KR100240645B1 - Packet error controller of multicast communication and method thereof - Google Patents

Abstract

The present invention relates to a packet error controller for multicast communication and a packet error control method using the same when the geographically dispersed users form a group to perform multimedia communication.

Reliable communication over a multicast channel causes congestion on the reverse network path from the sender because there are multiple receivers for the sender. Congestion of the reverse network path from the sender is caused by control packets, such as retransmission request packets, which the receivers reliably receive to receive data, and cause malfunctions of real-time applications that require constant data arrival rates. In order to perform multimedia communication such as audio and video data, a control method for packet errors during data transmission is required to make use of the continuity of media data to be transmitted.

Accordingly, the present invention consists of a packet error controller as a data transmission module, an error detection and recovery request module, and an error recovery module, thereby proactively coping with changes in the transmission environment, thereby ensuring a continuity for data and providing network load to users. We propose a method that can reduce and provide an optimal transmission quality of service.

Description

Packet Error Controller for Multicast Communication and Packet Error Control Method Using the Same

According to the present invention, when geographically dispersed entities (users) form a group to perform multimedia communication, the reliability of data packets transmitted on a channel when all the entities constituting the group communicate through a multicast channel is provided. The present invention relates to a packet error controller and a packet error control method using the same, which guarantees security, suppresses network overload, and improves transmission service quality in consideration of packet delay time required by a real-time application.

In the related art, each user belonging to a group is connected in a point-to-point manner to exchange messages. The one-to-one communication method required sending the same message repeatedly several times in order to send a message to all users belonging to the group. This has resulted in increased network traffic, reduced system throughput, and complex application programs. In an effort to solve this problem, in recent years, multicast at the network layer is used to transmit data by only one data transmission instead of the number of receiving systems in order to transmit data to a plurality of receivers. By doing so, the network efficiency and computer throughput have been greatly improved. However, since multicast at the network layer provides a network service that does not guarantee reliability, there is a need for a multi-party reliable transmission protocol to ensure the accuracy of the data transmitted to the application program. However, this also presents problems that become more complicated when the group composition entity changes dynamically.

Accordingly, the present invention is to reduce the traffic of the network and improve the processing capacity of the system by minimizing the generation of control packets transmitted to the channel when performing multimedia data communication in order to solve the above problems. In addition, it aims to facilitate communication between group members and improve transmission quality of service by collectively controlling the continuity of all data packet streams at the same time.

The packet error controller of the multicast communication of the present invention for achieving the above object forms a packet having a unique sequence number when a user requests a data transmission from a user and transmits a data packet to a communication network access device, and the network connection After receiving the data packet transmitted through the device to collect the transmission delay information and confirm the packet error, add the sequence number of the packet to the error packet list when the packet error, and activate the recovery request timer, the recovery request timer An error detection and recovery request module for transmitting an error recovery request packet through the network access device upon expiration of the network; and an error recovery request packet transmitted through the network access device, and transmitted from the data transmission module. According to the packet order, the packet is It is characterized in that it comprises an error recovery module for retrieving the transmission delay to update the transmission delay, and if necessary to retransmit the packet to accept the error recovery request.

Packet error control method using the packet error controller of the present invention multicast communication When the receiver detects a packet error, delaying the recovery request time through the measurement transmission delay, and recovering a plurality of packet errors occurred in the delayed time as one Encoding the request packet, transmitting the repair request packet by unicast instead of multicast, and suppressing error recovery by using a measurement transmission delay when the repair request is performed. .

1 is a block diagram illustrating a relationship with hardware associated with driving a packet error controller according to the present invention.

2 is a block diagram showing the configuration of a packet error controller according to the present invention;

3 is a flowchart illustrating an initial data transmission process, an error detection and recovery request, and an error recovery process according to the present invention.

<Description of Signs of Major Parts of Drawings>

10: host processor 20: main memory device

30: packet error controller 40: communication network access device

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

1 is a block diagram illustrating hardware (H / W) resources related to the operation of a packet error controller according to the present invention.

As shown, FIG. 1 shows a host processor 1, a main memory device 2, a packet error controller 3, a network connection device 4, an input / output bus 5 and a system bus 6, respectively.

The host processor 1 is connected to the main memory device 2 through the system bus 6, and is connected to the communication network connection device 4 via the input / output (I / O) bus 5 and embedded in the main memory device. The error controller 3 is executed to improve the transmission quality of service during multilateral data transmission. The packet error controller 3 is embedded in the main memory device 2 and detects a packet error for a packet stream transmitted or received from the network connection device 4, and when a packet error occurs, minimizes the generation of control packets. Ensures continuity of The network connection device 4 performs the reception of data from the external communication network and the data transmission service to the external communication network.

FIG. 2 is a block diagram showing the structure of the packet error controller 3 according to FIG.

The packet error controller 3 is composed of a data transmission module 10, an error detection and recovery request module 20, and an error recovery module 30. Operations are as follows.

