiment, the method includes sorting the HARQ feedback values ​​in ascending order of the transmission rate as determined by the modulation and coding scheme used for transmitting the HARQ feedback values, wherein the HARQ feedback values ​​correspond to multiple codewords for users (UEs) and multiple users (UEs) in subframes. Signal-to-noise ratio (SNR) values, if available at the transmitter (first communication device), can also be used to sort the HARQ values ​​in the list.

[0128] In another exemplary embodiment, based on any of the aforementioned criteria applied only to the first M subframes in the sorted list of multiple HARQ feedbacks or valid HARQ feedback values, the multiple HARQ feedback values ​​or valid HARQ feedback values ​​sorted according to the above example are combined by the first communication device to form a single valid NACK. This includes cases where M=1, such that the combined valid HARQ feedback value is the same as the value of the first HARQ feedback or valid feedback value in the list.

[0129] According to one exemplary embodiment, multiple HARQ feedback values ​​can be used in different ways for different groups of packets to determine how the contention window size changes. For example, for HARQ feedback values ​​corresponding to codewords in multiple-input multiple-output (MIMO) transmissions, different users in subframes, users in different subframes, and users in different transmission bursts, all HARQ feedback values ​​can be used in different ways to determine how the contention window size changes. Some exemplary implementations of this embodiment are described below.

[0130] (1) As described above, the HARQ feedback values ​​for different codewords transmitted to a single user are combined by the first communication device to form a single valid HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value per user. The HARQ feedback or valid HARQ feedback values ​​for all users within a subframe are further combined to generate a single valid HARQ feedback value for the subframe. The HARQ feedback values ​​for all subframes across all transmission bursts are sorted in a list in chronological order as described above, and are used as input to the algorithm for determining the random backoff contention window size, as will be described.

[0131] (2) In another example, HARQ feedback values ​​for different codewords used by a user are first combined by the first communication device, and then, as described in the embodiment above, the user's HARQ or valid HARQ feedback values ​​in a subframe are combined by the first communication device to generate a single valid HARQ feedback value for each subframe. Subsequently, as described above, the single values ​​for each subframe are further combined to generate a single valid HARQ feedback value for each transmission burst. The listed HARQ feedback values ​​or valid values ​​for each transmission burst are then sorted in a list according to their chronological order and used as input to the algorithm for determining the random backoff contention window size.

[0132] (3) In a variation of the above exemplary embodiment, the HARQ feedback values ​​in the list of HARQ feedback values ​​or valid values ​​for each transmitted burst can be combined by the first communication device to generate a single valid HARQ feedback value for determining the size of the random backoff contention window.

[0133] (4) According to another example, as described above, HARQ feedback values ​​for different codewords transmitted to a single user (UE) can be combined by the first communication device to form a single valid HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value per user. The HARQ feedback values ​​for all users (UEs) across all subframes and all transmission bursts are sorted in a list without any further combination and are used as input to the algorithm for determining the size of the random backoff contention window.

[0134] (5) In another example, the HARQ feedback values ​​for all codewords belonging to all users (UEs) in a subframe are combined by the first communication device to form a single valid HARQ feedback value as described, such that there is one HARQ feedback or valid HARQ feedback value per subframe. The HARQ feedback values ​​for all subframes across all transmission bursts are sorted in a list in chronological order and used as input to the algorithm for determining the random backoff contention window size.

[0135] (6) According to another example, as described above, the HARQ feedback values ​​for all codewords belonging to all users (UEs) in a subframe and for all subframes in a transmission burst are combined by the first communication device to form a single valid HARQ feedback value, such that there is one HARQ feedback or valid HARQ feedback value for each transmission burst. The HARQ feedback values ​​for all transmission bursts are sorted in a list in chronological order and are used as input to the algorithm for determining the random backoff contention window size.

