WO2002089413A1 - Appareil de communication et systeme de communication faisant appel a ce dernier - Google Patents

Appareil de communication et systeme de communication faisant appel a ce dernier Download PDF

Info

Publication number
WO2002089413A1
WO2002089413A1 PCT/JP2002/004107 JP0204107W WO02089413A1 WO 2002089413 A1 WO2002089413 A1 WO 2002089413A1 JP 0204107 W JP0204107 W JP 0204107W WO 02089413 A1 WO02089413 A1 WO 02089413A1
Authority
WO
WIPO (PCT)
Prior art keywords
communication device
unit
transmission
time
application data
Prior art date
Application number
PCT/JP2002/004107
Other languages
English (en)
French (fr)
Inventor
Hiroyuki Nakaoka
Yoshihiro Ohtani
Tohru Ueda
Tomonobu Tomaru
Original Assignee
Sharp Kabushiki Kaisha
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sharp Kabushiki Kaisha filed Critical Sharp Kabushiki Kaisha
Priority to JP2002586576A priority Critical patent/JP3703456B2/ja
Publication of WO2002089413A1 publication Critical patent/WO2002089413A1/ja

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00Network architectures or network communication protocols for network security
    • H04L63/04Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
    • H04L63/0428Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the data content is protected, e.g. by encrypting or encapsulating the payload
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/12Arrangements for detecting or preventing errors in the information received by using return channel
    • H04L1/16Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
    • H04L1/1607Details of the supervisory signal
    • H04L1/1614Details of the supervisory signal using bitmaps
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/004Arrangements for detecting or preventing errors in the information received by using forward error control
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/12Arrangements for detecting or preventing errors in the information received by using return channel
    • H04L1/16Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
    • H04L1/18Automatic repetition systems, e.g. Van Duuren systems

