KR19990081633A - Wireless Asynchronous Transfer Mode Medium Access Control Layer Protocol - Google Patents

Abstract

The present invention relates to a wireless asynchronous transmission mode medium access control layer protocol, comprising: generating a scheduling table through a W-DWEDF algorithm; A scheduling table alignment step of grouping cells to be transmitted as a virtual channel (VC) unit based on the schedule table information; A packing step of inserting a framehead, an MPDU overhead, and a switching slot for backward / forward switching in a timeframe for transmission of grouped cells; A forward traffic transmission algorithm and a CBR traffic transmission step comprising transmitting a timeframe to the terminal through the forward link; VBR traffic transmission step; ABR traffic transmission step; Based on the operation of the wired asynchronous transfer mode (ATM) switch in the forward transmission by including a reverse traffic transmission algorithm including a UBR traffic transmission step, the wireless side effectively supports the asynchronous transmission mode (ATM) service policy and also wireless physical Minimizing the overhead of the layer and efficiently allocating timeframes in reverse transmission.

Description

Wireless Asynchronous Transfer Mode Medium Access Control Layer Protocol

The present invention relates to a wireless asynchronous transmission mode medium access control (hereinafter referred to as "MAC") layer protocol, and in particular, based on the operation of a wired asynchronous transmission mode (ATM) switch in forward transmission, Packing that provides Wireless Dynamic Weighted Earliest Deading First (W-DWEDF) algorithm that effectively supports Asynchronous Service Mode (ATM) service policy and minimizes overhead of the wireless physical layer. In addition to providing an algorithm, it is possible to efficiently allocate the remaining bandwidth used by Constant Bit-Rate (CBR) and Variable Bit-Rate (VCR) to Available Bit-Rate (ABR) and Unspecified Bit-Rate (UBR) in reverse transmission. The present invention relates to a protocol that effectively supports variable VBR traffic through a dual scheduling method.

In general, a weighted round robin based polling algorithm used in a wired asynchronous transfer mode switch is an algorithm that extends the weighted round robin method to a wireless stage as follows.

Step 1: For each CBR source, the polling token is generated at the base station based on the average arrival time.

Step 2: For each CBR source, the first polling token is generated p seconds after call setup at the base station. Here, p is set to a value satisfying the delay.

Step 3: If the channel becomes empty after transmission, the base station performs the following operations.

Step 3-1: The base station checks the polling token buffer of the CBR source according to the preset priority. If a polling token is found, clear the polling token and poll the CBR source.

Step 3-2: If no polling token is found for the CBR source, check the polling token of the VBR source according to the preset priority. If a polling token for the VBR source is found, the base station polls without clearing the polling token.

Step 3-3: If there are no polling tokens in the CBR and VBR sources, the base station switches to the service for the ABR source or signaling information.

Step 4: When the CBR source is polled, send at least one packet in the RTT buffer.

Step 5: When the VBR source is polled, send one packet from the RTT buffer. If the RTT buffer is empty, an end-of-file signal is sent.

Step 6: When the endupfile signal is received from the VBR source, the base station removes all polling tokens of that source and generates a new token after p seconds.

Step 7: If the ABR source is polled, send at least one packet.

This conventional algorithm is an algorithm that extends the weighted round robin method proposed in the wired asynchronous transfer mode switch to the wireless end. All CBR and VBR traffic have a fixed variable, and the variable value is a token that can serve the link. It increases and decreases as (token) is created.

Since the algorithm operates based on polling, if the overhead due to polling is considered, a performance degradation is expected, and if the physical layer overhead is considered, a large amount of bandwidth may be lost. That is, since the physical layer overhead generally has 1 to 1/2 slots, cell-to-cell transmissions can yield up to 50-60% throughput.

Therefore, the weighted round robin-based polling algorithm cannot support VBR services with variable generation characteristics in asynchronous transmission mode environment due to fixed weights, and can only support CBR attributes. That is, the VBR should be able to represent the peak cell rate (PCR) and mean cell rate (MCR) in addition to the fixed variables. In addition, the weighted round robin method has a drawback that the consideration of ABR and UBR is very poor.

In addition, the following fixed band allocation method is known as a conventional protocol.

Policy 1: Based on the maximum transmission, bandwidth allocation is required from the user, sets up the call if it is acceptable, and blocks the call if the maximum requirement is not met.

Policy 2: A maximum rate based bandwidth allocation is required from the user, and if possible, sets up a call and negotiates with a lower band if the maximum requirement is not met.

Policy 3: A maximum rate based band allocation is required from the user, and if the call is acceptable, the call is set.