[0136] (7) In another example, all codewords belonging to all users in a subframe, HARQ feedback values ​​in a transmission burst and across multiple transmission bursts are combined by the first communication device to form a single valid HARQ feedback value as described, such that a generated HARQ feedback or valid HARQ feedback value exists. This single HARQ feedback or valid HARQ feedback value is used as input to the random backoff contention window size determination algorithm.

[0137] (8) In another example, all available HARQ values ​​can be directly used by the first communication device as input to the algorithm for determining the random backoff contention window size without any combination or sorting.

[0138] As mentioned above, the HARQ feedback value is used by the first communication device as input to the random contention window size determination algorithm of the first communication device to determine the window size CW. Below is an example of the method executed by the first communication device to determine CW.

[0139] According to an exemplary embodiment, as described above, the random backoff contention window size when performing LBT operation at the beginning of a transport burst containing one or more subframes is determined as a function of a sorted list of past HARQ feedback values. The window size CW can therefore be considered as the sorted list of HARQ feedback values ​​HARQ2, HARQ2, ..., HARQ2 according to the following equation. K The function f:

[0140] .

[0141] Where k is the number of the last HARQ feedback value in the list.

[0142] In another example, in a random backoff contention window size adaptation embodiment, whenever a NACK feedback is received, the random backoff contention window size CW is multiplied by a factor of two until it reaches a maximum value Cwmax; and whenever an ACK feedback is received, it is reset to CWmin. The size of the contention window used to perform LBT can be expressed in this case as:

[0143]

[0144] The multiplication factor a is 2, and n indicates the number of NACKs received after the last received ACK.

[0145] According to the example, n is set as follows:

[0146] n = floor((number of NACKs received after the last received ACK) / N)

[0147] The function floor() returns an integer value no greater than the input n. In other words, on average, for every N NACK feedback values, the window size CW increases by a multiplication factor a, where N is a parameter used to control the aggressiveness of the LBT algorithm.

[0148] In another example, the function for window resizing can be a non-exponential function, for example:

[0149] CW = CWmin × g(n)

[0150] in, g (n) can be an m-order polynomial:

[0151] g (n) = c m n m + c m-1 n m-1 + … +_c1 n + c0

[0152] In another example, the first communication device neither sorts nor combines the received HARQ feedback values. For example, the HARQ feedback values ​​are treated exactly the same in the contention window size determination / change algorithm. That is, there is no difference in how the HARQ values ​​are processed depending on whether the values ​​belong to the same user, multiple users, multiple subframes, etc. For example, the contention window CW(i) = f(HARQ ACK / NACK) for the i-th LBT operation can be defined as follows:

[0153] If NACK_ratio < T0

[0154] If NACK_ratio ≥ T0

[0155] Where the multiplication factor a is 2, NACK_ratio = (number of HARQ NACKs) / (total number of available unused HARQ feedback values), T0 is a threshold ranging from 0 to 1, and x is a function of NACK_ratio. A sample implementation for x is given below.

[0156] If x=1, and NACK_ratio ≥ T0

[0157] The following is another example implementation for x.

[0158] If x=0, and T0 ≤ NACK_ratio < T1

[0159] If x=1, and T1 ≤ NACK_ratio < T2

[0160] …

[0161] If x=m, and Tm ≤ NACK_ratio ≤ 1

[0162] Reference Figure 12 The diagram illustrates a block diagram of exemplary components of a first communication device 1200 according to the previously disclosed embodiments. The first communication device 1200 includes at least one antenna 1210, transmitter circuitry or transmitter module 1220, receiver circuitry or receiver module 1230, processor 1240 or processing module or processing circuitry, and memory 1250. As shown, the transmitter circuitry and receiver circuitry may be included in a transceiver circuitry or transceiver module 1260.

