KR20020049786A - The Transmission Interface Device Of Wireless High Speed Physical Layer For Wiress LAN System - Google Patents

Abstract

PURPOSE: An interface device and method for high-speed physical layer transmission suitable for a wireless LAN(Local Area Network) is provided to quickly and efficiently perform an interface with a wireless physical layer by composing a multiple transmission and receipt queue classified according to a transmission rate and processing data through N addressing using a module with a circular buffer type. CONSTITUTION: A demultiplexer(611) classifies inputted data into a plurality of packets according to a transmission rate and inputs the classified data to each transmission circular buffer(613a-613c) according to the control of a transmission buffer controller by the transmission rate. A multiplexer(612) transmits the data stored in each transmission circular buffer(613a-613c) to a wireless physical layer(200) according to the control of the transmission buffer controller. The transmission buffer controller controls the demultiplexer(611) and the multiplexer(612). A demultiplexer(621) classifies the data inputted through the wireless physical layer(200) into a plurality of packets according to the transmission rate, and inputs the classified data to each receipt circular buffer(623a-633c) according to the control of a receipt buffer controller(620) by the transmission rate. A multiplexer(622) transmits the data stored in each receipt circular buffer(623a-633c) to an MAC(Medium Access Control) controller(300) according to the control of the receipt buffer controller(620). The receipt buffer controller(620) controls the demultiplexer(621) and the multiplexer(622).

Description

Interface device and method for high speed physical layer transmission suitable for wireless LAN {The Transmission Interface Device Of Wireless High Speed Physical Layer For Wiress LAN System}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to high-speed wireless LAN data communication, and more particularly, to a wireless physical medium of a MAC device of a wireless mobile terminal device connected to Ethernet through a wireless LAN access point (AP). A wireless high speed physical layer transport interface device and method for interfacing with a layer are provided.

As the industrial society develops, the volume of information is increasing day by day, and the user's desire to use the vast amount of information quickly and accurately has also increased. Therefore, high-speed data transmission technology is inevitably required for rapid and accurate exchange of such a large amount of information. In addition, the recent advances in circuit and component technology, the availability of frequency bands that can be used without permission, and the proliferation of portable computers have increased interest in wireless LANs, which guarantee some mobility and transmit high-speed data. Therefore, many wireless LAN products are developed and distributed by vendors.

In the early stages of wireless LAN development, no standard architecture existed. In recent years, standardization based on the architecture recommended by the IEEE 802.11 Committee has been conducted as one of the standardization of wireless LANs. Wireless physical layer transmission technology of a wireless LAN can be broadly classified into a transmission method using a spread spectrum transmission method, an infrared transmission method, a frequency hopping method, and an OFDM modulation and demodulation method.

The Media Access Control (MAC) protocol of wireless LANs should be designed to be used in a variety of different transmission technologies. In general, a plurality of users share one wireless channel in a wireless LAN, and thus, if the user does not properly control the transmission between users, information collision between users on the wireless channel may result in information transmission or performance deterioration. . As described above, a technique for controlling the use authority for a transmission medium shared by multiple users is called a MAC.

The main functions of the general wireless LAN system are as follows.

First, it is a frame (MSDU: MAC Service Data Unit) transmission service. It must accommodate both asynchronous data transfer services that are insensitive to transmission delays and synchronous data transfer services requiring frame transmission delays within a certain time.

Second is the acceptance of multiple physical layers. When selecting the wireless physical media (spread spectrum transmission method, infrared transmission method, frequency hopping method, OFDM modulation and demodulation method) to be used, the transmission rate (1M ~ 56Mbps) and transmission error occurrence rate (the target is 10 Level 8) should be considered.

Third is the mobility of the terminal device. Consideration should be given to the scope of the service provision area and the type of communication of the terminal (fixed, semi-fixed, and moved) and the speed of travel (walking, traveling speed, etc.).

Fourth, the determinism and network interconnection of the service providing area. Basic configuration should be easy and extendable to Extended Service Set (ESS) using distribution system and Access Point (AP). In addition, the connection with the existing LAN and WAN should be easy.

Fifth, coexistence with other wireless LANs or wireless systems. Even when multiple wireless LANs are used in duplicate, each operation must be done independently. In this case, sufficient consideration should be given to contention of transmission channel acquisition and interference between service areas. As a security issue, information encryption methods (such as those performed over communication protocols and methods at the application layer) and access from unauthorized terminals should be prevented.