This method is proposed to use a fixed band allocation in call setup in a wireless asynchronous transmission mode, and it is divided into three types of fixed band allocation in initial call setup. The first is to request a fixed band by PCR and connect if possible, but to reject the call if it cannot be supported. The second method is to negotiate between the user and the switch between PCR and MCR when PCR cannot be supported. That's how. The third method is a method of accommodating new calls by coordinating allocation bands of other calls currently in communication when there is insufficient support for supporting new calls.

However, this fixed band allocation scheme is the best method in terms of QoS (Quality of Service) delay for traffic without dynamic band change such as CBR traffic, but has PCR and MCR like VBR. There is a problem in that it cannot effectively support the characteristics of traffic having burstness. In addition, in case of traffic where most bands are not used in the entire call setup time such as ABR and UBR, there is a problem that can cause a waste of a lot of bands.

In addition, the conventional PRADOS (Priority Regulated Allocation Dely Oriented Scheduling) scheduling algorithm is a MAC of the Magic WAND (Wireless ATM Network Demonstrator) project of European Advanced Communications Technologies and Services (ACTS). MASCARA (Mobile Access Scheme Based on Contention and Reservation for ATM; hereinafter referred to as "MASCARA") proposes a scheduler called PRADOS. Since PRADOS is basically designed for the construction of test bed systems, it is evaluated as the most realistic protocol among the proposed MAC protocols. An example of the PRADOS algorithm operation of ACTS Magic WAND MASCARA is shown in FIG.

As shown in Fig. 1, MASCARA's PRADOS is operated under the assumption that it has all the information in the reverse direction and the forward direction. Figure 1 (a) shows that PRADOS arranges forward and reverse traffic based on a deadline. This means that the cells are first placed so that the deadline of the link is satisfied as an end point for the buffer of the link in which the cell to be transmitted currently exists. In the case where two links overlap, the cells of the lower priority link are discarded. Priority is defined as the number of tokens of the link, which is created one by one as a fixed interval in the token pool of the link according to the average transfer rate, and decrements by one when one cell is transmitted.

Figure 1 (a) was created based on the deadline, it can be seen that empty slots appear everywhere. In this case, as shown in Fig. 1 (b), the cells are moved to the right in terms of satisfying the deadline. However, it can be seen that this process is performed only in front of the point where the UP and DOWN intersect. The frame structure of PRADOS is divided into frame header, forward transmitter, and backward transmitter, which gives time for up / down transition at the point where the forward transmitter and the reverse transmitter intersect, which is defined as a period of overhead. . In addition, a structure followed by the backward transmitter after the forward transmitter is defined as one frame. Accordingly, it can be seen from FIG. 1C that the overhead period is arranged at the start of the backward transmission unit. In FIG. 1 (d), it can be seen that the forward cells disposed after the backward transmission unit are moved ahead of the overhead time. This shows that the forward transmission cells interspersed within one frame are rearranged from the standpoint of supporting the deadline. Finally, in FIG. 1E, slot waste is suppressed by removing empty slots between the forward and the overhead periods. In addition, the frame header (FH) field, which indicates the start of a new frame, is inserted at the time when the next forward transmitter appears after the backward transmitter. In addition, an empty slot portion between the uplink transmitting end and the FH field is allocated as a contention slot, and finally, the backward slots intended to be transmitted in the next frame are moved to the backward portion of the current frame within the allowable range.

In the reverse case of the PRADOS scheduling algorithm, it is difficult to distinguish CBR / VBR / ABR / UBR traffic. Therefore, it is virtually impossible to effectively support the asynchronous transport mode traffic service policy. However, in the case of the forward transmission method of PRADOS, the token is generated based on the average transmission speed in the same way as the weighted round robin based algorithm, and the cell is transmitted according to the number of tokens. There is a problem with the poem. In addition, since VBR and CBR traffic always assume the arrival of certain traffic, there is a deficiency that cannot be prevented when congestion or uncertain system characteristics occur in the actual system. In addition, since CBR, VBR, and tokens are lowered below 0, there is a drawback that the ABR and UBR can not be transmitted due to the uncertain CBR and VBR. In conclusion, PRADOS has the most feasible structure, but in reality, there is a problem that it may not be able to support even legitimate traffic transmission of ABR and UBR by giving CBR and VBR traffic too high priority. In addition, the biggest feature of PRADOS is the method of allocating slots by calculating the deadline of each link in advance, but in this case, it has a structure similar to EDF (Earliest Deadline First) because it has a property that loses the advantage in the post-processing process. However, due to the cell-based calculation method, it requires too much complexity and thus has a low practicality.

