KR101992689B1 - QoS-aware Adaptive MPDU Aggregation Scheduler for Voice Traffic - Google Patents

Abstract

The present invention relates to a method for transmitting a voice over Internet protocol (VoIP) packet in real time at a communication device using a WLAN access network (ex: IEEE 802.11n or IEEE 802.11ac) provided with an aggregated MAC protocol data unit (A-MPDU) integrating function. According to the present invention, a method for processing a WLAN packet comprises the steps of: receiving and storing a VoIP MAC protocol data unit (MPDU) to be sent to a destination node in a first queue; generating A-MPDUs for a plurality of VoIP MPDUs stored in the first queue storage unit using end-to-end A-MPDU propagation time and jitter information obtained from a packet received from an external communication device; and transmitting the generated A-MPDUs to a destination.

Description

[0002] QoS-aware Adaptive MPDU Aggregation Scheduler for Voice Traffic (MPDU)

An embodiment of the present invention relates to a wireless LAN transmission apparatus that enables intelligent integration of an MPDU for voice over Internet protocol (VoIP) packet delivery so as to guarantee quality of service.

The contents described in this section merely provide background information on the embodiment of the present invention and do not constitute the prior art.

Real-time voice communication services require high processing speed and high bandwidth. However, since the network resources are not sufficient, there is a need for a technology capable of ensuring various QoS (Quality of Service) so that network resources can be efficiently used and solved. QoS technology refers to network capability that provides the ability to guarantee quality of service differentiated for selected network traffic or applications.

Recent improvements in wireless local area network (WLAN) technology have resulted in many improvements in both the physical layer (PHY) and the media access control (MAC) layer. For example, due to multiple input / multiple output (MIMO) technology, wider channels, and shorter guard intervals, the PHY transfer rate has reached 600 Mbps in IEEE 802.11n and 6.77 Gbps in the IEEE 802.11ac standard.

In addition, the MAC layer was able to utilize the high PHY data rate more efficiently through the aggregated MAC Protocol Data Unit (A-MPDU) integration and block ACK (Block ACK) mechanism. WLAN technology has become one of the most common means of broadband wireless Internet access due to its simple installation and wide bandwidth.

Since the actual network access delay and the relay transmission delay time in the wireless local area network (WLAN) in which the A-MPDU aggregation function can be used vary greatly according to the communication network conditions, the end- Real-time voice traffic in most wireless local area networks (WLANs) has an integrated MAC protocol data unit (A-B) in real-time, since it often happens that the requirements of end-to-end delay and jitter can not be met. MPDU) technology has not been applied.

When the number of Voice over Internet Protocol (VoIP) sessions increases in a WLAN where the A-MPDU function is not applied, when an isochronous real-time voice packet is transmitted separately, an increased collision Mobile station and access point (AP) nodes are faced with high voice packet loss rate (PLR) and / or queue overflow.

In order to improve the capacity of the VoIP session on the WLAN in which the A-MPDU function is available, an embodiment of the present invention integrates as many MPDUs as possible within a quality assurance range to generate a large unit of frames, And to provide an adaptive A-MPDU integrated scheduler capable of maximizing the bi-directional (uplink and downlink) transmission efficiency of a wireless LAN.

According to an embodiment of the present invention, there is provided a communication apparatus for real-time VoIP packet transmission in a wireless LAN access network provided with an A-MPDU integration function, comprising: a receiving unit for receiving a VoIP MPDU to be transmitted to a destination node, A MAC layer for generating A-MPDUs for a plurality of VoIP MPDUs stored in the first queue storage unit using information obtained from packets received from the device; And a physical layer for transmitting the A-MPDU generated by the MAC layer to a destination.

Herein, the MAC layer includes an end-to-end transmission delay estimation time of a head-of-line MPDU located at a head of the first queue, an inter-arrival delay interval of inter- -arrival time) and a time required to transmit one VoIP MPDU is equal to or greater than a predetermined end-to-end transmission delay time, the A-MPDU is generated including the corresponding VoIP MPDU.

Wherein the end-to-end propagation delay estimation time is determined based on a delay time after the head-of-line MPDU is transmitted from the upper layer of the communication apparatus and stored in the first queue to the present time, a medium access time of the communication apparatus, And the end-to-end transmission delay of the A-MPDU.

Wherein the transmission delay is calculated by obtaining an end-to-end propagation delay time from a packet received from a counterpart node that is the external communication apparatus and calculating an average time stored in a second queue after generation of the VoIP MPDU in the end- And the frame transmission duration average time.

Also, the end-to-end propagation delay time is acquired in a packet received from the counterpart node, and the MAC layer holds generation of the A-MPDU until receiving the first RR packet, And generates and transmits a frame in MPDU units.