The data transmission module 10 is a module that multicasts a data packet so that all members of a group formed of a plurality of entities can receive the same data transmission service, and a packet having a unique sequence number when a data transmission request is received from a user. Form 11 and transmit it to the communication network connection device 4 (12). The data transmission module 10 uses a transmission delay as an important quality of service factor. Transmission delay refers to the time it takes to transmit a packet and arrive at the receiver. It uses measurement transmission delay (Ds) and allowable transmission delay (Dm). Measurement transmission delay is a value that is measured during packet exchange and is a dynamically changing value. The allowable transmission delay is a maximum delay time allowed by the application, and is a value that does not change until the connection is terminated because a different value is determined for each application by negotiation in the connection process. The initial value of the measurement transmission delay for each receiver is zero and is recalculated and updated when an error recovery request arrives from the receivers. The updated measurement transmission delay value is known to the receivers via data packets. Packets sent to the recipients are created and stored in a packet list 13 in order to ensure reliability to the recipients. The packet list consists of the packet sequence number, transmission delay value, and data, and is referred to when a recovery request is made in the error recovery module.

When the data packet is transmitted by the data transmission module 10, the error detection and recovery request module of the receiver 20 receives the data packet 21 from the network connection device 4, collects transmission delay information, and sequence number. Check the packet error via (22). If there is a packet error, the sequence number of the packet is added to the error packet list and the recovery request timer is activated (23). When the recovery request timer expires, the error packet list at the moment is notified to the sender, and the error recovery request packet is transmitted through the network connection device 4 (24).

The error recovery module 30 operates when an error recovery request arrives from the network connection device 4 when the error detection and recovery request module 20 of the receiver issues a recovery request (31). The packet is retrieved from the packet list 13 stored in the data transmission module 10 to update the measurement transmission delay. After confirming the necessity of error recovery, the packet is retransmitted to accept the error recovery request if necessary.

3 is a flowchart illustrating a packet error control process of a packet error control method for reliable multicast communication according to the present invention.

Upon receiving a data transmission request from the user (301), a unique sequence number is generated (302), and the transmission delay value is inquired (303) in the transmission delay list 307 recorded for each receiver. The transmission time to be recorded in the packet is calculated (304), encoded into a packet to be transmitted, and stored in the packet list (305). The stored packet is multicasted to the receivers (306). At this time, a plurality of receivers exist, but they are transmitted to the network only once by multicast without repeatedly transmitting as many times as the number of receivers. The transmission delay value refers to the time taken to transmit a packet to reach a receiver. The transmission delay value is 0 when the first packet is transmitted and for a receiver that has not encountered an error.

Recipients, respectively, receive the data packet (309) and the operation of the packet error detection and repair request module is initiated. Upon receiving the packet, the transmission delay carried in the data packet is collected 310 to update the transmission delay list (311). The updated transmission delay list is used when an error occurs and an error recovery request is raised. Next, it determines whether there is an error by checking the continuity of the received packet (312). If the sequence number of the packet is not the next number that arrived before, the packet is considered lost. If a packet error does not exist (313) by receiving consecutive packets in the sequence number of the previous packet, the packet is deleted from the error packet list 315, and the next packet is prepared for reception. If the received packet has never been lost, that is, if it has never made a repair request, it will not be recorded in the error packet list. If a packet error exists (313), the sequence number of the lost packet is added to the packet error list (316). Next, it is checked whether the error recovery request timer is operating (317). If the error recovery request timer is running, it is ready to receive the next packet; if it is not running, it determines the error recovery request time (318). The recovery request time is determined by calculating the following Equation 1 with reference to the recorded transmission delay list 311.

T = min (Ds, Dm),

Timeout = Random (0, T)

(Where Ds is the measurement transmission delay and Dm is the maximum allowable transmission delay)

The above-described recovery request time determining method means determining a minimum value T among measurement transmission delay and maximum allowable transmission delay value and determining any value between 0 and T as the recovery request time. Here, in order to control not to issue a recovery request at the same time as another receiver for the same packet error, it is important to determine a random value between 0 and T. The reason is that if an error is detected at the same time with the same measurement transmission delay and the same maximum transmission delay, the receiver issuing a recovery request for the same packet in a group communication environment in which there are multiple receivers for one receiver. This is because the number may increase. If this happens, the network traffic is increased and the sender is burdened. The recovery request time is determined to start the timer according to the determined time and prepare to receive the next packet (319).

If the repair request timer expires (320), the repair request procedure is initiated. The repair request procedure refers to the error packet list 315. If there is no error list, the packet error is repaired by a previously issued repair request or a packet error is repaired by another receiver's repair request. Prepare (322). If an errored packet still exists in the error packet list (322), a recovery request is raised. The recovery request is transmitted by unicast unlike when transmitting a data packet by encoding the sequence number of all error packets generated until the timer expires into one recovery request packet. It is important to unicast the recovery request packet because the transmission of the recovery request packet by the multicast method reduces network efficiency and burdens other receivers. In addition, it is very important to encode an error of several packets generated while the timer is operating as one recovery request packet without encoding a separate recovery request packet for each error packet. The reason for delaying the repair request by determining the repair request time is to reduce the number of repair request packets even if there are a plurality of packet errors. After sending the repair request packet, it is ready to receive the next packet.