The main disadvantage of PRADOS of MASCARA is a contention area that UEs need to transmit a channel assignment signaling message to a base station for transmission of reverse traffic. Currently, MASCARA has not been able to suggest a concrete plan for this, and in the worst case, it is required to allocate the contention area as many as the number of terminals.

Particularly, in order to support effective performance, half of the number of terminals is allocated to the competition area, but one of the big problems related to this is that the size of the competition area becomes so large that it cannot be ignored. . For example, consider 5ms that MASCARA uses as the CDT (Cell Delay Tolerance; hereinafter referred to as "CDT") of the radio section of real-time CBR and VBR traffic, and the frame size, which is a cycle concept composed of multiple cells, is the minimum to satisfy the CDT. Assuming 5ms, the required level, a 5ms frame can transmit a total of 242 slots in one frame due to the MASCARA system configuration. In particular, the MASCARA requires about one slot size of physical layer overhead in order to transmit scheduling information of the MAC and the UE. Therefore, even in an environment where the number of channels is only 50, considering each contention slot overhead, in the worst case, about 100 contention areas are required, and the overhead reduction through the throughput reaches 40%. If the physical layer and MAC layer overhead incurred in the transmission of traffic slots are added to this, the performance of the system is reduced to a very low level. In particular, MASCARA has a very large complementary element because there is no specific plan for the composition of the competition area. In addition, it supports a variable length frame structure, so it is very difficult to support a function such as a sleep mode for power reduction.

The present invention has been invented to solve various defects and problems of the algorithm in the conventional asynchronous transmission mode in view of the above situation, and effectively implements the service policy of the wired asynchronous transmission mode network using the W-DWEDF algorithm for forward transmission. The purpose of the present invention is to provide a wireless asynchronous transmission mode MAC layer protocol that minimizes the overhead incurred during cell transmission by the packing algorithm and performs QoS support through dual scheduling for reverse transmission. There is this.

1 is a view for explaining the operation of the conventional PROADOS scheduling algorithm,

2 is a frame structure employed in the present invention,

3 is a system configuration employed in the present invention,

4 is a conceptual diagram of a forward MAC protocol algorithm based on the present invention WDWEDF and packing;

5 is a flowchart of a W-DWEDF scheduling algorithm according to the present invention;

6 illustrates an embodiment of a W-DWEDF scheduling algorithm according to the present invention;

7 illustrates an embodiment of a W-DWEDF scheduling algorithm according to the present invention;

8 is a diagram illustrating an embodiment of a packing algorithm according to the present invention;

9 is a conceptual diagram of a reverse MAC protocol algorithm according to the present invention;

10 is a view showing a transmission scheme of a reverse CBR trap pick according to the present invention,

11 is a diagram showing a method of transmitting reverse VBR traffic according to the present invention;

12 is a diagram illustrating a method of transmitting reverse ABR / UBR traffic according to the present invention.

<Description of the code | symbol about the principal part of drawing>

1: base station 2: terminal

3: wired network interface module 4: output buffer

5: Input buffer 6: Main controller

7: 2 level scheduler 8: Receive frame controller

9: Scheduler Table 10: W-DWEDF Scheduler

11: Packing Module 12: Time Frame Controller

13: wireless transmitter 14: wireless receiver

15: wireless transmitter 16: wireless receiver

17: time frame controller 18: receiving frame controller

19: Level 2 scheduler 20: Input buffer

21: output buffer 22: terminal interface module

23: main controller

The present invention for achieving the above object comprises the steps of generating a scheduling table through a W-DWEDF (Wireless Dynamic Weighted Earliest Deading First) algorithm in the forward traffic transmission information transmitting from the base station to the wireless terminal; A cell grouping step of grouping cells to be transmitted in units of a virtual channel (VC) based on the information of the schedule table; A packing step of inserting a framehead, a medium access control protocol data unit (MPDU) overhead, and a switching slot for switching a reverse / forward link in a timeframe for transmitting the cells grouped in the cell grouping step; And transmitting a timeframe to the terminal through the forward link.

In addition, the present invention, in the reverse traffic transmission for transmitting information from the wireless terminal to the base station, CBR (Constant Bit-Rate) traffic, VBR (Variable Bit-Rate) traffic, ABR (Available Bit-Rate) traffic, UBR (Unspecified Bit) A traffic selection step of selecting at least one of the same traffic or different traffic among the traffic; Characterized in that the information transmission step of transmitting information to the traffic selected in the traffic selection step.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

2 is a frame structure diagram employed in the present invention. The present invention basically operates based on a fixed length time-frame. By using a fixed size timeframe, power control of the terminal is easy, and scheduling at the access point is more effectively performed. The present invention also operates based on time division multiple access (TDMA).