In addition, the MAC layer receives VoIP MPDUs, each identified by a plurality of unique identifiers, from a plurality of source nodes, and the plurality of VoIP MPDUs are identified by one specific identifier included in the plurality of unique identifiers VoIP MPDUs. In this case, the end-to-end propagation delay estimation time is calculated using the generation time of the head-of-line MPDU located at the head of the first queue without acquiring from the packet received from the counterpart node.

Also, the end-to-end propagation delay estimation time may be determined based on a delay time from a time when a head-of-line MPDU located at the head of the first queue is generated at a source node of the plurality of source nodes to a present time, The maximum media access time of the device, and the transmission duration of the A-MPDU.

Also, the MAC layer suspends generation of the A-MPDU until a maximum medium access time of the communication device is obtained, and generates and transmits a frame in units of the VoIP MPDUs.

According to another embodiment of the present invention, there is provided a method for real-time VoIP packet transmission in a communication apparatus using a wireless LAN access network provided with an A-MPDU integration function, the method comprising: receiving a VoIP MPDU to be transmitted to a destination node, ; Generating A-MPDUs for a plurality of VoIP MPDUs stored in the first queue storage unit using information obtained from packets received from an external communication device; And transmitting the generated A-MPDU to a destination.

The present invention can be used by providing an Adaptive Aaggregation Scheduler for uplink voice traffic at a wireless station when a plurality of wireless stations compete for uplink voice and saturated low priority traffic transmissions It has an effect of improving the network throughput.

In addition, it can deliver up to 5.3 times the capacity of the existing implementation, while delivering all voice packets to the destination with an end-to-end delay of less than 150 ms.

1 is a diagram illustrating a network topology including a communication device in which an adaptive integrated A-MPDU scheduler according to the present embodiment is mounted.
FIG. 2 is a schematic diagram of the components participating in the adaptive integrated A-MPDU scheduling according to the present embodiment.
3 is a block diagram illustrating the functionality of the MAC layer 102 of an integrated scheduler operating in an 802.11n / ac VoIP node 100. As shown in FIG.
4 is a diagram showing an A-MPDU frame and a BlockACK frame.
FIG. 5 is a diagram illustrating an algorithm of an integrated scheduler operating in the wireless terminal 100. Referring to FIG.
6 is a diagram illustrating an algorithm of an integrated scheduler operating in AP 300. Referring to FIG.
FIG. 7 is a diagram illustrating an end-to-end delay performance at a capacity boundary for packets generated in mobile and wired stations using various VoIP codec speeds.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference symbols as possible even if they are shown in different drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

In this embodiment, a VoIP packet is transmitted in real time in a communication device using a wireless LAN (e.g., IEEE 802.11n or IEEE 802.11ac) access network provided with an A-MPDU integration function. MPDU transmission time and jitter information obtained from the packet received from the communication device, and transmits the generated A-MPDU to the destination as much as possible by integrating the VoIP MPDUs as much as possible to guarantee the quality of service And provides a method of processing a wireless LAN packet that can maximize the bi-directional (uplink and downlink) transmission efficiency of the wireless LAN.

The packet processing apparatus according to the present embodiment is installed in a communication apparatus that transmits VoIP packets in real time in a wireless LAN access network of IEEE 802.11n or IEEE 802.11ac.

Here, a wireless station and an AP may be used, but the present invention is not limited thereto.

1 is a diagram illustrating a network topology including a communication device in which an adaptive integrated A-MPDU scheduler according to the present embodiment is mounted.

The adaptive integrated A-MPDU scheduler according to the present embodiment can be mounted on the AP 300 or mounted on the wireless terminal 100.

The wireless terminal 100 may uplink the VoIP packet to the destination terminal 200 via the AP 300 and the AP 300 may transmit the VoIP packet transmitted from the destination terminal 200 to the wireless terminal 100 Downlink transmission can be performed.

FIG. 2 is a schematic diagram of the components participating in the adaptive integrated A-MPDU scheduling according to the present embodiment.

As shown in FIG. 2, the adaptive integrated A-MPDU scheduling includes a terminal 100, a destination terminal 200, and an AP 300 using an IEEE 802.11n / ac wireless LAN.

Here, the wireless terminal 100 and the destination terminal 200 provide VoIP (Voice over Internet Protocol) based real time voice communication service via the AP 300.

The wireless terminal 100 and the destination terminal 200 each have the following configuration.

The wireless terminal 100 and the destination terminal 200 may include a physical layer 802.11n / ac PHY layer 101, a MAC layer 102, a network layer 102, An IP layer 103 203, a UDP layer 104, an RTP / RTCP layer 105, and a VoIP And an application layer (App-VoIP layer) 106 (206).