Definitions

  • the present invention relates to a communication device and a communication system constituted by the communication device.
  • the present invention relates to a wireless communication system and apparatus for transferring real-time AV data, and particularly to an IEEE 802.11 wireless communication system and apparatus.
  • the IEEE 802.11.1 wireless communication system is a standard originally specified for wireless LAN.
  • This method is characterized in that a section where a collision may occur at a fixed period (Contention Period: CP) and a section where no collision occurs (Contention Free Period: CFP) are switched, and the area where a collision occurs is characterized.
  • CP Contention Period
  • CFP Contention Free Period
  • the main problem is that it can be used for LAN applications, and in areas where collisions do not occur, it can be used for communications that require communication bandwidth.
  • FIG. 30 is a diagram showing, on a time axis, a section in which CFP and CP occur when a conventional function of PIE (Point Coordination Function) of IEEE 8.2.11 is used.
  • PIE Point Coordination Function
  • a specific terminal called PC polls terminals using CFP to secure a communication band.
  • the PC transmits information (called TIM) for time adjustment of each terminal on a frame called Beacon and transmits the information at regular intervals.
  • a Beacon containing notification information (called DTIM) for a sleeping terminal is transmitted at a rate of once every few Beacons transmitted.
  • the sleeping terminal enters an active state around the time when Beacon including DTIM is transmitted, and enters the sleep state again when receiving the Beacon including DTIM and performing necessary communication.
  • the CFP occurs following the Beacon containing the DTIM and ends by the time the next Beacon is transmitted.
  • the transmission cycle of Beacon in IEEE802.11 may be, for example, about 10 Oms.
  • wireless media which are less reliable than wired media, are prone to errors and lost buckets on the communication path.
  • IEEE 802.11.1 MAC layer Since the current IEEE 802.11.1 MAC layer has neither an error correction code nor a packet erasure code, if an error occurs on the communication path, the packet must be retransmitted.
  • CFP occurs only once every several 10 Oms, and the real-time AV data continues to be input momentarily even during retransmission standby. Therefore, in order to perform retransmission, a buffer for holding data for several hundred milliseconds is required.
  • this period appears to the user in the form of a delay. For this reason, there is a problem that even if a user issues a command such as stop or rewind to the flowing AV data, it cannot be immediately reflected.
  • HCF Hybrid Coordination Function
  • a transmitting station that desires periodic data transmission sends a media access cycle to a control station (Hybrid Coordinator: HC).
  • the control station that requests the media access time and acknowledges the request, periodically transmits (polls) the bandwidth reservation frame to the transmitting station regardless of the CFP or CP, and determines the section where no collision occurs.
  • the HC also serves as the PC, so that the conventional IEEE 802.11 communication device equipped with the PCF does not hinder communication.
  • the conventional IE 82.1.1.1 communication device performs communication using a function called DCF (Distributed Coordination Function).
  • DCF Distributed Coordination Function
  • DIFS Distributed Interframe Space
  • DIFS Distributed Interframe Space
  • a control station (H C) equipped with an HCF can access the media regardless of CFP or CP by carrying out a carrier sense for a time interval of 25 ⁇ s called PIFFS (PCF Interframe Space).
  • PIFFS PCF Interframe Space
  • the controlled station polled by the control station (H C) can start transmission only by carrying out a carrier sense for a time interval of 16 ⁇ s called Short Interframe Space (SIFS).
  • SIFS Short Interframe Space
  • each transmitting station operates based on the Beacon interval determined by the PC, whereas in the currently proposed HCF, each transmitting station has a media access period (polling period) for the HC. And the time required for communication within the media access cycle. For this reason, the HC must perform complex media access scheduling for each transmitting station.
  • encryption is performed for privacy protection. This encoding is performed for each block, and an error correction code based on Read Solomon encoding is added to each encrypted block. Then, an error detection code is added to the entire data portion composed of a plurality of lead-Solomon coding blocks to which the error correction code has been added.
  • the ⁇ key (public key) is attached to the same Read Solomon encoding block as the MAC header, so if the error correction for this Lead Solomon encoding block is successful, the data It is possible to decrypt each Reed-Solomon encoding block using the encryption key. However, in order to guarantee the successful decoding of the packet itself, it is necessary that the error correction for all the Read-Solomon coding blocks constituting the data part be successful.
  • FIG. 31 shows the accuracy of the system timer of each station in the network system and the jitter generated at the source bucket output of the receiving station.
  • the transmitting station attaches the time information at the time of input indicated by its own timer to the AV source packet, and the receiving station reproduces the time sequence of the AV source bucket using the attached time information and its own timer. I do. However, the timers at each station do not have exactly the same accuracy, so that after a certain amount of time, the times will shift. The time lag between the transmitting station and the receiving station appears as jitter when the receiving station reproduces the time sequence of the AV source bucket. In other words, as shown in Fig.
  • the timing of the source bucket transmitted by the transmitting station becomes faster and the source The time information added to the bucket is earlier than the time indicated by the system timer of the control station.
  • the system timer of the receiving station is delayed, the timing of the source packet output from the receiving station is delayed, and the output interval of the source packet becomes shorter than the original output interval. It becomes longer and causes jitter.
  • the system timer of each station is matched by the beacon given periodically, so that the output interval of the source bucket at the receiving station where the time delay has occurred becomes shorter than the original output interval. Jitter occurs.
  • the output timing of each AV source packet is strictly determined, and must be output within a specified jitter range.
  • the current HCF only adjusts the time using beacons.
  • the timer accuracy in IEEE802.11 is 0.01%, but if the time is adjusted only by a beacon transmitted once every 100 ms, if There is a maximum time lag of about 20 ⁇ s between them.
  • the set wireless communication system is not a sufficient communication system for AV data transfer. Disclosure of the invention
  • an object of the present invention is to provide a wireless communication system capable of performing efficient data transmission suitable for AV data transfer.
  • a communication system includes a plurality of communication devices, and a communication system in which one of the plurality of communication devices is a control station.
  • the device adds a system timer for managing time and time information for time series management generated based on the input time of the application obtained from the system timer to the application data.
  • a data processing unit that sets a period in which data can be transmitted to the data processing unit, a buffer that temporarily stores the application data to which the time information has been given by the data processing unit, and that reads out the application data from the buffer.
  • An error correction code adding section for adding an error correction code to the application data, and a transmission packet generating section for generating a transmission packet from the application data to which the error correction code has been added by the error correction code adding section.
  • a transmission unit that transmits the transmission bucket created by the transmission packet generation unit; and controls reading of application data from the buffer, and the transmission bucket that is generated by the transmission packet generation unit. Transmission control for controlling transmission from the transmission unit at the time when the communication band is secured. And a receiving unit that receives a response signal indicating reception of one or a plurality of transmission buckets transmitted from the transmitting unit.
  • the transmission control unit When transmitting from the transmitting unit, the transmission control unit includes: Comparing the current time obtained from the system timer with the transmittable period set for the application data stored in the buffer, and when the current time exceeds the transmittable period, For the application data that has been transmitted, the error correction code adding unit does not read the data from the buffer, and the transmission unit transmits the transmission packet.
  • the transmission control unit confirms that a response signal has not been received by the reception unit, it is determined that the transmission bucket is not correctly received. It is characterized in.
  • another communication system of the present invention is a communication system configured with a plurality of communication devices and using one of the plurality of communication devices as a control station.
  • a communication unit configured to receive a transmission packet composed of application data to which time information for time-series management is added, and a reception unit receiving the transmission packet.
  • An error correction determination unit that determines whether an error correction code has been added to the transmission packet; and an error correction determination unit that determines whether an error correction code has been added to the transmission packets received by the reception unit.
  • An error correction processing unit that performs error correction; a first decoding unit that generates application data from the transmission bucket in which the error is corrected; and a transmission bucket received by the reception unit. Mistake A transmission bucket with no correction code added is used to generate the application data.
  • a second decryption unit configured to generate the application data; a first buffer configured to temporarily store the application data generated by the first decryption unit; and a second buffer configured to temporarily store the application data generated by the first decryption unit.
  • the data output unit reads out the application data from the first or second buffer and outputs the data, and the reception status of one or more transmission buckets.
  • a transmission unit that transmits a response signal indicating a state of the transmission unit, a first operation process performed by the error correction processing unit and the first decoding unit on the transmission packet received by the reception unit, and The second arithmetic processing by the two decoding units is performed simultaneously.
  • Another communication system of the present invention is configured by a plurality of communication devices, and in a communication system in which one of the plurality of communication devices is a control station, the communication device serving as the control station includes: A system timer that manages the time; a receiver that receives a bandwidth reservation request signal that requests the transmission of a communication band transmitted from another communication device; A band management unit to be managed; a time information signal for causing another communication device to recognize the current time obtained from the system timer; and a non-collision section for allowing the other communication device to recognize the start of the non-collision section.
  • FIG. 1 is a schematic configuration diagram showing one embodiment of a wireless communication network according to the present invention.
  • FIG. 2 is a block diagram of a transmission device according to the present invention.
  • FIG. 3 is a block diagram of a receiving device according to the present invention.
  • FIG. 4 is a block diagram of a control station according to the present invention.
  • FIG. 5 is a diagram showing an MPEG 2 -TS bucket input process according to the present invention.
  • FIG. 6 is a diagram showing a DVC bucket input process according to the present invention.
  • FIGS. 7A and 7B are diagrams showing a MAP format according to the present invention.
  • FIG. 8 is a diagram showing a flow of a bucket transmitting process according to the present invention.
  • FIG. 9 is a diagram showing a delay of an error correction circuit output with respect to an error detection circuit output according to the present invention.
  • FIG. 10 is a diagram showing a flowchart for determining the presence or absence of an error correction code in a packet according to the present invention.
  • FIG. 11 is a diagram showing a flow of a bucket receiving process according to the present invention.
  • FIG. 12 is a diagram showing an MPEG2-TS packet output process according to the present invention.
  • FIG. 13 is a diagram showing a DVC bucket output process according to the present invention.
  • FIG. 14 is a diagram showing a currently proposed delayed ACK frame format according to the present invention.
  • FIG. 15 is a diagram showing a delayed ACK frame format extended for retransmission of a Read Solomon coded block unit according to the present invention.
  • FIG. 16 is a diagram showing a delayed ACK frame according to the present invention.
  • FIG. 17 is a diagram showing the update of the transmission status of the transmitting station according to the present invention.
  • FIG. 18 is a diagram showing an AV transmission buffer in consideration of the lifetime according to the present invention.
  • FIG. 19 is a diagram showing a transmission cycle and a bucket configuration according to the present invention.
  • FIG. 20 is a diagram showing the number of bursts, the actual transfer rate, and the basic period when DVVCZLAN (24 Mbps) according to the present invention coexists.
  • FIG. 21 is a diagram showing a flowchart of band management according to the present invention.
  • FIG. 22 is a diagram showing a flow chart of band release according to the present invention.
  • FIG. 23 is a diagram showing a transmission cycle and a bucket configuration according to the present invention.
  • FIG. 24 is a diagram showing compensation for a band loss by the LAN packet according to the present invention.
  • FIG. 25 is a diagram showing a transmission cycle and a bucket configuration according to the present invention.
  • FIG. 26 is a diagram showing a flow chart of the band release according to the present invention.
  • FIG. 27 is a diagram showing a transmission cycle and a bucket configuration according to the present invention.
  • FIG. 28 is a diagram showing a CF—Pol1 frame according to the present invention.
  • FIG. 29 is a diagram showing a CF-End frame according to the present invention.
  • FIG. 30 is a diagram showing the relationship between C FP and C P of PCF.
  • FIG. 31 is a diagram showing conventional timer accuracy and time adjustment. BEST MODE FOR CARRYING OUT THE INVENTION
  • FIG. 1 is an example of a wireless network configuration for implementing the present invention.
  • Fig. 1 is a control station that controls access to wireless media by a transmitting station in a section where no collision occurs.
  • HCF is provided as a media access control function.
  • This control station 1 is usually called an H C (Hybrid Coordinator).
  • Fig. 1 2 is a transmitting station for real-time AV data
  • 3 is a receiving station for real-time AV data.
  • the media access in the section (CFP) where no collision occurs between the real-time AV data transmitting station 2 and the real-time AV data receiving station 3 is managed by the control station 1.
  • the control station 1, the transmitting station 2, and the receiving station 3 each include an HCF in order to function the media access management in the CFP.
  • Fig. 1 is a conventional station that can access media only in the section (CP) where a collision may occur.
  • the station 4 has only DCF.
  • each station has only one function, such as transmission, reception, control, and conventional media access, but this simplifies the explanation. For this reason, one station may actually implement a plurality of functions.
  • FIG. 2 is a block diagram showing a configuration of the transmitting station 2.
  • 5 in Fig. 2 is a system timer that manages the time information of the entire network system.
  • the system timer 5 always adjusts the time based on the current time information included in the frame that stores the time information such as Beacon issued by the control station 1, so that the time is almost the same as the system timer in the control station 1.
  • the media access time is managed based on the current time information.
  • MAP data block generation information for the AV source bucket
  • time information measured by the system timer 5 when the AV source packet is input is measured.
  • AV data pack There is also a method of separately preparing a timer for managing the input time of the AV source bucket to the data block generator 6, but in this embodiment, the system timer 5 is used for simplicity of description. .
  • the set of the removable blocks to which the MAP and the time information are added is referred to as an AV data pack in the present embodiment.
  • Reference numeral 7 in FIG. 2 denotes an AV transmission buffer, which sequentially stores the AV data packs generated by the data block generation circuit 6.
  • the AV transmission buffer 7 manages buffer addresses in units of AV data packs, and includes an identifier indicating a buffer address in a transmission standby state. This identifier is 1-bit data provided corresponding to the buffer address for each AV data pack.
  • the identifier corresponding to the buffer address of the input AV data pack is set to 1, and the AV data pack is output from the AV transmission buffer 7.
  • the identifier corresponding to the buffer address of the output AV data pack is ().
  • the buffer size of the AV transmission buffer 7 is set to a size that can store the AV data pack for the lifetime described later.
  • the AV transmission buffer 7 also includes a retransmittable identifier that indicates that retransmission is possible, which will be described later.
  • MAC Medium Access Control
  • MAC bucket generating circuit for storing data in buffer 10 for PDU (LLC Protocol Data Unit) 10 in MAC bucket .
  • Reference numeral 11 in FIG. 2 denotes a receiving unit for receiving a packet transmitted from another station.
  • the receiver 11 receives the Beacon transmitted by the control station 1, the receiver 11 notifies the system timer 5 of the time information stored in the Beacon, and the system timer 5 uses the time information to transmit its own time information. To update.
  • the receiving unit 11 receives the bandwidth securing frame from the control station 1 for AV transmission, it notifies the MAC unit 8 of that fact.
  • the notified MAC section 8 checks the state of the AV transmission buffer 7 and, if an AV data pack in the transmission standby state exists in the buffer, sends the MAC packet to the MAC packet generation circuit 9. Request to start generating packets.
  • receiving section 11 receives an ACK from receiving station 3, it determines whether receiving station 3 requires retransmission, and if so, notifies MAC section 8 of the fact. Then, necessary data can be transmitted at the next transmission.
  • the MAC packet generation circuit 9 includes a selector 12 for selecting input data, an error detection code addition circuit 13 for adding an error detection code to input data, and a state in which an error detection code is added.
  • An encryptor 14 for encrypting the data of the MAC, a MAC packet header generating circuit 15 for generating the MAC packet header, and a selector 16 for selecting the MAC packet header and the encrypted data.
  • An error correction code adding circuit 17 for adding an error correction code to the output of the selector 16 and a selector 18 for selecting whether or not to add the error correction code in the error correction code adding circuit 17 to the data.
  • an error detection code adding circuit 19 for adding an error detection code to the entire input packet.
  • the error correction code adding circuit 17 uses a lead-Solomon code (224, 208). That is, a 16-byte Reed-Solomon code is added to the 208-byte data, and 224-byte data is output as a whole.
  • FIG. 3 is a block diagram showing the configuration of the receiving station 3.
  • the MAC bucket reception circuit 20 in FIG. 3 performs error correction by an error detection circuit 21 that detects errors of the entire packet and an error correction code added to the packet.
  • It comprises a MAC packet analysis circuit 26 for analyzing a bucket, a decoder 27 for decrypting the encryption applied to the data, and an error detection circuit 28 for detecting an error in the decrypted data.
  • the error correction circuit 22 uses a lead-Solomon code (224, 208). That is, it outputs 224 bytes of data, and the last 16 bytes are used as a Read-Solomon code, and outputs error-corrected 208 bytes of data.
  • the receiving station 3 in FIG. 3 has two MAC packet analysis circuits, two decoders, and two error detection codes, but this is because the lead-solomon decoding process takes time. This is to prevent the processing of the conventional MAC packet from being disabled when a conventional MAC packet is received during Dosolomon decoding. Therefore, if such a function is not required, only one MAC packet analysis circuit, one decoder, and one error detection code may be provided. Also, the LLC-PDU buffer may be one LLC-PDU buffer as long as it has a control circuit capable of writing data from two systems at the same time.
  • the data output from the error detection circuit 25 is an AV data pack
  • the data is supplied to the AV transmission buffer 32 through the selector 29 and stored in the AV transmission buffer 32.
  • the AV transmission buffer 32 performs address management for each AV data pack, and includes an identifier indicating the storage state of the AV data pack with respect to the buffer address. This identifier is 1-bit data provided corresponding to the buffer address for each AV data pack.
  • the identifier corresponding to the buffer address of the input data pack is set to 1 and the AV data pack is output from the AV transmission buffer 32. Then, the identifier corresponding to the buffer address of the output data pack is set to 0.
  • the buffer size of the AV transmission buffer 32 is a size that can store an AV data pack for the life time described later.
  • 33 in FIG. 3 is a source bucket output circuit.
  • the source packet output circuit 33 reconstructs the source packet from the AV data pack stored in the AV transmission buffer 32, and calculates the source packet from the time information or the source packet attached to the source packet. The time sequence is reproduced according to the time information and the system timer 34, and the source packet is output.
  • the MAC bucket receiving circuit 20 When the MAC bucket receiving circuit 20 receives a frame storing time information such as Beacon, it notifies the system timer 34 of the time information, and the system timer 34 receives its own time according to the notified time information. Update information.
  • time information such as Beacon
  • Reference numeral 35 in FIG. 3 denotes a MAC section for controlling media access of the receiving device.
  • the MAC section 35 determines the time information of the system timer 34 and whether or not the own station can access the medium based on the media sense, that is, whether or not the ACK can be transmitted.
  • Reference numeral 36 in FIG. 3 denotes an ACK generator. This one! The generating unit 36 generates an ACK for the transmitting station 2 according to the status of the MAC bucket receiving unit 20 and the state of the AV transmission buffer 32.
  • FIG. 4 is a block diagram showing the configuration of the control station 1.
  • reference numeral 37 denotes a system timer
  • the MAC section 38 controls the media access based on the time information of the system timer 37.
  • the MAC unit 38 requests the transmitting unit 39 to transmit the Beacon or the bandwidth securing frame.
  • Reference numeral 40 in FIG. 4 denotes a receiving unit for receiving a packet transmitted from another station, and receives a band securing request frame or the like from a transmitting station that wants to newly request a band securing.
  • the band management unit 41 manages stations that need to periodically secure a band and media access time called TXOP. If there is enough bandwidth, the bandwidth management unit 41 permits this request and stores necessary information therein.
  • the MAC unit 38 requests the transmission unit 39 to transmit a bandwidth securing frame for the station from an appropriate time. .
  • the existing station 4 the existing station equipped with IEEE 802.11 PCF / DCF The description of the configuration will be omitted.
  • the data block generation circuit 6 converts the input AV source packet into a read-solomon. Generates an AV data pack reconstructed so that it is easy to use with codes (224, 208), that is, data is within 208 bytes, and stores it in the AV transmission buffer 7. .
  • FIG. 5 is a diagram showing the flow until MPPEG 2 -TS is stored in the AV transmission buffer 7 of the transmitting station 2.
  • the size of one MPEG2—TS bucket is 188 bytes, and it is intermittently transferred at a rate of about 6 Mbps for normal picture quality and 24 Mbps for high picture quality.
  • the data block generation circuit 6 attaches 4-byte time information based on the system timer 5 at the time of input to the input MPEG 2-TS packet as a source packet header (SPH).
  • SPH source packet header
  • an 8-byte AV data pack is generated by attaching 8-byte MAP information indicating the contents of the data block.
  • the AV data packs generated in this manner are sequentially supplied to the AV transmission buffer 7, where the AV data packs are output from the data block generation circuit 6 in the order described later. Is added and stored.
  • the input time is confirmed by the system timer 5.
  • an AV data pack is generated by attaching S ⁇ and MA ⁇ information based on the time information based on the time confirmed by the system timer 5
  • this AV data pack becomes the k-th data in the AV transmission buffer 7.
  • Add and store Tag Further, the AV data pack generated by the (n + 1) th source packet is stored in the (k + 1) th AV transmission buffer 7 to which Tag described later is added.
  • Fig. 6 shows the situation until the DVC bucket is stored in the AV transmission buffer 7 of the transmitting station 2. It is a figure showing the flow of.
  • One DVC bucket is 480 bytes, and is continuously transferred at a transfer rate of 30 Mbps.
  • the bucket head is determined by the rising edge of the frame pulse input.
  • the data block generation circuit 6 divides the input DVC bucket into five blocks, and attaches 8-byte MAP information indicating the contents of the data block to each of the two blocks.
  • the AV data packs generated in this way are sequentially supplied to the AV transmission buffer 7, and the AV data packs are output from the data block generation circuit 6 in the AV transmission buffer 7 in the order described later. ag is added and stored.
  • MAP information of the DVC two bytes of time information based on the system timer 5 at the time of the rising of the frame pulse are stored.
  • the input time is confirmed by the system timer 5.
  • the system packet is confirmed by the system timer 5 into two blocks F0n and Fin.
  • Generates an AV data pack by attaching MAP information with time information based on the current time.
  • This AV data pack is stored in the AV transmission buffer 7 with the k-th Tag described later added thereto.
  • an AV data pack is generated by attaching MAP information having the same time information to two blocks F 2 n and F 3 n, and this AV data pack is k + Add the first Tag described later and store.
  • the data block generation circuit 6 divides the data into five blocks F0n + 1 to F4n + 1. Then, MAP information including time information based on the input time of the (n + 1) th source packet confirmed by the system timer 5 is attached to the two blocks F4n and F0n + 1. To generate an AV data pack.
  • This AV data pack is stored in the AV transmission buffer 7 with the addition of Tag + which will be described later, which is the (k + 2) th. After that, an AV data pack based on the blocks Fln + 1 and F2n + 1 and an AV data pack based on the blocks F3n + 1 and F4n + 1 are added in order with Tag described later. And stored in the AV transmission buffer 7.
  • the 4-byte or 2-byte time information based on system timer 5 It may indicate the input time of the source packet confirmed by the timer 5, and may be the time obtained by adding the processing time at the transmitting station 2 and the transmitting time to the receiving station 3 to this input time. No problem.
  • FIGS. 7A and 7B are examples of a packet header format for storing MAP information for each of M PEG 2 —TS and D VC.
  • Table 1 is a table showing the number of bits of each field in MAP information
  • Table 2 is a table showing information of each field with respect to each of M PEG 2 —TS and D VC.
  • F N is a 3-bit field indicating the number of data blocks divided by the source bucket as an integer. 1 indicates no division, and 0 indicates division into eight. M PEG 2 — 1 is input because T S is not divided, and 5 is input if D V C is divided into 5.
  • DBI is a 3-bit field, and contains a numerical value indicating the number of the block stored at the beginning of the AV data pack. In the case of MPEG2-TS, DBI is always 0 because it is not divided, but in the case of DVC, since the source packet is divided, the order of the blocks generated by dividing the source packet is shown. Enter an integer from 0 to 4.
  • the SPC is an 8-bit field, which is a power counter value of a source bucket counter indicating the number of the source bucket stored in the AV data pack, and includes a plurality of divided blocks. When a block is stored, it is used as the counter value of the first block.
  • QPC is an 8-bit field that indicates the amount of padding to be applied in bytes if the total number of stored proxies is less than 200 bytes.
  • DVC packets were not input continuously.
  • it is possible that 96 bytes of padding may be added, and in such a case, it becomes 96 (
  • 0 1 1 0 0 0 0 0 I 2 , and in other cases, 0 (
  • RSV is a 1-bit field that is reserved for future use.
  • SPH is a 1-bit field that indicates whether or not a source bucket was added when the AV data pack was created.
  • 1 is added to add a 4-byte source bucket header as time information, and in the case of DVC, 0 is stored.
  • FMT is a 6-bit field indicating the format of the source packet. This FMT has upper 1-bit power; 1 indicates that it has the SYT field. Immediately In the case of MPEG2-TS, I 0 0 0 0 0
  • FDF is an 8-bit or 24-bit field in which more detailed information is input than in the format specified by FMT.
  • This FDF has an 8-bit field in the case of DVC because it has 2 bytes of SYT, and has a 24-bit field in the case of MPEG2-TS where there is no SYT.
  • SYT is a 16-bit field, which is the time information input by the source bucket.
  • time information is used as a source packet header, so that it is not added.
  • DVC the above-mentioned 2-byte time information is stored here. If the head block of the source packet is included in the AV data block, the SYT stores time information when the source bucket of the block was input.
  • FMT, FDF, and SYT are relayed in accordance with IEC 61883, which is a method for transferring real-time AV data in IEE 1394.
  • the MPEG2-TS source packet and the DVC source packet are both converted into a 200-byte AV data pack and stored in the AV transmission buffer 7.
  • FIG. 8 is a diagram showing a process up to generation of a MAC packet from an AV data pack stored in the AV transmission buffer 7.
  • the AV transmission buffer 7 outputs the 200-byte AV data pack generated by the data block generation circuit 6 by attaching Tag, which is 4-byte additional information including an 8-bit identifier, to the AV data pack. I do.
  • Tag which is 4-byte additional information including an 8-bit identifier
  • the 8-bit identifier in Tag which is this additional information, corresponds to the value indicating the order of storage in the AV transmission buffer 7 as described above.
  • FCS 4-byte error detection code
  • the 8-bit identifier in the Tag indicates which coding block is used when the receiving station 3 returns an ACK in communication in which retransmission is performed on a Read Solomon coding block basis. Used to indicate successful reception. Therefore, T ag
  • the 8-bit identifier in the above may be a format that the receiving station 3 can determine, for example, a continuous number.
  • the buffer address of the AV transmission buffer 7 is used as an 8-bit identifier in order to facilitate management and eliminate the need for another buffer used for the management. .
  • the Tag information is added to the AV data pack to be transmitted.
  • a 208-byte AV data pack to which a 4-byte Tag and a 4-byte error detection code are added is encrypted by the encryptor 14 using this as a unit.
  • an encryption method based on a public key method called "basic WEP" which is currently proposed in IEEE802.lie is used.
  • a secret key is determined between the transmitting station 2 and the receiving station 3 in advance. Then, the transmitting station 2 encrypts with the secret key and the public key. In addition, the public key is changed for each transmission packet, added to the packet, and notified to the receiving station 3. The receiving station 3 decrypts the received public key and private key using the notified public key and private key.
  • the public key is generated by the MAC packet header generation circuit 15, it is given to the encryptor 14.
  • a MAC header and a public key are added to the encrypted data, and a 6-byte Reed-Solomon code is added by the error correction code adding circuit 17.
  • the error correction code adding circuit 7 the MAC header and the public key are used as a single Read-Solomon coding block, and the data is equivalent to a 208-byte AV data pack.
  • the AV data pack for each encrypted data unit, which has been decoded by the encryptor 14, is selected as a data portion by the selector 16, and a 16-byte read circuit is provided by the error correction code adding circuit 17. Dosolomo Is added. Then, when an error correction code adding circuit 17 adds a read-Solomon code to each data portion corresponding to each of the predetermined number of AV data packs and outputs the result, a selector 16 generates a MAC packet header generating circuit 1. The connection is switched so that the output of 5 is supplied to the error correction code adding circuit 17.
  • the error correction code adding circuit 17 stores 0 or 1 padding in 1 to 172 bytes for the data block (header portion) using this MAC header and public key. At the same time, processing is performed as data in which the MAC header and the public key are stored in 173 to 208 bytes.
  • the error correction code adding circuit 17 may be initialized so as to be in an internal state after the processing of padding of 0 or 1 for 172 bytes.
  • the selector 18 selects the output of the error correction code addition circuit 17 and supplies it to the error detection code addition circuit 19. Therefore, the error detection code addition circuit 19 This adds a 4-byte FCS. That is, FCS is added to error-correction-coded data consisting of one header portion and a predetermined number (n) of data portions that have been subjected to Reed-Solomon encoding. Therefore, error detection encoding is performed on a predetermined number (n + 1) of data with error correction codes by the Read Solomon encoding block. This is a process for maintaining compatibility with the conventional IEEE 802.11 bucket.
  • the reconfiguration of the AV data pack so that it can fit within 208 kbytes without waste involves suppressing an increase in the band generated by padding.
  • padding of 0 or 1 up to the 209th byte and the 239th byte is removed, and a 16-byte lead Solomon code is added to the AV data pack until the 209th byte.
  • padding may occur in the Read-Solomon coding protocol depending on the size of the AV data pack and the size of the source bucket. Since the padding is removed in step 7, no padding occurs in the packet transferred on the medium, and there is an advantage that the band use efficiency is improved.
  • the larger the amount of padding the greater the processing delay at the padding point during Lead Solomon encoding, and to compensate for the padding delay, the prefetch ahead of the padding is performed. Need to go and process. For this reason, a faster operation clock and a large number of buffers are required to store the read-ahead read-Solomon-encoded data.
  • the MAC header and the public key are stored by 1 to 36 bytes, and the remaining 37 to 208 bytes are padded with 0 or 1 for processing.
  • a buffer is required to store data that has been read-ahead and has undergone Lead Solomon encoding, as in the case of variable-length Lead Solomon encoding.
  • the selector 12 selects a de-blocking port from the LLC-PDU buffer 10
  • the blocks 13 to 19 perform the same operation as described above. Therefore, the MAC packet is generated by processing the data for the Internet stored in the LLC-PDU buffer 10.
  • the selector 18 selects the output from the selector 16 and supplies the output to the error detection code addition circuit 19
  • the read-solomon encoding process in the error correction code addition circuit 17 is not performed, and the MAC packet is not processed.
  • the header part generated by the header generation circuit 15 and the data part obtained by encrypting the AV data pack by the encryptor 14 are provided to the error detection code addition circuit 19. Therefore, the addition of the Lead Solomon code An FCS is added to data consisting of a head part and a predetermined number of data parts.
  • the receiving station 3 needs to determine whether or not the received bucket has been subjected to error correction coding.
  • this discrimination method there is a method of discriminating from an identification field provided in the MAC header that indicates the presence or absence of error correction coding processing. May also occur. Therefore, the reliability is low if the discrimination is performed only by the identification field indicating whether or not the error correction coding process is performed.
  • Fig. 9 shows the output and error of the error detection circuit 21 when the processing speed for the same amount of data is the same in the error correction circuit 22 and the error detection circuit 21. The time relationship between the outputs of the correction circuit 22 is shown. With reference to FIG. 9, the following describes the temporal processing relationship in each of the error detection circuit 21 and the error correction circuit 22.
  • the error correction circuit 22 starts the Euclidean operation after buffering the input data once. After that, in the error correction circuit 22, when the Euclidean operation is performed over a time corresponding to the length of the lead-Solomon encoding block, a Chien search is performed next.
  • the output of the MAC header after error correction is started only after the input of the MAC header to the error detection circuit 21 in the MAC bucket receiving circuit 2 () starts from the input of the 276 bytes. Minutes (data block including MAC header 52 bytes + Lead Solomon coding block 222 bytes) after, and then another 52 bytes of data processing time After elapse, it is confirmed whether or not the error correction for the MAC header is successful, and it is also confirmed whether or not to perform the error correction process. When this error correction is not performed normally, the input of the MAC header to the error detection circuit 21 is opened. An error correction error signal is output at the timing when the data processing time of 328 bytes has elapsed from the beginning.
  • the receiving station 3 stores the MAC packet in the MAC packet receiving circuit 20 so that the MAC packet subjected to the error correction processing and the MAC packet without the normal error correction can be processed independently and in parallel. Equipped with a block for processing the MAC bucket of the system.
  • FIG. 10 shows a flow chart for performing various processing on the MAC packet after determining whether or not the MAC packet has been subjected to error correction processing in the MAC packet receiving circuit 2 (). The operation will be described below.
  • the error detection circuit 21 checks the destination address of the MAC bucket (STEP 1). At this time, if the destination address is not addressed to the own station or is not a broadcast without destination designation (No), the normal processing at the subsequent stage of the MAC packet analysis circuit 26 is stopped and the error correction circuit is executed. 22 Perform only error correction processing in the second stage (STEP 2). That is, if the destination address is wrong, the error is detected by the error detection circuit 21 and cannot be normally received by normal processing. Therefore, there is no need to send back an ACK for this MAC packet. Also, at this time, if the destination address correctly indicates the address of another station, the communication is irrelevant to the receiving station 3.
  • the destination address of the corrected MAC packet is checked again by the MAC bucket analysis circuit 23 (STEP 3). At this time, if the destination address of the MAC packet is addressed to the own station or broadcast (Yes), processing is performed in the subsequent stage of the decoder 24, and a delayed ACK described later is sent to the ACK generation circuit 36. Sent (STEP 4). Conversely, if the destination address of the MAC packet is not addressed to its own station or broadcast (No), the current error correction processing is stopped (STEP 5).
  • step 1 if the destination address of the MAC bucket confirmed by the error detection circuit 21 is addressed to the own station or broadcast (Yes), the error detection circuit 21 sends From the error correction processing identification field, it is determined whether or not error correction coding has been performed (STEP 6). If the identification field indicates that the error correction coding has been performed (Yes), the processing shifts to STEP 2 and only the error correction processing of STEP 2 to STEP 5 is performed. At this time, if the error correction processing identification field indicates error correction processing even though error correction processing has not been performed, this indicates that an error has occurred in the MAC packet to be processed, and this MAC packet Does not need to return ACK in normal processing.
  • the error correction processing in the subsequent stage of the error correction circuit 22 and the MAC Bucket analysis circuit 26 Performs normal processing in the subsequent stage in parallel (STEP 7).
  • the error correction processing identification field of the error-corrected MAC header output from the error correction circuit 22 is set to MA.
  • the data is analyzed by the C bucket analysis circuit 23 (STEP 9).
  • the error correction processing identification field indicates that the error correction encoding is performed (Y es)
  • the state of the error correction error signal is confirmed (STEP 10).
  • the normal processing described above is stopped, and the error correction processing described above is performed. It is performed continuously (STEP 11).
  • the delayed ACK is generated by the ACK generation circuit 36 and transmitted.
  • the MAC header If the error correction processing identification field does not indicate that the error correction processing identification field is error-correction coded in STEP 9 (No), or in step 10, the MAC header If the processing is not performed normally (No), the above-described error correction processing is stopped and only the above-described normal processing is continued (STEP 14). When this normal processing is performed, an ACK is generated by the ACK generation circuit 36 and transmitted.
  • the error detection circuit 21 confirms that the error correction processing has priority. However, this method cannot be used when a variable-length Lead Solomon code is used.
  • the probability that each of the error correction circuit 22 and the error detection circuit 21 malfunctions is compared, and the one that stochastically does not malfunction is prioritized. Note that the probability of erroneously correcting an error with a Read-Solomon code is 1/6000, and the probability that an error detection code cannot correctly detect an error varies depending on the data length.
  • the output of the error detection circuit 21 is buffered once, and if no error is detected, normal processing is performed according to the MAC header content.If an error is detected, error correction processing is performed. There is also a way to do it. In this case, since it is not necessary to perform the processing in parallel, the determination algorithm can be further simplified as compared with the flowchart of FIG. 10, but a delay due to a buffer occurs. Hereinafter, the error correction encoding processing will be described.
  • Fig. 11 shows the flow from the normal determination of a bucket subjected to error correction coding processing to the storage of the AV data pack in the bucket in the AV transfer buffer. .
  • the MACC bucket When the MACC bucket is given from the error detection circuit 21 to the error correction circuit 22, first, error correction processing is performed on the header part. At this time, the header part obtained by the normal correction by the error correction circuit 22 is given to the MAC bucket analysis circuit 23, and when the MAC header and the public key are extracted, the public key is decoded. Given to. Also, regarding the data portion included in the MAC packet, when the error correction circuit 22 corrects the error, it is confirmed whether or not a correction error signal has been output. Further, the data portion on which the error correction is performed by the error correction circuit 22 is provided to the decoder 24 through the MAC bucket analysis circuit 23.
  • the decryption unit 24 decrypts the data part using the secret key set in advance and the public key obtained by the MAC packet analysis circuit 23, and the result is decoded by the error detection circuit 25.
  • the FCS determines whether the data part is correctly decoded.
  • the error detection circuit 25 performs Reed-Solomon decoding, and also determines whether or not the error has been corrected normally.
  • the AV data pack obtained by removing the FCS by the error detection circuit 25 is stored in the buffer address indicated by the 8-bit identifier in Tag of the AV transmission buffer 32.
  • the error correction error signal output from the error correction circuit 22 is observed, and if it is indicated that this signal was successfully received, the buffer address storing the AV data pack is stored in the AV transmission buffer 32.
  • the storage status display identifier of the AV transmission buffer 32 is used for the data portions # 1 and # 3 from which the correction error signal is output. It becomes 0.
  • the storage status display identifier of the AV transmission buffer 32 is 1. That is, an AV data pack in which the 8-bit identifier in Tag obtained from the data portions # 1 and # 3 decoded by the decoder 24 and the error detection circuit 25 is k—1, k + 1 Is stored in the buffer for AV transmission 32.
  • the storage status display identifier corresponding to the dress is 0.
  • the storage status identifier corresponding to the buffer address in the AV transmission buffer 32 in which the AV data pack in which the 8-bit identifier in Tag is Tag is stored is 1.
  • the output of the error correction circuit 22 is buffered with a buffer having a length corresponding to the Read Solomon coding block, and a correction error occurs.
  • the packet may be output to the MAC packet analysis circuit 23. In this case, the analysis by the MAC packet analysis circuit 23 is delayed, and more buffers are provided, but the control of the AV transmission buffer 32 can be simplified.