The FH is a scheduling information area including information on which UEs have fixed length slots belonging to a timeframe, reverse / forward link information, and allocation information for CBR, VBR, ABR, and UBR services. Accordingly, the terminals acquire information related to the slot of the corresponding timeframe through the FH. Next, after the FH is transmitted, a forward link section for transmitting information from a base station (AP or BS) to terminals is followed, and slots associated with the reverse link appear after the forward link. In the reverse link interval, slots supporting CBR appear first, followed by slots allocated to VBR. In this case, they are used for the minimum bandwidth allocation of the VBR. After the slot allocated to the VBR, additional scheduling information is transmitted from the access point to the terminals according to the additional slot allocation request requested by the terminal in the form of piggyback when transmitting VBR traffic. To reschedule slots from the time frame to the end of the timeframe.

Therefore, additional VBR reverse link transmission slots according to rescheduling are allocated, and then the remaining band is allocated for transmission of ABR and UBR traffic. Finally, a control slot section for requesting call establishment of a new call and channel reservation of ABR and UBR traffic appears. At this time, the interface between all slots has a variable size that can be moved, and the information transmission in the terminal is made of a unit of MPDU (MAC Protocol Date Unit; MPDU), which combines several cells, and the MPDU is a physical layer and It consists of a plurality of asynchronous transmission mode cells with a header containing some information of the MAC for carrying piggyback information.

3 is a system configuration diagram employed in the present invention, which is defined as elements in a base station (AP or BS; 1) and a terminal (2). In the base station 1, there is a wired network interface module (Wired Network Interface Module) 3 for communicating with a wired network. Information transmitted and received through this is stored in the buffers (input buffer 5 and output buffer 4) allocated to each link.

And the main controller (Main Controller) (6) that performs the actual work is composed of the following elements in the base station (1).

That is, in the reception frame controller 8 and the base station 1 performing the function of transmitting the frames received at the second level scheduler 7 and the terminal 2 to the network interface module for controlling the two-level scheduling for reverse transmission. When the traffic is transmitted to the terminal 2 in the forward direction, a scheduler table prepared by the W-DWEDF scheduler 10 and the W-DWEDF scheduler 10 that performs the W-DWEDF algorithm to create a scheduler table 9 It is composed of a packing module 11 for performing a packing process (Packing Table) and a time frame controller 12 for transmitting and receiving a fixed length of the time frame as an easy form for transmission.

In addition, the base station 1 includes a radio transmitter 13 and a radio receiver 14 for transmitting and receiving radio terminals. The terminal 2 also includes radio transceivers 15 and 16 for transmission and reception of the main radio terminal, and has a terminal interface module 22 for interfacing with the mobile terminal. In addition, the input and output buffers 20 and 21 for buffering traffic with the terminal interface module 22 are provided, and the main controller 23 of the terminal 2 is connected to the main controller 6 of the base station 1. Compared with a relatively small scale, the W-DWEDF module and the packing module, which are used in the forward transmission, are not included as components, and are received by the two-level scheduler 19 and the terminal that control the two-level scheduling for the reverse transmission. It consists of a receiving frame controller 18 for transmitting the frames to the network interface module 18 and a time frame controller 17 for transmitting and receiving a fixed length of the time frame.

4 is a conceptual diagram of a forward MAC protocol algorithm based on the present invention WDWEDF and packing, in which the cells are stored in a temporary buffer of each link as cells are received from the network, and the stored cells are periodically generated. Is selected as cells to be transmitted according to the W-DWEDF algorithm at the time of generation, which in turn undergoes a packing procedure to reduce the physical transmission overhead.

In the packing procedure, groups cells to be transmitted as a unit of a virtual channel (hereinafter referred to as "VC") based on scheduling table information created through W-DWEDF, and the grouped cells are again framed for actual transmission. Insert the FH), the MPDU overhead, and the change-period for up / downlink switching. The generated timeframe is transmitted to the terminals through the forward link.

In the present invention, the forward traffic transmission scheme uses a deadline based EDF (Earliest Deadline First) algorithm to minimize delay and reduce the size of the buffer required by the algorithm. This suppresses the occurrence of congestion in the wireless asynchronous transmission mode network environment, efficiently supports the CBR / VBR / ABR / UBR traffic, and guarantees the reliability of the asynchronous transmission mode traffic service. In addition, the W-DWEDF algorithm in the present invention is to support VBR, CBR, ABR and UBR traffic, and each connection state is divided into five types. Description of each state is as follows. Here, P means a Peak Cell Ratio (PCR) value of a connection mapped to a cycle, and M means a Mean Cell Rate (MCR) value.

○ State 1 (S1): P> 0. When M> 0, if there is a cell requesting transmission, the cell is transmitted in the next round.