The AP 300 includes a physical layer (802.11n / ac PHY layer) 301 and a MAC layer (802.11n / ac MAC layer) 302.

The embodiment of the present invention is directed to a QoS-aware adaptive A-MPDU aggregation scheduler operating in the AP 300 as well as the wireless terminal 100. The operation of the integrated scheduling technique described in this embodiment is somewhat different depending on the type of the installed wireless station.

In the following description, the names of the integrated scheduling apparatus, the integrated scheduler, and the A-MPDU integrated scheduler are used in combination, but they all refer to the adaptive integrated A-MPDU scheduling apparatus according to the present embodiment. In the following description, even if the terms are used for a transmitting terminal, a transmitting node, a transmitting VoIP node, a wireless VoIP station, etc., they all refer to the wireless terminal 100, unless otherwise stated.

The integrated scheduler of the wireless VoIP station requires periodic end-to-end delay feedback from the VoIP counterpart node 200 located in another network, while the integrated scheduler of the AP 300 may require specific explicit feedback from the other node 200 There is a difference between the two.

Hereinafter, the integrated scheduling apparatus according to the present embodiment will be described in detail. In this embodiment, it is assumed that VoIP nodes 100 and 200 and AP node 300 are synchronized.

- First Embodiment -

In the first embodiment, it is assumed that the communication device on which the integrated scheduler according to the present embodiment is mounted is a wireless station.

In this case, the wireless station corresponds to the wireless terminal 100.

The Real-time Transport Protocol (RTP) layer 105 of the wireless terminal 100 encapsulates voice data into RTP packets. The header of the RTP packet includes a sequence number field and a time stamp field, the sequence number field has a unique sequence number, and the time stamp field has a value of a time at which the packet is generated.

The RTP layer 205 of the destination terminal 200 measures the end-to-end delay of each received voice packet using the value of the timestamp field and periodically (at approximately one second intervals) (RR) packet. A Receiver Report (RR) packet contains the maximum end-to-end delay (De2e) value observed during the previous period.

After the wireless terminal 100 receives the RR packet, the RTP layer delivers the maximum end-to-end delay information to the MAC layer 102 of the 802.11n / ac. In the MAC layer 102 according to the present embodiment, an interval between a point at which a frame to be transmitted becomes a head-of-line (a point at the front in a SwQueue) Stores statistics on media access delays defined by time. Here, the statistics on the media access delay include an average value, a maximum value, and the like.

Here, the frame refers to a single VoIP MPDU when transmitting a single VoIP MPDU, and refers to an A-MPDU when an A-MPDU is transmitted. In addition, the MAC layer 102 of the present embodiment can be configured to have a maximum end-to-end delay that varies with time, And adaptation of A-MPDU aggregation to voice traffic based on QoS requirements (i.e., end-to-end delay of less than 150 ms). The VoIP MPDU includes an APP-VoIP layer 106 data, an RTP layer 105 data, a UDP layer 104 data, and an IP protocol layer 103 data, Means a data unit.

3 is a block diagram illustrating the functionality of the MAC layer 102 operating in an 802.11n / ac VoIP node 100. As shown in FIG.

 In this embodiment, the following new functions are introduced by reusing some basic features of the reference ath9k driver.

Software Queue ( SwQueue ): This queue holds the newly arrived MPDU ( MPIPU ) and the retry MPDU from the upper layer. Herein, the retry MPDU means an MPDU reported in the BlockACK frame that the destination node 200 has not properly received.

4, the transmitting station transmits a maximum of 64 MAC PDUs (MPDUs) in one A-MPDU frame, and transmits the same to the A-MPDU frame. Referring to FIG. 4, can do. The receiving node responds with a BlockACK frame after a short inter-frame space (SIFS) period after receiving the A-MPDU. The BlockACK frame has a compressed bitmap field for informing the transmission status of each MPDU transmitted to the A-MPDU.

Hereinafter, the operation of the adaptive integrated A-MPDU scheduler according to the present embodiment will be described with reference to FIG.

There is one SwQueue per Access Class (AC). When an IP packet arrives from an upper layer (Layer 3) 104, the MAC layer 102 has information on a differentiated services code point (DSCP) field of the packet (having the layer 3 priority information for the packet ) Is mapped to the AC table to identify the type of traffic delivered, and adds the necessary headers to generate MPDUs and store them in the SwQueue of each access type.

On the other hand, the access type is classified into 'AC_VO', 'AC_VI', 'AC_BK' and 'AC_BE', and each of them is prioritized to deal with different types of traffic. Since each access type has a SwQueue, the traffic with the highest priority is 'AC_VO' (VoIP traffic) and SwQueue [VO] (31) as the corresponding SwQueue.