Sixth, the low power consumption that can be driven by the battery in consideration of the operation in the mobile terminal, for this purpose should consider the power saving mode.

In the standard specification of wireless LAN determined by the IEEE 802.11 committee, the MAC layer controls competition for transmitting data through co-transmission, that is, to efficiently use a channel shared by multiple terminals with minimum interference and maximum performance. It controls and detects the abnormality of the transmission path.

Transmission of a frame (MSDU: MAC Layer Service Data Unit) using the bit serial transmission function provided by the physical layer is no different from the wired LAN, which is the most basic function of the media access control method. In a wired LAN, the problem of viewing a cable as one channel and sharing it with a plurality of terminals is a major problem of the MAC layer. By the way, it is assumed that a wireless medium has a plurality of channels by using a modulated signal. Therefore, in the MAC layer of the wireless LAN, not only a single channel sharing method but also multiple channel management are required. Like a wired LAN, it is necessary for one user to send a frame to a terminal, broadcast to all terminals, and broadcast to a terminal belonging to a specific group.

The time it takes for a transmitting terminal to transmit a frame to be received at a user terminal is called a frame transmission delay. In case of transmission delay, there are asynchronous service and synchronous service for transmitting MAC layer frame of wireless LAN. In addition, when a signal collides, a signal having a large reception strength is normally received, and a signal having a small reception strength disappears. Since the quality of wireless media is inferior to wired media, the frequency of frame transmission errors including bit errors is high.

Therefore, a more powerful error detection function is required in the MAC layer of the wireless LAN. The MAC layer of a wired LAN generally has an error detection function, and recovery functions such as retransmission are performed at a higher protocol layer. However, the MAC layer of the wireless LAN requires error recovery as well as error detection. Whether the error recovery function is included in the MAC layer plays an important role in improving the transmission quality in the physical layer.

The MAC method of the wireless LAN includes a TDMA-based method used in satellite communication and a packet radio-based method called ALOHA or Carrier Sense Multiple Access with Collision Avoidance (CSMA).

The MAC of IEEE 802.11 wireless LAN basically operates with CSMA / CA, which is a frame interval control method, and uses RTS / CTS and network location vector to solve the hidden terminal problem. Based on distributed control method, central control method by polling is also used as an option.

The protocol selected by IEEE 802.11 is a CSMA / CA protocol, which is different from the CSMA / CD scheme used in a wired local area network. In wireless communication, unlike wired, collision between packets cannot be detected, so a CSMA / CA scheme for avoiding collision is required.

The MAC protocol is divided into contention service and contention free service. Competitive service is a general computer communication service that handles asynchronous data such as file transfer, and is processed by DCF (Distributed Coordination Function). DCF is based on CSMA / CA method, and non-competitive service is a service for delay sensitive transmission such as voice and video, and is processed by PCF (Point Coordination Function). The MAC protocol of IEEE 802.11 is designed to coexist with competing services and non-competitive services.

In the case of a wireless channel, collision detection is difficult because the received energy is tens of dBs smaller than the transmit power. Therefore, the basic operation of CSMA / CA is the same as CSMA / CD, and can be classified into three types as follows in a manner that avoids collision in advance.

First, p-persistent CSMA changes p according to traffic conditions.

Second, there is a difference in frame interval or retransmission interval according to priority.

Thirdly, it is determined whether or not a collision occurs by examining whether or not a pulse from a node which is trying to send a packet is sent out in a random pulse sending method.

In general, the CSMA / CA method may cause a hidden node problem and cause a rapid decrease in data throughput. Therefore, a countermeasure is required. Hidden Node is a case where the transmission of each other is not detected, which means that a serious collision occurs at this time. The IEEE 802.11 MAC protocol employs a protocol that adds collision avoidance by inter-frame space (IFS) control based on 1-persistent CSMA.

This is the first step, in which a terminal having a frame to be transmitted checks whether another terminal is transmitting by carrier sense. In the second step, the collision avoidance function confirms that the medium is idle, and then selects a random time interval to verify that the medium is idle even within this time interval. When this condition is met, the packet is sent. If not, the carrier sense is continued. The third step is the priority between frames. This is to adjust the priority of frames by the length of the frame transmission time interval. That is, several internal frame spacings (IFS) are used. DCF transmits the highest priority in order of priority.