  • the source bucket output circuit 33 sequentially reads out the valid AV data packs and analyzes the contents of the AV data packs. That is, in the AV transmission buffer 32, the AV data pack stored in the buffer address having the storage status display identifier of 1 is read out, and analyzed by the source packet output circuit 33.
  • the type of the AV source packet stored in the AV data pack is confirmed by the FMT and FDF fields of the MAP.
  • the size of the removable block is confirmed by the DBS of the MAP, and the number of data blocks stored in the AV data pack is recognized from this value and the value of the padding confirmed by the QPC.
  • an AV source packet is reconstructed according to the data block number indicated by SPC and DBI in each MAP and the number of divisions indicated by FN.
  • M Outputs the reconstructed AV source packet while reproducing the time sequence based on the SYT in the AP or the time information indicated by the source bucket header.
  • FIG. 12 shows a processing flow of the source bucket output circuit 33 when the AV data pack is an AV data pack of an AV source packet based on MPEG2-TS.
  • the source packet output circuit 33 analyzes the MAP as described above. At this time, since the MAP of the MPEG2-TS has the value shown in Table 2, the source packet output circuit 33 outputs the AV data pack read from the AV transmission buffer 32 according to the MPEG2-TS. It is confirmed that the AV source bucket is stored without division with the source bucket header added.
  • the source packet output circuit 33 further divides the source packet header and the AV source packet, and adds the time information in the source packet header to an offset time to be described later and the system timer of the own station. 34 Compares the current time information given by 4 and outputs an AV source bucket when they match.
  • FIG. 13 shows a processing flow of the source bucket output circuit 33 when the AV data pack is an AV data pack of an AV source packet by DVC.
  • the source packet output circuit 33 analyzes the MAP as described above. Since the DVC MAP has the value shown in Table 2, the source packet output circuit 33 indicates that this AV data pack is based on DVC, and that the AV source bucket is divided into five parts. It is confirmed. Also, it is confirmed which order data block of which order bucket is stored in this AV data pack.
  • the source packet output circuit 33 combines the AV data blocks F0 to F4 in the order of FO, F1, F2, F3, and F4 to form one AV source packet. To be reconfigured. Furthermore, the time information is checked from the SYT of the MAP of the AV data pack storing the AV data block of index 0, and the time information obtained by adding an offset time described later to this time information is obtained. The offset time information is compared with the current time information provided by the system timer 34 of the own station, and when they match, the AV source bucket is output.
  • the method of calculating the offset time is described below. Due to the delay generated by the circuits configured inside the transmitting station 2 and the receiving station 3 ⁇ ⁇ Due to the delay caused by the timing of media access, the AV source bucket generates It is not possible to output the source packet output circuit 33 of the receiving station 3 using the time input to the circuit 6 as it is. For this reason, it is necessary to offset the time information of the AV source packet at the transmitting station 2, the receiving station 3, or both.
  • the calculation of the offset time is performed by the source packet output circuit 33 of the receiving station 3. This is because if there are multiple stations that receive the same packet, there is a time difference between the packet outputs at each station, but the transmitting station 2 performs transmission without considering the delay due to the internal circuit of the receiving station 3. There is an advantage that it can be performed.
  • the delay difference between the receiving stations 3 depends on the time required for the Reed-Solomon decoding process, but if all the stations perform the decoding process at the same speed, such a time difference can be suppressed. Can be.
  • the offset time may be calculated only by the data block generation circuit 6 of the transmitting station 2. In this way, even when there are a plurality of receiving stations 3, the AV source bucket outputs of all the receiving stations can be set at almost the same time.
  • the amount of delay due to the internal circuit configuration of the transmitting / receiving station in the network is determined in advance within a range where the system does not fail. Which method is better depends on what the network system equipped with the device of the present invention requires.
  • the AV source bucket input to transmitting station 2 as described above is output from receiving station 3.
  • the receiving station 3 if the entire MAC packet is lost or if an error occurs in the MAC header and the MAC packet is discarded, the receiving station 3 does not return an ACK to the transmitting station 2. For this reason, transmitting station 2 retransmits spontaneously at the next transmission opportunity.
  • the read-solomon code of the data part in the MAC packet is read. If an error occurs in the encryption block, receiving station 3 requests retransmission from transmitting station 2 using ACK. At this time, since decoding of the Lead Solomon code in the error correction circuit 22 takes time, it is impossible to return a normal AC ⁇ . Therefore, a delayed ACK that can return ACKs for multiple packets at once after decoding is used.
  • Fig. 14 shows the frame format of the delay ACK currently proposed in IEEE 802.111e.
  • the two byte FrammeControll field and the two byte Durattion field declare a delayed ACK.
  • the address of the transmitting station 2 on the receiving side is indicated by a 6-byte RA field
  • the address of the receiving station 3 on the transmitting side is indicated by a 6-byte TA field.
  • An 8-bit identifier in the 2-byte RecOrdCoult field sets the number of ACKRecord fields in 4-byte units.
  • the 4-byte ACKRecord field is configured as one field.
  • a 4-byte FCS is added as an error detection code.
  • the sequence number (Sequence No. for TC-bitmap # 0 bit) in the ACKRecord field indicates a bucket requesting retransmission. That is, the 4-bit TC ID in TC-seq indicates the identifier of the video file to be retransmitted, and the sequence number indicates the number of the AV source bucket to be retransmitted.
  • the 16-bit TC—bitmap contains the positional relationship on the time axis between the AV source bucket indicated by the sequence number and the AV source bucket indicated by the sequence number on a bit-by-bit basis. Is shown in In TC — bitmap, 1 is given to the bit indicating the position of the AV source bucket requested to be retransmitted, and 0 is given to the bit indicating the position of the AV source bucket that has been successfully received.
  • a delayed ACK is configured to enable not only retransmission in packet units but also retransmission in Lead Solomon coding blocks (AV data pack units).
  • the method of storing the AV data pack in the present embodiment is such that even if only a part of the Lead Solomon encoded block in the packet can be received normally and the AV data pack is not stored in the AV transmission buffer, Reproduction of this normally received part is possible.
  • Each AV data pack can be distinguished by an 8-bit identifier in Tag. For this reason, there is no need to retransmit in packet units, and only the AV data packs that could not be obtained due to error correction could be retransmitted.
  • Fig. 15 is an example of the Reed-Solomon coded block-based retransmission extension of the ACK frame format shown in Fig. 14, and the retransmission is performed in units of packets that perform retransmission in units of packets.
  • a Bitmap Object bit for identifying whether to perform the operation is provided in the 2-byte RecodCount field.
  • the BitmapObject bit is used to identify whether the ACKRecoord field indicates the sequence number of the packet or whether it indicates an 8-bit identifier in Tag. This bitmapObjecct bit is set to 1 when retransmission is performed on a lead-solomon coded block unit.
  • BitmapObject bit power When 1, the information in the ACKRecOrd field indicates the value of the 8-bit identifier in Tag corresponding to the buffer address in the AV transmission buffer 32. Then, the 16-bit TC-bitmap can indicate the subsequent reception status from the 8-bit identifier in Tag indicated by TC-Seq. In TC — bitmap, 1 is given to the bit indicating the position of the AV data pack requested to be retransmitted, and 0 is given to the bit indicating the position of the AV data pack that has been successfully received.
  • the identifier of the video file to be retransmitted is indicated by the 4-bit TCID in the TC-seq
  • the 8-bit identifier of the AV data pack to be retransmitted is indicated by the sequence number.
  • the positional relationship on the time axis between the AV data pack after the AV data pack indicated by the sequence number and the AV data back indicated by the sequence number is stored for each bit. Show Is done.
  • FIG. 16 shows an ACK frame generated by the ACK generation circuit 36 with respect to the reception result shown in FIG.
  • the storage status of each buffer in the AV transmission buffer 32 is confirmed by the storage status display identifier given from the AV transmission buffer 32.
  • the 8-bit identifier in Tag of the AV data pack received and processed by the MAC packet receiving unit 20 is recognized by the MAC unit 35 and transmitted to the ACK generating circuit 36.
  • a retransmission request is made according to the storage status display identifier given from the AV transmission buffer 32.
  • An 8-bit identifier in Tag added to the AV data pack is recognized, and a delayed ACK is generated based on the recognized 8-bit identifier.
  • the 8-bit identifier in Tag of the AV data pack normally received in the previous reception is stored in the ACK generation circuit 36.
  • the Bitmap Object bit power in the R ecord Count field is set to 1;
  • the number of ACKR ECORD field since the only one in this case, is the R ECORD C ount in full I Lumpur de R ECORD C ount force s l.
  • the sequence number in TC-seq of the ACKR ecord field is the 8-bit identifier next to the 8-bit identifier k-2 corresponding to the latest buffer address that was successfully received last time, that is, this time received. 8 corresponds to the buffer address where the AV data pack included in the data portion at the beginning of the AV source packet is stored.
  • the bit identifier is k_1.
  • the reception status of each AV data pack is checked. Confirmed by the storage status display identifier provided by the transmission buffer 32.
  • the error correction processing has not been properly performed on the AV data pack to be stored in the buffer address corresponding to the 8-bit identifiers k-11 and k + 1. Therefore, the storage status display identifier in the buffer address corresponding to the 8-bit identifier k-1, k + 1 is given as 0.
  • the buffer address stored in the TC-Seq does not need to be the next address of the latest buffer address received normally previously as in this embodiment, and the buffer address in which a reception error is considered to have occurred is not necessary. It does not matter.
  • the receiving station 3 may generate a delayed ACK of the received packet halfway and return it to the transmitting station 2.
  • the delay ACK for the packet for which the delay ACK could not be returned may be returned at any time as long as the lifetime is not exceeded.
  • the transmitting station 2 it is necessary for the transmitting station 2 to be able to determine the order of the packet from which the receiving station 3 has returned the delayed ACK. This is because, for example, the time until the delayed ACK is returned is set in advance between the transmitting station 2 and the receiving station 3 or as a parameter of the entire system, and after that time, However, the bucket buffer address for which no delayed ACK is returned can be dealt with by retransmitting.
  • the ACK generation circuit 36 when the MAC section 35 confirms that the PIFS time has elapsed after receiving CF-E ⁇ d described later transmitted from the control station 1, the ACK generation circuit 36 generates And sends the delayed ACK.
  • the receiving station 3 since the delayed ACK is transmitted after the PIFS time shorter than the DIFS time + the back-off time, which is the time until other stations can transmit, the receiving station 3 transmits the media before the other stations. Can be accessed.
  • the delayed ACK may be transmitted after the elapse of the SIFS time shorter than the PIFS time, or the control of the timing of the delayed ACK transmission may be performed by the control station 1.
  • ACK is generated by the ACK generation circuit 36 and transmitted.
  • the ACK transmitted when this normal processing is performed is transmitted to the ACK generation circuit 36 by the MAC section 35 when the error detection circuit 28 confirms that the MAC packet has been normally received.
  • the transmission of the ACK is instructed, and the ACK is transmitted from the ACK generation circuit 36. Therefore, the transmission of the ACK is performed after the reception of the MAC packet is completed and before the SIFS time elapses.
  • the MAC state is determined by the transmission standby state identifier and retransmittable identifier in AV transmission buffer ⁇ and the ACKR ecord field in the delayed ACK.
  • the section 8 controls the operation of the MAC bucket generation circuit 9 by setting the AV data pack read out by the MAC bucket generation circuit 9 from the AV transmission buffer 7.
  • transmission is performed in consideration of the life time, which is the life time when transmitting and receiving the AV source packet.
  • the life time which is the life time when transmitting and receiving the AV source packet.
  • the lifetime is set in advance in consideration of delays in the system, etc. Transmission is prohibited for AV source packets.
  • the AV source packet output from the receiving station 3 cannot be played back even if the AV source bucket whose life time has passed is retransmitted even if a retransmission request is received from the receiving station 3. Becomes Therefore, receiving station 3 When a retransmission request is made or when the transmitting station 2 actually starts retransmission, the time information of the AV source bucket is compared with the lifetime, and retransmission is prohibited if the lifetime exceeds the lifetime.
  • Fig. 17 shows the transmission standby state identifier for each buffer address in the AV transmission buffer 7 of the transmitting station 2 that has received the delayed ACK of Fig. 16 in the information in the ACKR ecord field of the delayed ACK. The following describes the operation of updating in response.
  • the AV transmission buffer 7 includes, for each buffer address, a transmission standby state identifier indicating a transmission standby state and retransmittable (that is, the AV data pack stored in the buffer address). Resendable identifier indicating that the lifetime has not been exceeded).
  • the bitmap object bit power of the delayed ACK is confirmed to be 1.
  • the ACACK ecord field is confirmed to be 1
  • the ACKR ecord field relates to the 8-bit identifier in Tag.
  • make the MAC part 8 recognize that it is information.
  • the sequence number of the delayed ACK in the ACKRecord field is confirmed in the MACC unit 8, the 8-bit identifier indicated by the sequence number is recognized.
  • the transmission wait state identifier for the buffer address for the 8-bit identifier k—1 and subsequent 8-bit identifiers in Tag is Confirmed by AV transmission buffer 7.
  • the AV data pack stored in the buffer address for each of the 8-bit identifiers k + 2 to k + 4 is in the transmission standby state, as shown in the bit map in the second row of FIG.
  • the 16-bit transmission standby state identifier of I 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 I 2 is given to the MAC unit 8.
  • the AV transmission buffer 7 confirms the retransmittable identifiers for the buffer addresses corresponding to the 8-bit identifiers after the 8-bit identifier k-1. Assuming that the time information of the AV data pack stored in the buffer address for each of the 8-bit identifiers k-1 to k + 4 is within the lifetime, the bit map in the third row in FIG. Thus, the retransmission enable identifier of 16 bits of I 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
  • 2 is given to the MAC section 8 from the AV transmission buffer 7. .
  • the bitmap of the identifier to be transmitted obtained above and the bitmap of this retransmittable identifier are ANDed, and the bitmap of the transmission standby state identifier is used as the result by using the result.
  • 2 The updated 16-bit transmission wait state identifier bitmap and the 8-bit identifier of the AV data pack confirmed to be transmitted at the next transmission based on the sequence number are the MAC packet generation circuit. Given to 9. That is, in the case of FIG. 17, an 8-bit identifier k-1, k + 1 to k + 4 is given.
  • the AV data pack of the buffer address corresponding to the 8-bit identifier given by the MAC unit 8 is read out from the AV transmission buffer 7 by the MAC packet generation circuit 9, A MAC packet is generated. Therefore, in the case of FIG. 17, the AV data pack of the buffer corresponding to the 8-bit identifier k-1, k + 1 to k + 4 is read out from the AV transmission buffer 7, and the MAC bucket is read. After the packet is created, this MAC bucket is sent.
  • the above-mentioned resendable identifier is obtained by examining the time information originally attached to all AV data packs and comparing it with the current time information by the system timer 5 and the time information recognized by the life time. Will be updated. That is, the time information obtained by subtracting the life time from the current time information confirmed by the system timer 5 is compared with the time information added to the AV data pack, and the time information added to the AV data pack is compared. If the time indicates an earlier time, resendable The offspring is 0.
  • the retransmittable identifier and the address buffer in the AV transmission buffer 7 of the transmitting station 2 are used.
  • the relationship shown in FIG. 18 may be made as shown in FIG.
  • the media access cycle is counted by the system timer 5, and when the AV data pack is input to the AV transmission buffer 7, the count value of the media access cycle counted by the system timer 5 is 18th. It is stored in the cycle field as shown. Further, the transmission standby state display identifier and the resendable identifier for the buffer address of the AV data pack input to the AV transmission buffer 7 are set to 1.
  • the resendable identifier for the buffer is set to 0.
  • the lifetime is three times the media access cycle, and the count value of the media access cycle counted by the system timer 5 is 23. . Accordingly, at this time, when the AV data pack is stored in the buffer address for the 8-bit identifier k + 4 in Tag in the AV transmission buffer 7, the count value 23 is stored in the cycle field for this buffer address. Is stored.
  • FIG. 19 is an example of a basic packet exchange between the control station 1 having the HCF, the transmitting station 2 and the receiving station 3 on a time axis.
  • one basic cycle includes a section C F P where no collision occurs and a section C P where a collision occurs.
  • the control section 1 transmits a band securing frame CF—Po 11 from the transmitting section 39.
  • the interval from the transmission of C F—P ol 1 to the transmission of C F—End from the control station 1 is secured as a section C F P where no collision occurs.
  • the transmitting station 2 that has received C F—P o 1 confirms that it is the transmitting station from C F—P o 11 1, generates RTS to the receiving station 3, and transmits it.
  • Receiving station 3 that has received RTS confirms that it is a receiving station from RTS, and generates and transmits CTS to transmitting station 2.
  • the transmitting station 2 that has received the CTS confirms that the receiving station 3 is in a receivable state, and sends the receiving station 3 a packet # 1 including the MAC bucket generated by the MAC packet generating circuit 9. Perform single transfer or burst transfer of ⁇ #n.
  • the MAC section 38 When the control station 1 confirms that the transmitting station 2 has released the band of the wireless medium using the algorithm of FIG. 22 described later, the MAC section 38 generates CF-End. By transmitting from the transmission unit 39, the section CFP where no collision occurs will be terminated.
  • the intervals of CF—Poll1, RTS, CTs, packets # 1 to #n, and CF—End are SIFS (16 ⁇ s).
  • the section CP where a collision may occur starts.
  • the delay A CK is returned in this section.
  • the response is sent after the PIFS time. It should be noted that the method of returning the delay A CK may be based on controlling the transmission time of the delay A CK by the control station 1 as described above.
  • the control station 1 sends a CF-Pol 1 signal from the control station 1 so that the section CP is terminated and the next basic cycle is started. Is sent.
  • the LAN packet transmitted in the section CP in the basic cycle ends transmission, after the CF-Po 11 transmitted first when the basic cycle at the current time starts is transmitted.
  • CF_Po 11 is transmitted to start the next basic cycle.
  • the IEEE 802.11a physical layer is a physical layer with a maximum communication speed of 54 Mbps and an OFDM system in a frequency band of 5 GHz. It is a physical layer that can transfer modes.
  • Preamble, Signal, Service, Tail, and PADS are added to the PDU that is a MAC bucket by the IEEE 802.11a physical layer, and are preamble and Signa.
  • the transfer time forces are S 16 S and 4 s, respectively.
  • the 16-bit Service, the 6-bit Tai1, the PADS, and the PDU, which is a MAC packet, can be transferred by the OFDM scheme with a maximum communication speed of 54 Mbps. When the maximum communication speed is 54 Mbps, the symbol length in the OFDM system is 2 16 bits. PADS is added so that the symbol length of the last symbol including T ai 1 is 2 16 bits.
  • the lower part of FIG. 19 shows the configuration of the MAC packet, which is the PDU in the above-mentioned IEEE 802.11a physical layer.
  • the maximum value of the field called the frame frame of the protocol, excluding the 32-byte MAC header and 4-byte FCS, is limited to 2 312 bytes. Have been. Therefore, it is possible to store up to 10 224-byte lead-Solomon coding blocks.
  • one of a plurality of lead solomon encoding blocks included in each packet #l to #n is included in the AV data pack requested to be retransmitted.
  • the bandwidth for the lead-Solomon coding block is secured.
  • this block is used to facilitate arithmetic processing.
  • the configuration is as follows. Therefore, the data size that can store the regular (not retransmitted) AV source bucket by DVC is 1728 bytes.
  • Delayed ACK is output after PIFS (25 ⁇ s) after CF-End transmission is completed.
  • the Reed-Solomon encoding process involves a delay of two Reed-Solomon encoding blocks.
  • the processing speed of the internal circuitry of the receiving station 3 is the same as the speed of the physical transfer path, the delay becomes 6 8 mu 5.
  • Station 4 may wait for DIFS time (34 ⁇ s) + random back-off time for media access in CP where collision may occur No. Therefore, when the CF-End is transmitted and the section CFP ends and the section CP starts, a delayed ACK is transmitted after a PIFS time shorter than the DIFS time elapses. It will not be disturbed by the transmission operation of this.
  • the delayed ACK is transmitted following the bandwidth securing frame CF—Po 11 transmitted by the control station 1. You may send it. In this case, the transmission time of the bandwidth securing frame and the waiting time until the transmission of the delayed ACK are further required before the transmission of the delayed ACK, but the delayed ACK can be transmitted more reliably. .
  • FIG. 20 shows the number of bucket bursts indicating the number of buckets included in the non-colliding section CFP and the actual transfer rate of the DVC source bucket, taking into account the transfer time of each control packet described above.
  • the graph shows the result of calculating the relationship between the length and the basic period.
  • the graph described as having no RAN is the case where the section in which no collision occurs is minimized
  • the graph described as having RAN is the transfer rate of 24 Mbps and the frame body length of 2.
  • the value when coexisting with a 312-byte RAN bucket is shown.
  • the actual transfer rate is based on the assumption that one of the 10 Reed-Solomon coding blocks constituting the data part of the packet is dedicated to retransmission, and the remaining It shows the transfer amount that can be transferred by the Dosolomon coding block. In the example shown in Fig.
  • the minimum period at which the actual transfer rate of DVC becomes 30 Mbps or more when coexisting with LAN about 6.8 ms (15 bursts, 15 actual transfers) 30.2 Mbps) as the basic period.
  • the basic period can be designed to be compatible with DVC transfer, which has a much higher transfer rate, so that it is possible to transmit MPEG2—TS, which has a lower transfer rate than DVC.
  • FIG. 21 is a flowchart for the control station 1 to determine whether or not to permit the media access to the transmission station 2 which newly requests a periodic media access.
  • the media access cycle of each station belonging to the communication network is set to an integral multiple of the basic cycle.
  • the upper limit of the total media access time of each station periodically polled in the basic cycle shall be the same as the media access time in the case of DVC transfer in this basic cycle.
  • control station 1 stores the media access time and the media access cycle within the basic cycle of each station that has permitted the periodic media access in the bandwidth management unit 41, and calculates the total time of the stored media access times. The media access time required for DVC transfer is subtracted, and the time obtained in this way is stored as the remaining bandwidth.
  • the control station 1 receives, at the receiving section 40, a band securing request frame for obtaining a new periodic media access permission transmitted from a station other than the control station 1 (hereinafter referred to as a transmitting station) ( STEP 5 1). At this time, the media access cycle and media access time required by the transmitting station are confirmed as bandwidth information from the bandwidth reservation request frame.
  • the band management unit 41 sets the media access cycle required by the transmitting station requesting the new periodic media access to the basic cycle. It is checked whether the value is an integral multiple (STEP 52). That is, it is confirmed by the band management unit 41 whether or not the transmission cycle of the band securing frame C F—Po 11 transmitted to this transmitting station is an integral multiple of the basic cycle.
  • the bandwidth management unit 41 determines that the media access time required by the transmitting station that transmitted the The remaining bandwidth is compared (STEP 53). Then, when it is confirmed that the media access time is shorter than the remaining bandwidth (Y es), the bandwidth management unit 41 compares the media access time with the basic cycle (STEP 54).
  • the bandwidth management unit 41 confirms that the media access time confirmed from the bandwidth information is shorter than the basic cycle (Yes)
  • the media access cycle and the media access time requested by the transmitting station are set to the bandwidth.
  • the media access time is stored in the management unit 41, and the time of the remaining bandwidth is updated by subtracting the media access time from the remaining bandwidth (STEP 55). After completing the operation of STEP 55, STE Proceeding to P51, the reception confirmation of the band securing request frame is performed.
  • the media access cycle is not an integral multiple of the basic cycle in step 52 (N o), or if the media access time is longer than the remaining bandwidth in step 53 (N o), or In 4, when the media access time is longer than the basic cycle (No), the request by the bandwidth reservation request frame received in STEP 51 is rejected, and the media access cycle confirmed by the bandwidth management unit 41 is rejected. And the media access time is discarded (STEP 56). When the operation of STEP 56 is completed, the process returns to STEP 51 again as in STEP 55, and the reception confirmation of the bandwidth securing request frame is performed.
  • control station 1 When the control station 1 permits the media access of the transmission station 2 in this way, the control station 1 detects the end of transmission within the period of the transmission station 2 and releases the band used by the transmission station 2. In order to achieve this, the operation is performed according to the flowchart shown in FIG.
  • the control station 1 receives the packet transmitted from the transmission station 2 and the ACK or the delayed ACK transmitted from the reception station in the reception section 40. Then, the received signal is confirmed by the MAC unit 38 to monitor the media (STEP 101). Then, in the MAC section 38, the state of the transmitting station 2 is monitored from the NF (Non Final) bit included in the MAC header in the packet transmitted from the transmitting station 2, and the NF bit is set to 0. The transmission completion is confirmed by confirming whether or not the transmission has been completed (STEP 102). That is, in the case of FIG. 19, the NF bit included in the MAC header in the packet #n is () and the NF bit included in the MAC header in the packet # 1 to # n-1 is (). The number is 1.
  • the process of confirming the end of transmission using the NF bit of the MAC header is compatible with PCF.
  • PCF communication the communication between the transmitting station 2 and the receiving station 3 always goes through the control station 1, so that the NF bit can be monitored. If the control station 1 could not receive the packet and could not return an ACK to the transmitting station 2, the packet was immediately retransmitted from the transmitting station. It is possible to confirm the completion of the transmission of the bucket from the transmitting station 2 by the processing of 102.
  • step 102 if the NF bit is 1 (No), the transmitting station 2 or the receiving station must wait until the time equal to or longer than the SIFS time has elapsed after the transmission of the packet from the transmitting station 2 has ended. Whether or not the signal from the station 3 has been received by the receiving unit 40 is confirmed by the MAC unit 38 from the time information given from the system timer 37 (STEP 103). That is, the packet transmitted from the transmitting station 2 is the last packet, and it is confirmed whether or not a time equal to or more than the SIFS time has elapsed since this packet was transmitted.
  • the MAC section 38 checks whether or not the TXOP time, which is the media access time during which the transmitting station 2 can transmit, has elapsed by referring to the time information from the system timer 39 (STEP 11). 104). Then, when the passage of the TXOP time is not confirmed (No), the processing operations of STEP101 and thereafter are performed again.
  • step 102 When it is confirmed in step 102 that the NF bit camera is S0 (Yes), and in step 103, it is confirmed that the SIFS time has passed since the end of packet transmission from the transmitting station 2.
  • step 103 polling is performed to another station in the section CFP where the current collision does not occur.
  • Whether or not there is a scheduling schedule is confirmed by the MAC section 38 'based on the bandwidth securing time information obtained from the bandwidth management section 41 (STEP 105).
  • CF—Pol 1 to allow After transmitting from the transmitting unit 39 (STEP 106), the operation after STEP 101 is performed. Conversely, if it is confirmed that there is no polling schedule (No), it is confirmed that the section CFP where no collision occurs has been completed, so that the section where no collision occurs will be terminated. As described above, the CF-End is transmitted from the transmitting section 39 (STEP 107).
  • the control station 1 transmits CF—Po 11 to poll the station set as the next transmitting station according to the schedule. After that, similarly, when the RTS is transmitted from the next transmitting station, and then the CTS is transmitted from the next receiving station, the CTS is transmitted to the transmitting station and the receiving station. Communication takes place between them. Then, when the last packet #n is transmitted from the transmitting station, the control station 1 confirms that the time equal to or longer than the SIFS time has elapsed after the transmission of the packet #n has been completed. Next, since the next polling schedule is not confirmed, the control station 1 transmits CF-End to end the section CFP.
  • jitter allowed in the media access cycle is provided. That is, in a section CP where a collision may occur, a jitter is set that is allowed to be transmitted even if the RAN bucket transmitted and received between the stations deviates from the basic cycle.
  • jitter for ending the next basic cycle earlier is required. Is set.
  • Figure 24 shows the communication operation of each station when the station 4 is transmitting a LAN bucket at a transfer rate of 24 Mbps at the CF-Pol 1 transmission timing of the control station 1. This is a timing chart for
  • the station 4 When the station 4 completes the transmission of the LAN packet after the end of the basic cycle, the station receiving the LAN packet from the station 4 transmits the LAN packet to the LAN packet at an SIFS time interval after the transmission of the LAN packet. ACK is returned. Then, the station 4 waits for DIFS + random back-off time to receive the ACK. At this time, when the control station 1 confirms the end of the LAN packet transmission by the MAC section 38, the control section 1 waits for a PIFS time and then transmits the band securing frame CF-Po 11 from the transmission section 39. Therefore, between the end of the basic cycle and the transmission of C F—Po 11, jitter occurs, which is a delay due to the LAN packet.
  • the jitter resulting from such a delay is stored in the MAC section 38. Then, in the MAC section 38, the time obtained by subtracting the delay jitter from the length of the basic cycle is set as the length of the next basic cycle. Then, after operating according to the flowchart of FIG. 22, C F — End is transmitted to end the section C F P where no collision occurs. After that, the section CP where a collision may occur may occur.However, when the control station 1 reaches the basic cycle set by the MAC section 38 as described above, the control station 1 ends this section CP and proceeds to the next basic cycle. To start the operation, CF-Pol 1 is transmitted from the transmitter 39.
  • the length of the next basic cycle to be started at this time is the length of the original cycle.
  • the control station 1 generates the length of the next basic cycle in which the jitter that is delayed occurs.
  • the length should be shorter by the length of the jitter. Therefore, in the next basic cycle in which jitter that occurs as a delay occurs, it is confirmed that the time obtained by subtracting the jitter from the length of the basic cycle has elapsed since CF-Po 11 was transmitted. Send CF—Po 11 to start the basic cycle. At this time, the jitter corresponding to the early termination to compensate for the band is generated.
  • the interference by station 4 may occur again, but if it is a 24 Mbps LAN packet, it will There is no problem because the transmission time of the data does not exceed the range of jitter.
  • control station 1 may have a collision. In the section CP where there is a possibility, it waits for PIFS time after transmission of the delayed ACK. Then, it transmits the bandwidth reservation frame C F—Po 11 and restarts the partition C F P in which no collision occurs. By continuing such an operation until the lost band is compensated, the interference of the station 4 can be prevented, and the band can be compensated promptly. Thus, transmission of AV source packets by DVC and coexistence with existing LAN transmitted at 24 Mbps becomes possible.
  • FIG. 25 shows a timing chart in the basic bucket exchange between the control station 1, the transmitting station 2, and the receiving station 3 provided with HCF when this section CP is not provided in the basic cycle.
  • CF—Po 11 which is a band securing frame
  • CF—P o RTS is transmitted from transmitting station 2 that has received 11
  • CTS is transmitted from receiving station 3 that has received this RTS.
  • the transmitting station 2 and the receiving station 3 are determined in this way, the transmitting station 2 transmits packets # 1 to #n including the MAC packet to the receiving station 3.
  • the transmission section 39 transmits CF-End.
  • the delay ACK is transmitted from the receiving station 3 to the transmitting station 2 after the lapse of the SIFS time from the end of the receiving of CF-End.
  • the intervals of C F—Poll 1, R T S, C T S, packets #l to #n, and C F—E n d are SIFS, as in the example of FIG.
  • control station 1 transmits CF-Pol 1 so that the next basic cycle starts. Note that after the transmission of the last delay ACK transmitted in the basic cycle, the transmission waiting time SIFS or more must be elapsed before the next basic cycle is started.
  • transmitting section 39 After waiting for 1 FS time, transmitting section 39 transmits CF—Po 11.
  • control station 1 detects the end of transmission within the period of the transmission station 2 and releases the band used by the transmission station 2 according to the flowchart shown in FIG. Works.
  • steps for performing the same processes as those in the flowchart of FIG. 22 are denoted by the same reference numerals, and detailed description thereof will be omitted.
  • control station 1 checks the signal received by the receiving section 40 with the MAC section 38 and monitors the media (STEP 101), and checks the signal transmitted from the transmitting station 2 in the packet. Check whether the NF bit in the MAC header is 0 (STEP 10
  • the CF-End is transmitted from the transmission unit 39 (STEP 15) 1). Thereafter, the receiving unit 40 checks whether or not the delayed ACK from the transmitting station 3 has been received (step 1552), and if the delayed ACK has not been received (No), the CF -SIFS time has passed after End transmission Whether it is or not is confirmed from the time information from the system timer 37 (STEP 153). At this time, if the elapse of the SIFS time is not confirmed (No), the operation after STEP 15 2 is performed again.
  • the transmitting unit 39 transmits CF-End (STEP155). Thereafter, it is confirmed whether or not a delayed ACK has been received (STEP 15 6). If no delayed ACK has been received (Yes :), it is confirmed whether or not the SIFS time has elapsed (STEP 15). 7). At this time, if the elapse of the SIFS time is not confirmed (No), the operation of STEP156 and thereafter is performed again.
  • the first CF in the basic cycle is After confirming from the time information from the system timer 37 that the time of the basic cycle has elapsed since the transmission of Po 11, CF—Po 11 Is transmitted from the transmitting unit 39 (STEP 158), and then the operations of STEP 101 and thereafter are performed.
  • the control station 1 When the last packet #n is transmitted from the transmitting station, the control station 1 confirms that the time longer than the SIFS time has elapsed since the completion of transmission of this packet #n. Then, CF-End is transmitted from one control station. After the SIFS time elapses after the end of this CF-End transmission, it is confirmed that the delayed ACK has been transmitted from the receiving station. After the SIFS time elapses after the delayed ACK transmission, the polling schedule is used. Next, the control station 1 transmits CF—Po 11 to poll the station set as the transmitting station.
  • the communication between the transmitting station and the receiving station becomes Is used for communication.
  • the control station 1 confirms that a time longer than the SIFS time has elapsed after the completion of transmission of this packet #n. Since the next polling schedule is not confirmed, first, CF-End is transmitted from the control station 1.
  • the SIFS time elapses, and when it is confirmed that the delayed ACK has been transmitted from the receiving station, the basic cycle is started and the CF-Po 11 is transmitted. It is checked whether the basic period has elapsed since transmission. Then, when it is confirmed that the time corresponding to the basic cycle has elapsed, the control station 1 transmits CF-Po 11 to start the next basic cycle.
  • CF-End After transmitting CF-End from control station 1, if it is confirmed that the time equal to or longer than the SIFS time has elapsed, if there is a polling schedule, CF-Pol 1 is immediately transmitted, If there is no polling schedule, CF_Po 11 is transmitted after confirming that the basic cycle has ended.
  • the life cycle of the media access cycle is an integral multiple of the basic cycle.
  • the time can be set to an integral multiple of the basic period. In this way, the lifetime of the AV source packet can be measured in the basic cycle, so that the management of the lifetime is facilitated.
  • a multiple of the media access cycle with respect to the life time corresponds to the count value of the retransmittable media access cycle.
  • a multiple of the basic period can be considered as a count value of the retransmittable basic period.
  • AV source The greater the number of retransmissions per packet, the greater the resistance to errors, but the longer the delay in playing back the AV source bucket in the network system, the higher the transmission station 2 and the higher the reception station 3 The size of the AV transmission buffers 7, 32 provided for each also increases.
  • the amount of source packets is 25.5 Kbytes, and the amount of MPEG2-TS high quality mode (24 Mbps) source packets is 200.4. K bytes.
  • the total amount of AV source buckets transferred during the lifetime is proportional to a multiple of the fundamental period with respect to the lifetime.
  • the total amount of AV source packets transferred in the question of the lifetime is the size required for the AV transmission buffers 7 and 32 of the transmitting station 2 and the receiving station 3, respectively.
  • the system timer 37 gives CF-Pol 1 and CF_End transmitted from the control station 1 every basic cycle.
  • the time information is added by the transmitting unit 39 and transmitted.
  • Such a packet format of CF—Pol 1 and CF—End is shown in FIGS. 28 and 29.
  • g (l, CF-Pol1 is given by a 32-byte MAC header to indicate that it is CF-Pol1 and a system timer 37, as shown in Fig. 28. It consists of a two-byte time stamp that serves as the obtained time information, data that specifies the station to be the transmitting station and includes a media access time, and an FCS that is a four-byte error detection code.
  • CF-End is a 32-byte MAC header for indicating that it is CF-End and time information given by the system timer 37.
  • Byte timestamp and 4-byte error detection It is composed of the code FCS.
  • the timer accuracy of IEEE 802.11 is 0.01% as described above, when the time is adjusted once every 6.8 ms, the jitter is reduced to about 1.3 s. Can be suppressed.
  • every time CF-Po 11 having the configuration shown in FIG. 28 and CF-End having the configuration shown in FIG. — Pol 1 and CF — End The time is adjusted by the system stamp provided for each. Therefore, since the system timer accuracy of IEEE 802.11 is on the order of ⁇ s, the jitter in the time sequence reproduction of the AV source packet can be almost ignored.
  • the control station 1 may transmit a frame dedicated to time adjustment separately from CF-Poll 1 and CF-End.
  • transmission may be started without issuing CF-Poll1.
  • a frame dedicated to time adjustment may be transmitted separately from CF-Poll1 and CF-End.
  • the basic cycle for transferring the DVC has been described as an example. However, for example, in a certain system, the transmission of the DVC or the MPEG2-TS high image quality mode is abandoned, and the normal MPEG2-TS. In some cases, you may want to shunt several channels, or in some systems you want to prioritize LAN as much as possible.
  • the life time is set to be an integral multiple of the basic period.
  • this AV transmission buffer is used.
  • the maximum time in which real-time AV data can be stored may be used as the lifetime. In this case, or when the lifetime of the system is set, the actual transfer rate is larger than the transfer amount of the AV data to be transferred, and the length of the cycle is equal to the integer of the lifetime.
  • One of The basic cycle may be set as shown in FIG.
  • the media access time given when each transmitting station performs transmission is the media access time when the real-time AV data is transferred at the highest transfer rate.
  • a new entry is made as a transmitting station so that the total value of the media access time does not exceed the media access time required to transfer the real-time AV data having the fastest transfer speed of the system.
  • each transmitting station notifies the control station i of the queuing information which is the free memory information of the AV transmission buffer at every basic cycle or every integral multiple of the basic cycle.
  • the transmitting station does not always transmit a variable-rate real-time AV data such as MPEG2-TS at the maximum transfer rate.
  • the total media access time required by each transmitting station at the maximum transfer rate cannot exceed the media access time required to transfer the real-time AV data with the fastest transfer rate.
  • the total estimated media access time does not exceed the media access time required to transfer the real-time AV data with the fastest transfer speed.
  • control station 1 can perform media access control for each transmitting station.
  • the memory size buffer of the AV transmission buffer must be increased in order to absorb the discrepancy caused by the difference between the transfer rate predicted by the control station 1 and the transfer rate actually required by each transmitting station. At the same time, the life time can be extended.
  • a tag containing a buffer address as an identifier is used in order to identify a lead solomon coded block to be retransmitted.
  • the position of the lead-solomon coding block that could not be corrected in the bucket may be notified.
  • the receiving station 3 stores, for example, the SPC and DBI in the AV transmission buffer 32 by referring to the SPC and DBI in the MAP so that the SPC and DBI are continuous.
  • the transmitting station 2 needs to have a buffer for storing the sequence number of the transmitted packet and the storage status of the AV data pack. Further, in the receiving station 3, not only the source packet output circuit 33 but also the AV transmission buffer 32 needs to have a function for analyzing the MAP. By doing so, it is possible to reduce the overhead of the Tag field as additional information.
  • the size of the AV transmission buffer in the transmitting station and the receiving station can be optimized, and the lifetime of the AV source bucket can be easily managed.
  • the basic period Media access time, so that it can coexist with existing stations in any situation.
  • the band used by the existing station can be increased.
  • the media access control of each station can be realized with a simple algorithm in the control station by determining in advance the basic period at which each station can perform media access in the network system. .
  • the lifetime By setting the lifetime to be an integral multiple of the basic period, the size of the AV transmission buffer at the transmitting station and receiving station can be optimized, and the life time of the AV source bucket can be easily managed. There is.
  • the control station aggressively performs bandwidth release processing, thereby distributing surplus bandwidth generated in sections where collisions do not occur to sections where collisions may occur. Can be increased.
  • the jitter of the AV source bucket generated between the transmitting station and the receiving station is minimized, and it is possible to watch the AV source at any station without discomfort. Become.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Hardware Design (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mobile Radio Communication Systems (AREA)