Function of SwQueue : When each MPDU arrives at the SwQueue, the MAC layer 102 records the arrival time in the timestamp field of the skb buffer where the MPDU is stored. At this time, the difference between the timestamp of the skb buffer and the timestamp of the RTP packet included in the corresponding skb buffer is set so that the RTP packet is transmitted to the MAC layer 102 immediately after the UDP, IP and logical link control (LLC) And therefore can be ignored.

In addition, the MAC layer 102 continuously records an inter-arrival period (Tarr), which is a time between new MPDUs continuously arriving at the SwQueue. For example, Tarr for 64 Kbps PCM VoIP traffic with an 80 byte payload size is always 10 ms.

Transmission queue ( TxQueue ): A queue for transmission where only one frame consisting of A-MPDUs or one frame consisting of a single MPDU can be stored. Immediately after one frame is put into the TxQueue, a new back-off window is randomly selected and the back-off process for media access begins. Like SwQueue, each access type has its own TxQueue, and in Figure 3, the TxQueue corresponding to VoIP is TxQueue [VO] (32).

Function of TxQueue : In this embodiment, the MAC layer 102 calculates the medium access delay (Dma) for each successfully transmitted frame. Dma is the elapsed time between the moment the frame is queued in the TxQueue and the moment the non-collision frame transmission begins. Dma is a time-varying parameter and depends on the level of channel contention in the WLAN.

Dma is calculated by the following equation (1).

[Equation 1]

Dma = t _BA - t _TQ - T _F - T _SIFS - T _BA

T _BA is the instant when the BlockACK / ACK frame is received, t _TQ is the instant when the data frame is queued to the TxQueue (when BlockACK / ACK is received), T _SIFS is the duration of the Short Inter-Frame Space (SIFS) _F represents the frame transmission duration (including the PHY header and preamble), and T _BA represents the BlockACK / ACK frame transmission duration. Also, when the RTS / CTS (Request to Send / Clear to Send) access mode is used, the TF includes the time required for the RTS / CTS exchange (including RTS and CTS duration and two TSIFS).

The calculation of Dma depends on whether the basic access or RTS / CTS access mode is used as the transmission mode. If the MAC frame in the TxQueue is greater than the length specified in the dot11RTSThreshold setting of the 802.11n / ac MAC, transmit the frame using the RTS / CTS access mode. Otherwise, the default access mode is used.

The MAC layer 102 according to the present embodiment accumulates and stores Dma for all successful transmission frames and transmits a new RR packet to the destination terminal 200 through the RTP / RTCP (Real-time Transport Control Protocol The average medium access delay (D'ma) is updated.

FIG. 5 is a diagram illustrating an algorithm of an integrated scheduler operating in the wireless terminal 100. Referring to FIG.

Table 1 shows the algorithm of FIG. 5, and the parameters used in the equations and their brief description.

Algorithm 1 of FIG. 5 represents a pseudo-code that implements the MAC layer 102 of the QoS aware adaptive A-MPDU unified scheduler for uplink voice traffic at the wireless station, 102) is divided into two steps, an initialization step (uplink) and an A-MPDU generation step.

i) initialization step ( Uplink )

The initialization phase typically lasts for a fraction of a second. More precisely, the initialization step lasts until receipt of the first RR packet from the called party RTP / RTCP, and the first RR packet contains the value of the first maximum end-to-end delay previously observed by the called party RTP / RTCP. For reference, the RR packet is one of the data transmission status check packets transmitted by the RTCP, and notifies the sender of information necessary for estimating the jitter, packet loss, and round trip delay of the received media stream.

During the initialization phase, the MAC layer 102 is not allowed to integrate A-MPDUs for voice traffic because it has not yet received an RR packet from the destination RTP / RTCP and does not know the maximum end-to-end delay value, Real-time VoIP MPDUs are transmitted separately. Each time the new MPDU or retry MPDU arrives from the upper layer to the SwQueue, the MAC layer 102 calls the ScheduleTransmission () function, and the ScheduleTransmission () function calls SendMpdu () when the TxQueue is idle.

SendMpdu () sends the MPDUs to their respective destinations. Thus, SendMpdu () queues the MPDU at the head of the SwQueue to the TxQueue and starts the backoff process for media access. The MAC layer 102 may call the ScheduleTransmission () function when the SwQueue has pending MPDU (s) even after receiving the ACK / BlockACK frame.

If the new MPDU is stored in the TxQueue, the MAC layer 102 records the software delay Dsw of the MPDU. This delay is defined as the elapsed time until this MPDU is stored in the TxQueue since it first arrived at the SwQueue. During the initialization phase, since most VoIP MPDUs arrive at the SwQueue, the TxQueue is idle, so the VoIP MPDU is stored in the TxQueue as soon as it arrives at the SwQueue, and Dsw is usually 0 ms. However, if contention for media access increases, the TxQueue may be busy when the MPDU arrives in the SwQueue.