Priority 1 is used to make an immediate response (ACK and CTS frames, burst transmission of segmented data frames) as a short priority, and this IFS is called SIFS. The second priority is used for centralized control as PCF priority and is used for synchronous data communication. This IFS is called PIFS. The third priority is DCF priority, which is used for asynchronous data communication (data or RTS transmission), and this IFS is called DIFS.

To recap, first there is the shortest short Inter Frame Space (SIFS). SIFS is the shortest delay and is used for data transmission of voice, video, etc. having the highest priority, and intermediate length PIFS (Point Coordination Function Inter Frame Space) is used for polling of a terminal in a non-competitive service.

The basic CSMA / CA algorithm first detects the state of the channel with a delay of DIFS when a packet is generated. If the channel is empty at this time, the packet is transmitted. If the channel is in use, it waits until the channel is empty, and then retransmits with a delay caused by the backoff algorithm.

The MAC function transmits an MPDU from the MAC layer to the physical layer during transmission, and the physical layer used in the present invention uses an OFDM multi-carrier modulation and demodulation scheme (Orthogonal Frequency Division Multiplexing).

In contrast to the single-carrier method, which compensates for the multipath as an extension of the conventional reception technology, the multicarrier method essentially makes the received waveform itself strong against the multipath. The problem of multipath means that the higher the transmission speed, the greater the impact. To this end, in order to reduce the effects of multipath, the high-speed signal is divided into low-speed signals and multiplexed by frequency division multiplexing (FDM) on the frequency axis, thereby producing the same high-speed signal as a whole.

If multiple low speed channels are used, each channel is low speed and thus is hardly affected by multipath. By using multiple channels, the amount of data transmitted is the same as that of one high-speed channel, so that high-speed transmission is possible as a whole.

Orthogonal Frequency Division Multiplexing (OFDM) makes multicarrier technology stronger for multipath. Channel arrangement of OFDM in the transmitter uses an Inverse Fast Fourier Transform (IFFT).

In order to make the bit streams input in series on the time axis into N-channel multicarriers, the signal sequence is first converted to N bits in parallel. It is then input to N modulators. At this time, in general, an interleaver for randomizing a signal and correcting an error during conversion from serial to parallel is also performed. Then, the output of each modulator is input to the IFFT.

In order to improve the characteristics of the multipath by separating the sample signal bundle sequentially output in 1 / N (bps) unit from the other output bundle by performing IFFT, the signal after IFFT is called 'Guard Interval'. Add time. In OFDM, when the reflected wave is shorter than the added guard interval, it has excellent characteristics for multipath. The guard interval uses a portion of the waveform generated by the IFFT. As a result, while the processing time of the IFFT is kept constant, the input signal and the guard interval are applied during the IFFT output. During this time, the operation speed increases and the speed for valid data is increased. The guard interval has a characteristic that the next symbol point is not affected because the multipath signal delayed in the multipath appears within the guard interval.

When modulation and transmission are performed, the reception is performed at the destination apparatus via a wireless channel, and is received at the MAC layer through a demodulation process. Since the transmission rate of OFDM currently recommended by IEEE802.11a is 6 to 54 Mbps, it is essential to speed up the MAC layer. The interface between the physical layer and the MAC layer generally processes queued MAC layer frames by placing queue elements (FIFOs or buffers).

In the future, the technical problem is how to improve the access between mobile terminals with various traffic. The most basic access method in IEEE 802.11 MAC is DCF, which uses CSMA / CA algorithm. This CSMA / CA takes the hidden terminal problem seriously.

Negotiations such as TDMA are suitable for media access of long-holding audio and video systems, which means from start to end of communication, and random access systems are used to access packet data of computer systems with short hold times. Tends to be suitable for Since IEEE 802.11 is a computer-based access, packet transmission is mainly performed. In order to overcome this disadvantage, it has a two-layered structure.

The two-layered structure of IEEE 802.11 PCF and DCF was defined as a part of the entire MAC layer, even though DCF alone functions as a MAC layer, and added PCF to layer each.

The hybrid model of PCF and DCF, the two-layered structure of IEEE 802.11, mentions only non-collision-based access and DCF-conflict access. The PCF function, which improves the accuracy of the PCF repetition period, can also deliver audio or video.

This is an example of non-conflict sequence access by the PCF and sends out a beacon to repeatedly secure the access area. Of course, beacons are sent by DCF as they perform CSMA / CA, but IFS used by PCF has high priority because it uses PIFS. Repeatedly sending out a beacon has the same function as a frame making a repeating period of TDMA.