Description

明細書
通信装置及びこの通信装置によって構成される通信システム 技術分野
本発明は、 リ アルタイム A Vデータの転送を行う無線通信方式および装置、 特 に I E E E 8 0 2. 1 1無線通信方式および装置に関わる。 背景技術
I E E E 8 0 2. 1 1無線通信方式は、 本来無線 L ANのために規格化された 標準である。
この方式では、 一定周期で衝突が発生する可能性がある区間 (Contention Per iod : C P ) と衝突が生じない区間 (Contention Free Period: C F P ) が入れ替 わることを特徴と し、 衝突が発生する区問は主に L A Nの用途に、 衝突が生じな い区間は通信帯域の確保を必要とする通信のために使う こ とができる。
第 3 0図は、 従来の I E E E 8 0 2. 1 1 の P C F (Point Coordination Fun ction) という機能を使用したときに C F P と C Pの発生する区間を時間軸で示し た図である。
C F Pでは P C (Point Coordinator) と呼ばれる特定の端末が、 C F Pを利用 する端末をポーリ ングするこ とによ り、 通信帯域の確保を行う。 P Cは、 各端末 の時刻合わせなどを行うための情報 (T I Mと呼ばれる) を B e a c o n と呼ば れるフ レームに載せて一定間隔で送信する。 送信される B e a c o nのう ち数回 に 1 回の割合で、 ス リープ状態の端末に対する通知情報 (D T I Mと呼ばれる) を含む B e a c o nが送られる。 ス リープしている端末は D T I Mを含む B e a c o nが送信される時刻のあたり になるとアクティブ状態になり 、 D T I Mを含 む B e a c o nを受信し必要な通信を行う と再びス リープ状態に入る。
C F Pは D T I Mを含む B e a c o nに続いて発生し、 次の B e a c o nが送 信されるまでには終了する。 I E E E 8 0 2. 1 1 における B e a c o nの送信 周期は、 例えば 1 0 O m s程度の値の場合がある。
しかしながら、 C F Pを使って、 D V Cまたは M P E G 2— T Sなどのリ アル タイム A Vデータを送信する場合、 この周期が問題になる。 D V Cの標準画像は 3 0 M b p s 、 M P E G 2— T Sは標準で 6 M b p s、 高画質モー ドで 2 4 M b p s の転送レー トで連続的または継続的に送信局に入力されつづける。 しかし、 送信の機会、 即ち、 従来の I E E E 8 0 2. 1 1 において C F Pが発生する間隔 は数 1 0 0 m s に一回と非常に長く 、 その間に入力される膨大なソースパケッ ト をバッファ リ ングしなければならない。 更に、 数回の B e a c o n周期のうち 1 周期にしか C F Pが存在しないため、 たとえ現在規格化が進められている I E E E 8 0 2. 1 1 a 5 GH z物理層 (物理転送速度最大 5 4 M b p s ) を用いて も、 A Vデ一タの転送に必要な帯域を確保するこ とが困難である。
又、 有線メディアに比べ信頼性が劣る無線メディアでは、 通信路上でエラーや バケツ ト消失が発生しゃすい。 現在の I E E E 8 0 2. 1 1 MA C層は誤り訂正 符号もパケッ ト消失符号もないため、 通信路上でエラーが発生した場合、 バケツ トの再送が必要である。 しかしながら、 C F Pは数 1 0 O m s周期に一回しか発 生せず、 且つ再送待機中にも リ アルタイム A Vデータは刻々 と入力されつづける 。 従って、 再送を行う には更に数 1 0 0 m s分のデータを保持するバッファが必 要になる。
更に、 この周期は、 ユーザーに対し、 遅延という形で現れる。 このため、 ユー ザ一が流れてく る A Vデータに対して、 停止、 巻き戻しなどのコマン ドを発行し ても、 即座に反映できないという問題がある。
上記問題を解決するために B e a c o nの間隔を短くすることは、 即ち、 ス リ ープしている端末に対してよ り多く のアクティブ状態を強いるこ とになるため、 難しい。
このため、 リアルタイム A Vデータ伝送のよ う な Q o S (Quality of Service ) 通信を I E E E 8 0 2. 1 1 において実現するための規格化を行っている I E E E 8 0 2. l i e では、 2 0 0 1 年 3月現在、 H C F (Hybrid Coordination Function) という手法が提案されている。 P C Fでは数回に 1 回の B e a c o n 周期でしか衝突が生じない区間がなかったのに対し、 H C Fでは自由に衝突が生 じない区間を設けるこ とができる。 この方式では、 周期的なデータ送信を希望す る送信局が制御局 (Hybrid Coordinator : H C ) に対し、 メディアアクセス周期 及びメディアアクセス時間の要求を行い、 それを認めた制御局は、 C F P、 C P に関わらず、 帯域確保フ レームを定期的に送信局に送信 (ポーリ ング) するこ と で衝突が発生しない区間を生成する。
従来の I E E E 8 0 2. 1 1 の C F Pにおいては、 一般に H Cが P Cを兼ねる ため、 P C Fを搭載した従来の I E E E 8 0 2. 1 1 通信機器に通信を妨害され るこ とはない。
従来の I E E E 8 0 2. 1 1 の C Pにおいては、 従来の I E E E 8 0 2. 1 1 通信装置は D C F (Distributed Coordination Function) という機能を使って通 信を行う。 D C Fでは、 送信を開始する前に D I F S (Distributed Interframe Space) と呼ばれる 3 4 μ s の待機時間とランダムバックオフで定められた時間 だけキヤ リ アセンスする C S MA/C A方式が用いられる。
これに対し、 H C Fを備えた制御局 (H C) は C F P、 C Pに関わらず P I F S (PCF Interframe Space) と呼ばれる 2 5 μ s の時間間隔だけキャ リ アセンス すればメディアアクセスが可能になる。 又、 制御局 ( H C ) にポ一リ ングされた 被制御局は S I F S (Short Interframe Space) と呼ばれる 1 6 μ sの時間間隔 だけキャ リ アセンスするだけで送信を開始できる。 このため、 H C F機能を搭載 した機器が通信を開始すると、 D C Fで動作する従来の局はメディアアクセスが できなく なり、 H C Fを備えた制御局 (H C) 、 被制御局は従来の局に妨害され るこ となく通信できる。
しかしながら、 P C Fでは P Cが決めた B e a c o n間隔に基づき各送信局が 動作していたのに対し、 現在提案されている H C Fでは各送信局が、 それぞれ、 H Cに対しメディアアクセス周期 (ポーリ ング周期) 及びメディアアクセス周期 内での通信を行うための時間を要求する。 このため、 H Cは、 複雑な各送信局の メディアアクセススケジューリ ングをしなければならない。
又、 無線メディア上で発生した誤りに対し再送を行う場合、 再送用の帯域を別 途確保しておく必要がある。 この再送による負担を減らすために、 現在誤り訂正 符号 (リー ドソロモン符号) を使用する提案も行われている。 このリー ドソロモ ン符号が用いられる場合、 パケッ トが複数のブロ ックに分割されて、 各ブロ ック 毎にリー ドソロモン符号化が成される。 よって、 送信されるデータのパケッ トは 、 複数のリー ドソロモン符号化ブロ ックによって構成される。 しかし、 現在の提 案では、 バケツ ト内のほとんどのリ ー ドソロモン符号化ブロ ックでエラ一訂正に 成功した場合であっても、 一つでも リー ドソロモン復号に失敗すればバケツ ト全 体を再送する必要がある。
更に、 実際の通信においてはプライバシ保護のため暗号化が行われる。 この喑 号化が各プロ ック毎に行われ、 暗号化が行われた各ブロ ックにリー ドソロモン符 号化による誤り訂正符号が付加される。 そして、 この誤り訂正符号が付加された 複数のリー ドソロモン符号化ブロ ックによって成るデータ部全体に対し、 誤り検 出符号が付加される。
現在の提案では M A Cヘッダと同じリ ー ドソロモン符号化ブロ ックに喑号鍵 ( 公開鍵) が添付されているため、 このリ ードソロモン符号化ブロ ックに対する誤 り訂正が成功すれば、 データ部の各リー ドソロモン符号化ブロ ックの暗号鍵によ る復号は可能となる。 しかし、 パケッ ト 自体の復号の成功を保証するには、 デー タ部を構成する全てのリー ドソロモン符号化ブロ ックに対する誤り訂正が成功す る必要がある。
上記にあげた 2つの状況は、 ともに誤り訂正に成功したリー ドソロモン符号化 プロ ックまでも再送せねばならず、 そのために余計な通信帯域を確保せねばなら ない。 即ち、 再送するデータ量が、 本来再送を必要とするデータ量に比べて大き く なり、 データ転送の効率化を図るべき無線通信において、 非効率的なデータ転 送を行う こと となる。
上記とは別の問題と して、 システム全体の時刻合わせの問題がある。 第 3 1 図 が、 ネッ ト ワークシステムにおける各局のシステムタイマの精度と、 受信局のソ ースバケツ ト出力で発生するジッタを示す図である。
送信局は A Vソースパケッ トに対し自局のタイマが示す入力時の時刻情報を添 付し、 受信局は添付された時刻情報と 自局のタイマを使って A Vソースバケツ ト のタイムシーケンスを再生する。 しかし、 各局のタイマは全く 同じ精度ではない ため、 ある程度の時間が経過すると、 各々の時刻がずれてく る。 この送信局、 受 信局の間の時刻ずれが、 受信局が A Vソースバケツ 卜のタイムシーケンス再生を する際にジッタ と して現れる。 即ち、 第 3 1 図のよ うに、 制御局のシステムタイマに比べて、 送信局のシステ ムタイマが進むと、 送信局で送信されるソースバケツ トのタイ ミ ングが早く なる と と もに、 ソースバケツ トに付加される時刻情報が制御局のシステムタイマの表 す時刻よ り早いものとなる。 逆に、 第 3 1 図のよ うに、 受信局のシステムタイマ が遅れると、 受信局で出力されるソースパケッ トのタイ ミ ングが遅れて、 ソース パケッ トの出力間隔が本来の出力間隔と比べて長く なり、 ジッタが生じる。 更に 、 定期的に与えられるビーコンによって、 各局のシステムタイマが一致させられ るので、 時間遅れが発生している受信局におけるソースバケツ トの出力間隔が本 来の出力間隔と比べて短く なり 、 同じく ジッタが生じる。
一方、 個々の A Vソースパケッ トは出力タイ ミ ングが厳しく決まっており、 定 められたジッタ範囲内で出力されなければならない。 しかしながら、 現在の H C Fではビーコンによる時刻合わせしか行われない。 I E E E 8 0 2 . 1 1 におけ るタイマ精度は 0 . 0 1 %であるが、 仮に 1 0 0 m s に一回送信されるビーコン でしか時刻合わせを行わないとすると、 送信局と受信局の間では最大 2 0 μ s程 度の時刻ずれが発生する。
上述したバケツ ト毎のデータ再転送によるデータ転送の非効率化や、 許容範囲 外となるジッタの発生によるデータの欠落及び出力異常などによ り、 I Ε Ε Ε 8 0 2 . 1 1 などで設定される無線通信システムは、 A Vデータの転送に十分な通 信システムではない。 発明の開示
よって、 本発明は、 A Vデータ転送に適切となる効率的なデータ送信を行う こ とができる無線通信システムを提供することを目的とする。
本発明の通信システムは、 複数の通信装置によって構成されると ともに、 該複 数の通信装置の内の 1 つを制御局とする通信システムにおいて、 前記制御局によ つて送信局と される通信装置が、 時刻を管理するシステムタイマと、 アプリ ケー シヨ ンデータに対して前記システムタイマよ り得られた該アプリ ケーショ ンの入 力時刻に基づいて生成した時系列的管理用の時刻情報を付加すると ともに、 前記 アプリ ケーショ ンデータのうち保持期間が定められたアプリ ケーショ ンデータに 対して送信可能な期間を設定するデータ処理部と、 該データ処理部で前記時刻情 報が与えられた前記アプリ ケーショ ンデータを一時的に保持するバッファ と、 該 バッファから前記アプリ ケーショ ンデータを読み出すと もに該アプリ ケーショ ン データに誤り訂正符号を付加する誤り訂正符号付加部と、 該誤り訂正符号付加部 で誤り訂正符号が付加されたアプリ ケーショ ンデータから送信パケッ トを作成す る送信パケッ ト生成部と、 該送信パケッ 卜生成部で作成した前記送信バケツ トを 送信する送信部と、 前記バッファからアプリ ケーショ ンデータを読み出すことを 制御すると ともに、 前記送信バケツ ト生成部で生成された前記送信バケツ トを通 信帯域が確保された時間に前記送信部から送信するこ とを制御する送信制御部と 、 前記送信部から送信した一つ又は複数の送信バケツ トの受信を示す応答信号を 受信する受信部と、 を備え、 前記送信部から送信する際に、 前記送信制御部にお いて、 前記システムタイマから得られる現在時刻と前記バッファ内に格納された アプリ ケ一ショ ンデータに対して設定された前記送信可能な期間とを比較し、 該 現在時刻が前記送信可能な期間を超過しているアプリ ケーショ ンデータについて は、 前記誤り訂正符号付加部が前記バッファから読み出さず、 又、 前記送信部よ り前記送信パケッ トを送信した後、 所定時間以上経過しても、 前記送信パケッ ト に対する前記応答信号を前記受信部で受信しなかったことが前記送信制御部にお いて確認されると き、 該送信バケツ トが正しく受信されていないと判定すること を特徴とする。
又、 本発明の別の通信システムは、 複数の通信装置によって構成されると と も に、 該複数の通信装置の内の 1 つを制御局とする通信システムにおいて、 前記制 御局によって受信局と される通信装置が、 時刻を管理するシステムタイマと、 時 系列的管理用の時刻情報が付加されたアプリ ケーショ ンデータから成る送信パケ ッ トを受信する受信部と、 該受信部で受信された送信バケツ 卜に誤り訂正符号が 付加されているか否かを判定する誤り訂正判定部と、 前記受信部で受信された送 信バケツ トのう ち誤り訂正符号が付加された送信パケッ トに対して誤り訂正を施 す誤り訂正処理部と、 誤り訂正が成された送信バケツ 卜よ り アプリ ケーショ ンデ ータを生成する第 1復号部と、 前記受信部で受信された送信バケツ トのう ち誤り 訂正符号が付加されていない送信バケツ トょ り前記アプリ ケ一ショ ンデータを生 成する第 2復号部と、 前記第 1及び第 2復号部で生成された前記アプリ ケーショ ンデータを一時的に格納する第 1及び第 2バッファ と、 前記システムタイマによ る時刻が前記アプリ ケーショ ンデータに付加された時刻情報による時刻と一致し たこ とを確認すると、 前記アプリ ケーショ ンデータを前記第 1又は第 2バッファ から読み出して出力するデータ出力部と、 1 つ又は複数の送信バケツ トの受信状 態を示す応答信号を送信する送信部と、 を備え、 前記受信部で受信された送信パ ケッ トに対して、 前記誤り訂正処理部及び前記第 1復号部による第 1演算処理と 、 前記第 2復号部による第 2演算処理とを同時に施すことを特徴とする。
本発明の別の通信システムは、 複数の通信装置によって構成されると と もに、 該複数の通信装置の内の 1 つを制御局とする通信システムにおいて、 前記制御局 となる通信装置が、 時刻を管理するシステムタイマと、 他の通信装置から送信さ れる通信帯域の確保を要求する帯域確保要求信号を受信する受信部と、 通信帯域 が確保され衝突が生じない非衝突区間を時分割で管理する帯域管理部と、 前記シ ステムタイマよ り得られる現在時刻を他の通信装置に認識させるための時刻情報 信号と、 前記非衝突区間の開始を他の通信装置に認識させるための非衝突区間開 始信号と、 前記非衝突区間の終了を他の通信装置に認識させるための非衝突区間 終了信号と、 を送信する送信部と、 を備え、 前記非衝突区間を発生させる周期を 基本周期と し、 前記受信部で受信された前記帯域確保要求信号よ り確認される通 信許可を求める期間の周期であるメディアアクセス周期が、 前記基本周期の整数 倍でないとき、 該他の通信装置の帯域確保の要求を拒否することを特徴とする。 図面の簡単な説明
第 1 図は、 本発明に係わる無線通信ネッ トワークの一実施例を示す概略構成図 である。
第 2図は、 本発明に係わる送信装置のプロ ック図である。
第 3図は、 本発明に係わる受信装置のブロ ック図である。
第 4図は、 本発明に係わる制御局のブロ ック図である。
第 5図は、 本発明に係わる M P E G 2 — T Sバケツ ト入力処理を示す図である 第 6図は、 本発明に係わる D V Cバケツ ト入力処理を示す図である。
第 7 A図及び第 7 B図は、 本発明に係わる MA Pフォーマツ トを示す図である 第 8図は、 本発明に係わるバケツ ト送信処理の流れを示す図である。
第 9図は、 本発明に係わる誤り検出回路出力に対する誤り訂正回路出力の遅延 を示す図である。
第 1 0図は、 本発明に係わるパケッ ト内の誤り訂正符号有無を判別するフロー チヤ一 トを示す図である。
第 1 1 図は、 本発明に係わるバケツ ト受信処理の流れを示す図である。
第 1 2図は、 本発明に係わる MP E G 2— T Sパケッ ト出力処理を示す図であ る。
第 1 3図は、 本発明に係わる D V Cバケツ ト出力処理を示す図である。
第 1 4図は、 本発明に係わる現在提案されている遅延 A C Kフ レームフォーマ ッ トを示す図である。
第 1 5図は、 本発明に係わる リー ドソロモン符号化プロ ック単位再送のために 拡張した遅延 A C Kフ レームフォーマツ トを示す図である。
第 1 6図は、 本発明に係わる遅延 AC Kフ レームを示す図である。
第 1 7図は、 本発明に係わる送信局の送信状況更新を示す図である。
第 1 8図は、 本発明に係わるライフタイムを考慮した AV伝送用バッファを示 す図である。
第 1 9図は、 本発明に係わる送信周期とバケツ ト構成を示す図である。
第 2 0図は、 本発明に係わる D V CZ L AN ( 2 4 M b p s ) 共存時のバース ト回数と実転送レー トおよび基本周期を示す図である。
第 2 1 図は、 本発明に係わる帯域管理のフローチャー トを示す図である。 第 2 2図は、 本発明に係わる帯域解放のフローチャー トを示す図である。 第 2 3図は、 本発明に係わる送信周期とバケツ ト構成を示す図である。
第 2 4図は、 本発明に係わる L A Nパケッ トによる帯域損失に対する補填を示 す図である。
第 2 5図は、 本発明に係わる送信周期とバケツ ト構成を示す図である。 第 2 6図は、 本発明に係わる帯域解放のフローチヤ一トを示す図である。
第 2 7図は、 本発明に係わる送信周期とバケツ ト構成を示す図である。
第 2 8図は、 本発明に係わる C F— P o l 1 フ レームを示す図である。
第 2 9図は、 本発明に係わる C F— E n d フ レームを示す図である。
第 3 0図は、 P C Fの C F P と C Pの関係を示す図である。
第 3 1 図は、 従来のタイマ精度と時刻合わせを示す図である。 発明を実施するための最良の形態
以下に本発明の実施例を示す。 本発明における通信ネッ トワークは有線、 無線 を特定しないが、 以下の実施例では I E E E 8 0 2. 1 1無線通信方式を用いた 場合を例にして説明を行う。
第 1 図は、 本発明を実施する無線ネッ トワーク構成例である。
第 1 図の 1 は衝突が発生しない区間において送信を行う局の無線メディアへの アクセスを制御する制御局である。 本実施形態ではメディアアクセス制御機能と して H C Fを備えるものとする。 この制御局 1 は、 通常、 H C (Hybrid Coordin ator) と呼ばれる。
第 1 図の 2はリ アルタイム AVデータの送信局、 3はリ アルタイム AVデータ の受信局である。 リ アルタイム AVデータ送信局 2及ぴリ アルタイム AVデータ 受信局 3の衝突が発生しない区間 ( C F P ) におけるメディアアクセスは制御局 1 によって管理される。 本実施形態では、 この C F Pにおけるメディアアクセス 管理を機能させるために、 制御局 1 及び送信局 2及び受信局 3はそれぞれ、 H C Fを備えるものとする。
第 1 図の 4は衝突が起きる可能性がある区間 (C P ) のみメディアアクセスで きる従来の局である。 本実施形態では、 局 4は D C Fのみを備えるものとする。 尚、 第 1 図において、 それぞれの局には送信、 受信、 制御、 従来のメディアァ クセスのよ うにひとつの機能しか実装されていないよ う に記載しているが、 これ は説明を簡潔にするためであり 、 実際には一つの局に複数の機能を実装するもの と してもよレ、。
第 2図は、 送信局 2の構成を示すプロ ック図である。 第 2図の 5はネッ トワークシステム全体の時刻情報を管理するシステムタイマ である。 システムタイマ 5は制御局 1 が発行する B e a c o nなどの時刻情報を 格納したフ レームに含まれる現在時刻情報に基づいて時刻合わせをするこ とで、 常に制御局 1 内のシステムタイマとほぼ同じ時刻になるよ うに動作すると と もに 、 この現在時刻情報に基づいてメディアアクセス時間を管理する。
第 2図の 6は入力された A Vソースパケッ トを転送に適した形に変換するため のデータブロ ック生成回路である。 又、 このデータプロ ック生成回路 6は、 複数 のデータブロ ックの集合体に対して、 A Vソースバケツ トに対するデータブロ ッ クの生成情報 (本実施形態ではこの情報を M A P と呼ぶこ とにする) 、 及び A V ソースパケッ トが入力されたときのシステムタイマ 5で測定される時刻情報を添 付する。
A Vソースバケツ トのデータプロ ック生成回路 6への入力時刻を管理するタイ マを別途用意する方法もあるが、 本実施形態では説明を簡単にするために、 シス テムタイマ 5を使用している。 この M A P と時刻情報を付加されたデ一タブロ ッ クの集合を、 本実施形態において、 A Vデータパック と呼ぶことにする。
第 2図の 7は A V伝送用バッファであり、 データブロ ック生成回路 6で生成さ れた A Vデータパックを順次格納する。 A V伝送用バッファ 7は、 A Vデータパ ック単位でバッファア ドレス管理を行い、 且つ、 送信待機状態にあるバッファァ ドレスを示す識別子を備える。 この識別子は、 A Vデータパック単位毎のバッフ ァア ドレスに対応して与えられる 1 ビッ トのデータである。 そして、 A V伝送用 バッファ 7 に A Vデータパックが入力されると、 入力された A Vデータパックの バッファア ドレスに対応する識別子を 1 と し、 又、 A V伝送用バッファ 7から A Vデータパックが出力されると、 出力された A Vデータパックのバッファァ ドレ スに対応する識別子を() とする。 A V伝送用バッファ 7 のバッファサイズは、 後 述するライフタイム分の A Vデータパックを格納できるサイズとする。 又、 A V 伝送用バッファ 7には、 後述する再送可能であるこ とを示す再送可能識別子をも 備える。
第 2図の 8は送信局 2のメディアアクセスを制御する M A C ( Medi a Access C ontro l ) 部であり、 第 2図の 9は A V伝送用バッファ 7から出力される A Vデ一 タパック、 または L L C (Logical Link Control) 層のデータをバッファ リ ング する L L C— P D U (LLC Protocol Data Unit) 用バッファ 1 0内のデータを M A Cバケツ 卜に格納するための MA Cバケツ ト生成回路である。
第 2図の 1 1 は、 他局から送信されたパケッ トを受信するための受信部である 。 受信部 1 1 が制御局 1 が送信した B e a c o nを受信した場合、 B e a c o n 内部に格納されている時刻情報をシステムタイマ 5に通知し、 システムタイマ 5 はその時刻情報を使って自身の時刻情報を更新する。 受信部 1 1 が制御局 1 から の AV伝送のための帯域確保フ レームを受信した場合、 MA C部 8にその旨を通 知する。 通知された MA C部 8は、 AV伝送用バッファ 7の状態を調べ、 バッフ ァに送信待機状態の A Vデ一タパックが存在しているときは、 MA Cパケッ ト生 成回路 9に対し MA Cパケッ トの生成を開始するよ う要求する。 受信部 1 1 が受 信局 3からの A C Kを受信したときは、 受信局 3が再送を必要と しているかどう かを判断し、 再送が必要であれば、 その旨を M A C部 8に通知し、 次回の送信時 に必要なデータを送信できるよ う にする。
MA Cパケッ ト生成回路 9は、 入力データを選択するセレク タ 1 2 と、 入力さ れたデータに対し誤り検出符号を付加する誤り検出符号付加回路 1 3 と、 誤り検 出符号を付加した状態のデータを暗号化する暗号化器 1 4 と、 MA Cパケッ トへ ッダを生成する MA Cパケッ トヘッダ生成回路 1 5 と、 MA Cパケッ トヘッダと 暗号化されたデータを選択するセレクタ 1 6 と、 セレクタ 1 6の出力に対し誤り 訂正符号を付加する誤り訂正符号付加回路 1 7 と、 誤り訂正符号付加回路 1 7で の誤り訂正符号をデータに付加するか否かを選択するセレクタ 1 8 と、 入力され たパケッ ト全体に対し誤り検出符号を付加する誤り検出符号付加回路 1 9 と、 で 構成される。 本実施実施形態において、 誤り訂正符号付加回路 1 7はリー ドソロ モン符号 ( 2 2 4, 2 0 8 ) を极ぅ ものとする。 即ち、 2 0 8ノくイ トのデータに 対し、 1 6バイ トのリ ー ドソロモン符号を付加し、 全体と しては 2 2 4バイ トの データを出力する。
第 3図は、 受信局 3の構成を示すプロ ック図である。
第 3図における MA Cバケツ ト受信回路 2 0は、 パケッ ト全体の誤り検出を行 う誤り検出回路 2 1 と、 バケツ 卜に付加された誤り訂正符号によ り誤り訂正を行 う誤り訂正回路 2 2 と、 H C Fに対応した MACバケツ トを解析する MACパケ ッ ト解析回路 2 3 と、 データに施された暗号を復号する復号器 2 4 と、 復号され たデータの誤り を検出する誤り検出回路 2 5 と、 データの出力先を L L C— P D U用バッファ 3 0及び AV伝送用バッファ 3 2のいずれにするかを選択するセレ クタ 2 9 と、 リー ドソロモン符号を付加されていないバケツ トを解析する MA C パケッ ト解析回路 2 6 と、 データに施された暗号を復号する復号器 2 7 と、 復号 されたデータの誤り を検出する誤り検出回路 2 8で構成される。 本実施形態では 、 誤り訂正回路 2 2はリー ドソロモン符号 ( 2 2 4, 2 0 8 ) を极う。 即ち 2 2 4バイ トのデータに对し、 末尾 1 6バイ トをリー ドソロモン符号と して极ぃ、 ェ ラー訂正を行った 2 0 8バイ トのデータを出力する。
尚、 第 3図の受信局 3は、 MA Cパケッ ト解析回路、 復号器、 誤り検出符号を それぞれ 2つずつ備えているが、 これはリー ドソロモン復号処理には時間がかか るため、 リー ドソロモン復号中に従来の M A Cバケツ トを受信したよ うな場合に 、 従来の M A Cパケッ トの処理ができなく なるこ とを防ぐためである。 従って、 このよ う な機能が必要ではない場合には、 MA Cパケッ ト解析回路、 復号器、 誤 り検出符号それぞれ 1 つのみを備えるものと しても構わない。 又、 L L C— P D U用バッファについても、 同時に 2系統からの書き込みが可能な制御回路をもつ バッファであれば L L C一 P D U用バッファを 1つと しても構わない。
誤り検出回路 2 5から出力されるデータが A Vデータパックであるならば、 セ レクタ 2 9を通じて AV伝送用バッファ 3 2に与えられて、 AV伝送用バッファ 3 2に格納される。 AV伝送用バッファ 3 2は、 AVデータパック単位にァ ドレ ス管理を行い、 バッファァ ドレスに対して A Vデータパックの格納状態を示す識 別子を備える。 この識別子は、 A Vデータパック単位毎のバッファア ドレスに対 応して与えられる 1 ビッ トのデータである。 そして、 AV伝送用バッファ 3 2に 正常な A Vデータパックが入力されると、 入力されたデータパックのバッファァ ドレスに対応する識別子を 1 と し、 A V伝送用バッファ 3 2から A Vデータパッ クが出力されると、 出力されたデータパックのバッファァ ドレスに対応する識別 子を 0 とする。 AV伝送用バッファ 3 2のバッファサイズは、 後述するライフタ ィム分の AVデータパックを格納できるサイズとする。 第 3図の 3 3はソースバケツ ト出力回路である。 このソースバケツ 卜出力回路 3 3は、 AV伝送用バッファ 3 2に蓄えられている AVデータパックからソース バケツ トを再構築し、 ソースパケッ トに添付されている時刻情報又はソースパケ ッ トから算出された時刻情報とシステムタイマ 3 4に従ってタイ ムシーケンスを 再現し、 ソースパケッ トを出力する。
M A Cバケツ ト受信回路 2 0が B e a c o nなどの時刻情報を格納したフ レー ムを受信すると、 その時刻情報をシステムタイマ 3 4に通知し、 システムタイマ 3 4は通知された時刻情報に従って自身の時刻情報を更新する。
第 3図の 3 5は、 受信装置のメディアアクセスを制御する MA C部である。 こ の MA C部 3 5によって、 システムタイマ 3 4の時刻情報や、 メディアセンスに よって自局がメディアアクセス可能であるかどう力、、 即ち、 A C K送信可能な時 であるかを判断する。
第 3図の 3 6は A C K生成部である。 この じ! 生成部 3 6は、 MA Cバケツ ト受信部 2 0の状況や AV伝送用バッファ 3 2の状態に応じて、 送信局 2に対し AC Kを生成する。
第 4図は、 制御局 1 の構成を示すプロ ック図である。
第 4図で 3 7はシステムタイマであり 、 MA C部 3 8はこのシステムタイマ 3 7の時刻情報をも とにメディアアクセス制御を行う。 システムタイマ 3 7が規定 の時刻を示すと、 MA C部 3 8 は B e a c o nや帯域確保フ レームを送信するよ う送信部 3 9に要求する。
第 4図の 4 0は、 他局から送信されたバケツ トを受信するための受信部であり 、 新規に帯域確保を要求したい送信局からの帯域確保要請フ レームなどを受信す る。 受信部 4 0が帯域要請フ レームを受信すると、 その情報は帯域管理部 4 1 に 与えられる。 帯域管理部 4 1 は周期的に帯域を確保する必要がある局と、 T X O P と呼ばれるメディアアクセス時間を管理している。 帯域管理部 4 1 は、 帯域に 余裕があれば、 この要求を許可し、 内部に必要な情報を記憶する。 MA C部 3 8 は帯域管理部 4 1 から帯域情報が変わったこ とを通知されると、 適当な時間から その局のための帯域確保フ レームを送信するよ う、 送信部 3 9に要求する。 既存の局 4については、 I E E E 8 0 2. 1 1 の P C F/D C Fを備えた従来 の装置であるため、 その構成の説明を省略する。
以下では送信局 2に入力された A Vソースパケッ トが受信局 3 にて出力される までの処理の流れについて順次説明する。 ここでは例と して、 A Vソースバケツ トが M P E G 2— T S又は D V C (スタ ンダー ド) である場合について説明する データブロ ック生成回路 6は、 入力された AVソースパケッ トを、 リー ドソロ モン符号 ( 2 2 4, 2 0 8 ) で极いやすい形式、 すなわちデータが 2 0 8バイ ト 以内になるよ う に構成しなおした A Vデータパックを生成し、 A V伝送用バッフ ァ 7 に格納する。
第 5図は M P E G 2— T Sが送信局 2の AV伝送用バッファ 7 に格納されるま での流れを示した図である。
一つの M P E G 2— T Sバケツ トの大きさは 1 8 8バイ トであり、 通常の画質 で平均 6 M b p s程度、 高画質のもので 2 4 M b p s の速度で断続的に転送され る。 データプロ ック生成回路 6では、 入力された M P E G 2— T Sパケッ トに対 し、 入力時のシステムタイマ 5 に基づく 4バイ 卜の時刻情報をソースパケッ トへ ッダ ( S P H ) と して添付すると と もにデータプロ ックの内容を示す 8バイ トの MA P情報を添付することで、 2 0 0バイ トの AVデータパックを生成する。 こ のよ う にして生成された A Vデータパックは順次 A V伝送用バッファ 7に与えら れ、 AV伝送用バッファ 7 において、 A Vデータパックがデータブロ ック生成回 路 6 よ り出力される順に後述する T a gが付加されて格納される。
即ち、 η番目 (ソースパケッ トカ ウンタによって与えられたカウント値に相当 する) に与えられたソースバケツ トがデータブロ ック生成回路 6に与えられると 、 入力された時刻がシステムタイマ 5で確認される。 このシステムタイマ 5で確 認された時刻に基づく 時刻情報による S Ρ Η及び M A Ρ情報を添付して A Vデ一 タパックを生成すると、 この A Vデータパックを A V伝送用バッファ 7において k番目 となる後述する T a gを付加して格納する。 又、 n + 1番目のソースパケ ッ トによ り生成される AVデータパックが、 k + 1番目 となる後述する T a gが 付加された A V伝送用バッファ 7 に格納される。
第 6図は D V Cバケツ トが送信局 2の A V伝送用バッファ 7に格納されるまで の流れを示した図である。
一つの D V Cのバケツ トは 4 8 0バイ トであり、 3 0 M b p s の転送速度で連 続的に転送される。 バケツ ト先頭はフ レームパルス入力の立ち上がりで判別され る。 データブロ ック生成回路 6は入力された D V Cバケツ トを 5つのブロ ックに 分割し、 2つのブロ ック毎にデータブロ ックの内容を示す 8バイ トの MA P情報 を添付して、 A Vデータパックを生成する。 このよ う にして生成された AVデー タパックは順次 A V伝送用バッファ 7に与えられ、 A V伝送用バッファ 7におい て、 A Vデータパックがデータプロ ック生成回路 6 よ り 出力される順に後述する T a gが付加されて格納される。 又、 D V Cの MA P情報には、 フ レームパルス 立ち上がり時のシステムタイマ 5に基づく 2バイ 卜の時刻情報を格納している。 即ち、 n番目に与えられたソースバケツ トがデータブロ ック生成回路 6に与え られると、 入力された時刻がシステムタイマ 5で確認される。 そして、 ソースパ ケッ トが 9 6ノくィ 卜の 5つのブロ ック F 0 n〜F 4 nに分割される と、 2つのブロ ック F 0 n, F i nに、 システムタイマ 5で確認された時刻に基づく 時刻情報を備 えた MA P情報を添付して AVデータパックを生成する。 この AVデータパック を A V伝送用バッファ 7 において k番目 となる後述する T a gを付加して格納す る。 又、 2つのプロ ック F 2 n, F 3 nに、 同一の時刻情報を備えた M A P情報を 添付して AVデータパックを生成し、 この AVデータパックを AV伝送用バッフ 了 7において k + 1 番目 となる後述する T a gを付加して格納する。