The MAC layer 102 calls the UpdateStats () function when a new RR packet is received after ACK / BlockACK is received. The UpdateStats () function updates the average software delay (D'sw), the average medium access delay (D'ma), and the average frame transmission duration (T ' _F ) according to the number of frame transmission successes (nTx). Next, the UpdateStats () function calculates the approximate transmission delay [Dt] of the frame using the following equation (2) using the maximum end-to-end delay value (De2e) obtained from this new RR packet .

&Quot; (2) "

[Dt] = De2e - D 'sw - D'ma - T' _F

[Dt] represents an approximate delay between the 802.11n / ac MAC layer 302 of the AP 300 and the RTP / RTCP layer 205 of the destination node 200.

ii) A- MPDU  Generation step ( Uplink )

After the first approximate transmission delay ([Dt]) is obtained, the A-MPDU generation step, which is the normal operation mode, is started in the MAC layer 102 according to the present embodiment. During this phase, two types of MPDUs may arrive at the SwQueue: i) new MPDUs coming in from the upper application layer; and ii) retry MPDUs that failed in previous A-MPDU transmissions due to noise in the wireless channel MPDUs). New MPDUs are queued at the end of the SwQueue queue while retry MPDUs are queued at the beginning of the SwQueue. Each time an MPDU (new MPDU or retry MPDU) arrives at the SwQueue, the MAC layer 102 calculates an approximate end-to-end delay ([De2e]) for the head-of-line MPDU in the SwQueue by the following equation As shown in FIG.

&Quot; (3) "

[De2e] = Dsw + D'ma + T F + [Dt]

In Equation (3), [De2e] denotes a software delay Dsw, an average medium access delay (D'ma), an A-MPDU (where A-MPDU is a SwQueue present in is obtained as the sum of the transmission duration (T _F) and approximately transmission delay ([Dt]) in composed of a current MPDU). [De2e] represents the expected value of the end-to-end delay of the head-of-line MPDU. Every time [De2e] is updated in the MAC layer 102, the following condition 1 is checked, and if it is true, the SendAmpdu () function is immediately called. The SendAmpdu () function immediately constructs the A-MPDU using the currently available MPDUs in the SwQueue and puts the A-MPDU into the TxQueue as soon as the A-MPDU is configured to transmit the configured A-MPDU.

Condition 1: [De2e] + Tarr + T _MPDU ≥ ≥ 150 ms.

(De2e) for SwQueue's head-of-line MPDU, the time required to wait for a new MPDU to arrive at the SwQueue (Tarr), and the time it takes to send one VoIP MPDU (TMPDU ) Denotes the future approximate end-to-end delay of the head-of-line MPDU when a new MPDU arrives at the SwQueue. If condition 1 is not true (i.e., the future approximate end-to-end delay of the head-of-line MPDU is expected to be less than 150 ms), the MAC layer 102 of the integrated scheduler waits for a new MPDU Recheck this condition. While condition 1 is not true, it continues to wait for another new MPDU.

However, when this condition 1 is true, the scheduler immediately constructs the A-MPDU using the MPDU currently available in the corresponding SwQueue and forwards it to the TxQueue. After storing the A-MPDU in the TxQueue, the scheduler in turn starts the backoff process for media access. Therefore, this integrated scheduler schedules the A-MPDU transmission to prevent the end-to-end delay from exceeding the 150 ms threshold when the head-of-line MPDU arrives at the destination RTP / RTCP.

Here, the generated A-MPDU is transmitted in a CSMA / CA (Carrier Sense Multiple Access / Collision Avoidance) method by a virtual collision handler 33.

Meanwhile, the physical layer 101 physically transmits the A-MPDU generated by the MAC layer 102.

- Second Embodiment -

In the second embodiment, it is assumed that the communication device on which the integrated scheduler according to the present embodiment is mounted is an AP.

The voice traffic that the AP 300 should provide on the downlink is equal to the sum of the traffic that the wireless stations must provide in the uplink. That is, when n wireless stations 100 are connected to one AP 300 and each of the wireless stations 100 transmits uplink VoIP traffic via the corresponding AP 300, The number of VoIP packets for link transmission is n times the number of uplink transmission VoIP packets of one single wireless station 100.

In IEEE 802.11n / ac, AP 300 and wireless stations 100 have the same priority for media access. Also, as the number n of wireless stations connected to the AP increases in the conventional art, downlink voice traffic in the AP does not apply to the downlink voice traffic. Therefore, the number of wireless stations connected to the AP increases, The amount of medium access delay at the time of downlink packet transmission is further increased. Consequently, as n increases, the queue utilization of the AP increases and as the queue of APs becomes full, new packets arriving at the AP are discarded (with packet loss due to exhaustion of the retry limit) resulting in increased packet loss.