MAC in wireless LAN has two layers of DCF and PCF. In general QOS (Quality of Service) control, it has a frame like TDMA, and the inside is divided into at least two areas, and one is divided into It is used for media access and the other for computer-based packet access. The two areas allocate dynamic areas such as TDMA and CSMA / CA, and the boundary between these areas is also dynamically changed in accordance with the magnitude of mutual traffic, thereby efficiently performing QOS control corresponding to various media. Is generally considered.

As the demand for the QOS control is required to speed up the physical layer, the conventional technology has the following problems.

First, the elements of the queue that process voice, image, and data packets are not separated. In particular, it is difficult to integrate the data of the audio and video series and the computer network series, which have extreme characteristics, and can degrade the system performance. The architecture of the conventional wireless LAN MAC processing apparatus has only one queue element (FIFO or buffer) that separates only transmission and reception, and there is no structure for processing such voice, image, and data packets separately. There is a noticeable difference in the Arrival Rate 1 / λ depending on the voice, image, and data packet. In view of the above problems, the present invention invents a device having multiple transmission / reception queues. The importance of the configuration and function of the queue can be considered a queue model modeling the interface between the RISC processor-cue-physical layer in the prior art (see FIG. 3).

Second, the prior art is configured as shown in Figure 5, as this structure is pointed out in the first, the separation of the buffer according to the transmission rate is not made. Detailed description of Figure 5 will be described later.

1 is a block diagram of a general wireless LAN system to which the present invention is applied. In FIG. 1, the base station 10 (also called an access point (AP)) connected to the Ethernet network 13 by wire is equipped with a wireless transceiver, and the service area 11 is schematically indicated by a dotted line. A total of N mobile terminals (also shown in FIG. 3 for simplicity) are also in communication with the base station 10 via the wireless channel 12. In this case, since the N mobile stations 20 and the base station 10 share one radio channel 12, a time slot or frequency may be assigned to each mobile station using a multiple access scheme such as TDMA / FDMA / CDMA. Assigning a specific code prevents information collisions on the wireless channel.

The base station roughly handles three functions: (1) an IP packet signal processor for interworking with an Ethernet network, (2) a MAC processor for processing a medium access control function, and (3) a wireless transmitter / receiver. After receiving the IP packet signal from the packet signal processor and processing it, the MSDU is loaded into the MAC frame and sent to the wireless transmitter. On the contrary, it is responsible for receiving the MAC frame and MSDU signal from the wireless receiver, extracting the IP packet signal, and sending it to the IP packet signal processor.

Next, FIG. 2 is a schematic diagram of the internal terminal configuration of the mobile terminal 20 in FIG. 1, which is a 5 GHz RF processing unit 24 for receiving a radio signal with the AP 10 and an OFDM modem unit capable of 25 Mbps processing. There is 23. There is a baseband signal processing processor 22 for converting the OFDM modem signal to a baseband, and there is a media access control processing apparatus 21 for processing a MAC frame from the baseband signal. The MAC-processed signal is interfaced with the external system bus 20. Here, the external system bus can be various system buses such as PCMCIA bus, CardBus, PCI bus, and EISA bus. At the time of transmission, it operates in the opposite signal path at the time of reception.

3 is an apparatus configured inside the media access processing apparatus 21 shown in FIG. As can be seen from FIG. 3, the apparatus for accessing media 21 is composed of a RISC processor 100, a transmission and reception queue 30, 31, and a wireless physical layer 200, and is a queue model modeling each interface. . The RISC processor 100 performs the MAC layer.

As can be seen from FIG. 3, in the conventional medium access control apparatus 21, the same as the transmit / receive buffers 30 and 31 between the wireless physical layer 200 having the OFDM demodulation scheme of the 5 GHz band regardless of the difference in data rates. Buffer was used. That is, it has a structure in which audio, image, and data packet processing having an Arrival Rate (1 / λ) indicating a significant difference is stored in one buffer 30 or 31. The structure of such a device causes a decrease in system performance.