又、 n + 1 番目のソースパケッ トが与えられると、 データブロ ック生成回路 6 で 5つのブロ ック F 0 n + l〜F 4 n+lに分割される。 そして、 2つのブロ ック F 4 n, F 0 n + 1に、 システムタイマ 5で確認された n + 1番目のソースパケッ トの入 力された時刻に基づく 時刻情報を備えた M A P情報を添付して A Vデータパック を生成する。 この A Vデータパックを A V伝送用バッファ 7において k + 2番目 となる後述する T a gを付加して格納する。 その後、 ブロ ック F l n+1, F 2 n+1 による A Vデータパック、 プロ ック F 3 n + l, F 4 n + 1による A Vデータパックが それぞれ、 順に、 後述する T a gが付加されて A V伝送用バッファ 7に格納され る。
尚、 システムタイマ 5に基づく 4バイ ト又は 2バイ トの時刻情報は、 システム タイマ 5によって確認されたソースパケッ トの入力時刻を表すものであっても構 わないし、 この入力時刻に送信局 2での処理時間及び受信局 3への送信時間を付 加した時刻と しても構わない。
第 7図 A及び第 7図 Bは、 M P E G 2 — T S及び D V Cそれぞれに対する MA P情報を格納するパケッ トヘッダフォーマッ トの一例である。 又、 表 1 は、 MA P情報における各フィール ドのビッ ト数を表した表であり、 表 2は、 M P E G 2 — T S及び D V Cそれぞれに対する各フィール ドの情報を示した表である。
表 1 フィール ド ビッ ト数
D B S 8
F N 3
D B I 3
S P C 8
Q P C 8
R S V 1
S P H 1
F M T 6
F D F 8 , 2 4
S Y T 1 6
表 2
Figure imgf000019_0001
第 7図 Aの M P E G 2— T Sに対する MA P情報のフ ィ ール ド構成と、 第 7図 Bの D V Cに対する M A P情報のフ ィ ール ド構成とカゝら明らかなよ うに、 M P E G 2— T Sに対する MA P情報には、 S Y Tフ ィ ール ドが無く 、 F D Fフ ィ 一ル ドと されてレ、る。 その他のフィールドについては、 M P E G 2— T S及び DV C のいずれに対する MA P情報にも含まれる。
以下に、 各フ ィ ール ドの詳細について説明する。 まず、 D B Sは 8 ビッ トのフ ィール ドで、 格納時の最小単位であるデータブロ ックのサイズを q u a d 1 e t (- 4バイ ト) 単位で示している。 即ち、 M P E G 2— T Sの場合は、 時間情報 と ソースパケッ トを合わせた 1 9 2バイ トが最小単位になるため、 4 8 (= I 0 0 1 1 0 0 0 0 I ) が、 D V Cの場合は、 フラグメ ン トされた 9 6バイ トが最小 単位になるため、 2 4 ( = I 0 0 0 1 1 0 0 0 I 2) が入力される。
F Nは 3 ビッ トのフィ一ルドで、 ソースバケツ トをいくつのデ一タブ口 ックに 分割したかを整数で示す。 1 は分割をしていないことを示し、 0は 8個に分割し たことを示す。 M P E G 2— T Sの場合は分割しないため 1 を、 D V Cの場合は 5つに分割するため 5が入力される。
D B I は 3 ビッ トのフィール ドで、 A Vデータパックの先頭に格納されたブロ ックが何番目のブロ ックであるを示す数値が入る。 M P E G 2— T Sの場合、 分 割しないため、 D B I は常に 0であるが、 D V Cの場合はソースパケッ トが分割 されるため、 ソースバケツ 卜が分割されて生成されるブロ ックの順番を示す 0〜 4までの整数が入る。
S P Cは 8 ビッ 卜のフィール ドで、 A Vデータパック内に格納されているソー スバケツ トが何番目のバケツ トであるかを示すソースバケツ トカウンタによる力 ゥンタ値であり、 分割された複数のブロ ックが格納されているときは、 その先頭 のブロ ックのカウンタ値と される。
Q P Cは 8 ビッ 卜のフィ ール ドで、 格納したプロ ックの合計が 2 0 0 トに 満たない場合に施すパデイ ングの量をバイ ト数で示す。 ここでは M P E G 2— T Sは、 パディ ングを行わないため、 常に 0 ( = 1 0 0 0 0 0 0 0 0 I 2 である。 しかし、 D V Cの場合は、 パケッ トが連続的に入力されなかった場合、 9 6バイ トのパデイ ングが付加される可能性があり、 このよ うなときは 9 6 (= | 0 1 1 0 0 0 0 0 I 2 となり 、 このよ うな場合以外では 0 (= | 0 0 0 0 0 0 0 0 I 2 ) となる。
R S Vは 1 ビッ トのフィ ール ドで、 後に利用可能とするための予約フィール ド である。 S P Hは 1 ビッ トのフィール ドで、 A Vデータパックを作成する際にソ ースバケツ トへッダが付加していたか否かを示す。 M P E G 2— T Sの場合は時 刻情報と して 4バイ トのソースバケツ トヘッダを付加するため 1 が、 D V Cの場 合は 0が格納される。
F MTは 6 ビッ トのフィ 一ノレ ドでソース ケッ トのフォーマツ トを示す。 この F M Tは、 上位 1 ビッ ト力; 1 のとき . S Y Tフィール ドを有することを示す。 即 ち、 M P E G 2— T Sの場合は I 0 0 0 0 0 0 | 2力;、 D V Cの場合は時刻情報と して 2バイ トの S Y Tを備えるため 1 1 0 0 0 0 0 I 2が格納される。
F D Fは 8 ビッ ト又は 2 4 ビッ トのフィール ドで、 F M Tで指定されたフォ一 マッ トよ り も詳しい情報が入力される。 この F D Fは、 D V Cの場合は 2バイ ト の S Y Tを備えるため 8 ビッ 卜のフィール ドとなり、 M P E G 2— T Sの場合は S Y Tがないので 2 4 ビッ トのフィ ール ドとなる。
S Y Tは 1 6 ビッ トのフィール ドで、 ソースバケツ 卜が入力した時刻情報であ る。 M P E G 2— T Sの場合は、 ソースパケッ トヘッダと して時刻情報をもって いるため付加せず、 D V Cの場合は先に述べた 2バイ トの時刻情報をこ こに格納 する。 尚、 この S Y Tには、 ソースパケッ トの先頭のブロ ックが AVデータブロ ックに含まれる場合は、 そのブロ ックのソースバケツ トが入力された時刻情報が 格納される。 又、 本実施形態では、 FMT、 F D F、 S Y Tは、 I E E E 1 3 9 4でリ アルタイム A Vデータを転送するための方式である I E C 6 1 8 8 3に従 つてレヽる。
以上の操作によ り、 M P E G 2— T S ソースパケッ ト、 D V C ソースパケッ ト は、 と もに 2 0 0バイ トの AVデータパックに変換され、 A V伝送用バッファ 7 に格納される。
第 8図は、 A V伝送用バッファ 7 に格納された A Vデータパックから M A Cパ ケッ トを生成するまでの処理を示す図である。
A V伝送用バッファ 7は、 データブロ ック生成回路 6で生成された 2 0 0バイ トの A Vデータパックに 8 ビッ トの識別子を含む 4バイ 卜の付加情報である T a gを添付して出力する。 この付加情報である T a g内の 8 ビッ トの識別子は、 上 述したよ うに A V伝送用バッファ 7 に格納された順序を表す値に相当する。 この T a gを付加された 2 0 4バイ 卜の AVデータノ ックがセ レク タ 1 2で選択され て誤り検出符号付加回路 1 3に与えられると、 誤り検出符号付加回路 1 3によつ て 4バイ トの誤り検出符号 ( F C S ) が付加される。
T a gにある 8 ビッ ト識別子は、 本実施形態において、 リ ー ドソロモン符号化 プロ ック単位での再送を行う通信において、 受信局 3が A C Kを返信する際に、 どの符号化ブロ ックを正常に受信したかを示すために使われる。 よって、 T a g にある 8 ビッ ト識別子は、 受信局 3が判別できる形式、 例えば連続的な番号でも よい。
本実施形態では、 管理を容易にすると と もにその管理に用いられる別のバッフ ァを不要とするために、 A V伝送用バッ ファ 7のバッファァ ドレスを 8 ビッ ト識 別子と して使用する。 このことによ り、 回路が簡略化される。 又、 本実施形態で は、 送信する A Vデータパックに T a g情報を付加しているが、 これは一例であ り、 T a g情報を付加せずに上位フォーマッ トヘッダによ り識別を行うなどの方 法も可能である。
4バイ トの T a g と 4バイ トの誤り検出符号が付加された 2 0 8バイ 卜の AV データパックは、 これを 1 単位と して、 暗号化器 1 4によって暗号化される。 こ の暗号化器 1 4において用いられる方法と して、 現在、 I E E E 8 0 2. l i e において提案されている b a s i c WE P と呼ばれる公開鍵法による暗号化方 法が用いられる。
この暗号化方法は、 予め、 送信局 2 と受信局 3の間において秘密鍵を決めてお く 。 そして、 送信局 2において、 この秘密鍵と公開鍵とによって暗号化する。 又 、 公開鍵は、 送信パケッ ト毎に変更され、 パケッ トに付加して受信局 3に通知さ れる。 受信局 3では、 通知された公開鍵と秘密鍵とによって復号化する。 尚、 公 開鍵は MA Cパケッ トヘッダ生成回路 1 5で生成されると、 暗号化器 1 4に与え られる。
暗号化されたデータに对し、 MA Cヘッダと公開鍵が付加され、 更に、 誤り訂 正符号付加回路 1 7によって ] 6バイ 卜のリー ドソロモン符号が付加される。 こ こで、 誤り訂正符号付加回路】 7において、 MA Cヘッダと公開鍵を一つのリ一 ドソロモン符号化プロ ック と し、 データに関しては 2 0 8バイ 卜の A Vデ一タパ ックに相当する暗号化データ単位でリー ドソロモン符号化プロ ック とする。 即ち、 まず、 M A Cパケッ トヘッダ生成回路 1 5で生成された M A Cヘッダと 公開鍵とがへッダ部と してセレク タ 1 6で選択されて誤り訂正符号付加回路 1 7 に与えられ、 1 6バイ トのリー ドソロモン符号が付加される。 その後、 暗号化器 1 4で喑号化された暗号化データ単位毎の A Vデータパックがデータ部と してセ レクタ 1 6で選択され、 誤り訂正符号付加回路 1 7で 1 6バイ トのリー ドソロモ ン符号が付加される。 そして、 誤り訂正符号付加回路 1 7で所定数の AVデータ パックそれぞれに相当する各データ部にリー ドソロモン符号が付加されて出力さ れると、 セ レク タ 1 6によって、 MA Cパケッ トヘッダ生成回路 1 5の出力を誤 り訂正符号付加回路 1 7に与えるよ うに接続を切り換える。
又、 3 2バイ 卜の MA Cヘッダと 4バイ トの公開鍵のデータ量は 3 6バイ ト し かなく 、 2 0 8バイ トに満たない。 このため、 誤り訂正符号付加回路 1 7は、 こ の MA Cヘッダと公開鍵とによるデータブロ ック (ヘッダ部) について、 1 〜 1 7 2バイ 卜に 0又は 1 のパデイ ングが格納されると と もに 1 7 3〜 2 0 8バイ ト に MA Cヘッダと公開鍵とが格納されているデータ と して、 処理を行う。 このへ ッダ部を処理する際、 誤り訂正符号付加回路 1 7が、 1 7 2バイ ト分の 0又は 1 のパディ ングの処理後の内部状態となるよ う に、 初期化すればよい。
そして、 セレクタ 1 8が誤り訂正符号付加回路 1 7の出力を選択して誤り検出 符号付加回路 1 9に与えるため、 リ ー ドソロモン符号化処理されたデータ全体に 対し、 誤り検出符号付加回路 1 9によ り 4バイ トの F C Sが付加される。 即ち、 リー ドソロモン符号化処理された 1 つのヘッダ部と所定数 ( n ) のデータ部によ る誤り訂正符号付データに対して、 F C Sが付加される。 よって、 所定数 ( n + 1 ) のリー ドソロモン符号化ブロ ックによる誤り訂正符号付きデータに対して、 誤り検出符号化が行われる。 これは従来の I E E E 8 0 2. 1 1 バケツ ト との互 換性を保っための処理である。
上述のよ う に、 A V伝送用バッファ 7 において、 A Vデータパックを 2 0 8ノく ィ ト以内に無駄なく収まるよ う に構成しなおすことには、 パディ ングによって発 生する帯域の増加を抑制できるという利点と、 後述する リー ドソロモン符号化ブ ロ ック単位で再送する際の処理が簡略化できるという利点がある。
又、 MA Cバケツ ト生成回路 9での MA Cバケツ 卜の生成処理において、 固定 長のリー ドソロモン符号を使った例を挙げたが、 他の方法と して、 ソースバケツ トの大き さにあわせてリー ドソロモン符号の符号長を可変させる方法もある。 例 えば、 リー ドソロモン符号と して ( 2 5 5、 2 3 9 ) を利用する場合において、 M P E G 2— T S ソースバケツ トを処理するときは、 先頭から 2 0 8バイ ト目ま でを誤り訂正符号付加回路 1 7 を通した後、 2 0 9バイ ト 目から 2 3 9バイ ト 目 まで 0又は 1 をパデイ ングしたものと して計算を行い、 1 6バイ トのリー ドソロ モン符号を得る。 そして、 2 0 9バイ 卜 目力、ら 2 3 9バイ ト 目までの 0又は 1 の パディ ングは除去して、 2 0 8バイ ト目まで A Vデータパックに 1 6バイ トのリ — ドソロモン符号を付加して誤り検出符号付加回路 1 9に送出する方法がある。 固定長のリードソロモン符号の場合は、 A Vデータパックやソースバケツ 卜の 大きさによってリー ドソロモン符号化プロ ック内にパデイ ングが発生する可能性 があるが、 この方法では誤り訂正符号付加回路 1 7においてパデイ ングが除去さ れるので、 メディア上を転送されるパケッ ト内にはパデイ ングが発生せず、 帯域 の使用効率が向上するという利点がある。
この場合、 パディ ングする量が多く なればなるほど、 リー ドソロモン符号化時 のパディ ング箇所の処理の遅延が大きく なり 、 パディ ング分の遅延を補償するた めに、 パディ ング分だけ先行した先読みを行って処理する必要がある。 このため 、 よ り速い動作ク ロ ック と、 先読みでリ ー ドソロモン符号化処理を施したデータ を保存するために多く のバッファが必要となる。
又、 ヘッダ部において、 1 〜 3 6バイ トまでに M A Cヘッダと公開鍵を格納し 、 残りの 3 7 〜 2 0 8バイ トを 0又は 1 のパディ ングとする形で処理をしても構 わないが、 その場合は、 符号化の際にパディ ング部で発生する遅延時間を考慮す る必要がある。 この場合、 可変長のリー ドソロモン符号化処理と同様、 先読みで リー ドソロモン符号化処理を施したデータを保存するためのバッファが必要とな る。
又、 セレクタ 1 2によって、 L L C— P D U用ノくッファ 1 0力 らのデ一タブ口 ックが選択されたとき、 ブロ ック 1 3〜 1 9が上述と同様の動作を行う こ とによ つて、 L L C— P D U用バッファ 1 0に格納されたイ ンタ一ネッ ト用のデータな ど処理されて M A Cバケツ 卜が生成される。
更に、 セ レク タ 1 8が、 セレクタ 1 6からの出力を選択して誤り検出符号付加 回路 1 9に与えるとき、 誤り訂正符号付加回路 1 7における リー ドソロモン符号 化処理が行われずに、 M A Cパケッ トヘッダ生成回路 1 5で生成されたヘッ ド部 及び暗号化器 1 4で A Vデータパックが暗号化されて得られたデータ部が、 誤り 検出符号付加回路 1 9に与えられる。 よって、 リー ドソロモン符号の付加されて いないへッ ド部及び所定数のデータ部よ りなるデータに対して、 F C Sが付加さ れる。
次に、 受信局 3が MA Cバケツ 卜を受信したときの処理について述べる。 まず、 受信局 3では、 受信バケツ トが誤り訂正符号化処理をされているかを判 別する必要がある。 この判別方法において、 MA Cヘッダに設けた誤り訂正符号 化処理の有無を示す識別フィールドよ り判別する方法があるが、 無線メディァ上 で発生したエラーによって、 誤り訂正符号化処理識別フィールドに誤り が発生す る場合もある。 そのため、 単純にこの誤り訂正符号化処理の有無を示す識別フィ —ル ドのみによる判別だけでは、 信頼性が低い。
よって、 誤り訂正回路 2 2によって誤り訂正符号化処理を行うか否かが確認さ れる。 今、 第 9図に、 誤り訂正回路 2 2 と誤り検出回路 2 1 とにおいて同一のデ ータ量のデータに対する処理速度が同じであるものと したときの、 誤り検出回路 2 1 の出力と誤り訂正回路 2 2の出力の時間的な関係を示す。 この第 9図を参照 して、 以下に、 誤り検出回路 2 1及び誤り訂正回路 2 2それぞれでの時間的な処 理関係を説明する。
まず、 誤り検出回路 2 1 に MA Cパケッ トが入力されると、 MACヘッダと公 開鍵と リ ー ドソロモン符号とによるデータブ口 ック長分の処理時間が経過すると 、 誤り訂正回路 2 2に MA Cパケッ トのデータが出力される。 そして、 誤り訂正 回路 2 2は、 入力されたデータを一度バッファ リ ングした後に、 ユーク リ ッ ド演 算を開始する。 その後、 誤り訂正回路 2 2において、 リー ドソロモン符号化プロ ック長分の時間をかけてユーク リ ッ ド演算が行われると、 次に、 チェンサーチが 行われる。
このため、 誤り訂正後の MA Cヘッダの出力が開始されるのは、 MA Cバケツ ト受信回路 2 ()内の誤り検出回路 2 1への MA Cヘッダの入力開始から 2 7 6バ ィ ト分 (MA Cヘッダを含むデータブロ ック 5 2バイ ト + リー ドソロモン符号化 ブロ ック 2 2 4バイ ト) のデータ処理時間後であり、 この後、 更に 5 2バイ ト分 のデータ処理時間を経過した後に、 M A Cヘッダに対する誤り訂正が成功したか 否かが確認されると と もに、 誤り訂正処理を行うか否かが確認される。 この誤り 訂正が正常に行われなかったと き、 誤り検出回路 2 1 への MA Cヘッダの入力開 始から 3 2 8バイ ト分のデータ処理時間が経過したタイ ミ ングで、 誤り訂正エラ —信号が出力される。
これに対し、 従来の I E E E 8 0 2. 1 1 a パケッ トにおいては、 受信後、 S I F S ( 1 6 M S ) の間隔で A C Kを返さなければならない。 しかしながら、 上 述のよ うな誤り訂正処理を待った上で誤り訂正処理を行うか否かを判断した場合 、 M A Cヘッダに対して施されるチェンサーチが終了するまでにかかる処理時間 が S I F S よ り長く なるため、 S I F Sの間隔で A C Kを返すこ とは難しい。 このため、 受信局 3は、 誤り訂正処理された MA Cパケッ ト と通常の誤り訂正 がない M A Cバケツ トをそれぞれ独立かつ並列に処理できるよ う に、 M A Cパケ ッ ト受信回路 2 0内に 2系統の MA Cバケツ ト処理用のブロ ックを備える。
第 1 0図に、 M A Cパケッ ト受信回路 2 ()において、 M A Cパケッ トが誤り訂 正処理をされているか否かを判断した後に、 MA Cバケツ トに対して各種処理を 施すためのフローチヤ一 ト示し、 以下でその動作を説明する。
最初に、 誤り検出回路 2 1 において MA Cバケツ 卜の宛先ァ ドレスをチェック する ( S T E P 1 ) 。 このとき、 宛先ア ドレスが自局宛か、 又は、 宛先指定のな いブロー ドキャス トでない場合は (N o ) 、 M A Cパケッ ト解析回路 2 6後段で の通常の処理は中止し、 誤り訂正回路 2 2後段での誤り訂正の処理のみを行う ( S T E P 2 ) 。 即ち、 宛先ア ドレスが間違っている場合は、 誤り検出回路 2 1 で 誤り検出されるために、 通常の処理では正常に受信できないので、 この MACパ ケッ トに対する A C Kを返信する必要がない。 又、 このとき、 宛先ア ドレスが正 しく他局のァ ドレスを示している場合は、 受信局 3にとつて無関係な通信である その後、 誤り訂正処理部 2 2で誤り訂正をした後、 誤り訂正された MACパケ ッ トの宛先ァ ドレスを MA Cバケツ ト解析回路 2 3で再度チェックする ( S T E P 3 ) 。 このとき、 MACパケッ トの宛先ア ドレスが自局宛又はブロー ドキャス トである場合 (Y e s ) 、 復号器 2 4後段での処理が行われて、 後述する遅延 A C Kが A C K生成回路 3 6 よ り送信される ( S T E P 4 ) 。 又、 逆に、 MA Cパ ケッ トの宛先ア ドレスが自局あて又はブロー ドキャス トでない場合 (N o ) 、 現 在行っている誤り訂正処理を中止する ( S T E P 5 ) 。 S T E P 1 において、 誤り検出回路 2 1 で確認された MA Cバケツ トの宛先ァ ドレスが自局宛又はブロー ドキャス トである場合は (Y e s ) 、 誤り検出回路 2 1 において、 MA Cヘッダ内の誤り訂正処理識別フィールドよ り 、 誤り訂正符号 化されているか否かの判断を行う ( S T E P 6 ) 。 この識別フィール ドで誤り訂 正符号化されているこ とが示されていたら (Y e s ) 、 S T E P 2に移行して、 S T E P 2〜 S T E P 5の誤り訂正処理のみを行う。 このとき、 誤り訂正処理を していないにも関わらず誤り訂正処理識別フィールドが誤り訂正処理を示してい る場合、 処理する M A Cパケッ トにエラーが発生しているこ とを示し、 この MA Cパケッ トは通常の処理では A C Kを返信する必要がない。
又、 S T E P 6において、 M A Cヘッダ内の誤り訂正処理識別フィールドにお いて誤り訂正符号化されていることが示されていない場合は、 誤り訂正回路 2 2 後段での誤り訂正の処理と、 MA Cバケツ ト解析回路 2 6後段での通常の処理と を平行して行う ( S T E P 7 ) 。
S T E P 7 を経た後、 誤り訂正回路 2 2で誤り訂正処理が行われると ( S T E P 8 ) 、 誤り訂正回路 2 2から出力される誤り訂正処理後の MA Cヘッダの誤り 訂正処理識別フィール ドが MA Cバケツ ト解析回路 2 3で解析される ( S T E P 9 ) 。 ここで、 誤り訂正処理識別フィールドが誤り訂正符号化されていることを 示している場合 (Y e s ) 、 誤り訂正エラー信号の状態を確認する ( S T E P 1 0 ) 。 このとき、 M A Cヘッダが誤り訂正回路 2 2において正常に誤り訂正処理 が成されたこ とが確認されると (Y e s ) 、 上述の通常の処理を中止すると と も に、 上述の誤り訂正処理が続けて行われる ( S T E P 1 1 ) 。 この誤り訂正処理 が行われるとき、 遅延 AC Kを A C K生成回路 3 6で生成して送信する。
S T E P 9において、 誤り訂正処理識別フィール ドが誤り訂正符号化されてい ることを示していない場合 (N o ) 、 又は、 S T E P 1 0において、 MA Cへッ ダが誤り訂正回路 2 2における誤り訂正処理が正常に成されていない場合 (N o ) 、 上述の誤り訂正処理を中止する と と もに、 上述の通常の処理のみを続けて行 う ( S T E P 1 4 ) 。 この通常の処理が行われるとき、 AC Kを AC K生成回路 3 6で生成して送信する。
S T E P 7 を経た後、 上述の通常の処理が行われたとき ( S T E P 1 2 ) 、 M A Cバケツ ト検出回路 2 1 において、 M A Cバケツ トヘッダにおけるフォーマツ ト的なエラーが確認される ( S T E P 1 3 ) 。 このとき、 MA Cパケッ トヘッダ におけるエラーが検出されると (Y e s ) 、 上述の通常の処理が中止されると と もに、 上述の誤り訂正処理を続けて行う ( S T E P 1 1 ) 。 又、 MA Cパケッ ト ヘッダにおけるエラーが検出されない限り (N o ) 、 上述の誤り訂正処理を中止 すると と もに、 上述の通常の処理を続けて行う ( S T E P 1 4 ) 。
S T E P 7における通常の処理と誤り訂正処理の平行動作は、 それぞれの判別 までにかかる時間に差があるため、 先に正常と判断された処理動作が優先されて 、 優先された処理動作に対して、 以降のパケッ ト処理を行う。 尚、 上述の両方の 処理において、 同時に正常と判断された場合は、 以下の 2通りの判断方法を用い るこ とができる。 一つの判断方法は、 物理層から報告されるパケッ ト長から判断 する方法、 残り の一つの判断方法は、 確率的に正しいと考えられる処理動作を選 択する方法である。
前者の判断方法では、 受信した M A Cバケツ トのバケツ ト長が誤り訂正符号処 理をされた場合のパケッ ト長 (= 3 2バイ ト (MA Cヘッダ) + 4バイ ト ( I V
: 公開鍵) + 1 6バイ ト ( R S符号) + n X 2 2 4ノくイ ト ( nは、 MACバケツ ト内に含まれる A Vデータパックの数に相当する) + 4バイ ト ( F C S ) ) であ るこ とが誤り検出回路 2 1 で確認されると、 誤り訂正処理を優先する。 しかしな がら、 可変長のリー ドソロモン符号を使用している場合は、 この判断方法を使用 するこ とはできない。
後者の判断方法では、 誤り訂正回路 2 2 と誤り検出回路 2 1 のそれぞれが誤動 作する確率を比較し、 確率的に誤動作をしにく い方を優先する。 尚、 リー ドソロ モン符号で誤り を誤訂正をする確率は 1 / 6 0 0 0 0であり、 誤り検出符号が誤 り を正常に検出できない確率はデータ長によって変化する。
上記判断方法以外に、 誤り検出回路 2 1 の出力を一度バッファ リ ングし、 誤り が検出されないときは、 その MACヘッダ内容に従って通常の処理を行い、 誤り が検出されたときは、 誤り訂正処理を行う方法もある。 この場合、 同時並行的に 処理する必要がないので、 第 1 0図のフローチャー ト と比べて、 その判定アルゴ リ ズムをよ り簡略化できる反面、 バッファによる遅延が発生する。 以下に、 誤り訂正符号化処理について説明する。 第 1 1 図は、 誤り訂正符号化 処理されたバケツ ト と して正常に判断された後、 バケツ ト内の A Vデータパック が A V転送用バッファに格納されるまでの流れを示したものである。
誤り検出回路 2 1 から誤り訂正回路 2 2に M A Cバケツ トが与えられると、 ま ず、 ヘッダ部について誤り訂正処理が行われる。 このと き、 誤り訂正回路 2 2で 正常に訂正されて得られたヘッダ部が M A Cバケツ ト解析回路 2 3に与えられ、 M A Cヘッダ及び公開鍵が取り 出されると、 公開鍵が復号器 2 4に与えられる。 又、 MA Cパケッ ト内に含まれるデータ部についても、 誤り訂正回路 2 2で誤り 訂正が成される際、 訂正エラー信号が出力されたか否かが確認される。 又、 誤り 訂正回路 2 2で誤り訂正が行われたデータ部は、 M A Cバケツ ト解析回路 2 3を 通じて復号器 2 4に与えられる。
復号器 2 4は、 予め設定していた秘密鍵と、 M A Cパケッ ト解析回路 2 3が得 られた公開鍵を用いて、 データ部の復号化を行い、 その結果を誤り検出回路 2 5 によって、 データ部が正しく復号化されているかどうかを F C Sによって判別す る。 ここで、 誤り検出回路 2 5 はリ ー ドソロモン復号が行われて、 正常に訂正で きたか否かの判別も行われる。
誤り検出回路 2 5によって F C Sが除去されて得られた A Vデータパックは、 A V伝送用バッファ 3 2の T a gにある 8 ビッ ト識別子が示すバッファァ ドレス に格納されていく。 同時に誤り訂正回路 2 2から出力される誤り訂正エラー信号 を観測し、 この信号が正常に受信できたことを示す場合には、 A V伝送用パッフ ァ 3 2において、 A Vデータパックを格納したバッファァ ドレスに対応する格納 状況表示識別子に対し 1 を書き込み、 誤り訂正エラ一信号が訂正エラーを示した 場合はその格納状況表示識別子に 0を書き込む。
即ち、 第 1 1 図の場合、 誤り訂正回路 2 2において、 訂正エラ一信号が出力さ れるデータ部 # 1, # 3に対しては、 A V伝送用バッファ 3 2の格納状況表示識 別子が 0 となる。 又、 データ部 # 2 に対しては、 A V伝送用バッファ 3 2の格納 状況表示識別子が 1 となる。 即ち、 復号器 2 4及び誤り検出回路 2 5で復号化さ れたデータ部 # 1 , # 3 よ り得られる T a g内の 8 ビッ ト識別子が k — 1, k + 1 となる A Vデータパックが格納される A V伝送用バッファ 3 2内のバッファァ ドレスに対応する格納状況表示識別子が 0 となる。 又、 同様に、 T a g内の 8 ビ ッ ト識別子が T a g となる A Vデータパックが格納される A V伝送用バッファ 3 2内のバッファア ドレスに対応する格納状況識別子が 1 となる。
A V伝送用バッファ 3 2 での処理をよ り簡単にするために、 誤り訂正回路 2 2 の出力をリー ドソロモン符号化ブロ ック分の長さのバッファでバッファ リ ングし 、 訂正エラーが発生しなかったこ とを確認した後、 M A Cパケッ ト解析回路 2 3 に出力させるよ うにしても構わない。 この場合、 M A Cパケッ ト解析回路 2 3に よる解析が遅延すると と もに、 よ り多く のバッファがいるが、 A V伝送用バッフ ァ 3 2の制御は簡略化でき る。
又、 本実施形態の場合、 A V伝送用バッファ 3 2に、 受信した M A Cパケッ ト よ り得られた A Vデータパックを書き込む際、 誤り訂正が正常に行われて、 対応 するバッファァ ドレスの格納状況表示識別子が既に 1 である場合は、 書き込みを 中止するよ う にしてもよい。 これは、 既に正しく書き込まれていたバッファア ド レスのデータ領域に、 誤り誤訂正や誤り訂正失敗している可能性があるデータが 書き込まれるこ とを防ぐためである。
次に、 A V伝送用バッファ 3 2内の A Vデータ ックがソースパケッ ト と して 再構成され出力されるまでの処理動作を説明する。
ソースバケツ ト出力回路 3 3は、 A V伝送用バッファ 3 2に有効な A Vデータ パックが格納されていると き、 この有効な A Vデータパック順次読み出して、 A Vデータパックの中身を解析する。 即ち、 A V伝送用バッファ 3 2において、 格 納状況表示識別子が 1 となるバッファァ ドレスに格納される A Vデータパックが 読み出されて、 ソースパケッ ト出力回路 3 3で解析される。
ソースパケッ ト出力回路 3 3 において、 M A Pの F M T、 F D Fフィール ドに よって A Vデータパックに格納されている A Vソースパケッ トの種類を確認する 。 又、 M A Pの D B Sによってデ一タブロ ックサイズを確認すると と もに、 この 値と Q P Cで確認されるパデイ ングの値とから A Vデータパックに格納されてい るデータブロ ックの個数が認識される。 そして、 各データブロ ックに対して、 そ れぞれの M A Pにおける S P C及び D B I で指示されたデータブロック番号と F Nで指示された分割数とに応じて、 A Vソースパケッ トを再構築する。 更に、 M A P内の S Y T又はソースバケツ トヘッダで指示された時刻情報に基づいて、 タ ィムシーケンスを再現しながら、 再構築した A Vソースパケッ トを出力する。 第 1 2図は、 AVデータパックが M P E G 2— T Sによる AVソースパケッ ト の AVデータパックである場合のソースバケツ ト出力回路 3 3の処理の流れを示 している。
まず、 ソースパケッ ト出力回路 3 3は、 上述のよ うに MA Pを解析する。 この と き、 M P E G 2— T Sの MA Pが表 2に示した値となるため、 ソースパケッ ト 出力回路 3 3は、 AV伝送用バッファ 3 2 よ り読み出される AVデータパックが M P E G 2— T Sによるものであり 、 AVソースバケツ トがソースバケツ トへッ ダが付加された状態で分割なしに格納されていることが確認される。
ソースパケッ ト出力回路 3 3は、 更に、 ソースパケッ トヘッダと AVソースパ ケッ トとを分割して、 ソースバケツ トヘッダ内の時刻情報に後述するオフセッ ト 時問を加算した時刻情報と 自局のシステムタイマ 3 4が与える現在時刻情報とを 照ら し合わせ、 両者が一致したときに A Vソースバケツ トを出力する。
第 1 3図は、 AVデータパックが D V Cによる AVソースパケッ トの AVデー タパックである場合のソースバケツ ト出力回路 3 3の処理の流れを示している。 まず、 ソースパケッ ト出力回路 3 3は、 上述のよ う に MA Pを解析する。 D V Cの MA Pが表 2に示した値となるため、 ソースパケッ ト出力回路 3 3は、 この A Vデータパックが D V Cによるものであり、 A Vソースバケツ トが 5つに分割 されているこ とが確認される。 又、 この A Vデータパックには、 何番目のソース バケツ トの何番目のデータプロ ックが格納されているかが確認される。
ソースパケッ ト出力回路 3 3は、 AVデータプロ ック F 0〜 F 4を F O, F 1 , F 2 , F 3 , F 4の順の並びになるよ う結合して、 一つの AVソースパケッ ト に再構成する。 さ らにイ ンデックス 0の AVデータプロ ックを格納していた AV データパックの M A Pの S Y Tから時刻情報を確認し、 この時刻情報に後述する オフセッ ト時間を加算した時刻情報を求める。 オフセッ トされた時刻情報と 自局 のシステムタイマ 3 4が与える現在時刻情報とを照ら し合わせ、 両者が一致した と きに AVソースバケツ トを出力する。
以下に、 オフセッ ト時問の計算方法について説明する。 送信局 2及び受信局 3の内部に構成される回路によって発生する遅延ゃメディ ァアクセスのタイ ミ ングによって発生する遅延のため、 A Vソースバケツ トが送 信局 2のデ一タブ口 ック生成回路 6に入力された時刻をそのまま利用して、 受信 局 3のソースパケッ ト出力回路 3 3から出力することはできない。 このため、 送 信局 2、 受信局 3 も しく はその両方において、 A Vソースパケッ トの時刻情報を オフセッ 卜する必要がある。