When the sender receives the repair request packet from the receivers, it initiates an error recovery process (324). First, the transmission delay for the receiver who has issued the recovery request is updated (325), and the packet is requested in the packet list 308 (326) to determine whether to accept the recovery request (326). Whether to accept the repair request is determined based on the minimum value Da of the elapsed time when the packet is transmitted and the measurement transmission delay of the receivers. If the packet is transmitted and has not passed Da (328), the recovery request is ignored (329). This is because error recovery is performed by a recovery request from another receiver, and when recovery is performed again, data packets are repeatedly transmitted unnecessarily. If the packet is sent and Da passes, error recovery is performed. Error recovery consists of updating the packet transmission time (330), sending the packet to the packet list (308), and then retransmitting by multicast. The retransmission means here for the receiver who has an error but is delaying the recovery request. Therefore, a receiver that detects an error for any packet and delays a recovery request has an opportunity to recover the packet error by a recovery request of another receiver.

As described above, the packet error control scheme devised in the present invention has the following excellent effects when applied to a multicast protocol providing reliability.

First, the number of repair request packets can be minimized. This is possible by delaying the request for recovery as much as the application allows, by encoding multiple packet errors that occurred during the delay into one recovery request packet.

Secondly, it minimizes the network traffic due to the repair request packet and does not burden the other receivers. When the recovery request packet is transmitted in a multicast manner, it is delivered to all receivers that do not have error recovery capability. However, since the recovery request packet is transmitted in a unicast manner, only a sender having recovery capability is transmitted.

Third, it does not reduce the performance of the transceiver and the network due to unnecessary error recovery. This is because even if the sender receives an error recovery request, unnecessary error recovery can be suppressed by using the transmission time of the packet and the minimum transmission delay time of the receivers.

As a result, the effect of the present invention is to control the packet error using the dynamically varying measurement transmission delay time, thereby providing the effect of providing an optimal transmission environment when a large number of participants form a group and exchange multimedia information. .

Claims (6)

A data transmission module for forming a packet having a unique sequence number when a data transmission request is made from a user and transmitting the data packet to a communication network access device; Receiving the data packet transmitted through the network access device to collect transmission delay information and confirming a packet error, adding a sequence number of the packet to an error packet list at the time of packet error, and operating a recovery request timer; An error detection and recovery request module for transmitting an error recovery request packet through the network access device when the recovery request timer expires; It operates according to the error recovery request packet transmitted through the network access device, updates the transmission delay by searching for the packet in the packet list according to the packet order transmitted from the data transmission module, and if necessary, recovers the packet. Packet error controller of the multicast communication, characterized in that it comprises a error recovery module for retransmitting the error recovery request to accept. Delaying the recovery request time through a measurement transmission delay when the receiver detects a packet error; Encoding a plurality of packet errors occurring at the delayed time into one repair request packet; Transmitting the repair request packet by unicast instead of multicast; And restoring error recovery by using a measurement transmission delay when the recovery request is made. 3. The method of claim 2, wherein delaying a recovery request time through the measurement transmission delay Receiving a data transmission request from a user, generating a unique sequence number and retrieving a transmission delay value from a transmission delay list; And calculating a packet transmission time and storing the packet in a packet list, and then multicasting the stored packet to a plurality of receivers only once. The method of claim 2, wherein the encoding of the plurality of packet errors occurring at the delayed time into one repair request packet comprises: When the receiver receives each data packet, collecting a transmission delay carried in the data packet and updating the transmission delay list; A second step of checking the continuity of the received packets to determine the presence of an error, As a result of the check, if the error is absent, the packet is prepared to be received by deleting the packet from the error packet list. If the error exists, the error recovery request timer is operated after adding the sequence number of the lost packet to the packet error list. A third step to check whether As a result of the checking, when the error recovery request timer is operated, a packet is prepared for reception, and when the error recovery request timer is operated, the error recovery request time is determined by referring to a transmission delay list, and then the timer is operated according to the determined time, and the next packet reception ready Packet error control method using a packet error controller of the multicast communication, characterized in that it comprises a fourth step. 3. The method of claim 2, wherein transmitting the repair request packet by unicast instead of multicast When the repair request timer expires, the first step of performing a repair request procedure by referring to an error packet list and checking whether an error list exists; As a result of the checking, a second step of preparing to receive the next packet in the presence of an error list, a request for recovery by unicast transmission in the presence of an error list, and then preparing to receive the next packet is included. Packet error control method using a packet error controller of the multicast communication. 3. The method of claim 2, wherein in the recovery request, suppressing error recovery by using a measurement transmission delay After receiving a repair request packet, a first step of updating a transmission delay for a receiver who has filed a repair request; A second step of determining whether to accept the repair request by referring to the repaired packet in the packet list, and confirming whether to expand the repair request; As a result of the check, a third step of disregarding the recovery request when the recovery is not necessary, updating the packet transmission time if necessary, transmitting the packet to the packet list, and then retransmitting by multicast Packet Error Control Method Using Packet Error Controller of Multicast Communication.

Images