On the other hand, when Little's Law is applied to the structure as shown in Fig. 3, in the steady state rather than the initial starting condition, the average number of input tasks is equal to the average number of output tasks output through the queue. It can be seen. Little's Law relates to the average number of tasks in the system, the average arrival rate of new tasks, and the task execution time. This relationship can be expressed as the relationship of the average number of tasks in the system = arrival rate x average response time. In this case, if the average time to perform the task in the physical layer is called Time _Phy and the average time per task in the _{queue is} called Time _queue , the total response time (Time _system ) is as follows. .

Through this analysis, it can be seen that the processing speed of the queue element interfaced with the RISC processor and the wireless physical layer in the structure of the wireless LAN MAC processing device is an important factor. This is because the time except for processing the physical layer is the processing time of the queue element.

However, the related art recommends that the wireless LAN MAC architecture satisfies the requirement for QOS through two-layered DCF and PCF. Although the recommendation of two-layering is important, the multiple transmission / reception queue element processing apparatus of the present invention enables efficient processing of voice, image, and data packets having a transmission rate (1 / λ) representing a significant difference.

As mentioned above, in the prior art as illustrated in FIG. 3, a transmission / reception buffer (B) between a RISC processor 100 performing a MAC layer and a wireless physical layer 200 having an OFDM demodulation scheme of 5 GHz band. 30, 31) the same buffer was used regardless of the difference in data rates. The structure of these devices contributed to the degradation of system performance.

Meanwhile, FIG. 5 is a detailed structural diagram of the conventional medium access control apparatus 21 shown in FIG. 2 performing the procedure as shown in FIG. 3. As can be seen from FIG. 5, the RISC performing the MAC layer. The exchange of data between the processor 100 and the wireless physical layer 200 having the OFDM demodulation scheme of 5 GHz band is performed in the transmission and reception buffers 500 and 510 as shown in FIG. 3, and the MAC controller 300 and the transmission / reception controller 520 controlling the same are illustrated in FIG. It is composed of

The MAC controller 300 performs the protocol of the media access control layer while having an interface with the RISC processor 100 and the memory 110 together with the transceiver controller 520. Determining system performance is very important in speeding up the time of the average time per task in the queue, as described in Little's Law, above.

In the prior art having the configuration as shown in Figure 5, the data transfer between the memory 110 and the transmission and reception buffers (500, 510) is configured by a memory copy, a DMA method or the like. At this time, the address, length, etc. of the MAC frame stored in the memory are stored in the buffer descriptor 530 so that the MAC controller 300 sees the contents and controls the transmission / reception controller 520 to process data transmission. At this time, the address controller 540 performs address determination between the memory 110 and the transmission / reception buffers 500 and 510.

In the conventional structure, as mentioned in the above-described problem in the prior art, the separation of the buffer according to the transmission rate is not made. In addition, the control of the transmission and reception buffers 500 and 510 is performed to transmit data to the memory 110 based on the contents stored in the buffer descriptor 530. The transmit / receive buffers 500 and 510 are generally configured as FIFOs. This operation is performed in the high-speed operation of the wireless physical layer 200 having the OFDM modulation and demodulation scheme of the 5 GHz band. 110) can be a problem if the memory copy or DMA function is not completed.

Data transmitted and received through the wireless physical layer 200 for high speed operation is stored in the data buffers 500 and 510 as temporary data and memory management is performed. Therefore, in the case of real-time media access control protocol processing that requires dynamic memory allocation, the size must be determined to fit the corresponding buffer memory capacity, and the method is largely divided into two methods.

The first is a FIFO buffer that performs linear addressing. This method requires bottlenecking of the system running in real time because it needs to check whether the read and write indicators have reached the end of the buffer. This problem may be a problem of the prior art.

Second, in order to eliminate this problem, the present invention uses the function of the circular buffer as shown in FIG. This buffer either dynamically calculates the buffer address when the pointer reaches the end of the buffer, or in some cases places the pointer at the beginning of the buffer. However, to find out when to reset the start of the buffer, which is a problem with these circular buffers, the next data pointer to be used must be examined, which can be an additional overhead. In order to solve this problem, the present invention employs a modulo addressing method, which will be described in detail later.

Korean Patent No. 211921 (Patent holder: Hyundai Electronics Co., Ltd.) has a high-speed data transmission apparatus for wireless LAN that introduces a demultiplexing technique applied to a wired LAN to a wireless LAN to enable high-speed data transmission and multimedia services through a wireless channel. In detail, the reception field strengths transmitted from a plurality of physical layers in the physical layer unit are matched, and an acquisition channel is determined according to the matching reception field strengths, and the user data is demultiplexed according to the determined number of acquisition channels. The technique is used.