○ State 2 (S2): When P> 0 and M = 0, if there is a cell requesting transmission, it shall be transmitted only when no connection exists in the S1 state.

○ State 3 (S3): When P = O, M = 0, all the allocated bandwidth is used and it has the transmission qualification in the next cycle.

State 0 (S0): State for ABR connections.

Status × (S ×): Status for UBR connections.

In the W-DWEDF algorithm, each virtual channel connection (hereinafter referred to as "VCC") has the following state information to support QoS related to bandwidth allocation. In this case, VCn is the nth VCC connection, and the variables fixed when the connection is established are as follows.

○ VCn. service: The service class of the VCC.

○ VCn. peak: PCR value of the VCC link.

○ VCn. mean: Mean Cell Rate (MCR) value of the VCC connection.

Unlike fixed values, state variables that change dynamically during the execution of the W-DWEDF algorithm are as follows.

O VCn.cpr: Current PCR value of the VCC based on cycle value.

○ VCn.cmr: Current MCR (Mean Cell Rate) value of VCC based on cycle value.

○ VCn. state: The current state of the VCC (S1, S2, S3, S0, S ×).

In this case, cpr and cmr are decremented by one each time the connection receives a cell transmission, and are initialized again with PCR and MCR (Mean Cell Rate) values based on the cycle size when the cycle (C) set in units of cells starts.

The cycle size in the W-DWEDF scheme can only support the effects of EDF considering VBR and CBR, and can guarantee the reliability of ABR traffic based on the total bandwidth. In the case of setting the cycle size and P and M for each VCC in consideration of the total bandwidth, an example of calculating cpr and cmr values for each VCC connection is as follows. That is, if C, the cycle size, is 100, the number of VCCs connected to the output link is N, and if the total bandwidth of the output link is given by χ bps, VCn.peak is ψ / if the PCR of the nth connection is ψ bps. Assign it to χ.

5 is a flow chart of the W-DWEDF algorithm, which divides each channel state into VBR, CBR, and ABR, and provides a limited service in one cycle by assigning P and M values to each channel. I'm trying to. Therefore, S1 (P> 0, M> 0), S2 (P> O, M = 0), S3 (P = 0, M = 0) states are defined through P and M values as in the above state classification. In case there is no connection in S1 state, S2 state connections are supported. However, the W-DWEDF algorithm ensures stable ABR traffic service through resource allocation based on the total bandwidth. In other words, when the cycle size is allocated around the entire bandwidth, the ABR service is supported to be reliable. Therefore, for this, we have SO and SX which are states for ABR and UBR services respectively.

As shown in Fig. 5, when the cells arrive from the input VCCs in the W-DWEDF algorithm, the cells are stored in a temporary buffer, and then the cells of the input VCC having the cells closest to the deadline according to the EDF technique in one cycle. Is stored first in the scheduling table. This can support efficient delay performance. The transition from the S1 state to the S2 state in relation to the state transition is most of the VBR traffic because the PCR value of the VBR is generally larger than the mean cell rate (MCR).

On the contrary, when the S1 state is converted to the S3 state, most of the CBR traffics are mostly because the PCR of the CBR traffic is the same as the MCR. Descriptions of functions appearing in relation to algorithms are as follows.

Bp (n): Passes the dynamic P (VCn.cpr) value of the nth VCC connection.

Bm (n): Passes the dynamic M (VCn.cmr) value of the nth VCC connection.

Sort (n): Sorts the cells waiting to be sent for connections in the S1, S2, S3, SO, and SX states in EDF order.

Size (n): conveys the number of VCCs in the n-state.

Pop (n): Delivers the connection number to be preferentially serviced among the VCCs in the n-state.

Visit (n): Stores one asynchronous transmission mode cell in the scheduling table in the n-th VCC.

Increase_rate (n): VCn of the nth VCC. cmr, VCn. Decrease cpr value by one.

As can be seen from the above algorithm, during the cycle, each connection is serviced according to P, M values. When there is a cell to be transmitted to one or more VCCs, the cell is selected according to the EDF. In VBR and CBR connections, the cell is started in the S1 state and waits until the MCR (Meam Cell Rate) is satisfied. If all the connections satisfy the MCR, additional support is performed for the nodes in the S2 state by the number of times (PCR-MCR) to expect the effect of statistical multiplexing. In addition, if there are no cells waiting for VBR and CBR connection, ABR and UBR support should be performed.If a cycle is set based on the total bandwidth, the remaining area other than the bandwidth allocated to VBR and CBR is displayed. Used to service UBR traffic.