本実施形態では、 オフセッ ト時間の計算は、 受信局 3のソースパケッ ト出力回 路 3 3が行っている。 これは、 仮に同じパケッ トを受信した局が複数台存在する 場合、 それぞれの局においてパケッ ト出力の時間差ができるが、 送信局 2が受信 局 3の内部回路による遅延を考慮せずに送信を行えるという利点がある。 各受信 局 3の間で遅延差が生じるのは、 リ ー ドソロモン復号処理にかかる時間によるも のであるが、 すべての局が同じ速度で復号処理をすれば、 このよ うな時間差を抑 制することができる。
又、 逆に、 送信局 2のデータブロ ック生成回路 6のみでオフセッ ト時間を計算 するものと しても構わない。 このよ う にするこ とで、 受信局 3が複数台存在する 場合でも、 すべての受信局の A Vソースバケツ ト出力をほぼ同じ時間とするこ と ができ る。
いずれの方法を用いると しても、 ネッ トワーク内の送受信局の内部回路構成に よる遅延量は、 システムが破綻しない範囲内であらかじめ決定しておく。 又、 本 発明の装置を備えたネッ トワークシステムが何を要求しているかによって、 いず れの方法がよいかは異なる。
以上のよ うにして送信局 2に入力された A Vソースバケツ トは、 受信局 3から 出力される。
次に無線通信路上でエラーが発生した場合の再送方法について述べる。
受信局 3において、 M A Cパケッ ト全体が消失したか、 又は、 M A Cヘッダ部 に誤りが発生してこの M A Cバケツ トを破棄した場合、 受信局 3から送信局 2に 対して A C Kは返信されない。 このため、 送信局 2は次の送信機会において自発 的に再送を行う。
次に、 受信局 3において、 M A Cパケッ ト内のデータ部のリー ドソ ロ モン符号 化ブロ ックにおいて誤りが発生した場合は A C Kを使って受信局 3から送信局 2 に対し再送を要求する。 このと き、 誤り訂正回路 2 2における リー ドソロモン符 号の復号には時間がかかるため、 通常の A C Κを返すこ とはできない。 よって、 復号後に複数のパケッ トに対し一括して A C Kを返すことができる遅延 A C Kを 使用する。
第 1 4図に、 現在 I E E E 8 0 2. 1 1 e において提案されている遅延 A C K のフ レームフォーマツ トを示 。
2バイ トの F r a m e C o n t r o l フィール ド及び 2バイ トの D u r a t i o nフィールドによって、 遅延 A C Kであるこ とが宣言される。 又、 6バイ ト の R Aフィールドによって受信側となる送信局 2のァ ドレスが示されると と もに 、 6バイ 卜の T Aフィール ドによって送信側となる受信局 3のア ドレスが示され る。 2バイ トの R e c o r d C o u n t フィール ド内の 8 ビッ トの識別子によ つて、 4バイ ト単位毎の A C K R e c o r d フィール ドの数が設定される。 第 1 4図では、 4バイ トの A C K R e c o r d フィールドを 1 つと して構成して レヽる。 又、 誤り検出符号と して、 4バイ トの F C Sが付加されている。
この A C K R e c o r d フィール ド内のシーケンス番号 (Sequence No. for TC-bitmap #0 bit) は再送を要求するバケツ トを示している。 即ち、 T C— s e q内の 4 ビッ トの T C I Dによって、 再送する映像ファイルの識別子が示される と ともに、 シーケンス番号によって再送する AVソースバケツ トのバケツ トの番 号が示される。 そして、 1 6 ビッ トの T C— b i t m a pには、 シーケンス番号 で示される A Vソースバケツ ト以降の AVソースバケツ トとシーケンス番号で示 されるパケッ ト との時間軸上の位置関係が 1 ビッ ト毎に示される。 T C— b i t m a pにおいて、 再送要求される A Vソースバケツ トの位置を示すビッ トに 1 が 与えられ、 受信が成功した A Vソースバケツ 卜の位置を示すビッ 卜に 0が与えら れる。
尚、 現在提案中の遅延 A C Kはバケツ ト単位での再送しか考慮されていない。 このよ う なパケッ ト単位での再送は、 課題でも述べたが、 パケッ ト内に格納され ている リー ドソロモン符号化ブロ ックのう ち一つでも誤り訂正エラーが発生する とすべてを再送しなければならず、 非常に効率が悪い。 それに対して、 本実施形態では、 パケッ ト単位の再送のみならず、 リー ドソロ モン符号化ブロ ック単位 ( A Vデータパック単位) の再送も可能とするよ うに、 遅延 A C Kを構成する。 本実施形態における A Vデータパック格納方法は、 パケ ソ ト内の一部のリー ドソロモン符号化ブロ ック しか正常に受信できず、 A V伝送 用バッファにその A Vデータパックが格納されなかった場合でも、 この正常に受 信された部分における再生は可能である。 又、 各 A Vデータパックは T a gにあ る 8 ビッ ト識別子によって区別することが可能である。 このため、 パケッ ト単位 で再送をする必要性はなく 、 誤り訂正ができずに得られなかった A Vデータパッ クのみ再送を行えばよい。
第 1 5図は、 第 1 4図に示した AC Kフ レームフォーマッ トのリー ドソロモン 符号化ブロ ック単位再送拡張の一例であり、 パケッ ト単位の再送を行うカ プロ ック単位の再送を行うかを識別するための B i t m a p O b j e c t ビッ トを 2バイ 卜の R e c o r d C o u n t フィールド内に設けてレ、る。 この B i t m a p O b j e c t ビッ トは、 A C K R e c o r d フィ—ノレドがパケッ トのシ —ケンスナンバーを示すの力、、 T a gにある 8 ビッ ト識別子を示すのかを識別す るためのものである。 リー ドソロモン符号化ブロ ック単位再送を行う場合は、 こ の B i t m a p O b j e c t ビッ トを 1 にする。
B i t m a p O b j e c t ビッ ト力、; 1 のと き、 A C K R e c o r d フィー ル ド内の情報は、 A V伝送用バッファ 3 2内のバッファァ ドレスに相当する T a gにある 8 ビッ ト識別子の値を示す。 そして、 1 6 ビッ トの T C— b i t m a p には、 T C一 S e qで示されている T a gにある 8 ビッ ト識別子から以降の受信 状況を示すこ とができる。 T C— b i t m a p において、 再送要求される AVデ ータパックの位置を示すビッ 卜に 1 が与えられ、 受信が成功した A Vデータパッ クの位置を示すビッ 卜に 0が与えられる。
即ち、 T C— s e q内の 4 ビッ トの T C I Dによって、 再送する映像ファイル の識別子が示されると ともに、 シーケンス番号によって再送する AVデ一タパッ クの 8 ビッ ト識別子が示される。 そして、 1 6 ビッ トの T C一 b i t m a pには 、 シーケンス番号で示される A Vデータパック以降の A Vデータパック とシーケ ンス番号で示される A Vデータバック との時間軸上の位置関係が 1 ビッ ト毎に示 される。
第 1 1 図では、 前回の受信においてバッファア ド レス N— 2まではすべて正常 な書き込みが行われものとすると と もに、 受信した MACバケツ ト内の 1番目 と 3番目のリ ー ドソ ロ モン符号化ブロ ックにおいて誤り訂正エラーが発生しており 、 AVデータパックが得られないものとする。 よって、 AV伝送用バッファ 3 2 の N番目のバッファア ドレスには正常にデータが書き込まれたが、 k— 1 、 k + 1番目のバッファァ ド レスはヌルになっている。
第 1 6図は、 第 1 1 図で示す受信結果に対し、 A C K生成回路 3 6が生成した A C Kフ レームを示す。 A C K生成回路 3 6では、 A V伝送用バッファ 3 2内の 各バッ フ ァァ ド レスの格納状態が、 AV伝送用バ ッ フ ァ 3 2から与えられる格納 状況表示識別子によって確認される。 又、 M A Cパケッ ト受信部 2 0で受信され て処理が成された AVデータパックの T a g内の 8 ビッ ト識別子が MA C部 3 5 で認識され、 A C K生成回路 3 6に伝えられる。
そして、 MA C部 3 5 よ り MA Cバケツ ト受信部 2 0で処理された AVデータ パックについての情報が与えられると、 AV伝送用バッファ 3 2から与えられる 格納状況表示識別子によって、 再送要求する A Vデータパックに付加された T a g内の 8 ビッ ト識別子が認識され、 この認識された 8 ビッ ト識別子に基づいて、 遅延 A C Kが生成される。 尚、 このとき、 前回の受信で正常に受信された A Vデ —タパックの T a g内の 8 ビッ 卜識別子は A C K生成回路 3 6内に記憶されてい る。
このとき、 まず、 A C K R e c o r d フィールド内の情報が T a gにある 8 ビッ ト識別子を示すために、 R e c o r d C o u n t フィーノレ ド内の B i t m a p O b j e c t ビッ ト力; 1 と される。 又、 A C K R e c o r d フィール ド の数がこの場合 1 つだけとなるので、 R e c o r d C o u n t フ ィ ール ド内の R e c o r d C o u n t 力 s l と される。
又、 A C K R e c o r d フ ィ ール ドの T C— s e qにおけるシーケンス番号 が、 前回正常に受信した最新のバッファァ ドレスに対応する 8 ビッ ト識別子 k - 2の次の 8 ビッ ト識別子、 即ち、 今回受信された A Vソースパケッ トの先頭のデ ータ部に含まれる AVデータパックが格納されるバッ フ ァァ ド レスに対応する 8 ビッ ト識別子 k _ 1 となる。
このと き、 MA C部 3 5 によって与えられる情報よ り、 MA Cパケッ ト受信部 2 0で受信されて処理された A Vデータパックが確認されると、 各 AVデータパ ックの受信状況を AV伝送用バッファ 3 2よ り与えられる格納状況表示識別子よ り確認される。 今、 第 1 1 図において、 8 ビッ ト識別子 k 一 1 , k + 1 に対応す るバッファア ドレスに格納されるべき AVデータパックが正常に誤り訂正処理が 成されていない。 よって、 8 ビッ ト識別子 k— 1, k + 1 に对応するバッファァ ドレスにおける格納状況表示識別子が 0 と して与えられる。
よって、 この格納状況表示識別子が 0 となるバッファァ ドレスに対する 8 ビッ ト識別子 k一 1 , k + 1 が確認されるため、 8 ビッ ト識別子 k— 1, k + 1 にあ たるビッ ト位置が 1 となるよ う に、 T C一 b i t m a pが生成される。 即ち、 1 6 ビッ トの T C一 b i t m a p力 S | 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 | 2となる。
尚、 T C一 S e qに格納されるバッファア ドレスは、 本実施形態のよ うに前回 正常に受信した最新のバッファア ドレスの次のァ ドレスである必要はなく 、 受信 異常が発生したと思われるバッファァ ドレス と しても構わない。
又、 同じ受信機会において受信したすべてのバケツ トに対する遅延 A C Kを生 成する必要もない。 即ち、 受信局 3が最後のパケッ トまでの A C Kが返せない場 合は、 途中までに受信したパケッ トの遅延 A C Kを生成し、 送信局 2に返すよ う にしてもよレ、。 又、 このと き遅延 A C Kを返せなかったパケッ トに対する遅延 A C Kは、 ライフタイムを超過しない限りいつ返してもよレ、。
但し、 この場合は、 送信局 2において、 受信局 3が何番目のパケッ トまでに対 する遅延 A C Kを返信したのかを判別できるよ うにする必要がある。 これは、 例 えば、 遅延 A C Kが返されるまでの時間を、 送信局 2及び受信局 3の間、 も しく はシステム全体のパラメ一ターと して、 予め設定しておき、 その時間を過ぎても 遅延 A C Kが返されないバケツ トゃバッファァ ドレスについては再送をするとい う方法で対処できる。
このよ うな遅延 A C Kについては、 他の局の送信によ り遅延 A C K自体が送信 不可能になるという こ とを防ぐために、 他の通信よ り優先させて送信する必要が ある。 このため、 本実施形態では、 制御局 1 から送信される後述する C F — E η dを受信した後、 P I F S時間が経過したことを MA C部 3 5が確認すると、 A C K生成回路 3 6で生成した遅延 A C Kを送信する。
よって、 他の局が送信可能となるまでの時間である D I F S時間 +バックオフ 時間よ り も短い P I F S時間経過後に遅延 A C Kが送信されるため、 他の局よ り も先に受信局 3がメディアにアクセスすることが可能となる。 尚、 P I F S時間 よ り も短い S I F S時間経過後に遅延 A C Kの送信を行う よ う にしても構わない し、 制御局 1 によ り遅延 A C K送信のタイ ミ ング制御を行う よ う にしても構わな い o
又、 上述したよ うに MA Cバケツ ト解析回路 2 6後段の各回路が動作を行う通 常処理が成されるとき、 A C Kが A C K生成回路 3 6で生成されて送信される。 この通常処理が成されると きに送信される A C Kは、 誤り検出回路 2 8で MA C パケッ トが正常に受信されたこ とを確認すると、 MA C部 3 5によって A C K生 成回路 3 6に A C Kの送信を指示し、 A C K生成回路 3 6から A C Kが送信され る。 よって、 この A C Kの送信は、 MA Cパケッ トの受信を完了してから S I F S時間が経過するまでに行われる。
送信局 2では、 このよ う な遅延 A C Kを受信部 1 1 で受信すると、 A V伝送用 バッファ Ίにおける送信待機状態識別子及び再送可能識別子と、 遅延 A C Kにお ける A C K R e c o r d フィール ドとによって、 MA C部 8が MA Cバケツ ト 生成回路 9が A V伝送用バッファ 7 よ り読み出す A Vデータパックを設定して、 MA Cバケツ ト生成回路 9 の動作を制御する。
又、 本実施形態では、 A Vソースパケッ トの送受信する際の寿命時間であるラ ィフタイムを考慮した送信を行う。 即ち、 時間が経過しすぎた A Vソースバケツ トを再生すると映像または音声に乱れが生じるため、 予めシステム内遅延などを 考慮したう えで、 ライフタイムを設定しておき、 このライフタイムを超えた A V ソースパケッ トについては送信を禁止する。
A Vソースバケツ 卜の再送の場合も同様、 受信局 3から再送要求されてもライ フタイムを経過した A Vソースバケツ トは再送しても、 受信局 3から出力された A Vソースパケッ トが再生不可能となる。 そのため、 受信局 3が送信局 2に対し て再送要求をするときや、 送信局 2が実際に再送を開始するときには、 AVソー スバケツ トの時刻情報とライフタイムを比較し、 ライフタイムを超えているもの については再送を禁止する。
このよ う に、 ライ フタイムが設定されるときに、 遅延 AC Kを受信した送信局 2における動作を、 以下に説明する。 第 1 7図は、 第 1 6図の遅延 AC Kを受信 した送信局 2の A V伝送用バッファ 7内において、 各バッファア ドレスに対する 送信待機状態識別子を、 遅延 A C Kの A C K R e c o r d フィール ド内の情報 に応じて更新する動作を示す。
上述したよ うに、 A V伝送用バッ ファ 7には、 各バッファア ドレスに対して、 送信待機状態を示す送信待機状態識別子と再送可能である (即ち、 バッファア ド レスに格納されている AVデータパックがライフタイムを超えていない) ことを 示す再送可能識別子とが備えられる。
まず、 受信部 1 1 で遅延 A C Kが受信されると、 遅延 A C Kの B i t m a p O b j e c t ビッ ト力; 1 であるこ とが確認されると、 A C K R e c o r d フィ ールドが T a g内の 8 ビッ ト識別子に関する情報であることを MA C部 8に認識 させる。 そして、 MA C部 8におレヽて、 遅延 A C Kの A C K R e c o r d フィ 一ル ドにおけるシーケンス番号が確認されると、 このシーケンス番号によって示 されれる 8 ビッ ト識別子が認識される。
そして、 シーケンス番号よ り 8 ビッ ト識別子が k一 1 であるこ とが認識される ため、 T a g内の 8 ビッ ト識別子 k — 1 以降の 8 ビッ ト識別子に対するバッファ ァ ドレスに対する送信待機状態識別子が A V伝送用バッファ 7 よ り確認される。 今、 8 ビッ ト識別子 k + 2〜 k + 4のそれぞれに対するバッファァ ドレスに格納 される A Vデータパックが送信待機状態にあるものとすると、 第 1 7図の 2段目 にあるビッ トマップよ うに、 A V伝送用バッファ 7 よ り I 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 I 2となる 1 6 ビッ ト分の送信待機状態識別子が M A C部 8に 与えられる。
又、 MA C部 8において、 遅延 A C Kの A C K R e c o r d フィール ドにお ける第 1 7図の 1段目のよ うな I 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 | 2と なる T C— b i t m a pが確認される。 そして、 MA C部 8において、 確認され た T C— b i t m a p と A V伝送用バッファ 7 よ り得られた 1 6 ビッ ト分の送信 待機状態識別子との O Rを取ることで、 再送を含めた送信対象となる識別子ビッ トマップ I 1 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 | 2が得られる。
更に、 M A C部 8において、 8 ビッ ト識別子 k— 1 以降の 8 ビッ ト識別子に対 するバッファア ドレスに対する再送可能識別子が A V伝送用バッファ 7 よ り確認 される。 8 ビッ ト識別子 k— 1〜 k + 4のそれぞれに対するバッファァ ドレスに 格納される AVデータパックの有する時刻情報がライフタイム以内にあるものと すると、 第 1 7図の 3段目にあるビッ トマップのよ う に、 AV伝送用バッファ 7 よ り I 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 | 2となる 1 6 ビッ ト分の再送可 能識別子が MA C部 8に与えられる。
そして、 上記で得られた送信対象となる識別子ビッ トマップと、 この再送可能 識別子によるビッ トマップとの A N Dを取り、 その結果を使って送信待機状態識 別子のビッ トマップを I 1 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 | 2と して更新 する。 この更新された 1 6 ビッ ト分の送信待機状態識別子のビッ トマップと、 シ 一ケンス番号よ り次回の送信時での送信が確認された A Vデータパックの 8 ビッ ト識別子が MACパケッ ト生成回路 9に与えられる。 即ち、 第 1 7図の場合、 8 ビッ ト識別子 k— 1, k + 1〜 k + 4が与えられる。
そして、 M A Cパケッ ト生成回路 9によって、 次回の送信機会に、 MA C部 8 によって与えられた 8 ビッ ト識別子に対応するバッファア ドレスの A Vデータパ ックが A V伝送バッファ 7 よ り読み出されて、 MA Cパケッ トが生成される。 よ つて、 第 1 7図の場合、 8 ビッ ト識別子 k一 1, k + l〜k + 4に対応するバッ ファァ ドレスの AVデータパックが AV伝送バッファ 7 よ り読み出されて、 MA Cバケツ 卜が生成された後、 この M A Cバケツ トが送信される。
又、 上述の再送可能識別子は、 本来すベての A Vデータパックに付加されてい る時刻情報を調べて、 システムタイマ 5 による現在時刻情報及びライ フタイムで 認識される時刻情報と比較を行った上で更新される。 即ち、 システムタイマ 5に よって確認される現在時刻情報からライ フタイムを減算した時刻情報と、 AVデ ータパックに付加されている時刻情報とを比較し、 A Vデータパックに付加され ている時刻情報の方が早い時刻を示す場合は、 送信を禁止するために再送可能識 別子を 0 とする。
又、 制御局 1 によって指示されるメディアアクセス周期が決まっており、 且つ ライ フタイムがそのメディアアクセス周期の整数倍である場合、 送信局 2の A V 伝送用バッファ 7における再送可能識別子とア ドレスバッファ との関係が第 1 8 図で示すよ うな関係となるよ う にしても構わない。 このとき、 メディアアクセス 周期がシステムタイマ 5によってカウン トされ、 A Vデータパックが A V伝送用 バッファ 7に入力されると き、 システムタイマ 5によってカウン ト されたメディ ァアクセス周期の計数値が第 1 8図のよ うにサイクルフィールドに格納される。 又、 この A V伝送用バッファ 7に入力される A Vデータパックのバッファァ ド レスに対する送信待機状態表示識別子及び再送可能識別子が 1 と される。 更に、 システムタイマ 5によってメディアアクセス周期がカウン トされる度に、 現在の 周期回数となる計数値とライフタイムをメディアアクセス周期で割った数との差 分を計算し、 サイクルフィールドがその計算値よ り小さく なつたとき、 そのバッ ファァ ドレスに対する再送可能識別子を 0にする。
例えば、 第 1 8図のよ う に、 ライフタイムがメディアアクセス周期の 3倍にな つており 、 又、 システムタイマ 5によってカ ウン トされるメディアアクセス周期 の計数値が 2 3であるものとする。 よって、 このとき、 A V伝送用バッファ 7 に おいて、 T a g内の 8 ビッ ト識別子 k + 4に対するバッファア ドレスに A Vデー タパックが格納されるとき、 このバッファァ ドレスに対するサイクルフィールド に計数値 2 3が格納される。
又、 8 ビッ ト識別子 k + 4のバッ ファア ドレスに対する送信待機状態表示識別 子及び再送可能識別子が 1 と される。 そして、 サイクルフィール ドに格納された 値が現在の周期回数となる計数値 2 3 と ライ フ リ ミ ツ トをメディアアクセス周期 で割った数 3 との差分 2 3 — 3 = 2 0 よ り も小さく なるバッファア ドレスに対す る再送可能識別子が 0 と される。
このよ う にすることで、 再送の際に送信局 2において、 A V伝送用バッファ 7 内のすべてのバッファア ドレスに格納された A Vデータパックの送信時間を調べ る必要がなく なるため、 上述した A Vデータパックに付加されている時刻情報を 比較する場合にく らべ回路を大幅に簡略化することができる。 次に、 制御局 1 による帯域確保の方法について、 基本周期内に衝突が発生する 区間を備える場合と備えない場合それぞれに対して説明する。
1 . 基本周期内に衝突が発生する区間を備える場合
まず、 メディアアクセスのための基本周期の決定方法について述べる。 第 1 9 図は、 H C Fを備えた制御局 1 と送信局 2 と受信局 3の基本的なパケッ ト交換を 時間軸で示した一例である。 この第 1 9図の例では、 1 つの基本周期に、 衝突が 発生しない区間 C F P と衝突が発生する区間 C P とが含まれる。
第 1 9図のよ うに、 まず、 制御局 1 から帯域確保フ レームである C F— P o 1 1 が送信部 3 9から送信される。 この C F— P o l 1 が送信されてから、 制御局 1 から C F— E n dが送信されるまでの問は衝突が発生しない区間 C F P と して 確保される。 C F— P o l 1 を受信した送信局 2は、 自局が送信局であることを C F - P o 1 1 よ り確認して、 受信局 3 に対し R T Sを生成して送信する。 又、 R T Sを受信した受信局 3は、 自局が受信局であるこ とを R T S よ り確認 して、 送信局 2に対し、 C T Sを生成して送信する。 C T Sを受信した送信局 2 は受信局 3が受信可能な状態であることを確認し、 受信局 3に対し、 MA Cパケ ッ ト生成回路 9で生成した MA Cバケツ トを含むバケツ 卜 # l 〜 # nの単一転送 又はバース ト転送を行う。
そして、 制御局 1 が、 後述する第 2 2図のアルゴリ ズムを利用 して送信局 2が 無線メディアの帯域を解放したことを確認すると、 C F— E n d を MA C部 3 8 で生成して送信部 3 9 より送信することで、 衝突が発生しない区間 C F Pを終了 させる。 この区間 C F P内での C F— P o l 1 、 R T S、 C T S、 パケッ ト # 1 〜 # n、 C F— E n dそれぞれの間隔は S I F S ( 1 6 μ s ) である。
次に、 衝突が起きる可能性のある区間 C Pが開始する。 第 1 9図の例において は遅延 A C Kをこの区間で返すものと している。 但し、 遅延 A C Kが従来の I E E E 8 0 2. 1 1 方式に従った動作を行う通信装置の妨害で返信されなく なるこ とを避けるため、 P I F S時間後に返信を行っている。 尚、 この遅延 A C Kの返 信方法については、 上述の通り制御局 1 による遅延 A C Kの送信時間を制御する ことによるものでも構わなレ、。
そして、 最初の C F— P o l 1 が送信されてから基本周期分の時間が経過する と、 ネッ トワークシステムを構成する局間で通信が行われていないときは、 区間 C Pを終了させると と もに次の基本周期が開始されるよ うに、 制御局 1 から、 C F— P o l 1 が送信される。 尚、 基本周期内で区間 C Pで送信される L ANパケ ッ トが送信終了する場合、 現時点における基本周期が開始されたときに最初に送 信される C F— P o 1 1 が送信されてから基本周期に相当する時間が経過すると 、 次の基本周期を開始するために C F _ P o 1 1 が送信される。
第 1 9図の中段において、 区間 C F Pにおいて送信局 2から送信されるバケツ トの物理層に I E E E 8 0 2. 1 1 a物理層 ( 5 GH z ) を使用した場合の物理 層パケッ トの構成を示す。 I E E E 8 0 2. 1 1 a物理層は最大通信速度が 5 4 M b p s で、 周波数帯域 5 G H z における O F DM方式を備えた物理層であり、 先に述べた D V Cや M P E G 2 — T S高画質モー ドの転送を行える可能性のある 物理層である。
P r e a m b l e , S i g n a l , S e r v i c e、 T a i l , P AD Sは、 I E E E 8 0 2. 1 1 a物理層によって MA Cバケツ トである P D Uに付加され るものであり 、 P r e a m b 1 e と S i g n a 1 については、 それぞれの転送時 間力 S 1 6 S 、 4 s と決まってレ、る。 1 6 ビッ トの S e r v i c e と 6 ビッ ト の T a i 1 と P AD S と MA Cパケッ トである P D Uとは、 最大通信速度が 5 4 M b p s となる O F DM方式で転送することができる。 最大通信速度 5 4 M b p s となる時の O F D M方式での 1 シンボル長は 2 1 6 ビッ トである。 そして、 P A D Sは、 T a i 1 を含む最後のシンボルのシンボル長が 2 1 6 ビッ トになるよ う に付加される。
第 1 9図の下段において、 上述の I E E E 8 0 2. 1 1 a物理層における P D Uである MA Cパケッ トの構成を示す。 I E E E 8 0 2. 1 1 MAC層では、 3 2バイ トの MA Cヘッダ及び 4バイ トの F C Sを除いたプロ トコルのフ レームボ ディ と呼ばれるフィール ドの最大値が 2 3 1 2バイ 卜に制限されている。 従って 、 2 2 4バイ 卜のリー ドソロモン符号化ブロ ックを 1 0個まで収めることが可能 である。
よって、 本実施形態で挙げた AVソースバケツ トのう ち最も速い転送速度をも つ D V Cの場合について考えると、 上 iポの 1 パケッ トで 9 6 (データブロ ックサ ィズ) X 2 (データブロック個数) X I 0 (リー ドソロモン符号化ブロ ック個数 ) = 1 9 2 0バイ トを転送することができる。
又、 第 1 9図に示すよ う に、 各パケッ ト # l 〜 # nに含まれる複数のリー ドソ ロモン符号化ブロ ックの内の 1 ブロ ックを再送要求された A Vデータパックによ る リー ドソロモン符号化ブロ ックに対する帯域と して確保している。 実際にデー タ送信を行う場合は、 このプロ ックを必ず再送要求された A Vデータパックによ る リー ドソロモン符号化ブロ ック とする必要はないが、 演算処理を容易にするた めにこのよ うな構成と している。 よって、 正規の (再送ではない) D V Cによる A Vソースバケツ トを格納できるデータサイズは 1 7 2 8バイ ト となる。
又、 M A Cパケッ ト全体の大きさは、 3 2 (M A Cヘッダ) + 4 ( I V ) + 1 6 (M A Cヘッダ部の R S符号) + 2 2 4 X 1 0 (リ ー ドソロモン符号化ブロ ッ ク) + 4 ( F C S ) = 2 2 9 6バイ トである。 これに、 1 6 ビッ トの S e r v i c e と 6 ビッ トの T a i 1 とを加え、 2 1 6 ビッ ト ( O F D Mシンボル長) で割 つて小数点以下を切り上げた値 (= 8 6 ) が P A D Sを含む O F D Mシンボル数 となる。
1 つの O F D Mシンボルの転送時間は 4 μ s であるので、 4 X 8 6 ( O F D M シンポノレ数) + 1 6 ( P r e a m b l e ) + 4 ( S i g n a l ) = 3 6 4 /x s が 一つの物理層パケッ トを転送するのに必要な時間である。
又、 遅延 A C Kは、 C F — E n dが送信完了した後、 P I F S ( 2 5 μ s ) 経 過した後に出力される。 リ 一 ドソロモン符号化処理は、 第 9図に示すよ うにリ 一 ドソロモン符号化ブロ ック 2個分の遅延が発生する。 受信局 3の内部回路の処理 速度が物理転送路の速度と同じとすると、 この遅延は 6 8 μ 5 となる。 これに対 し、 第 1 9図における最後のパケッ ト # η送信終了後からの遅延 A C Kを送信す るまでの間隔は、 1 6 S ( S I F S ) + 2 7 M S ( C F - E n d ) + 2 5 μ s ( P I F S ) = 6 8 x s となり 、 リ ー ドソロモン復号処理と同等の時間を確保で きる。 実際に遅延 A C Kを送信する際には物理層バケツ 卜の 1 6 μ s のプリ アン ブル区間があるため、 遅延 A C Κ内のデータ生成に更に時間的な余裕がある。 衝突が起きる可能性がある区問 C Pで局 4がメディアアクセスをするには D I F S時間 ( 3 4 μ s ) +ランダムバックオフ時間による期間だけ待機しなければ ならない。 よって、 C F— E n dが送信されて、 区間 C F Pが終了すると と もに 区間 C Pが開始すると、 D I F S時間よ り も短い P I F S時間経過後に遅延 A C Kが送信されるため、 遅延 A C Kの送信を局 4の送信動作によって妨害されるこ とはない。
尚、 遅延 AC Kを局 4の送信動作によって妨害されないための方法と しては、 上記以外にも、 制御局 1 が送信する帯域確保フ レーム C F— P o 1 1 に続いて遅 延 A C Kを送信するよ うにしても構わない。 この場合、 遅延 A C Kを送信するま でに、 帯域確保フ レーム送信時間及び遅延 A C Kの送信までの待機時間が更に必 要となるが、 よ り確実に遅延 A C Kを送信するこ とが可能になる。
又、 第 2 0図は、 上述した各制御パケッ トの転送時間などを考慮して、 衝突し ない区間 C F Pに含まれるバケツ ト数を表すバケツ トバース ト回数と、 D V C ソ ースバケツ 卜の実転送レー ト及び基本周期の長さ との関係を計算した結果をグラ フで示したものである。
第 2 0図で L ANなしと記述されているグラフは衝突が発生しない区間を最小 にした場合であり、 L AN有り と記述されているグラフは転送速度 2 4 M b p s 、 フ レームボディ長 2 3 1 2バイ トの L ANバケツ 卜 と共存したときの値を示す 。 実転送レー トは、 再送用の帯域はパケッ トのデータ部を構成する 1 0個のリー ドソロモン符号化ブロ ックの う ち一つを再送専用と して割り 当てたと して、 残り のリー ドソロモン符号化ブロ ックで転送できる転送量を示したものである。 第 1 9図の例では、 例えば L ANと共存したときに D V Cの実転送速度が 3 0 M b p s以上となる最小の周期、 約 6. 8 m s (バース ト回数 1 5回、 実転送レ — ト 3 0. 2 M b p s ) を基本周期とする。 このよ う にすることで、 基本周期を もっと も転送速度が速い D V C転送に適合するよ う設計できるため、 D V Cよ り も転送速度が低い M P E G 2— T Sの送信も可能である。
次に、 上述のよ う にして各局間で通信が行われているときにおいて、 制御局 1 がメディァアクセスを許可する送信局を決定する際の動作について、 以下に説明 する。 第 2 1 図は、 制御局 1 が新規に周期的なメディアアクセスを要求する送信 局 2に対し、 メディアアクセスを許可するか否かを判別するためのフローチヤ一 トである。 尚、 本実施形態において、 通信ネッ トワーク に属する各局のメディアアクセス 周期を基本周期の整数倍とする。 又、 基本周期内における周期的にポーリ ングさ れる各局のメディアアクセス時間の合計の上限は、 この基本周期における D V C 転送の場合にかかるメディアアクセス時問と同等とする。 更に、 制御局 1 は、 周 期的なメディアアクセスを許可した各局の基本周期内におけるメディアアクセス 時間及びメディアアクセス周期を帯域管理部 4 1 に記憶し、 記憶したメディアァ クセス時間を合計した時間を D V C転送時にかかるメディアアクセス時間から減 算し、 このよ う にして得られた時間を残り帯域と して記憶する。