In addition, Korean Patent No. 138001 (Patent Holder: International Business Machines Corporation) is a patent for a medium access control method suitable for a wireless local area network. Specifically, among the time-division fixed frame structures having time division in the MAC protocol, the fixed frame structures are respectively It consists of three periods according to the header of the three periods, which can be varied using a variable boundary technology, which allows the base station to predict the number of mobile stations to control transmission attempts to the mobile stations. Is disclosed.

However, the two Korean patents, as will be described later, is distinguished from the features of the present application that implements the interface of the high-speed wireless physical layer by adopting a unique hardware means for the implementation of the MAC function.

5 is a configuration diagram of a media access control processing apparatus 21 of the prior art.

Upon reception, the data received from the baseband signal processor 22 of FIG. 2 becomes the wireless physical layer 200, and the media access control processing apparatus 21 of FIG. 5 receives the MAC controller 300 via the reception FIFO 510 upon reception. MAC frame is made through the transmit / receive controller 520. At this time, the RISC processor 100 performs an operation of converting into an IP packet. Temporary parameter values required for the operation of the RISC processor 100 are stored in the memory 110 to be accessed through the RISC processor 100 and the memory controller 120.

On the contrary, the IP packet is converted into a MAC frame and transmitted to the wireless physical layer 200 through the transmission FIFO 500 through the MAC controller 300 and the transmission / reception controller 520. This radio physical layer 200 sends a MAC frame to the baseband signal processor 22 of FIG. An interface between the external system bus 20 (eg, PCMCIA, PCI, ISA, EISA, CardBus, etc.) and the RISC processor 100 of FIG. 2 is provided. Once created, the IP packet is stored in the external shared memory, and this data is also loaded on the external bus.

Memory copy and DMA control used in the MAC controller 300 plays a role of directly inserting the contents of the transmission / reception FIFOs 500 and 510 into the external volatile memory regardless of the operation of the RISC processor 100. In addition, during transmission, the contents of the external shared volatile memory are transmitted to the transmission / reception FIFOs 500 and 510 via the MAC frame generation slave. The external nonvolatile memory stores firmware data for operating the media access control processor 21.

The MAC controller 300 performs the protocol of the media access control layer while having an interface with the RISC processor 100 and the memory 110 together with the transceiver controller 520. Data transmission between the memory 110 and the transmission and reception buffers 500 and 510 may be performed by a memory copy, a DMA method, or the like. At this time, the address, length, and the like of the MAC frame stored in the memory are stored in the buffer descriptor 530 so that the MAC controller 300 sees the contents and controls the transmission / reception controller 520 to process data transmission. At this time, the address controller 540 performs address determination between the memory 110 and the transmission / reception buffers 500 and 510.

The present invention was derived to solve the above-mentioned disadvantages of the interface method using a single transmit / receive queue, and employs multiple transmit / receive queues classified according to data rates and processes them through N addressing in a circular buffer type module. An object of the present invention is to provide an interface device and method for enabling high speed and efficient interface with a physical layer.

1 is a schematic block diagram of a wireless LAN to which the present invention is applied;

2 is a schematic configuration diagram of a mobile terminal MT shown in FIG. 1;

3 is a queue model configuration diagram having a conventional single transmit / receive queue element (FIFO or buffer);

4 is a queue model diagram having multiple transmit and receive queue elements in accordance with a preferred embodiment of the present invention;

5 is a media access control processing apparatus for controlling a conventional single transmit / receive queue element;

6 is a media access control processing apparatus for controlling multiple transmit / receive queue elements in accordance with a suitable embodiment of the present invention; And

7 is a diagram illustrating a modular N wraparound addressing method of a circular buffer for controlling multiple transmit / receive queue elements according to the present invention.

* Brief description of the main symbols in the drawings

40: split send queue 41: split receive queue

100: RISC processor 200: wireless physical layer

300: MAC controller 610: transmit buffer controller

620: Receive buffer controllers 613a to 613c, 623a to 623c: circular buffer

According to a first aspect of the present invention, a medium access control apparatus used in a mobile terminal device connected to an Ethernet through a wireless LAN access point (AP) and configured to interface between a RISC processor and a wireless physical layer for processing MAC frame data. And a transmission queue for processing the transmission data and a reception queue for processing the reception data are configured with a plurality of circular buffers controlled by modulo N addressing by storing a plurality of packets which are divided and processed according to transmission rates. An access control device is provided.