[0163] Antenna 1210 includes one or more antennas for transmitting and / or receiving radio frequency (RF) signals via an air interface. Antenna 1210 may, for example, receive RF signals from transceiver / transmitter circuitry 1260 and transmit RF signals via the air interface to one or more second communication devices, such as UEs or radio network nodes, i.e., radio base stations, such as eNodeBs or eNBs, and receive RF signals from said one or more second communication devices, such as radio base stations or UEs, via the air interface, and provide RF signals to transceiver circuitry 1260.

[0164] Processing module / circuit 1240 includes a processor, microprocessor, application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or similar. Processor 1240 controls the operation of the first communication device 1200 and its components. Memory (circuit or module) 1250 includes random access memory (RAM), read-only memory (ROM), and / or another type of memory to store data and instructions usable by processor 1240. The first communication device 1200 may include... Figure 12 Other components not shown.

[0165] As described above, the first communication device 1200 is configured to transmit at least one burst comprising one or more subframes to one or more second communication devices via transmitter circuitry / module 1230; wherein a successful Listen-Before-Speak (LBT) procedure is performed prior to the transmission of the at least one burst, i.e., the channel is determined by the first communication device 1200 to be free and capable of transmitting the at least one burst. A receiver module or receiving circuitry 1220 is configured to receive corresponding HARQ feedback values ​​for at least one subframe in the burst. HARQ feedback is received from the at least one or more second communication devices. A processing module or processing circuitry or processor 1240 is configured to determine a random backoff contention window size based on each received HARQ feedback, and also based on previously unused HARQ feedback available to the first communication device 1200 when performing the LBT procedure to access the channel.

[0166] Memory 1250 may contain instructions executable by processor 1240, thereby enabling first communication device 1200 to perform the method steps described above. A computer program including computer-readable code components is also provided, which, when executed in first communication device 1200, for example by means of processor 1240, cause the first communication device to perform the method steps described above, the steps including: transmitting at least one burst comprising one or more subframes to one or more second communication devices for reception; receiving corresponding HARQ feedback values ​​for at least one subframe in the burst; and determining a random backoff contention window size based on each received HARQ feedback, and also based on previously unused HARQ feedback available to first communication device 1200 when performing LBT procedures to access the channel.

[0167] When the computer-readable code component is running in the first communication device, it also causes the first communication device 1200 to combine the received HARQ values ​​by means of the processor 1240 as described above, and also to determine the random backoff contention window size by means of the processor 1240 as described above.

[0168] Throughout this disclosure, the terms "comprise" or "comprising" are used in a non-limiting sense, meaning "consisting of at least...". While specific terms may be used herein, they are used only in a general and descriptive sense and not for limiting purposes. Specifically, it should be noted that although terms from 3GPP LTE have been used in this disclosure to illustrate the invention, this should not be construed as limiting the scope of the invention to only the systems mentioned above. Other wireless systems, including LTE-A (or LTE-Advanced), UMTS, WiMax, and WLAN, may also benefit from developing and utilizing the ideas covered within this disclosure.

[0169] Explanation of Abbreviations

[0170] CCA Clearance Channel Assessment

[0171] DCF Distributed Coordination Function

[0172] DIFS DCF inter-frame spacing

[0173] DL downlink

[0174] DRS detects reference signal

[0175] eNB evolves into NodeB, base station

[0176] LAA Licensed Assisted Access

[0177] LBT Listen before you speak

[0178] PDCCH (Physical Downlink Control Channel)

[0179] PIFS PCF inter-frame spacing

[0180] PCell main cell

[0181] PUSCH Physical Uplink Shared Channel

[0182] QoS (Quality of Service)

[0183] SCell Auxiliary Community

[0184] SIFS Short Inter-Frame Interval

[0185] UE User Equipment

[0186] UL uplink

Claims

1. A method for adapting a random back-off contention window size performed by a first communication device (1200), the method comprising: - transmitting a plurality of transmissions to at least one second communication device; - receiving a plurality of Hybrid Automatic Repeat Request, HARQ, feedback values corresponding to the plurality of transmissions from the at least one second communication device; and - adapting a random back-off contention window size based on the received plurality of HARQ feedback values. The transmissions are preceded by a successful Listen Before Talk, LBT, procedure.