In the operation of the W-DWEDF algorithm, it is assumed that the buffer state of each link at the time of generating the time frame is as shown in FIG. 6. In this example, the total number of links is assumed to be five, and it can be seen that each has the characteristics of CBR / VBR / VBR / ABR / UBR. In this case, 5/6/3/3/2 cells are stored in the buffer of each link and are waiting for transmission to the terminal. Assuming that the deadline time of each link is the same, a scheduling table using the W-DWFDF algorithm can be created based on the given example.

Information required for scheduling table generation requires the number of traffic slots that can be transmitted through the forward link, which may be determined in consideration of the proper split ratio of the forward and reverse links when generating a timeframe in the scheduler. That is, it may be determined based on the allocation request and allocation cycle characteristics of the reverse link and the transmission traffic latency of the forward link.

FIG. 6 is a diagram illustrating an embodiment of a W-DWEDF scheduling algorithm according to the present invention. Assuming that the number of traffic slots that can be transmitted through a forward link is 22, buffers in FIG. 6 implement the W-DWEDF algorithm as shown in FIG. Through scheduling table can be. In this case, it is assumed that P and M values of CBR / VBR traffic are 6 or more. It is assumed that VC1 / 2/3 are all in the S1 state. First, since the W-DWEDF algorithm supports the VC1, VC2, and VC3 links of the S1 state, the cells for each link in the table can be seen to be stored in the scheduling table. If support for ABR traffic is available after support for CBR and VBR traffic, the W-DWEDF algorithm transmits cells of VC4, which is ABR traffic. If there is still time to transmit a traffic slot within the timeframe, the W-DWEDF algorithm may transmit cells of UBR traffic VC5.

When creating the scheduling table, it is necessary to check whether the number of available forward traffic slots is exceeded as the number of available slots is exceeded as the currently transmitted cell is added. The capacity of the head and the cell should be calculated. Therefore, in FIG. 7, it can be seen that up to 21.5 slots are used when the number of available forward slots is 22.

The scheduling table created by the W-DWEDF algorithm needs to go through a packing process that is rearranged to generate a timeframe. The reason why the packing is performed may be that smooth performance is not supported due to the overhead of the radio physical layer when transmitting at each cell unit, such as a wired end, in the wireless end. This is because the scheme of the present invention has a physical slot overhead of 1/2 slot size when transmitting one cell in the worst case. Therefore, the transmission for the same UE transmitted in one timeframe may minimize physical layer overhead when multiple cells are transmitted together.

FIG. 8 illustrates an example of packing of the scheduling table created in FIG. 7. The first step of the packing algorithm is to group cells of the scheduling table created by the W-DWEDF algorithm according to a link. The grouped cells are formed in a structure corresponding to the beginning of the time frame by inserting a frame header, an MPDU header, and an up / down link transition section through a second step. Then, one part of the time frame is completed by adding a portion for reverse transmission, and is transmitted through the forward link.

FIG. 9 is a conceptual diagram of a reverse wireless asynchronous transmission mode MAC layer protocol algorithm according to the present invention. The present invention is basically designed to transparently extend a wired end asynchronous transmission mode traffic support scheme to a wireless end. First, to support CBR service, a fixed bandwidth is periodically allocated. In the case of CBR, fixed channel allocation is appropriate due to the characteristics of the service as it uses a fixed band. FIG. 10 is a diagram illustrating a CBR traffic transmission scheme according to the present invention. In the case of CBR traffic, CBR slot allocation information from an AP is transmitted to a UE through an FH field of a timeframe. Information is transmitted to a base station (AP) by using a sized and periodically generated reverse link transmission slot.

For VBR service, a dual scheduling scheme is used to support PCR. In this scheme, after allocating only the minimum number of slots required by the VBR, an additional traffic slot allocation is performed once more by using the piggyback information of the traffic transmitted on the reverse link. This prevents channel wasting of unused slots caused by unpredictable traffic volume. In addition, it effectively prevents congestion intervals that occur when PCR is concentrated. In addition, by assigning slots through scheduling for resource allocation in the AP, it is possible to guarantee priority for traffic to be processed quickly. Accordingly, the present invention supports the 'pre-minimum slot allocation, post-add slot allocation' scheme to support the bandwidth requirement of MCR or more to be suitable for QoS in a short time and to minimize the slot waste caused by the bandwidth requirement of MCR or less. Support.

FIG. 11 is a diagram illustrating a VBR traffic transmission method according to the present invention, wherein a UE to transmit VBR traffic acquires information related to a minimum allocation band through an FH from a base station (AP), and selects a corresponding reverse link slot. The traffic is transmitted to the base station (AP) through. In this case, the UE, which needs to additionally transmit additional information, includes an acknowledgment information in the transmitting MPDU to request an additional slot allocation from the base station (AP). The base station AP receives the additional scheduling information again. Accordingly, the terminals requesting the additional slots transmit additional traffic to the base station AP. In the case of VBR, the base station (AP) must manage information related to the slot size and the traffic generation period used by the channel for the minimum slot allocation.