According to a second aspect of the present invention, there is provided an interface device between a RISC processor and a wireless physical layer for processing MAC frame data in a medium access control device used for a mobile terminal device connected to Ethernet through a wireless LAN access point (AP). A transmitter configured to process the input MAC frame data for each packet classified according to a transmission rate and transmit the received MAC frame data to the wireless physical layer; And a receiver configured to process MAC frame data received from the radio physical layer for each packet classified according to a transmission rate.

In addition, according to a third aspect of the present invention, an interface method between a RISC processor and a wireless physical layer for processing MAC frame data in a medium access control apparatus used for a mobile terminal device connected to Ethernet through a wireless LAN access point (AP) The method includes: dividing the MAC frame data into a plurality of packets for each transmission rate, storing each of the divided packets in a plurality of circular buffers for each transmission rate, and then controlling the circular buffer through modular N addressing. Characterized in that.

Reference will now be made in detail to the preferred embodiments of the present invention, by way of example only, with reference to the accompanying drawings.

Referring to FIG. 4, transmission and reception buffers 4-2 and 4-3 are provided between a RISC processor 4-1 performing a media access control layer and a wireless physical layer 4-4 having an OFDM demodulation scheme of 5 GHz band. ), Which separates the transmission and reception buffer according to the difference of transmission rates, and divides the transmission / reception buffer into multiple transmission queues 4-2 and multiple reception queues 4-3, respectively, and has a significant difference (Arrival Rate) of 1 / λ, Image and data packet processing can be efficiently performed.

FIG. 4 is used interchangeably with the device configured inside the medium access processing device 2-4 of FIG. As shown in FIG. 4, a configuration of a media access processing apparatus according to a preferred embodiment of the present application is a RISC processor 4-1-multiple transmit / receive queues 4-2 and 4-3-a wireless physical layer 4-4. It is composed of a queue model modeling each interface. The RISC processor 4-1 performs a media access control layer. Multiple buffers having separate buffers according to the transmission rate difference between the transmission and reception buffers 4-2 and 4-3 between the wireless physical layer 4-4 having OFDM modulation and demodulation scheme of 5GHz band were used. That is, it has a structure in which audio, image, and data packet processing having an Arrival Rate 1 / λ representing a significant difference is stored in each buffer 4-2 or 4-3. The advantage of this structure is that the performance of the system can be improved.

6 is an internal configuration diagram of a media access processing apparatus according to a preferred embodiment of the present invention. As can be seen from the figure, the structure of the media access processing apparatus according to the preferred embodiment of the present application as compared to Figure 5 has no buffer descriptor and offset adjuster. Upon reception of the transmit and receive buffer controllers, the data received by the baseband signal processor 2-3 in FIG. 2 becomes the radio physical layer 6-7.

The operation upon reception in the media access control processing apparatus according to FIG. 6 is as follows. According to the type of data received, the reception buffer controller 6-6 divides the received data into voice, image, and data packets according to the Arrival Rate 1 / λ, and the input data is selected by the rate selector DMUX 6-9. Input the circular buffer (6-15, 6-16, 6-17) for each buffer. The input MAC frame is inputted to the MUX 6-8 through the control of the reception buffer controller 6-6 and sent to the MAC controller 6-4. That is, without the buffer descriptor (5-8) and the offset adjuster (5-9) used in the prior art, it is transmitted to the memory (6-1) by the control of the MAC controller (6-4) through the operation of the circular buffer immediately. Becomes

On the contrary, when transmitting, IP packet is converted into MAC frame, and the transmission buffer controller 6-5 divides the transmission data into voice, image, and data packets according to the Arrival Rate 1 / λ according to the type of transmitted data. The rate selector DMUX (6-10) inputs the input data into the circular buffers (6-12, 6-13, 6-14) for each buffer. The input MAC frame is input to the MUX 6-11 through the control of the transmission buffer controller 6-5 and sent to the radio physical layer 6-7. The operation of the circular buffer is described in FIG.

Next, referring to FIG. 7, it is a schematic diagram illustrating the concept of modulo addressing used in a media access control processing apparatus according to a preferred embodiment of the present invention.

The circular buffer derived by the inventor of the present invention for high speed data communication is a method for having a high speed physical layer and a high speed interface. The circular buffer is a form connecting the start and end of the buffer. When the pointer reaches the end of the buffer, it either dynamically calculates the buffer address or, in some cases, places the pointer at the beginning of the buffer.

However, in such circular buffers, the data pointer to be used next needs to be examined to know when to reset the start of the buffer, which can be some additional overhead. The present inventors modulo addressing to solve this problem.

If the effective address of the wraparound operation, which is the operation of the circular buffer, is Base address = B, most-significant buffer size = L, and least-significant buffer size = I, the memory space from addresses B to B + L-1 is valid. The address operation is executed in the following way.

The offset is determined by the command, and the size of the offset during wraparound should be smaller than the buffer size. In the present invention, the address area of the transmission and reception buffer as shown in FIG. 7 is set to Base address to Base address + K * N + (K-1). Where K is an integer multiple of modulo N and N is the modulo number of modulo addressing. At this time, modulo N's dynamic memory allocation for wraparound operation is made. In this way, the transmitted and received MAC frames are sent to the MAC controller 6-4 and the radio physical layer 6-7 according to their respective transmission rates.

According to the high-speed physical layer transmission interface apparatus and method according to the present invention, in the case of transmitting and receiving data having a remarkable difference in transmission rate (1 / λ) according to voice, image, and data packet, multiple transmission / reception queues ( Queue) can be used to represent efficient system performance in high-speed wireless communication, and the control of such a queue is carried out using the circular buffer method of N addressing with the proposed module in the present invention, thereby providing an efficient interface with a high-speed wireless physical layer. This can dramatically increase system performance.

Claims (9)

A medium access control apparatus used for a mobile terminal device connected to an Ethernet through a wireless LAN access point (AP) and having an interface between a RISC processor and a wireless physical layer for processing MAC frame data, Each of the transmission queue processing the transmission data and the reception queue processing the reception data each comprise a plurality of circular buffers controlled by modulo N addressing by storing a plurality of packets which are divided and processed by transmission rate. Control unit. An interface device between a RISC processor and a wireless physical layer for processing MAC frame data in a media access control device used for a mobile terminal device connected to Ethernet through a wireless LAN access point (AP), A transmitter for processing the input MAC frame data for each packet classified according to a transmission rate and transmitting the received MAC frame data to the wireless physical layer; And And a receiver configured to process MAC frame data received from the radio physical layer for each packet classified according to a transmission rate. The method of claim 2, The transmitting unit: A plurality of transmit circular buffers; A transmission rate selector for dividing the input data into a plurality of packets according to the transmission rate, and inputting the divided data into each of the transmission circular buffers according to the transmission rate according to the control of the transmission buffer control means; A MUX for transferring data stored in each of the transmission circular buffers to the wireless physical layer under control of transmission buffer control means; And And a transmission buffer controller for controlling the rate selector and the MUX. The method of claim 2, The receiving unit: A plurality of receive circular buffers; A rate selector for dividing data input through the wireless physical layer into a plurality of packets according to a transmission rate, and inputting the divided data into each of the reception circular buffers according to a transmission rate according to a control of a reception buffer control unit; A MUX for transferring data stored in each of the reception circular buffers to a MAC controller under control of a reception buffer control means; And And a reception buffer controller for controlling the rate selector and the MUX. The method according to claim 1, 3 or 4, Each of the circular buffers has a wraparound operation valid address from B to B + L-1 when the base address is B, the start buffer address is L, and the end buffer address is I, And performing an address operation such that the offset is smaller than the circular buffer size. The method of claim 5, The address area of each of the transmit and receive circular buffers is set from a base address to a base address + K * N + (K-1), where N is a modulo number integer of modulo addressing, and K is an integer multiple of modulo N. Interface device characterized in that. The method according to claim 1, 3 or 4, The packets classified by the data rates are voice, image and data. In an interface method between a RISC processor and a wireless physical layer for processing MAC frame data in a media access control device used for a mobile terminal device connected to Ethernet through a wireless LAN access point (AP), And dividing the MAC frame data into a plurality of packets for each transmission rate, storing each of the divided packets in a plurality of circular buffers for each transmission rate, and then controlling the circular buffer through modulo N addressing. The method of claim 8, In each of the circular buffers, when the base address is B, the start buffer address is L, and the end buffer address is I, at a wraparound operation valid address from B to B + L-1, And an offset operation is smaller than the circular buffer size.