Conventionally, in the AP, the VoIP capacity of the WLAN is limited to a very small number of VoIP sessions due to excessive voice PLR. In this embodiment, to increase the downlink VoIP stream capacity of the AP 300, an integrated scheduler for use in the downlink VoIP traffic is proposed in the AP 300, and the MAC layer of the integrated scheduler to be used for the downlink VoIP traffic here The operation of the MAC layer 302 is similar to the operation of the MAC layer 102 of the unified scheduler at the wireless station, but the method for calculating the end-to-end delay is somewhat different. The operation of the AP unified scheduler according to the present embodiment is divided into two steps, an initialization step and an A-MPDU generation step.

i) initialization step ( Downlink )

Updates the maximum media access delay (Dma_max) in the MAC layer 302 of the AP 300 approximately every second. The initialization step lasts until the first Dma_max is obtained. Here, unlike the first embodiment, the maximum medium access delay is calculated because the channel contention increases as the number of the wireless terminals 100 increases, and there is a high possibility that the change is large.

During this phase, the AP's MAC layer 302 separately transmits downlink VoIP MPDUs without applying A-MPDU aggregation. At the end of the initialization phase, the A-MPDU generation phase begins as the AP's MAC layer 302 begins applying A-MPDU integration to the VoIP MPDUs.

FIG. 6 is a diagram illustrating an algorithm 2 of an integrated scheduler operating in the AP 300; FIG.

Each time the MPDU arrives, the MAC layer 302 of the AP 300 stores the destination in the corresponding SwQueue according to the destination of the arriving MPDU.

Here again, the access types are classified as 'AC_VO', 'AC_VI', 'AC_BK', and 'AC_BE', and priority is given to processing in different ways according to categories. The traffic having the highest priority is 'AC_VO' (VoIP traffic) and has a TxQueue [VO] 32 as a TxQueue corresponding to SwQueue [VO] 31 as a corresponding SwQueue.

The MAC layer 302 receives the MPDUs respectively identified by the plurality of unique identifiers from the plurality of source nodes 200 and the MPDUs stored in the corresponding SwQueue are allocated to one specific identifier included in the plurality of unique identifiers Gt; MPDUs < / RTI > This identifier is identified according to where the destination is, and this identification information is obtained from the packet information contained in the arriving MPDU.

The MAC layer 302 maintains the SwQueue (i) of the corresponding access class for each destination node i, and calls the corresponding ScheduleTransmission (i) function for each destination node. The ScheduleTransmission (i) function is called when an MPDU arrives at SwQueue (i) or a BlockACK or ACK is received from destination node i.

The MAC layer 302 receives the MPDUs respectively identified by the plurality of unique identifiers from the plurality of source nodes 200 and the MPDUs stored in the corresponding SwQueue are allocated to one specific identifier included in the plurality of unique identifiers Gt; MPDUs < / RTI > This identifier is identified according to where the destination is, and this identification information is obtained from the packet information contained in the arriving MPDU.

The ScheduleTransmission (i) function waits for MPDU if there is no MPDU in SwQueue (i), and checks if TxQueue is idle when MPDU arrives. The ScheduleTransmission (i) function checks whether the current MAC layer mode is in operation mode when the TxQueue is idle. Here, the operation mode is a mode in which it is confirmed that the A-MPDU generation is possible, and when the first Dma_max is obtained, the operation mode is established.

If the MAC layer mode is not the operation mode, it calls SendMpdu () to immediately store the MPQU of the SwQueue in the TxQueue and transmit it to the destination.

ii) A- MPDU  Generation step ( Downlink )

When the first Dma_max is obtained, the MAC layer mode enters an operation mode, i.e., an A-MPDU generation step.

In the A-MPDU generation step, whenever a new packet or retry packet for stream i (stream to wireless station i) arrives from the wireless station i (200) to the SwQueue of the AP (300) The scheduler MAC layer 302 calculates the approximate end-to-end delay ([De2e [i]) for the head-of-line packet of the SwQueue belonging to the stream i by the following equation (4).

&Quot; (4) "

[De2e [i]] = D [i] + Dma_max + T F [i]

D [i] is the delay amount of the first MPDU of the SwQueue belonging to the stream i, and is the elapsed time until reaching the current AP 300 after the encapsulated RTP packet is generated in the MPDU. The MAC layer 302 of the AP 300 can obtain D [i] by referring to the RTP packet in the first MPDU of the SwQueue and subtracting the timestamp value stored in the header of the RTP packet from the current time. Here, T _F [i] can be calculated from the size and data rate of the A-MPDU as a transmission duration of the A-MPDU can be of a MPDU is currently pending in SwQueue belonging to stream i.

For reference, in calculating the end-to-end propagation delay estimation time ([De2e [i]]), the AP does not use the maximum end-to-end delay value acquired from the RR packet, and the first MPDU of the SwQueue belonging to the stream i is Note that the elapsed time until reaching the AP 300 is considered.

If the following Equation (5) is satisfied, the AP 300 calls SendAMpdu () to construct an A-MPDU using currently available pending MPDUs belonging to stream i.

&Quot; (5) "

[De2e [i]] + Tarr [i] + T _MPDU ≥ 150 ms

Here, Tarr [i] is the arrival interval between VoIP MPDUs destined for the wireless station i, which can be obtained by the AP as the difference of the timestamp values of consecutive MPDUs belonging to the voice packet stream i And T _MPDU represents the time taken to transmit one VoIP MPDU. The left sum of equation (5) provides a future approximate end-to-end delay estimate of the first MPDU when the next new MPDU that belongs to the same voice stream arrives.

Here, the generated A-MPDU is transmitted in a CSMA / CA (Carrier Sense Multiple Access / Collision Avoidance) method by a virtual collision handler 33.

Meanwhile, the physical layer 301 physically transmits the A-MPDU generated by the MAC layer 302.

- Performance evaluation -

The network configuration for performance evaluation is set as follows.

A VoIP session is established between clients located on an 802.11n WLAN and a Gigabit Ethernet LAN. 802.11n APs and Gigabit Ethernet switches are interconnected via a 1 Gbps point-to-point link with configurable transit delay. The wireless stations are located at a distance of 5 m from the AP and the channel conditions are set to ideal. The PHY transmission rate of the AP and the wireless station is set to 600 Mbps. In both the wireless station and AP, the SwQueue size is set to 5000 packets. Each wireless station and wired station pair uses one of the following VoIP codec rates in each experimental scenario.

- 64 Kbps Pulse Code Modulation (PCM): Generates an 80 byte voice payload at 10 ms intervals.

- G.723.1 (5.3 Kbps): 20 byte payload with 30 ms interval.

Generally, in a WLAN, one packet / MPDU is lost / discarded for three reasons.

(1) Some MPDUs are discarded when their corresponding retry counter is exhausted. In the NS-3 simulator's WiFi module, the default retry limit is 7.

(2) If the SwQueue is full, new packets arriving from the upper layer are deleted.

(3) The IEEE 802.11n standard specifies the dot11EDCATableMSDULifetime parameter, which is basically 500 ms. The MPDU of the queue whose lifetime has expired is deleted.

The above three reasons correspond to packets that are not part of an A-MPDU session (i.e., A-MPDU aggregation is not applied). However, when a packet is part of an A-MPDU session, it is not discarded due to exhausted retry limits, but is deleted due to end of life or queue overflow.

FIG. 7 is a diagram illustrating an end-to-end delay performance at a capacity boundary for packets generated in mobile and wired stations using various VoIP codec speeds.

FIG. 7 shows the end-to-end delay performance of the technique according to the present embodiment for 64 Kbps and 5.3 Kbps VoIP codec rates.

In FIG. 7, it can be seen that even for capacity boundaries, all VoIP packets are delivered with an end-to-end delay of less than 150 ms.

The upper limit of a VoIP session with guaranteed QoS depends on the transmission delay and data rate of the VoIP codec.

A 64 Kbps VoIP codec allows 160 VoIP sessions, while a 5.3 Kbps VoIP codec allows approximately 270 VoIP sessions (for example, in a university campus VoIP communication with a transmission delay of less than 5 ms) Is allowed.

Almost 90 VoIP sessions are allowed using the 64 Kbps VoIP codec when the transmission delay is very long, such as intercontinental communication with a transmission delay of almost 100 ms. Almost 100 VoIP sessions are allowed when using the 5.3 Kbps VoIP codec do.

Figure 7 also shows that the end-to-end delay measured at end terminals (i.e., WiFi nodes and wired nodes) is within the end-to-end delay limit of 150 ms, which is a strictly required limitation in QoS guaranteed real-time VoIP communication

As described above, in this embodiment, the A-MPDU integration is adaptively applied to the voice traffic in consideration of the dynamic medium access delay, the transmission delay in the backbone network, and the QoS requirement of the real-time voice traffic in both the wireless station and the AP, As a result of evaluation, all real-time VoIP packets are delivered to the destination with an end-to-end delay of less than 150 ms, but VoIP capacity is 5.3 times better than the existing implementation.

As described above, the adaptive integrated A-MPDU scheduling method according to the present embodiment can be implemented by a program and recorded on a computer-readable recording medium. A program for implementing the adaptive integrated A-MPDU scheduling method according to the present embodiment is recorded, and a computer-readable recording medium includes all kinds of recording devices for storing data that can be read by a computer system. Examples of such computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like. The computer readable recording medium may also be distributed over a networked computer system so that computer readable code is stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the present embodiment can be easily inferred by programmers in the technical field to which the present embodiment belongs.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit and scope of the invention as defined by the appended claims. will be. Therefore, the embodiments of the present invention are not intended to limit the scope of the present invention but to limit the scope of the technical idea of the present invention.

100: wireless terminal
200: destination terminal
300: AP
101, 201, 301: physical layer
102, 202, 302: Mac layer
103, 203: Network layer
104, 204: UDP layer
105, 205: RTP / RTCP layer
106, 206: VoIP application layer
31: SqQueue [VO]
32: TxQueue [VO]
33: Virtual conflict handler

Claims (11)

A communication apparatus for transmitting a VoIP packet in real time in a wireless LAN access network provided with an A-MPDU integration function,
A-MPDU for a plurality of VoIP MPDUs stored in the first queue, using the information obtained from the packet received from the external communication device, and receiving the VoIP MPDU to be transmitted to the destination node, MPDU generation means; And
MPDU transmission means for transmitting the A-MPDU generated by the A-MPDU generation means to a destination
≪ / RTI >
Wherein the A-MPDU generation means comprises:
An end-to-end transmission delay estimation time of a head-of-line MPDU located at a head of the first queue, an inter-arrival time of a VoIP MPDU arriving at the first queue, And generates the A-MPDU when the sum of the time required for transmitting the VoIP MPDU is equal to or larger than the predetermined end-to-end transmission delay time. delete 2. The method of claim 1, wherein the end-to-
MPDU is transmitted from an upper layer of the communication device and is stored in the first queue to a current time, a communication access time of the communication device, a transmission duration of the A-MPDU, Is calculated from the sum of the transmission delays of the first and second antennas. 4. The method of claim 3,
An average time to acquire an end-to-end delivery delay time from a packet received from the external communication apparatus, an average time stored in a second queue after generation of a VoIP MPDU in the end-to-end delivery delay time, And subtracting the continuous average time from the result of subtracting the continuous average time. 5. The method of claim 4, wherein the end-
And a data transfer status checking packet received from the external communication device. A communication apparatus for transmitting a VoIP packet in real time in a wireless LAN access network provided with an A-MPDU integration function,
A-MPDU for a plurality of VoIP MPDUs stored in the first queue, using the information obtained from the packet received from the external communication device, and receiving the VoIP MPDU to be transmitted to the destination node, MPDU generation means; And
MPDU transmission means for transmitting the A-MPDU generated by the A-MPDU generation means to a destination
≪ / RTI >
Wherein the A-MPDU generation means comprises:
Wherein the generation of the A-MPDU is suspended until the first RR packet is received, and the frame is generated and transmitted in units of the VoIP MPDUs stored in the first queue. The method according to claim 1,
Wherein the A-MPDU generation means receives VoIP MPDUs each identified by a plurality of unique identifiers from a plurality of source nodes,
Wherein the plurality of VoIP MPDUs are VoIP MPDUs identified by a specific identifier included in the plurality of unique identifiers. 8. The method of claim 7, wherein the end-
Of-line MPDUs located at the head of the first queue, instead of being acquired in the data transmission status checking packet received from the external communication apparatus. 8. The method of claim 7, wherein the end-
Wherein the head-of-line MPDU located at the head of the first queue comprises a delay time since the generation of the MPDU at any one of the plurality of source nodes to the present, a maximum media access time of the communication device, Is calculated from the sum of the transmission duration times of the first and second communication devices. 8. The A-MPDU generating apparatus according to claim 7,
Wherein the generation of the A-MPDU is suspended until the maximum media access time of the communication device is obtained, and the frame is generated and transmitted in units of the VoIP MPDUs stored in the first queue. A method for real-time VoIP packet transmission in a communication apparatus using a wireless LAN access network provided with an A-MPDU integration function,
Receiving a VoIP MPDU to be transmitted to a destination node and storing the VoIP MPDU in a first queue;
Generating an A-MPDU for a plurality of VoIP MPDUs stored in the first queue using information obtained from a packet received from an external communication apparatus; And
Transmitting the generated A-MPDU to a destination
, ≪ / RTI &
In the step of generating the A-MPDU,
An end-to-end transmission delay estimation time of a head-of-line MPDU located at a head of the first queue, an inter-arrival time of a VoIP MPDU arriving at the first queue, Wherein the A-MPDU is generated when the sum of the time taken to transmit the VoIP MPDU is equal to or greater than the predetermined end-to-end delivery delay time.

Images