制御局 1 は、 制御局 1 以外の局 (以下、 送信局とする) から送信される新規に 周期的なメディァアクセスの許可を得るための帯域確保要請フ レームを受信部 4 0で受信する ( S T E P 5 1 ) 。 このとき、 帯域確保要請フレームよ り、 帯域情 報と して、 送信局が必要とするメディアアクセス周期及びメディアアクセス時間 が確認される。
そして、 この帯域情報が帯域管理部 4 1 に送出されると、 帯域管理部 4 1 にお いて、 この新規に周期的なメディアアクセスを要求する送信局が必要とするメデ ィァァクセス周期が基本周期の整数倍であるか確認される ( S T E P 5 2 ) 。 即 ち、 この送信局に対して送信する帯域確保フ レーム C F— P o 1 1 の送信周期が 、 基本周期の整数倍であるか、 帯域管理部 4 1 で確認される。
このとき、 メディアアクセス周期が基本周期の整数倍であるこ とが確認される と (Y e s ) 、 帯域管理部 4 1 において、 帯域確保要請フレームを送信した送信 局が必要とするメディアアクセス時間と上述した残り帯域とを比較する ( S T E P 5 3 ) 。 そして、 メディアアクセス時間が残り帯域よ り も短いことが確認され ると (Y e s ) 、 帯域管理部 4 1 において、 このメディアアクセス時間と基本周 期とを比較する ( S T E P 5 4 ) 。
この帯域管理部 4 1 で、 帯域情報から確認されたメディァアクセス時間が基本 周期よ り短いことが確認されると (Y e s ) 、 送信局から要求されたメディアァ クセス周期及びメディアアクセス時間が帯域管理部 4 1 に格納されると と もに、 このメディアアクセス時間を残り帯域から減算することで、 残り帯域の時間を更 新する ( S T E P 5 5 ) 。 この S T E P 5 5の動作を終了すると、 再び、 S T E P 5 1 に移行して、 帯域確保要請フ レームの受信確認が行われる。
又、 S T E P 5 2において、 メディアアクセス周期が基本周期の整数倍でない とき (N o ) 、 又は、 S T E P 5 3 において、 メディアアクセス時間が残り帯域 よ り長レヽとき (N o ) 、 又は、 S T E P 5 4において、 メディアアクセス時間が 基本周期よ り長いとき (N o ) 、 S T E P 5 1 で受信された帯域確保要請フ レー ムによる要求を拒否し、 帯域管理部 4 1 における確認されたメディアアクセス周 期及びメディアアクセス時間が破棄される ( S T E P 5 6 ) 。 そして、 この S T E P 5 6の動作を終了すると、 S T E P 5 5の時と同様、 再び、 S T E P 5 1 に 移行して、 帯域確保要請フ レームの受信確認が行われる。
以上のよ うにメディアアクセス時間の設定を行う ことによって、 既存 L ANの ための帯域を保護しつつ、 D V Cまたは D V C以下の転送速度をもつリ アルタイ ム AVデータの帯域を容易に確保することができる。
このよ う に制御局 1 が送信局 2のメディアアクセスの許可を行う とき、 制御局 1 が送信局 2の周期内での送信終了を検出し、 送信局 2が利用していた帯域を解 放するために、 第 2 2図に示すフローチャー トに従って動作する。
まず、 現在の I E E E 8 0 2. 1 1 と同様、 制御局 1 は、 受信部 4 0で送信局 2から送信されるパケッ ト及ぴ受信局から送信される A C K又は遅延 A C Kが受 信されると、 受信された信号を M A C部 3 8で確認してメディアの監視を行う ( S T E P 1 0 1 ) 。 そして、 MAC部 3 8において、 送信局 2から送信されるパ ケッ ト内の MA Cヘッダに含まれる N F (Non Final) ビッ トよ り送信局 2の状態 を監視し、 N F ビッ トが 0 となったか否かを確認することで、 送信終了を確認す る ( S T E P 1 0 2 ) 。 即ち、 第 1 9図の場合、 パケッ ト # n内の MACヘッダ に含まれる N F ビッ トが()であると ともに、 ノ ケッ ト # 1 〜 # n— 1 内の M A C ヘッダに含まれる N F ビッ 卜が 1 である。
このよ う に MA Cヘッダの N F ビッ トによ り送信終了を確認する処理は P C F に適合したものである。 この P C Fの通信では、 送信局 2 と受信局 3 との間で通 信を行う ときには必ず制御局 1 を経由するので、 この N F ビッ トを監視するこ と ができる。 又、 制御局 1 がパケッ トを受信できず、 送信局 2へ A C Kの返信でき なかった場合には、 送信局から即座にパケッ トが再送されていたため、 S T E P 1 0 2の処理による送信局 2からのバケツ トの送信終了の確認が可能である。
S T E P 1 0 2で、 N F ビッ トが 1 である場合 (N o ) 、 送信局 2からのパケ ッ トの送信が終了してから S I F S時間以上の時間が経過するまでに送信局 2又 は受信局 3からの信号を受信部 4 0で受信したか否かが、 MAC部 3 8でシステ ムタイマ 3 7 よ り与えられる時刻情報よ り確認される ( S T E P 1 0 3 ) 。 即ち 、 送信局 2から送信されるパケッ トが最後のパケッ トであり、 このパケッ トが送 信されてから S I F S時間以上経過したか否かが確認される。
このと き、 送信局 2からのバケツ トが送信終了してから S I F S時間が経過す るまでに、 送信局 2から送信される次のバケツ ト又は受信局から送信される A C Kを受信すると (N o ) 、 送信局 2が送信可能と されるメディアアクセス時間で ある T X O P時間が経過したか否かが、 M A C部 3 8によってシステムタイマ 3 9による時刻情報を参照するこ とで確認される ( S T E P 1 0 4 ) 。 そして、 T X O P時間の経過が確認されなかったと きは (N o ) 、 再度 S T E P 1 0 1 以降 の処理動作を行う。
このよ う な S T E P 1 0 3及び S T E P 1 0 4の処理動作が成されるこ とで、 送信局 2及び受信局 3の間で制御局 1 を経由 しない通信を行う と ともに遅延 A C Kの送信が可能となる H C Fにおいても、 送信局 2の帯域解放処理を行う ことが できる。 即ち、 N F ビッ トの確認を行わない場合でも、 送信局 2から送信される 最後となるパケッ トを制御局 1 において確認することができる。
S T E P 1 0 2において N F ビッ トカ S 0であることが確認されたとき (Y e s ) 、 又、 S T E P 1 0 3において送信局 2からのパケッ トの送信終了後 S I F S 時問以上経過したことが確認されたとき (Y e s ) 、 又、 S T E P 1 0 4におい て T X O P時間が経過したことが確認されたとき (Y e s ) 、 現在の衝突が発生 しない区間 C F Pにおいて別の局へポーリ ングを行う ポーリ ングスケジュールが あるかどうかが、 MA C部 3 8'によって、 帯域管理部 4 1 よ り得られる帯域確保 時間情報よ り確認される ( S T E P 1 0 5 ) 。
そして、 MA C部 3 8によって、 帯域確保時間情報よ り ポーリ ングスケジュ一 ルがあることが確認されると (Y e s ) 、 このポ一リ ングスケジュールによって 指定される次の局に対してメディアアクセスを許可するために、 C F— P o l 1 を送信部 3 9から送信した後 ( S T E P 1 0 6 ) 、 S T E P 1 0 1以降の動作を 行う。 又、 逆に、 ポ一リ ングスケジュールがないことが確認されると (N o ) 、 衝突が発生しない区間 C F Pが終了 したことが確認されるため、 衝突が発生しな い区間を終了させるよ うに、 C F— E n dを送信部 3 9から送信する ( S T E P 1 0 7 ) 。
このよ うなフローチヤ一 卜に従って動作するとき、 区間 C F P内で 2つの局に 対してメディアアクセスが許可されている場合、 第 2 3図のよ う な通信が行われ る。 即ち、 区間 C F Pの開始が確認されて、 制御局 1 よ り C F— P o 1 1 が送信 されると、 C F— P o 1 1 によつて送信局と して設定される局から R T Sが送信 されると と もに、 R T Sによって受信局と して設定される局から C T Sが送信さ れる。 その後、 送信局となる局と受信局となる局との間で通信が行われる。 そして、 送信局となる局から最後のパケッ ト # nが送信されると、 このバケツ ト # nの送信終了後 S I F S時間以上の時間が経過したこ とが制御局 1 で確認さ れ、 ポーリ ングスケジュールによ り次に送信局と して設定される局をポーリ ング するための C F— P o 1 1 が制御局 1 よ り送信される。 その後、 同様に、 次に送 信局となる局から R T Sが送信された後、 次に受信局となる局から C T Sが送信 されると、 この送信局となる局及び受信局となる局との間で通信が行われる。 そして、 送信局となる局から最後のパケッ ト # nが送信されると、 このバケツ ト # nの送信終了後 S I F S時間以上の時間が経過したこ とが制御局 1 で確認さ れると と もに、 次のポーリ ングスケジュールが確認されないので、 区間 C F Pを 終了するための C F— E n dが制御局 1 よ り送信される。
又、 第 1 9図の例において、 L A Nと共存するために確保した区間を補償する ために、 メディアアクセス周期で許容されるジッタが設けられている。 即ち、 衝 突が起きる可能性がある区間 C Pにおいて、 各局間で送受信される L ANバケツ 卜が基本周期から外れても送信許容されるジッタが設定される。 又、 このよ う に 区間 C Pにおいて L ANパケッ トが基本周期から外れることで生じた基本周期の ずれを解消して帯域補填を行うために、 次の基本周期を早く終了するためのジッ タが設定される。
このよ うな第 1 9図に示すジッタを用いた帯域補填動作について、 以下に説明 する。 第 2 4図は、 制御局 1 の C F— P o l 1 の送信タイ ミ ングにおいて局 4が 2 4 M b p s の転送速度で L A Nバケツ トを送信中であったときの、 各局の通信 動作を示すためのタイ ミ ングチャー トである。
局 4が、 基本周期終了後、 L ANパケッ トの送信を終了すると、 この局 4から L A Nバケツ ト受信した局よ り 、 L A Nバケツ トの送信終了後 S I F S時間の間 隔で L ANパケッ トに対する A C Kが返信される。 そして、 局 4は、 この A C K を受信するために、 D I F S +ランダムバックオフ時間だけ待機する。 このとき 、 制御局 1 は、 L A Nパケッ ト送信終了を M A C部 3 8で確認すると、 P I F S 時間だけ待機した後に、 送信部 3 9 よ り帯域確保フ レーム C F - P o 1 1 を送信 する。 よって、 基本周期終了から C F— P o 1 1 の送信までの間に、 L ANパケ ッ トによる遅延となるジッタが生じる。
このよ うに生じた遅延となるジッタは、 MA C部 3 8内に格納される。 そして 、 MA C部 3 8において、 この遅延となるジッタを基本周期の長さから減算した 時間を、 次の基本周期の長さと して設定する。 そして、 第 2 2図のフローチヤ一 トに従って動作した後、 C F— E n dを送信して衝突が発生しない区間 C F Pを 終了させる。 その後、 衝突が起きる可能性がある区間 C Pになるが、 制御局 1 は 、 上述のよ うにして MAC部 3 8で設定した基本周期となると、 この区間 C Pを 終了させて、 次の基本周期を開始するために、 C F— P o l 1 を送信部 3 9から 送信する。 このとき開始させる次の基本周期の長さは、 元の周期の長さ と される 即ち、 制御局 1 では、 遅延となるジッタが発生した次の基本周期については、 その長さを、 発生したジッタの長さ分短い長さ とする。 よって、 遅延となるジッ タが発生した次の基本周期では、 C F— P o 1 1 が送信されてから、 基本周期の 長さからジッタを減算した時間が経過したこ とを確認すると、 次の基本周期が開 始されるよ うに、 C F— P o 1 1 を送信する。 このとき、 帯域補填するために早 期終了させた分のジッタが発生する。
このよ うに、 局 4による妨害によって遅延した分だけ早く衝突が起きる可能性 がある区間 C Pを終了させ、 次の基本周期を開始する。 この場合、 局 4による妨 害が再び発生する可能性があるが、 2 4 M b p s の L ANパケッ トであれば、 そ の送信時間がジッタの範囲を超えることはないため、 問題はない。
又、 局 4の L A Nパケッ トの転送速度が、 本例で第 2 0図を利用して基本周期 を計算した際に使用した転送速度以下である場合は、 制御局 1 は、 衝突が起きる 可能性がある区間 C Pにおいて、 遅延 A C Kの送信終了後 P I F S時間だけ待機 する。 そして、 帯域確保フ レーム C F— P o 1 1 を送信して、 衝突が発生しない 区問 C F Pを再開する。 このよ うな動作を、 失った帯域を補填するまで続けるこ とによって、 局 4の妨害を防ぐと と もに、 速やかに帯域を補填することができる このよ う な帯域補填を行う こ とによ り、 D V Cによる AVソースパケッ トの送 信と 2 4 M b p s で送信される既存 L ANとの共存が可能になる。
2. 基本周期内に衝突が発生する区間を備えない場合
上記では、 第 1 9図のよ うな衝突が起きる可能性がある区間 C Pが基本周期内 に設けられるときの例を挙げたが、 この区間 C Pが基本周期内に設けられないも のと しても構わない。 この区間 C Pが基本周期内に設けられないときの、 H C F を備えた制御局 1 と送信局 2 と受信局 3の基本的なバケツ ト交換におけるタイ ミ ングチャー トを、 第 2 5図に示す。
第 2 5図の場合においても、 第 1 9図の場合と同様、 まず、 制御局 1 から帯域 確保フ レームである C F— P o 1 1 が送信部 3 9から送信された後、 C F— P o 1 1 を受信した送信局 2よ り R T Sが送信され、 この R T Sを受信した受信局 3 よ り C T Sが送信される。 このよ う にして、 送信局 2及び受信局 3が決定される と、 送信局 2から受信局 3に対して、 MA Cパケッ トを含むパケッ ト # l 〜 # n が送信される。
そして、 制御局 1 において、 パケッ ト # nの送信が終了したこ とが確認される と、 C F— E n dが送信部 3 9から送信される。 この C F— E n dを受信局 3が 受信すると、 C F— E n dの受信が終了してから S I F S時間が経過後に、 受信 局 3から送信局 2に対して遅延 A C Kが送信される。 尚、 C F— P o l 1 、 R T S、 C T S、 パケッ ト # l 〜 # n、 C F— E n dそれぞれの間隔は、 第 1 9図の 例と同様、 S I F Sである。
そして、 最初の C F— P o l 1 が送信されてから基本周期分の時間が経過する と、 次の基本周期が開始されるよ う に、 制御局 1 から、 C F— P o l 1 が送信さ れる。 尚、 基本周期内で最後に送信される遅延 AC Kが送信終了した後、 送信待 機時間 S I F S以上の時間が経過してから次の基本周期が開始されるよ うに、 S
1 F S時間分待機した後に送信部 3 9から C F— P o 1 1 が送信される。
このよ う に基本周期が設定されるとき、 制御局 1 がメディアアクセスを許可す る送信局を決定する際の動作については、 第 1 9図の例と同様、 第 2 1 図のフロ 一チヤ一卜に従って、 周期的なメディアアクセスを要求する送信局 2に対して、 メディアアクセスの許可を行う。 よって、 その詳細な説明は省略する。
又、 本例において、 制御局 1 が送信局 2の周期内での送信終了を検出し、 送信 局 2が利用していた帯域を解放するために、 第 2 6図に示すフローチャー トに従 つて動作する。 尚、 第 2 6図のフローチャー トにおいて、 第 2 2図に示すフロー チヤ一ト と同一の処理を行うステップについては、 同一の符号を付してその詳細 な説明は省略する。
まず、 制御局 1 は、 受信部 4 0で受信された信号を MA C部 3 8で確認してメ ディアの監視を行い ( S T E P 1 0 1 ) 、 送信局 2から送信されるパケッ ト内の M A Cヘッダにおける N F ビッ トが 0 となつたか否かを確認する ( S T E P 1 0
2 ) 。 このとき、 N F ビッ トが 1 である場合 (N o ) 、 次に、 送信局 2から最後 のパケッ トが送信されてから S I F S時間以上経過したか否かが確認される ( S T E P 1 0 3 ) 。 更に、 送信局 2から S I F S時間以内にパケッ トが送信される と (N o ) 、 T X O P時間が経過したか否かが確認される ( S T E P 1 0 4 ) 。 この S T E P 1 0 2〜 S T E P 1 0 4の処理動作がおこなわれているとき、 N F ビッ トが 0であること、 又は、 最後のパケッ トが送信局 2から送信されたこ と 、 又は、 T X O P時間が経過したこ とのそれぞれが確認されると、 次のポーリ ン グスケジュールがあるか確認される ( S T E P 1 0 5 ) 。
このとき、 帯域管理部 4 1 よ り与えられる帯域確保時間情報から、 次のポーリ ングスケジュールが確認されると (Y e s ) 、 送信部 3 9から C F— E n dを送 信する ( S T E P 1 5 1 ) 。 その後、 受信部 4 0で送信局 3からの遅延 A C Kを 受信したか否かが M A C部 3 8で確認され ( S T E P 1 5 2 ) 、 遅延 A C Kが受 信されなかつたとき (N o ) 、 C F - E n d送信終了後 S I F S時間が経過した か否かが、 システムタイマ 3 7からの時刻情報よ り確認される ( S T E P 1 5 3 ) 。 このとき、 S I F S時間の経過が確認されなかったとき (N o ) 、 再び、 S T E P 1 5 2以降の動作を行う。
又、 S T E P 1 5 2において遅延 AC Kの受信が確認されたとき (Y e s ) 、 又は、 S T E P 1 5 3において C F— E n d送信終了後 S I F S時間の経過が確 認されたとき (Y e s;) 、 S T E P 1 0 5で確認されたポーリ ングスケジュール によって指定される次の局に対してメディアアクセスを許可するために、 C F— P o l 1 を送信部 3 9から送信した後 ( S T E P 1 0 6 ) 、 S T E P 1 0 1 以降 の動作を行う。
又、 S T E P 1 0 5において、 次のポーリ ングスケジュールが確認されなかつ たとき (N o ) 、 送信部 3 9力、ら C F— E n d を送信する ( S T E P 1 5 5 ) 。 その後、 遅延 A C Kを受信したか否かが確認され ( S T E P 1 5 6 ) 、 遅延 A C Kが受信されなかったとき (Y e s:) 、 S I F S時間が経過したか否かが確認さ れる ( S T E P 1 5 7 ) 。 このとき、 S I F S時間の経過が確認されなかったと き (N o ) 、 再び、 S T E P 1 5 6以降の動作を行う。
又、 S T E P 1 5 6において遅延 A C Kの受信が確認されたとき (N o ) 、 又 は、 S T E P 1 5 7 において S I F S時間の経過が確認されたとき (Y e s ) 、 基本周期における最初の C F— P o 1 1 を送信してから基本周期分の時間が経過 したこ とをシステムタイマ 3 7からの時刻情報よ り確認した後、 次の基本周期を 開始するために、 C F— P o 1 1 を送信部 3 9から送信した後 ( S T E P 1 5 8 ) 、 S T E P 1 0 1 以降の動作を行う。
このよ うなフローチヤ一卜に従って動作するとき、 区間 C F P内で 2つの局に 対してメディアアクセスが許可されている場合、 第 2 7図のよ うな通信が行われ る。 即ち、 基本周期の開始が確認されて、 制御局 1 よ り C F— P o l 1 が送信さ れると、 送信局となる局から R T Sが送信されると と もに、 受信局となる局から C T Sが送信される。 その後、 送信局となる局と受信局となる局との間で通信が 行われる。
そして、 送信局となる局から最後のパケッ ト # nが送信されると、 このバケツ ト # nの送信終了後 S I F S時問以上の時間が経過したことが制御局 1 で確認さ れ、 C F— E n dが制御局 1 カゝら送信される。 この C F— E n dの送信終了後 S I F S時間が経過して、 遅延 A C Kが受信局となる局から送信されたこ とが確認 された後、 遅延 A C K送信後 S I F S時間経過すると、 ポーリ ングスケジュール によ り次に送信局と して設定される局をポーリ ングするための C F— P o 1 1 が 制御局 1 より送信される。
その後、 同様に、 次に送信局となる局から R T Sが送信された後、 次に受信局 となる局から C T Sが送信されると、 この送信局となる局及び受信局となる局と の間で通信が行われる。 そして、 送信局となる局から最後のパケッ ト # nが送信 されると、 このバケツ ト # nの送信終了後 S I F S時問以上の時間が経過したこ とが制御局 1 で確認される と と もに、 次のポーリ ングスケジュールが確認されな いので、 まず、 C F— E n dが制御局 1 から送信される。
この C F— E n dの送信終了後 S I F S時問が経過して、 受信局となる局よ り 遅延 A C Kが送信されたこ とが確認されると、 基本周期を開始して C F— P o 1 1 を送信してから、 基本周期となる時間が経過したか確認される。 そして、 基本 周期となる時間が経過したこ とが確認されると、 次の基本周期を開始するために 、 制御局 1 から C F— P o 1 1 が送信される。
尚、 制御局 1 から C F— E n dを送信した後、 S I F S時間以上の時間が経過 したこ とが確認されると、 ポーリ ングスケジュールがある場合は、 すぐに C F— P o l 1 が送信され、 又、 ポ一リ ングスケジュールがない場合は、 基本周期が終 了したことが確認された後に C F _ P o 1 1 が送信される。
又、 本実施形態において、 第 1 8図で示したよ うに、 AVソースパケッ トのラ ィフタイムをメディアアクセス周期の整数倍とするこ とによって、 メディアァク セス周期が基本周期の整数倍であるため、 ライフタイムを基本周期の整数倍とす ることができる。 このよ う に、 A V ソースパケッ トのライフタイムを基本周期で 計測することができるので、 このライ フタイムの管理が容易になる。
よって、 上述したよ うに、 ライフタイムが AVソースパケッ トを送受信するこ とのできる時間を示すため、 ライ フタイムに対するメディアアクセス周期の倍数 が再送可能なメディアァクセス周期の計数値に相当するのと同様、 基本周期の倍 数は再送可能な基本周期の計数値と して考えるこ とができる。 又、 AVソースパ ック毎の再送回数が多ければ多いほどエラ一に対する耐性は向上するが、 ネッ ト ワークシステムにおいて A Vソースバケツ トを再生する際の遅延が大き く なると と もに、 送信局 2及び受信局 3それぞれに備えられる A V伝送用バッファ 7, 3 2のサイズも増大する。
上述のよ う に、 基本周期を 6. 8 m s と した場合、 この間に転送される D V C
(スタンダー ド、 3 0 M b p s ) ソースパケッ トの量は、 2 5. 5 Kバイ トであ り 、 M P E G 2— T S高画質モー ド ( 2 4 M b p s ) ソースパケッ ト量は 2 0. 4 Kバイ トである。 ライフタイムの間に転送される全 AVソースバケツ ト量は、 ライフタイムに対する基本周期の倍数に比例する。 このライフタイムの問に転送 される全 AVソースパケッ ト量は、 即ち、 送信局 2及び受信局 3それぞれの A V 伝送用バッファ 7, 3 2に必要と されるサイズになる。
よって、 本実施形態における送信局 2の機能及び受信局 3の機能を一つの L S
I で実現する場合には、 備える A V伝送用バッファのバッファサイズを適切なも のと して、 このバッファサイズに適合するよ う基本周期を決める必要がある。 又 、 基本周期が小さいほど、 A V伝送用バッファのバッファサイズを節約しながら 再送回数を増やすこ とができ、 同時にライフタイムによるシステム遅延も抑える こ とができる。
又、 本実施形態では高い頻度で時刻合わせを行うために、 基本周期毎に必ず制 御局 1 から送信される C F— P o l 1 や C F _ E n d に対して、 システムタイマ 3 7から与えられる時刻情報が送信部 3 9で付加されて送信される。 このよ うな C F— P o l 1 や C F— E n dのパケッ トフォーマッ トが、 第 2 8図及び第 2 9 図のよ う に表される。
g(lち、 C F— P o l 1 は、 第 2 8図のよ うに、 C F— P o l 1 であるこ とを示 すための 3 2バイ 卜の MA Cヘッダと、 システムタイマ 3 7 よ り与えられた時刻 情報となる 2バイ トのタイムスタンプと、 送信局となる局を指定すると ともにメ ディアアクセス時間などを備えるデータ と、 4バイ トの誤り検出符号である F C S とによって構成される。 又、 C F— E n dは、 第 2 9図のよ うに、 C F— E n dであることを示すための 3 2バイ トの MA Cヘッダと、 システムタイマ 3 7 よ り与えられた時刻情報となる 2バイ トのタイムスタ ンプと、 4バイ トの誤り検出 符号である F C S とによって構成される。
又、 I E E E 8 0 2. 1 1 のタイマ精度は、 上述したよ う に 0. 0 1 %である ため、 6. 8 m s に一回で時刻合わせをする場合、 ジッタを 1 . 3 s程度に抑 えることができる。 本実施形態では、 基本周期毎に送信される第 2 8図のよ うな 構成の C F— P o 1 1 及び第 2 9図のよ うな構成の C F— E n dそれぞれが受信 される度に、 C F— P o l 1 及び C F— E n dそれぞれに備えられたシステムス タンプによって時刻合わせが行われる。 よって、 I E E E 8 0 2. 1 1 のシステ ムタイマ精度が μ s オーダーであるため、 AVソースパケッ トのタイムシーケン ス再生におけるジッタがほとんど無視できる。
但し、 基本周期を長く した場合は、 制御局 1 は、 C F— P o l 1 及び C F— E n d とは別に時刻合わせ専用のフ レームを送信するよ う にしても構わない。 又、 制御局 1 自身が送信局を兼ねる場合は、 C F— P o l 1 を出さずに送信を開始す る場合がある。 この場合についても、 C F— P o l 1 及び C F— E n d とは別に 時刻合わせ専用のフ レームを送信するよ うにしても構わない。 尚、 本実施形態では、 D V Cを転送するための基本周期を例にしたが、 例えば 、 あるシステムでは D V Cや M P E G 2 — T S高画質モー ドの送信をあき らめ、 通常の MP E G 2 — T Sを数チヤネル分流したいという場合もあるだろう し、 ま たあるシステムではできるかぎり L ANを優先したいという場合もある。 衝突が 生じない区間においてどのよ う な A Vデータを扱うカヽ 衝突が起きる可能性があ る区間においてどのよ うな機器と共存するかによって、 基本周期と実転送レー ト が大きく変わる。 そのよ う なシステムの構築例をいくつか取り決めておけば、 ュ 一ザ一が望む最適な無線ネッ トワークシステムを簡単に実現できる。
又、 本実施形態では、 ライ フタイ ムは基本周期の整数倍になるよ うに設定した が、 送信局及び受信局それぞれの A V伝送バッファのバッファサイズに制限があ る場合は、 この AV伝送バッファにおいて、 リ アルタイム AVデータを格納でき る最大時間をライフタイムと しても構わない。 この場合や、 又は、 システムと し てのライフタイムが設定されている場合、 実転送レー 卜が転送したい AVデータ の転送量よ り も大きく 、 且つ、 その周期の長さがライフタイムの整数分の 1 にな るよ う に、 基本周期を設定しても構わない。
更に、 本実施形態では、 各送信局が送信を行う際に与えるメディアアクセス時 間が、 リ アルタイム A Vデータが最速の転送速度で転送されるときのメディアァ クセス時間と される。 又、 このメディアアクセス時間の値の合計がシステムが极 う最も転送速度が速いリ アルタイム A Vデータを転送するために必要なメディア アクセス時間を超えないよ うに、 送信局と して新規に参入する局の制限を行う。 これに対して、 基本周期毎又は基本周期の整数倍毎に、 各送信局が A V伝送用 バッファの空きメモリ情報となるキューイング情報を制御局 i に通知し、 制御局
1 は、 このキュ一イング情報を用いて、 次回以降の基本周期において各送信局が 必要とするメディアアクセス時問を予想し、 この予想に基づいてメディアァクセ ス制御を行う よ う にしても構わない。
又、 M P E G 2— T Sのよ うな可変レ一 トのリ アルタイム A Vデータを极ぅ送 信局は、 常に最大の転送速度で送信することがない。 本実施形態では、 最大転送 速度時に各送信局が必要とするメディアアクセス時間の合計が最も転送速度が速 いリ アルタイム A Vデータを転送するために必要なメディアアクセス時間を超え るこ とはできない。 しかしながら、 上述のキューイング情報からメディアァクセ ス時間を予想する場合、 予想したメディアアクセス時間の合計が最も転送速度が 速いリ アルタイム A Vデータを転送するために必要なメディアアクセス時間を超 えない限り、 制御局 1 は各送信局に対するメディアアクセス制御を行う ことがで さる。
しかし、 転送レー トの可変量はリ アルタイム A Vデータの内容によって変化す るため、 予測が不可能な場合もある。 このよ うな場合には、 制御局 1 が予測した 転送速度と各送信局が実際に必要と した転送速度の差分によつて発生した齟齬を 吸収するため、 A V伝送用バッファ のメモリサイズバッファを大きくすると と も に、 ライ フタイムを長くすればよい。
又、 再送すべき リー ドソ ロモン符号化ブロ ックを識別するために、 本実施形態 では、 バッファア ドレスを識別子と して収めた T a gを使用したが、 受信したパ ケッ トのシーケンス番号及びそのバケツ ト内で誤り訂正できなかったリー ドソロ モン符号化プロ ックの位置を知らせるよ うにしても構わない。 この場合、 受信局 3は、 例えば M A Pにある S P Cと D B I を参照して、 A V伝送用バッファ 3 2 に A Vデータパックを格納する際に S P Cと D B I が連続的になるよ う に格納す る。
又、 このとき、 送信局 2 においては、 送信済みのパケッ トのシーケンス番号及 び A Vデータパックの格納状況を保存するバッファを備える必要がある。 更に、 受信局 3においては、 ソースパケッ ト出力回路 3 3だけではなく 、 A V伝送用バ ッファ 3 2においても、 M A Pを解析するための機能を備える必要がある。 この よ う にすることによ り付加情報となる T a g フィール ドのオーバーへッ ドを減ら すこ と も可能である。
又、 本実施形態では説明を省略したが、 L L C P D Uデータに対しても本実 施形態で説明した処理と同じ処理を行う ことができる。 但し、 L L C P D Uの 場合は、 複数のリー ドソロモン符号化ブロ ックに分割できても、 各リー ドソロモ ン符号化ブロ ック単位ではデータ処理不可能なため、 必ずバケツ トがすべてそろ つた後に上位層に渡されて処理される。 しかし、 リー ドソロモン符号化ブロ ック 単位での再送はパケッ ト単位再送よ り も再送に必要とする帯域が少ないため、 帯 域を有効に使う意味で有効である。 産業上の利用可能性
以上よ り、 ネッ トワークシステムで各局がメディアアクセスできる基本周期を 予め設定するこ とで、 制御局において簡単な処理動作と小規模の回路で各局のメ ディアアクセス制御を実現できる。 又、 誤り訂正符号と誤り訂正符号化ブロ ック 単位の再送および暗号化を行う ことによ り、 送信局及び受信局において、 最小の 帯域で信頼性の高く かつプライバシを保護する通信を実現できる。 又、 帯域制御 フ レームによる時刻合せによって、 送信局と受信局の間で発生する A Vソースパ ケッ トのジッタを最小におさえ、 どの局で A V ソースが再生されても違和感なく 鑑赏することが可能になる。
更に、 ライフタイムを基本周期の整数倍にするこ とで、 送信局及び受信局にお ける A V伝送用バッファを最適なサイズにできるだけでなく 、 A Vソースバケツ トのライフタイムの管理が容易になる。 又、 基本周期の計算において既存局のメ ディアアクセス時間を確保しているため、 どのよ うな状況でも既存局との共存を 実現できる。 又、 制御局の帯域解放処理を行う ことによ り、 既存局が使用する帯 域を多く とれる。
以上説明したよ う に、 本発明によれば、 ネッ トワークシステムで各局がメディ ァアクセスできる基本周期をあらかじめ決めてしま う ことで制御局において簡単 なアルゴリ ズムで各局のメディ ァアクセス制御を実現できる。
誤り訂正符号と誤り訂正符号化ブロ ック単位の再送を使用することで、 再送の 発生頻度を抑え、 かつ再送が発生した場合でも再送に必要と される帯域を抑えつ つ信頼性の高い通信を実現でき る。
プライバシ保護のため、 暗号化を行った場合でも、 上記誤り訂正化ブロ ック単 位の再送を実現できる。
ライフタイムを基本周期の整数倍とするこ とで、 送信局、 受信局における A V 伝送用バッファを最適なサイズにできるだけでなく 、 A Vソースバケツ トのライ フタイムの管理が容易になるとレ、う効果がある。
基本周期の計算において既存局のメディアアクセス時間を確保していることに よ り、 どのよ うな状況でも既存局との共存を実現できる。 さ らに既存局が想定し たメディアアクセス時間以上の通信を行った場合でも、 失った帯域を補填するま で既存局にメディアアクセス権を与えないメディア制御をするこ とによ り、 衝突 が発生しない区間における通信路の品質を落と さない通信を実現できる。
制御局が積極的に帯域解放処理を行う ことによ り、 衝突が発生しない区間にお いて発生した余剰帯域を衝突が起きる可能性のある区間に分配することで、 既存 局のメディアァクセス機会を増やすことが可能になる。
さ らに帯域制御フ レームによる時刻合せによって、 送信局と受信局の間で発生 する A Vソースバケツ トのジッタを最小におさえ、 どの局で A Vソースが再生さ れても違和感なく鑑賞が可能になる。