2. The method of claim 1, wherein, The adapting of the random back-off contention window size is based on the received HARQ feedback values and further based on any previously unused HARQ feedback available at the first communication device when performing the LBT procedure to access a channel.

3. The method of claim 2, wherein, The adapting of the random back-off contention window size comprises increasing the random back-off contention window size as a result of determining that a proportion of Negative Acknowledgements, NACKs, in the received plurality of HARQ feedback values is greater than or equal to a predetermined threshold.

4. The method of claim 1 or 2, wherein, The random back-off contention window size CW is adapted as an exponential function with base 2 whenever receiving NACK feedback up to a maximum value CWmax.

5. The method of claim 1 or 2, wherein, The adapting of the random back-off contention window size, here denoted as CW, is performed according to:

6. The method of claim 5, wherein, where CWmin is a default size of the random back-off contention window, the multiplication factor a is 2, and n represents the number of NACKs received after the last received ACK, and where n is: CW = CWmin x a n n = floor((number of NACKs received after the last received ACK) / N) where the function floor() returns an integer value not greater than the input of the function, i.e. on average the window size CW is increased by the multiplication factor a for every N NACK feedback values, where N is a parameter that can be used to control the aggressiveness of the LBT algorithm. The adapting of the random back-off contention window size comprises resetting the random back-off contention window size to a minimum value CWmin as a result of determining that the received plurality of HARQ feedback values comprises Acknowledgements, ACKs.

7. The method of claim 1 or 2, wherein, 8. An apparatus for adapting a random back-off contention window size comprising a first communication device (1200), the apparatus comprising a processor (1240) and a memory (1250), said memory (1250) containing instructions executable by said processor (1240), whereby said apparatus is operative to: - transmit a plurality of transmissions to at least one second communication device; - receive a plurality of Hybrid Automatic Repeat Request, HARQ, feedback values corresponding to the plurality of transmissions from the second communication device; and - adapt a random back-off contention window size based on the received plurality of HARQ feedback values. The transmissions are preceded by a successful Listen Before Talk, LBT, procedure.

9. The apparatus of claim 8, wherein, 10. The apparatus according to claim 9, operative to adapt the random back-off contention window size based on the received HARQ feedback values and further based on any previously unused HARQ feedback available at the first communication device when performing the LBT procedure to access a channel. ​ 11. The device of claim 8 or 9, operable to increase the random back-off contention window size as a result of determining that a proportion of negative acknowledgements, NACKs, in the received plurality of HARQ feedback values is greater than or equal to a predetermined threshold.

12. The device of claim 8 or 9, operable to adapt the random back-off contention window size, CW, as an exponential function with base 2 whenever receiving NACK feedback up to a maximum value, CWmax.

13. The device of claim 12, operable to perform the adaptation of the random back-off contention window size, here denoted as CW, as follows: CW = CWmin * a"n where CWmin is a default size of the random back-off contention window, the multiplication factor, a, is 2, and n represents the number of NACKs received after the last received ACK, and where n is: n = floor((number of NACKs received after the last received ACK) / N) where the function floor() returns an integer value not greater than the input of the function, i.e. on average the window size, CW, is increased by the multiplication factor, a, for every N NACK feedback values, where N is a parameter that can be used to control the aggressiveness of the LBT algorithm. CW = CWmin x a n The device is operable to adapt the random back-off contention window size to a minimum value, CWmin, as a result of determining that the received plurality of HARQ feedback values comprises acknowledgements, ACKs.

15. A storage medium having stored thereon a computer program which, when executed by a processor (1240) in a device, causes the device to perform the method of any one of claims 1-7. ​ 14. The device (1200) according to claim 8 or 9, wherein, ​ ​

Images