In support of ABR service, it supports Demand Assignment. In this case, as a scheme for transmitting the control channel allocation request from the mobile station MT to the base station AP, a contention scheme and a polling scheme may be considered. In the present invention, a competition based solution is considered first. The ABR transmission request is stored in the scheduling information tabled in the AP, and the base station (AP) supports the transmission of ABR traffic in a scheme of accommodating QoS such as delay tolerance after the service of CBR and VBR traffic. . In addition, it provides reliable service by supporting crash-sensitive ABR traffic in a non-collision manner.

In the case of ABR traffic shown in FIG. 11, when traffic arrives at the terminal, the terminal stores the ABR traffic in its own buffer. After that, the UE that needs to transmit the ABR traffic transmits the transmission reservation of the ABR traffic to the base station (AP) through the frame header, and waits for a channel allocation command from the base station (AP). In this case, as a scheme for requesting allocation, a scheme based on competition of reservation information transmission and polling may be considered. When slot allocation information for transmitting ABR traffic is received from the base station (AP), the terminal transmits traffic through the corresponding slot. To this end, there is a table for scheduling ABR traffic in the base station (AP), and the slots are allocated according to the priority while checking the delay and QoS of the traffic through the table.

In supporting UBR service, the scheduling policy of asynchronous transmission mode with lower priority than ABR is supported. UBR is also supported as a similar structure to ABR. As a scheme for transmitting a control channel allocation request from a mobile station (MT) to a base station (AP), a contention scheme and a polling scheme are considered, and priority transmission of reservation information is considered. Similar to ABR, the UBR transmission request is supported in a scheme of accommodating QoS of the UBR service through information tabled in a base station (AP).

12 is a diagram illustrating a method of transmitting UBR traffic according to the present invention, and UBR traffic is also supported as a method similar to ABR. That is, like ABR, reservation information is transmitted to the base station AP, and information is transmitted through a slot allocated by the base station AP according to the corresponding information. In the base station (AP), separate tables are managed to support UBR, and the remaining bands used by CBR, VBR, and ABR are allocated to UBR.

The present invention can be used as a method of minimizing the transmission overhead and delay of the asynchronous transmission mode cells in the wireless interface of the wireless asynchronous transmission mode environment, as well as the third generation movement such as IMT2100, FPLMTS, UMTS as well as the wireless asynchronous transmission mode environment In the case of providing a higher level of services transparently to support wired and wireless interworking in a communication system, it can be effectively used as a transmission method for improving the efficiency of a wireless channel. It is supported through the exchange of signal information.

In addition, the present invention can be used as a method of multiplexing multiple cells or packets when applying a wireless environment of a packet or cell based system.

As described above, the W-DWEDF scheme of the present invention effectively supports not only VBR and CBR, but also ABR, which has flexibility in bandwidth, and improves performance by reducing delay and reducing queue size by scheduling cells based on deadlines. ABR traffic, which is sensitive to loss and requires a large amount of buffers in the switch, is pushed out of priority due to excessive VBR and CBR traffic, thereby preventing unserviced situations and ensuring reliable service. have. In addition, stable support of ABR traffic is loss-sensitive, but has a lower priority than VBR and CBR, thus reducing the buffer size required in the switch.

In addition, the wireless asynchronous transmission mode MAC according to the present invention supports a packing algorithm for effectively using limited radio resources by minimizing the physical layer overhead during wireless transmission, thereby providing an optimal link as in the wired end in case of high burst traffic. The wireless asynchronous transmission mode forward link transmission algorithm can effectively and transparently extend the asynchronous transmission mode service from the wired asynchronous transmission mode to the wireless end. It is effective to support effectively. In particular, the wireless asynchronous transmission mode reverse link transmission algorithm can be effectively used for wireless data service using LAN and ABR. It has the advantage of providing QoS guarantee of VBR and priority support method for ABR and UBR.

In addition, the algorithm of 'pre-minimal allocation and post-additional allocation' of the reverse transmission scheme of the present invention eliminates wasted slots and bandwidth consumption due to collisions due to dynamic characteristics of the VBR traffic and reverse transmission scheme. In the case of ABR and UBR, there is an advantage in that it can effectively support the service policy of the asynchronous transmission mode to use the remaining bandwidth used by the CBR and VBR by performing band allocation through scheduling of the base station according to the request of the terminal.

Claims (9)

In the forward traffic transmission for transmitting information from the base station to the wireless terminal, generating a scheduling table through a Wireless Dynamic Weighted Earliest Deading First (W-DWEDF) algorithm; A cell grouping step of grouping cells to be transmitted in units of a virtual channel (VC) based on the information of the schedule table; A packing step of inserting a framehead, a medium access control protocol data unit (MPDU) overhead, and a switching slot for switching a reverse / forward link in a timeframe for transmitting the cells grouped in the cell grouping step; And transmitting the time frame to the terminal through the forward link. The method of claim 1, wherein the generating of the scheduling table comprises changing each channel state into a constant bit rate (CBR), a variable bit rate (VBR), an available bit rate (ABR), and an unspecified bit rate (UBR). Dividing step; Define the PCR (Peak Cell Rate) value mapped to the cycle as P and the MCR (Mean Cell Rate) value as M, so that S1 (P> 0, M> 0), S2 (P> O, M = 0), S3 (P = 0, M = 0) and the status of each of the ABR and UBR services to define the S0 state and SX state, characterized in that the wireless asynchronous transmission mode medium access Control Layer Protocol. 3. The method of claim 2, wherein the S1 state transmits a cell in the next round when there is a cell requesting transmission, and the S2 state is a cell only when there is no connection existing in the S1 state when there is a cell requesting transmission. The S3 state is a state in which the allocated bandwidth is used up, and has a transmission qualification in the next cycle. The S0 state and the SX state are states for ABR connections and UBR connections, respectively. Asynchronous Transfer Mode Medium Access Control Layer Protocol. 2. The method of claim 1, wherein the cell grouping step uses an Earliest Deadline First (EDF) algorithm point to group cells to be transmitted in units of virtual channel VCs, with transmission deadlines prioritizing them in a fast cell order. Asynchronous Transfer Mode Medium Access Control Layer Protocol. In reverse traffic transmission that transmits information from a wireless terminal to a base station, CBR (Constant Bit-Rate) traffic, VBR (Variable Bit-Rate) traffic, ABR (Available Bit-Rate) traffic, UBR (Unspecified Bit-Rate) traffic A traffic selection step of selecting at least one of the same traffic or different traffic; Wireless asynchronous transmission mode medium access control layer protocol, characterized in that the information transmission step of transmitting information to the traffic selected in the traffic selection step. The method of claim 5, wherein when the traffic selected in the traffic selection step is CBR traffic, transmission of information transfers CBR slot allocation information from the base station (AP) to the mobile station (MT) through a frame header (FH) field of a time frame. Making a step; And transmitting the information to the base station (AP) using a reverse link traffic slot periodically assigned to the mobile station (MT) and having a fixed size. The mobile terminal (MT) of claim 5, wherein when the traffic selected in the traffic selection step is VBR traffic, information related to a minimum VBR slot allocation band is transmitted from a base station (AP) through a frame header (FH) field. Delivering to; Transmitting information to a base station (AP) by using a traffic slot of a reverse link in which the mobile station MT is fixed size and periodically allocated; Requesting, by the mobile station MT, additional slot allocation from the base station AP, which additionally needs to transmit additional information; Transmitting VBR slot allocation information, which is additional scheduling information, from the base station (AP) to the mobile station (MT); The mobile terminal (MT) is a dual scheduling comprising the step of transmitting information to the base station (AP) by using the additionally allocated traffic slot of the reverse link, the wireless asynchronous transmission mode medium access control layer protocol. The method of claim 5, wherein when the traffic selected in the traffic selection step is ABR traffic, traffic is transmitted to the mobile station MT, and the ABR slot allocation information is transmitted from the base station AP. Transmitting to the mobile terminal MT through the mobile terminal; Requesting, by the mobile station MT, a base station (AP) for an allocated reverse traffic slot; Transmitting ABR slot allocation information from the base station (AP) to the mobile station (MT) through a frame header (FH) field; The mobile terminal (MT) comprises a step of transmitting information to the base station (AP) through the ABR slot, the wireless asynchronous transmission mode medium access control layer protocol. 6. The method of claim 5, wherein when the traffic selected in the traffic selection step is UBR traffic, the information is transmitted to the mobile station MT, and the UBR slot allocation information is transmitted from the base station AP to the frame header (FH) field. Transmitting to the mobile terminal MT through the mobile terminal; Requesting, by the mobile station MT, the base station (AP) for the request allocation reverse traffic slot; Transmitting the UBR slot allocation information from the base station (AP) to the mobile station (MT) through a frame header (FH) field; The mobile terminal (MT) comprises the step of transmitting information to the base station (AP) through the corresponding UBR slot, the wireless asynchronous transmission mode medium access control layer protocol.