Claims

請求の範囲
1 . 時刻を管理するシステムタイマと、
アプリ ケーショ ンデータに対して前記システムタイマよ り得られた該アプリ ケ ーショ ンの入力時刻に基づいて生成した時系列的管理用の時刻情報を付加すると と もに、 前記アプリ ケーショ ンデータのうち保持期間が定められたアプリ ケーシ ヨ ンデータに対して送信可能な期間を設定するデータ処理部と、
該データ処理部で前記時刻情報が与えられた前記アプリ ケーショ ンデータを一 時的に保持するバッファ と、
該バッファから前記アプリケーショ ンデータを読み出す.と もに該アプリ ケーシ ヨ ンデータに誤り訂正符号を付加する誤り訂正符号付加部と、
該誤り訂正符号付加部で誤り訂正符号が付加されたアプリ ケーショ ンデータか ら送信パケッ トを作成する送信パケッ ト生成部と、
該送信パケッ ト生成部で作成した前記送信バケツ トを送信する送信部と、 前記バッファからアプリ ケ一ショ ンデータを読み出すことを制御すると と もに 、 前記送信バケツ ト生成部で生成された前記送信バケツ トを通信帯域が確保され た時間に前記送信部から送信するこ とを制御する送信制御部と、
前記送信部から送信した一つ又は複数の送信パケッ トの受信を示す応答信号を 受信する受信部と、
を備え、
前記送信部から送信する際に、 前記送信制御部において、 前記システムタイマ から得られる現在時刻と前記バッファ内に格納されたアプリ ケーショ ンデータに 対して設定された前記送信可能な期間とを比較し、 該現在時刻が前記送信可能な 期間を超過しているアプリ ケーショ ンデータについては、 前記誤り訂正符号付加 部が前記バッファから読み出さず、
又、 前記送信部よ り前記送信パケッ トを送信した後、 所定時間以上経過しても 、 前記送信パケッ トに対する前記応答信号を前記受信部で受信しなかったことが 前記送信制御部において確認されるとき、 該送信バケツ 卜が正しく受信されてい ないと判定することを特徴とする通信装置。
2 . 前記データ処理部において、 前記アプリ ケーシ ョ ンデータを複数のブロ ック に分割して前記バ ッ フ ァ内に格納し、
前記誤り訂正符号付加部において、 前記バ ッ フ ァ内から各プロ ック毎に前記ァ プリ ケーシ ヨ ンデータを読み出すと ともに、 各ブロ ック毎に誤り訂正符号を付与 することを特徴とする請求の範囲 1 に記載の通信装置。
3 . 前記送信バケツ トのバケツ トヘッダ部を生成するヘッダ生成部を備え、 前記誤り訂正符号付加部において、 前記ヘッダ生成部で生成したバケツ トへッ ダ部及び前記バッファよ り読み出した前記アブリ ケーショ ンデータそれぞれに誤 り訂正符号を付加した後、
前記送信バケツ ト生成部において、 前記誤り訂正符号付加部で誤り訂正符号が 付加された前記パケッ トへッダ部及び前記アブリ ケ一ショ ンデータによつて送信 パケッ トを生成することを特徴とする請求の範囲 2に記載の通信装置。
4 . 前記ブロ ック単位での再送要求を示す前記応答信号を前記受信部で受信した とき、 再送要求されたプロ ックの前記アプリ ケ一ショ ンデータのみを再送するこ とを特徵とする請求の範囲 3に記載の通信装置。
5 . 前記バッファに格納される前記アプリ ケーショ ンデータのブロ ック毎の格納 位置情報を、 前記アプリ ケーショ ンに対して、 各ブロ ック毎に付加することを特 徴とする請求の範囲 4に記載の通信装置。
6 . 前記アプリ ケーショ ンデータに対して前記暗号キーを用いて暗号化を施す暗 号化部を備え、
前記ヘッダ生成部において、 暗号キーをバケツ トヘッダ部に付与すると と もに 前記暗号化部で暗号化したアプリ ケ一ショ ンデータ部に対して、 前記誤り訂正 付加部で誤り訂正符号を付加するこ とを特徴とする請求の範囲 5に記載の通信装 置。
7 . 前記アプリ ケーショ ンデータに対して、 ブロ ック毎に誤り検出符号を付加す る誤り検出付加部を備え、
該誤り検出付加部で誤り検出符号が付加されたブロ ックのアプリ ケーショ ンデ —タに対して、 前記暗号化部で暗号化することを特徵とする請求の範囲 6に記載 の通信装置。
8 . 前記送信可能な期間を、 制御局となる通信装置によって通信路における衝突 が発生しない非衝突区間が設定される周期である基本周期の整数倍と し、 前記アプリ ケ一ショ ンデータの保持期間を前記基本周期毎に管理することを特 徵とする請求の範囲 7に記載の通信装置。
9 . 前記制御局となる通信装置によって、 通信が許可される時間となるメディア アクセス時間と、 該メディ アアクセス時間の与えられる周期であるメディアァク セス周期とが、 設定され、
該メディアアクセス周期が、 前記基本周期の整数倍と されることを特徴とする 請求の範囲 8に記載の通信装置。
1 0 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると ともに、 該時刻情報に基づいて前記システムタイマによる時刻を修正することを特微と する請求の範囲 9に記載の通信装置。
1 1 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する こ とを特徴とする請求の範囲 1 0に記載の通信装置。
1 2 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す ることを特徴とする請求の範囲 1 0に記載の通信装置。
1 3 . 前記プロ ック単位での再送要求を示す前記応答信号を前記受信部で受信し たと き、 再送要求されたブロ ックの前記アプリ ケーショ ンデータのみを再送する こ とを特徴とする請求の範囲 2に記載の通信装置。
1 4 . 前記バッファに格納される前記アプリ ケーショ ンデータのブロック毎の格 納位置情報を、 前記アプリ ケーショ ンに対して、 各ブロ ック毎に付加するこ とを 特徵とする請求の範囲 2に記載の通信装置。
1 5 . 前記アプリ ケーショ ンデータに対して前記暗号キーを用いて暗号化を施す 暗号化部を備え、
前記ヘッダ生成部において、 暗号キーをバケツ トヘッダ部に付与すると と もに 前記暗号化部で暗号化したアプリ ケーショ ンデータ部に対して、 前記誤り訂正 付加部で誤り訂正符号を付加するこ とを特徴とする請求の範囲 2に記載の通信装 置。
1 6 . 前記アプリ ケーショ ンデータに対して、 ブロ ック毎に誤り検出符号を付加 する誤り検出付加部を備え、
該誤り検出付加部で誤り検出符号が付加されたブロ ックのアプリ ケーショ ンデ ータに対して、 前記暗号化部で暗号化するこ とを特徴とする請求の範囲 1 5に記 載の通信装置。
1 7 . 前記送信可能な期間を、 制御局となる通信装置によって通信路における衝 突が発生しない非衝突区間が設定される周期である基本周期の整数倍と し、 前記アプリ ケーショ ンデータの保持期間を前記基本周期毎に管理するこ とを特 徴とする請求の範囲 2に記載の通信装置。
1 8 . 制御局となる通信装置によって、 通信が許可される時間となるメディアァ クセス時間と、 該メディアアクセス時間の与えられる周期であるメディアァクセ ス周期とが、 設定され、
該メディアアクセス周期が、 前記制御局となる通信装置によって通信路におけ る衝突が発生しない非衝突区間が設定される周期である基本周期の整数倍と され るこ とを特徴とする請求の範囲 2に記載の通信装置。
1 9 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると と もに、 該時刻情報に基づいて前記システムタイマによる時刻を修正することを特徴と する請求の範囲 2に記載の通信装置。
2 0 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する ことを特徴とする請求の範囲 1 9に記載の通信装置。
2 1 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す るこ とを特徴とする請求の範囲 1 9 に記載の通信装置。
2 2 . 前記送信バケツ トのバケツ トヘッダ部を生成するヘッダ生成部を備え、 前記誤り訂正符号付加部において、 前記ヘッダ生成部で生成したバケツ トへッ ダ部及び前記バッファよ り読み出した前記アプリ ケーショ ンデータそれぞれに誤 り訂正符号を付加した後、
前記送信バケツ ト生成部において、 前記誤り訂正符号付加部で誤り訂正符号が 付加された前記バケツ トヘッダ部及び前記アプリ ケ一ショ ンデータによって送信 パケッ トを生成するこ とを特徵とする請求の範囲 1 に記載の通信装置。
2 3 . 前記アプリ ケーショ ンデータに対して喑号キーを用いて暗号化を施す喑号 化部を備え、
前記ヘッダ生成部において、 前記暗号キーをバケツ トヘッダ部に付与すると と もに、
前記暗号化部で暗号化したアプリ ケ一ショ ンデ一タ部に対して、 前記誤り訂正 付加部で誤り訂正符号を付加するこ とを特徵とする請求の範囲 1 に記載の通信装
2 4 . 前記送信可能な期間を、 制御局となる通信装置によって通信路における衝 突が発生しない非衝突区間が設定される周期である基本周期の整数倍と し、 前記アプリ ケーショ ンデータの保持期問を前記基本周期毎に管理することを特 徴とする請求の範囲 1 に記載の通信装置。
2 5 . 制御局となる通信装置によって、 通信が許可される時間となるメディアァ クセス時間と、 該メディアアクセス時間の与えられる周期であるメディアァクセ ス周期とが、 設定され、
該メディアアクセス周期が、 前記制御局となる通信装置によって通信路におけ る衝突が発生しない非衝突区間が設定される周期である基本周期の整数倍と され るこ とを特徴とする請求の範囲 1 に記載の通信装置。
2 6 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると と もに、 該時刻情報に基づいて前記システムタイマによる時刻を修正することを特徴と する請求の範囲 1 に記載の通信装置。
2 7 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する こ とを特徴とする請求の範囲 2 6に記載の通信装置。
2 8 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す ることを特徴とする請求の範囲 2 6に記載の通信装置。
2 9 . 時刻を管理するシステムタイマと、
時系列的管理用の時刻情報が付加されたアプリ ケーショ ンデータから成る送信 バケツ トを受信する受信部と、
該受信部で受信された送信パケッ トに誤り訂正符号が付加されているか否かを 判定する誤り訂正判定部と、 前記受信部で受信された送信パケッ トのう ち誤り訂正符号が付加された送信パ ケッ トに対して誤り訂正を施す誤り訂正処理部と、
誤り訂正が成された送信パケッ トよ り アプリ ケーショ ンデータを生成する第 1 復号部と、
前記受信部で受信された送信パケッ トのう ち誤り訂正符号が付加されていない 送信バケツ トよ り前記アブリ ケ一シヨ ンデータを生成する第 2復号部と、 前記第 1及び第 2復号部で生成された前記アプリ ケーショ ンデータを一時的に 格納する第〗 及び第 2バッファ と、
前記システムタイマによる時刻が前記アブリ ケーショ ンデータに付加された時 刻情報による時刻と一致したこ とを確認すると、 前記アプリ ケーショ ンデータを 前記第 1 又は第 2バッファから読み出して出力するデータ出力部と、
1 つ又は複数の送信パケッ トの受信状態を示す応答信号を送信する送信部と、 を備え、
前記受信部で受信された送信バケツ 卜に対して、 前記誤り訂正処理部及び前記 第 1復号部による第 1演算処理と、 前記第 2復号部による第 2演算処理とを同時 に施すことを特徴とする通信装置。
3 0 . 前記誤り訂正判定部で、 前記受信部で受信された送信パケッ トが他の通信 装置宛の送信バケツ トであるこ とを確認すると、 前記第 2演算処理を停止して、 前記第 1演算処理のみを行う こ とを特徴とする請求の範囲 2 9に記載の通信装置
3 1 . 前記誤り訂正判定部で、 前記受信部で受信された送信パケッ トに含まれる バケツ トヘッダに誤り訂正符号が付加されているこ とを確認すると、 前記第 2演 算処理を停止して、 前記第 1 演算処理のみを行う ことを特徴とする請求の範囲 3 0に記載の通信装置。
3 2 . 前記第 1及び第 2演算処理を平行して行う ときに、 前記誤り訂正処理部で 誤り訂正が施された送信パケッ トに含まれるパケッ トへッダに誤りが確認された とき、 前記第 1演算処理を停止して、 前記第 2演算処理のみを行う ことを特徴と する請求の範囲 3 1 に記載の通信装置。
3 3 . 前記第 1及び第 2演算処理を平行して行う ときに、 前記誤り訂正判定部で 前記送信パケッ トに誤りが確認されたと き、 前記第 2演算処理を停止して、 前記 第 1 演算処理のみを行う こ とを特徴とする請求の範囲 3 1 に記載の通信装置。
3 4 . 前記通信パケッ トがプロ ック毎に分割されたアプリ ケーショ ンデータに対 して誤り訂正符号化が成され、
正常に復号できなかった前記アプリ ケーショ ンデータにおいて前記ブロ ック単 位で再送の要求を行う前記応答信号を前記送信部よ り送信するこ とを特徴とする 請求の範囲 3 1 に記載の通信装置。
3 5 . 前記アプリ ケ一ショ ンデータの各プロ ックを一意に判別できるプロ ック確 認情報を前記各ブロ ック毎に前記アプリ ケーショ ンデータに付与して前記第 1又 は第 2バッファに格納し、
前記プロ ック確認情報を用いて正常に復号できなかった前記アプリ ケーショ ン のプロ ックを認識して、 前記プロ ック単位の再送要求を行う応答信号を生成する ことを特徴とする請求の範囲 3 4に記載の通信装置。
3 6 . 前記第 1又は第 2バッフ ァにおいて、 前記アプリ ケーショ ンデータの各ブ ロ ックを、 付加された前記ブロ ック確認情報に応じたバッファァ ドレスに格納す るこ とを特徴とする請求の範囲 3 5に記載の通信装置。
3 7 . 制御局となる通信装置によって通信路における衝突が発生しない非衝突区 間が終了すること示す非衝突区間終了信号を前記受信部で受信すると、
非衝突区間終了信号の受信して所定時間経過後に、 前記プロ ック単位の再送要 求を行う応答信号を前記送信部よ り送信することを特微とする請求の範囲 3 6に 記載の通信装置。
3 8 . 前記送信パケッ 卜がアプリケーショ ンデータが前記ブロ ック単位毎に暗号 化されたデータ部と暗号化するための暗号キーを有するパケッ トヘッダ部を備え ると ともに、
前記第 1及び第 2復号部において、 前記バケツ トヘッダよ り暗号キーを抽出し た後、 前記データ部を該暗号キーで前記ブロ ック単位毎に復号することを特徴と する請求の範囲 3 7に記載の通信装置。
3 9 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると と もに、 該時刻情報に基づいて前記システムタイマによる時刻を修正することを特徴と する請求の範囲 3 8に記載の通信装置。
4 0 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する ことを特徴とする請求の範囲 3 9に記載の通信装置。
4 1 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す るこ とを特徴とする請求の範囲 3 9 に記載の通信装掣。
4 2 . 前記送信バケツ トがアプリケーショ ンデータが喑号化されたデータ部と喑 号化するための喑号キ一を有するバケツ トヘッダ部を備えると ともに、
前記第 1及び第 2復号部において、 前記バケツ トヘッダよ り暗号キーを抽出し た後、 前記データ部を該喑号キーで復号することを特徴とする請求の範囲 3 1 に 記載の通信装置。
4 3 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると と もに、 該時刻情報に基づいて前記システムタイマによる時刻を修正するこ とを特徴と する請求の範囲 3 1 に記載の通信装置。
4 4 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する こ とを特徴とする請求の範囲 4 3に記載の通信装置。
4 5 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す ることを特徴とする請求の範囲 4 3 に記載の通信装置。
4 6 . 前記通信バケツ トがブロ ック毎に分割されたアプリ ケーショ ンデータに対 して誤り訂正符号化が成され、
正常に復号できなかった前記アプリ ケーショ ンデータにおいて前記ブロ ック単 位で再送の要求を行う前記応答信号を前記送信部よ り送信するこ とを特徴とする 請求の範囲 2 9に記載の通信装置。
4 7 . 前記アプリ ケ一ショ ンデータの各プロ ックを一意に判別できるプロ ック確 認情報を前記各プロ ック毎に前記アブリ ケーショ ンデータに付与して前記第 1又 は第 2バッファに格納し、
前記ブロ ック確認情報を用いて正常に復号できなかった前記アプリ ケ一ショ ン のブロ ックを認識して、 前記ブロ ック単位の再送要求を行う応答信号を生成する こ とを特徴とする請求の範囲 4 6に記載の通信装置。
4 8 . 前記第 1 又は第 2バッファにおいて、 前記アプリ ケーショ ンデータの各ブ ロ ックを、 付加された前記ブロ ック確認情報に応じたバッファァ ドレスに格納す ることを特徴とする請求の範囲 4 7 に記載の通信装置。
4 9 . 制御局となる通信装置によって通信路における衝突が発生しない非衝突区 間が終了すること示す非衝突区間終了信号を前記受信部で受信すると、
非衝突区間終了信号の受信して所定時間経過後に、 前記プロ ック単位の再送要 求を行う応答信号を前記送信部よ り送信するこ とを特徴とする請求の範囲 2 9に 記載の通信装置。
5 0 . 前記送信バケツ 卜がアプリ ケーショ ンデータが暗号化されたデータ部と喑 号化するための喑号キーを有するバケツ トへッダ部を備えると と もに、 前記第 1及び第 2復号部において、 前記バケツ トヘッダよ り暗号キーを抽出し た後、 前記データ部を該暗号キーで復号するこ とを特徴とする請求の範囲 2 9に 記載の通信装置。
5 1 . 前記受信部において、 時刻情報を備えるパケッ トを受信すると と もに、 該時刻情報に基づいて前記システムタイマによる時刻を修正することを特徴と する請求の範囲 2 9に記載の通信装置。
5 2 . 前記受信部において、 前記時刻情報を備えるパケッ トを定期的に受信する ことを特徴とする請求の範囲 5 1 に記載の通信装置。
5 3 . 前記受信部において、 前記時刻情報を備えるパケッ トを準定期的に受信す ることを特徴とする請求の範囲 5 1 に記載の通信装置。
5 4 . 時刻を管理するシステムタイマと、
他の通信装置から送信される通信帯域の確保を要求する帯域確保要求信号を受 信する受信部と、
通信帯域が確保され衝突が生じない非衝突区間を時分割で管理する帯域管理部 と、
前記システムタイマよ り得られる現在時刻を他の通信装置に認識させるための 時刻情報信号と、 前記非衝突区間の開始を他の通信装置に認識させるための非衝 突区間開始信号と、 前記非衝突区問の終了を他の通信装置に認識させるための非 衝突区間終了信号と、 を送信する送信部と、
を備え、
前記非衝突区間を発生させる周期を基本周期と し、
前記受信部で受信された前記帯域確保要求信号よ り確認される通信許可を求め る期間の周期であるメディアアクセス周期が、 前記基本周期の整数倍でないとき 、 該他の通信装置の帯域確保の要求を拒否することを特徴とする通信装置。
5 5 . 前記基本周期を、 最も転送速度の速いデータを送信局となる通信装置から 受信局となる通信装置へ転送するために必要な時間間隔とするこ とを特徴とする 請求の範囲 5 4に記載の通信装置。
5 6 . 前記受信部で受信された前記帯域確保要求信号よ り確認される通信許可を 求める期間であるメディアアクセス時間が、 前記非衝突区間よ り既に通信許可を 行った通信装置のメディアアクセス時間の合計を減算して得た残り帯域よ り短い と き、 前記帯域確保要求信号を送信した通信装置の帯域確保を行う ことを特徴と する請求の範囲 5 5に記載の通信装置。
5 7 . アプリ ケーショ ンデータのタイムシーケンス再生時のジッタ許容値と、 送 信局となる通信装置と受信局となる通信装置との間で発生するタイマ誤差の値か ら、 前記時刻情報信号の送信間隔を決定するこ とを特徴とする請求の範囲 5 6に 記載の通信装置。
5 8 . 帯域確保が行われて前記帯域管理部で帯域が管理された他の通信装置に対 して、 メディァアクセス時間が開始されたこ とを示す帯域確保信号を送信部よ り 送信すると と もに、 前記非衝突区間開始信号が該帯域確保信号の一つであるこ と を特徴とする請求項 5 7に記載の通信装置。
5 9 . 前記帯域確保信号が、 前記時刻情報信号と しての機能を備えるこ とを特徴 とする請求項 5 8に記載の通信装置。
6 0 . 前記非衝突区間終了信号が、 前記時刻情報信号と しての機能を備えるこ と を特徵とする請求項 5 9に記載の通信装置。
6 1 . 前記非衝突区間開始信号の送信が遅延した場合、 該非衝突区間開始信号に よって開始する基本周期の長さを本来の周期間隔よ り も短くすることを特徴とす る請求の範囲 6 0に記載の通信装置。
6 2 . 前記非衝突区間に送信されたパケッ トを受信した他の通信装置の応答状態 を示す応答信号を該他の通信装置が送信するための帯域確保を、 他のバケツ トょ り優先的に行う こ とを特徴とする請求項 6 1 に記載の通信装置。
6 3 . 前記送信部が、 前記時刻情報信号を準定期的にも送信することを特徴とす る請求項 6 2に記載の通信装置。
6 4 . 前記基本周期に、 通信路における衝突が発生する可能性がある衝突可能性 区間が、 前記非衝突区間の後に設けられ、
前記非衝突区間終了信号を送信部から送信したとき、 前記衝突可能性区間が開 始するこ とを特徴とする請求項 5 4に記載の通信装置。
6 5 . 前記受信部で受信された前記帯域確保要求信号よ り確認される通信許可を 求める期間であるメディアアクセス時間が、 前記非衝突区間よ り既に通信許可を 行った通信装置のメディアアクセス時間の合計を減算して得た残り帯域よ り短い とき、 前記帯域確保要求信号を送信した通信装置の帯域確保を行う ことを特徴と する請求の範囲 6 4に記載の通信装置。
6 6 . アプリ ケーショ ンデータのタイムシーケンス再生時のジッタ許容値と、 送 信局となる通信装置と受信局となる通信装置との問で発生するタイマ誤差の値か ら、 前記時刻情報信号の送信間隔を決定することを特徴とする請求の範囲 6 4に 記載の通信装置。
6 7 . 前記送信部が、 前記時刻情報信号を準定期的にも送信することを特徴とす る請求項 6 6に記載の通信装置。
6 8 . 帯域確保が行われて前記帯域管理部で帯域が管理された他の通信装置に対 して、 メディアアクセス時間が開始されたこ とを示す帯域確保信号を送信部よ り 送信すると と もに、 前記非衝突区間開始信号が該帯域確保信号の一つであるこ と を特徵とする請求項 6 4に記載の通信装置。
6 9 . 前記帯域確保信号が、 前記時刻情報信号と しての機能を備えることを特徴 とする請求項 6 8に記載の通信装置。
7 0 . 前記非衝突区間終了信号が、 前記時刻情報信号と しての機能を備えること を特徴とする請求項 6 4に記載の通信装置。
7 1 . 前記非衝突区間開始信号の送信が遅延した場合、 該非衝突区間開始信号に よって開始する基本周期の長さを本来の周期間隔よ り も短くすることを特徵とす る請求の範囲 6 4に記載の通信装置。
7 2 . 前記非衝突区間に送信されたパケッ トを受信した他の通信装置の応答状態 を示す応答信号を該他の通信装置が送信するための帯域確保を、 他のバケツ トょ り優先的に行う ことを特徵とする請求項 6 4に記載の通信装置。
7 3 . 前記送信部が、 前記時刻情報信号を準定期的にも送信することを特徴とす る請求項 6 4に記載の通信装置。
7 4 . 前記受信部で受信された前記帯域確保要求信号よ り確認される通信許可を 求める期間であるメディアアクセス時間が、 前記非衝突区間よ り既に通信許可を 行った通信装置のメディアアクセス時間の合計を減算して得た残り帯域よ り短い とき、 前記帯域確保要求信号を送信した通信装置の帯域確保を行う こ とを特徴と する請求の範囲 5 4に記載の通信装置。
7 5 . アプリ ケーショ ンデータのタイムシーケンス再生時のジッタ許容値と、 送 信局となる通信装置と受信局となる通信装置との間で発生するタイマ誤差の値か ら、 前記時刻情報信号の送信間隔を決定するこ とを特徴とする請求の範囲 5 4に 記載の通信装置。
7 6 . 前記送信部が、 前記時刻情報信号を準定期的にも送信するこ とを特徴とす る請求項 5 4に記載の通信装置。
7 7 . 帯域確保が行われて前記帯域管理部で帯域が管理された他の通信装置に対 して、 メディアアクセス時間が開始されたことを示す帯域確保信号を送信部よ り 送信すると と もに、 前記非衝突区間開始信号が該帯域確保信号の一つであるこ と を特徴とする請求項 5 4に記載の通信装置。
7 8 . 前記帯域確保信号が、 前記時刻情報信号と しての機能を備えることを特徴 とする請求項 7 8に記載の通信装置。
7 9 . 前記非衝突区間終了信号が、 前記時刻情報信号と しての機能を備えるこ と を特徵とする請求項 5 4に記載の通信装置。
8 0 . 前記非衝突区間開始信号の送信が遅延した場合、 該非衝突区間開始 h しん ごうによって開始する基本周期の長さを本来の周期間隔よ り も短くすることを特 徴とする請求の範囲 5 4に記載の通信装置。
8 1 . 前記非衝突区間に送信されたパケッ トを受信した他の通信装置の応答状態 を示す応答信号を該他の通信装置が送信するための帯域確保を、 他のバケツ トょ り優先的に行う ことを特徴とする請求項 5 4に記載の通信装置。
8 2 . 前記送信部が、 前記時刻情報信号を準定期的にも送信するこ とを特徴とす る請求項 5 4に記載の通信装置。
8 3 . 複数の通信装置によって構成されると ともに、 該複数の通信装置の内の 1 つを制御局とする通信システムにおいて、
前記制御局によって送信局と される通信装置が、
時刻を管理するシステムタイマと、
アプリ ケーショ ンデータに対して前記システムタイマよ り得られた該アプリ ケ —シヨ ンの入力時刻に基づいて生成した時系列的管理用の時刻情報を付加すると と もに、 前記アプリ ケーシ ョ ンデータのうち保持期間が定められたアプリ ケーシ ヨ ンデータに対して送信可能な期間を設定するデータ処理部と、
該データ処理部で前記時刻情報が与えられた前記アプリ ケーショ ンデータを一 時的に保持するバッファ と、 .
該バッファから前記アプリ ケーシ ョ ンデータを読み出すと もに該アプリ ケーシ ヨ ンデータに誤り訂正符号を付加する誤り訂正符号付加部と、
該誤り訂正符号付加部で誤り訂正符号が付加されたアプリ ケーショ ンデータか ら送信パケッ トを作成する送信パケッ ト生成部と、
該送信バケツ ト生成部で作成した前記送信パケッ トを送信する送信部と、 前記バッファからアプリ ケーショ ンデータを読み出すこ とを制御すると と もに 、 前記送信バケツ ト生成部で生成された前記送信バケツ トを通信帯域が確保され た時間に前記送信部から送信するこ とを制御する送信制御部と、
前記送信部から送信した一つ又は複数の送信パケッ トの受信を示す応答信号を 受信する受信部と、
を備え、
前記送信部から送信する際に、 前記送信制御部において、 前記システムタイマ から得られる現在時刻と前記バッファ内に格納されたアプリ ケーショ ンデータに 対して設定された前記送信可能な期間とを比較し、 該現在時刻が前記送信可能な 期間を超過しているアプリ ケーショ ンデータについては、 前記誤り訂正符号付加 部が前記バッファから読み出さず、
又、 前記送信部よ り前記送信パケッ トを送信した後、 所定時間以上経過しても 、 前記送信バケツ トに対する前記応答信号を前記受信部で受信しなかったことが 前記送信制御部において確認されるとき、 該送信バケツ 卜が正しく受信されてい ないと判定する、 ことを特徴とする通信システム。
8 4 . 複数の通信装置によって構成されると と もに、 該複数の通信装置の内の 1 つを制御局とする通信システムにおいて、
前記制御局によって受信局と される通信装置が、
時刻を管理するシステムタイマと、
時系列的管理用の時刻情報が付加されたアブリ ケーショ ンデータから成る送信 バケツ トを受信する受信部と、
該受信部で受信された送信パケッ トに誤り訂正符号が付加されているか否かを 判定する誤り訂正判定部と、
前記受信部で受信された送信パケッ トのう ち誤り訂正符号が付加された送信パ ケッ トに対して誤り訂正を施す誤り訂正処理部と、
誤り訂正が成された送信パケッ トよ り アプリ ケーショ ンデータを生成する第 1 復号部と、
前記受信部で受信された送信バケツ トのう ち誤り訂正符号が付加されていない 送信バケツ トょ り前記アプリ ケーショ ンデータを生成する第 2復号部と、 前記第 1及び第 2復号部で生成された前記アプリ ケーショ ンデータを一時的に 格納する第 1 及び第 2バッファ と、
前記システムタイマによる時刻が前記アプリ ケーショ ンデータに付加された時 刻情報による時刻と一致したことを確認すると、 前記アプリ ケーショ ンデータを 前記第 1 又は第 2バッファから読み出して出力するデータ出力部と、
1つ又は複数の送信パケ ッ トの受信状態を示す応答信号を送信する送信部と、 を備え、
前記受信部で受信された送信バケツ トに対して、 前記誤り訂正処理部及び前記 第 1復号部による第 1 演算処理と、 前記第 2復号部による第 2演算処理とを同時 に施す
ことを特徴とする通信システム。
8 5 . 複数の通信装置によつて構成されると と もに、 該複数の通信装置の内の 1 つを制御局とする通信システムにおいて、
前記制御局となる通信装置が、
時刻を管理するシステムタイマと、
他の通信装置から送信される通信帯域の確保を要求する帯域確保要求信号を受 信する受信部と、
通信帯域が確保され衝突が生じない非衝突区間を時分割で管理する帯域管理部 と、
前記システムタイマよ り得られる現在時刻を他の通信装置に認識させるための 時刻情報信号と、 前記非衝突区間の開始を他の通信装置に認識させるための非衝 突区間開始信号と、 前記非衝突区間の終了を他の通信装置に認識させるための非 衝突区間終了信号と、 を送信する送信部と、
を備え、
前記非衝突区間を発生させる周期を基本周期と し、
前記受信部で受信された前記帯域確保要求信号よ り確認される通信許可を求め る期間の周期であるメディアアクセス周期が、 前記基本周期の整数倍でないとき 、 該他の通信装置の帯域確保の要求を拒否する
こ と を特徴とする通信システム。
PCT/JP2002/004107 2001-04-26 2002-04-24 Appareil de communication et systeme de communication faisant appel a ce dernier WO2002089413A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2002586576A JP3703456B2 (ja) 2001-04-26 2002-04-24 通信装置及びこの通信装置によって構成される通信システム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2001-128657 2001-04-26
JP2001128657 2001-04-26

Publications (1)

Publication Number Publication Date
WO2002089413A1 true WO2002089413A1 (fr) 2002-11-07

Family

ID=18977320

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2002/004107 WO2002089413A1 (fr) 2001-04-26 2002-04-24 Appareil de communication et systeme de communication faisant appel a ce dernier

Country Status (2)

Country Link
JP (1) JP3703456B2 (ja)
WO (1) WO2002089413A1 (ja)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005067213A1 (ja) * 2003-12-26 2005-07-21 Matsushita Electric Industrial Co., Ltd. 無線アクセスシステム
JP2006197032A (ja) * 2005-01-11 2006-07-27 Sony Corp データ伝送システム
JP2006521074A (ja) * 2003-02-28 2006-09-14 トムソン ライセンシング Wlan排他的ダウンリンクチャネルのための方法
JP2007043353A (ja) * 2005-08-02 2007-02-15 Sony Corp 送受信システムおよび方法、送信装置および方法、受信装置および方法、並びにプログラム
JP2008079150A (ja) * 2006-09-22 2008-04-03 Canon Inc 通信機器及びデータ転送方法
JP2008109471A (ja) * 2006-10-26 2008-05-08 Nec Corp Lanシステム、送信装置、受信装置、lanシステム制御方法、プログラム、及び複数のフレーム
JP2009523371A (ja) * 2006-01-10 2009-06-18 コネクサント システムズ、インク 対称送信機会(txop)切断処理
JP2009239451A (ja) * 2008-03-26 2009-10-15 Nec Electronics Corp 到着確認及び中継処理確認型ネットワーク装置及びシステム、フレーム転送方法
US7680148B2 (en) 2004-08-11 2010-03-16 Kabushiki Kaisha Toshiba Communication apparatus and communication method
JP2011172283A (ja) * 2006-01-04 2011-09-01 Interdigital Technology Corp Wlanシステムにおいて複数モードの効率的な動作を提供するための方法およびシステム
US8909286B2 (en) 2006-01-04 2014-12-09 Interdigital Technology Corporation Methods and systems for providing efficient operation of multiple modes in a WLAN system
US11044195B1 (en) 2008-08-21 2021-06-22 United Services Automobile Association (Usaa) Preferential loading in data centers

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07107078A (ja) * 1993-10-05 1995-04-21 Nec Corp 時刻設定方式
JPH09130447A (ja) * 1995-10-31 1997-05-16 Nec Corp 無線データ伝送装置
JPH09191314A (ja) * 1996-01-10 1997-07-22 Mitsubishi Electric Corp 連続データ伝送方法および連続データ伝送装置
JPH11213044A (ja) * 1998-01-26 1999-08-06 Nippon Telegr & Teleph Corp <Ntt> カード情報転送方法及びシステム装置
JP2000253017A (ja) * 1999-03-02 2000-09-14 Nippon Telegr & Teleph Corp <Ntt> 無線パケット制御局
JP2001094537A (ja) * 1999-09-17 2001-04-06 Toshiba Corp エラー処理機能を備えた伝送装置及びエラー処理方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07107078A (ja) * 1993-10-05 1995-04-21 Nec Corp 時刻設定方式
JPH09130447A (ja) * 1995-10-31 1997-05-16 Nec Corp 無線データ伝送装置
JPH09191314A (ja) * 1996-01-10 1997-07-22 Mitsubishi Electric Corp 連続データ伝送方法および連続データ伝送装置
JPH11213044A (ja) * 1998-01-26 1999-08-06 Nippon Telegr & Teleph Corp <Ntt> カード情報転送方法及びシステム装置
JP2000253017A (ja) * 1999-03-02 2000-09-14 Nippon Telegr & Teleph Corp <Ntt> 無線パケット制御局
JP2001094537A (ja) * 1999-09-17 2001-04-06 Toshiba Corp エラー処理機能を備えた伝送装置及びエラー処理方法

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006521074A (ja) * 2003-02-28 2006-09-14 トムソン ライセンシング Wlan排他的ダウンリンクチャネルのための方法
WO2005067213A1 (ja) * 2003-12-26 2005-07-21 Matsushita Electric Industrial Co., Ltd. 無線アクセスシステム
US7440432B2 (en) 2003-12-26 2008-10-21 Matsushita Electric Industrial Co., Ltd. Wireless access system
US7869418B2 (en) 2004-08-11 2011-01-11 Kabushiki Kaisha Toshiba Communication apparatus and communication method
US7903632B2 (en) 2004-08-11 2011-03-08 Kabushiki Kaisha Toshiba Communication apparatus and communication method
US7680148B2 (en) 2004-08-11 2010-03-16 Kabushiki Kaisha Toshiba Communication apparatus and communication method
US7738441B2 (en) 2004-08-11 2010-06-15 Kabushiki Kaisha Toshiba Communication apparatus and communication method
JP2006197032A (ja) * 2005-01-11 2006-07-27 Sony Corp データ伝送システム
JP4736434B2 (ja) * 2005-01-11 2011-07-27 ソニー株式会社 データ伝送システム
JP2007043353A (ja) * 2005-08-02 2007-02-15 Sony Corp 送受信システムおよび方法、送信装置および方法、受信装置および方法、並びにプログラム
US8108674B2 (en) 2005-08-02 2012-01-31 Sony Corporation Transmitting/receiving system and method, transmitting apparatus and method, receiving apparatus and method, and program used therewith
JP4596256B2 (ja) * 2005-08-02 2010-12-08 ソニー株式会社 送受信システムおよび方法、送信装置および方法、受信装置および方法、並びにプログラム
JP2011172283A (ja) * 2006-01-04 2011-09-01 Interdigital Technology Corp Wlanシステムにおいて複数モードの効率的な動作を提供するための方法およびシステム
JP2013165505A (ja) * 2006-01-04 2013-08-22 Interdigital Technology Corp Wlanシステムにおいて複数モードの効率的な動作を提供するための方法およびシステム
US8909286B2 (en) 2006-01-04 2014-12-09 Interdigital Technology Corporation Methods and systems for providing efficient operation of multiple modes in a WLAN system
JP2009523371A (ja) * 2006-01-10 2009-06-18 コネクサント システムズ、インク 対称送信機会(txop)切断処理
JP2008079150A (ja) * 2006-09-22 2008-04-03 Canon Inc 通信機器及びデータ転送方法
US8126012B2 (en) 2006-09-22 2012-02-28 Canon Kabushiki Kaisha Communication apparatus and method of transferring data
JP2008109471A (ja) * 2006-10-26 2008-05-08 Nec Corp Lanシステム、送信装置、受信装置、lanシステム制御方法、プログラム、及び複数のフレーム
JP2009239451A (ja) * 2008-03-26 2009-10-15 Nec Electronics Corp 到着確認及び中継処理確認型ネットワーク装置及びシステム、フレーム転送方法
US11044195B1 (en) 2008-08-21 2021-06-22 United Services Automobile Association (Usaa) Preferential loading in data centers
US11683263B1 (en) 2008-08-21 2023-06-20 United Services Automobile Association (Usaa) Preferential loading in data centers

Also Published As

Publication number Publication date
JPWO2002089413A1 (ja) 2004-11-11
JP3703456B2 (ja) 2005-10-05

Similar Documents

Publication Publication Date Title
US9749091B2 (en) Method and device for data communication in a communication network
KR101400658B1 (ko) Tdma mac 계층에서의 mac 프로토콜 데이터 유닛 집합
JP4440037B2 (ja) 通信装置及び通信方法
US8228889B2 (en) Communication apparatus, communication system and communication control program
EP1913723B1 (en) Method and apparatus to transmit and/or receive data via wireless network and wireless device
AU2004310448B2 (en) Method of operating in a network in which a plurality of stations communicate over a shared medium
US7978734B2 (en) Multichannel MAC data stream for wireless communication
JP4724928B2 (ja) 無線伝送装置及び無線伝送方法
JP4374001B2 (ja) 通信装置、通信方法、および通信システム
US8718089B2 (en) Aggregation and fragmentation of multiplexed downlink packets
JP3730245B2 (ja) 通信管理方法、通信管理プログラム、通信管理プログラムを記録した記録媒体、通信装置、中央管理装置、およびネットワークシステム
JP4421651B2 (ja) 無線lanシステムおよびその送信局
WO2004036840A1 (ja) パケット送受信装置
JP2010051024A (ja) 無線伝送方法、有線伝送方法、無線伝送装置及び有線伝送装置
JP3703456B2 (ja) 通信装置及びこの通信装置によって構成される通信システム
EP1553718B1 (en) Methods and system of error control with feedback resource allocation scheme
JP4176402B2 (ja) 通信管理方法、通信管理プログラム、通信管理プログラムを記録した記録媒体、ならびに通信局
JP2007267082A (ja) データ送受信装置及びデータ送受信方法
CN101238674A (zh) 经由无线网络和无线装置发送和/或接收数据的方法和设备
US8391301B2 (en) Method for micro-controller and DSP based data communication within a transceiver
JP4033860B2 (ja) データ通信方法及びデータ送信装置
JP2009010552A (ja) フレーム制御方法及び通信システム
Ohtani et al. Wireless digital video transmission system using IEEE802. 11b PHY with error correction block based ARQ protocol
JP2002135273A (ja) 無線伝送方法および無線伝送装置
Ohtani et al. Implementation of wireless MPEG2 transmission system using IEEE 802.11 b PHY

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ OM PH PL PT RO RU SD SE SG SI SK SL TJ TM TN TR TT TZ UA UG US UZ VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
WWE Wipo information: entry into national phase

Ref document number: 2002586576

Country of ref document: JP

REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase