JP2007336191A - Parallel transmission method and system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable the utilization through network of application requiring mass data transfer and its service provision, while enabling data communications using a plurality of highly flexible communication paths. <P>SOLUTION: Between transceiving devices which perform time synchronization, a transmitting set attaches time information to a frame and transmits it, and a receiving set determines the processing order of the frame by using the time information attached to the frame. Further, the propagation delay time and transmitting interval are measured from the time information given to the frame. Moreover, the guard time is computed from the propagation delay time, and provided to every communication path so as to perform the frame processing in the right order without fail. Then, the frame processing is performed according to the computed transmitting interval. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a system for performing data transmission via a plurality of communication paths. In particular, the present invention relates to a method for controlling the processing order of frames. In the present invention, the term “frame” includes the meanings of packets, cells, and datagrams.

When it is difficult to secure a required transmission band with a single communication path, a method of virtually combining a plurality of communication paths into a single broadband communication path has been proposed and put into practical use. For example, in Ethernet (registered trademark), which is a layer 2 communication method, link aggregation technology that bundles a plurality of physical links to form a single broadband link is standardized (for example, see Non-Patent Document 1), commercially available switches, etc. Has been implemented. Also in ATM (Asynchronous Transfer Mode), IMA (Inverse) which performs data transmission using a plurality of lines of the same speed.
Multiplexing for ATM) technology has been standardized (for example, see Non-Patent Document 2).

  A method for performing data transmission with high real-time property in a best effort type network has been proposed. For example, there is a method in which a desired delivery time is assigned to an IP (Internet Protocol) packet, and the remaining time (allowance time) from the current time to the desired delivery time is used as a priority control parameter or a route selection metric value to suppress the maximum delay time. (For example, refer to Patent Document 1).

IEEE Std. 802.3ad-2000, Mar. 2000; “Amment to Carrier Sense Multiple Access with Collision of Physical Specimen Sig. ATM Spec. AF-PHY-0086.001, Mar. 1999; “Inverse Multiplexing for ATM Specification Version 1.1.” Japanese Patent No. 3563278

  The Ethernet link aggregation technology distributes the frames to be transmitted to the physical links in the virtual link based on the transmission destination address, the transmission source address, or both addresses of the frame. For this reason, the same flow always passes through the same physical link, so that it is not necessary to reconstruct the frame order in the receiving apparatus, but load distribution may not function sufficiently.

  On the other hand, since the ATM IMA technology distributes the cells in the same flow (synonymous with “frame” used in the text) in order to all the lines, the load distribution always functions. However, the IMA needs to transmit cells at the same transmission timing of all lines for cell rearrangement processing in the receiving apparatus, and lacks the flexibility of transmission.

  In addition, since it is difficult to measure the propagation delay time for each cell in IMA, if the propagation delay time fluctuates frequently, the cells may not be rearranged in the correct order. Furthermore, these methods are technologies closed to P2P (Point-to-Point) transmission, and cannot be applied to M2P (Multipoint-to-point) and M2M (Multipoint-to-Multipoint) transmission.

  The method of assigning a desired delivery time to an IP packet and using the remaining time from the current time to the desired delivery time as a priority control parameter or routing metric value is an effective method for quality assurance of applications limited to the maximum delay time. Yes, it can be applied to packet rearrangement by using the remaining time. However, in the non-congested state, processing is performed for each arrival order regardless of the remaining time, and therefore, when data is transmitted using a plurality of communication paths, packets may not be processed in the correct order.

  The present invention has been made under such a background, enables data transmission using a plurality of highly flexible communication paths, and uses an application requiring large-capacity data transfer via a network. Another object of the present invention is to provide a parallel transmission method and system that can provide the service.

  Between the time-synchronized transmission / reception devices, the transmission device transmits the frame with time information, and the reception device uses the time information attached to the frame to determine the processing order of the frames. Also, by measuring the propagation delay time and transmission interval from the time information given to the frame, and calculating the guard time from the propagation delay time and providing this for each communication path, the frame processing is performed in the correct order. . At this time, frame processing is performed according to the calculated transmission interval.

  For example, when the propagation delay times of the communication channels # 1 and # 2 are 1 second and 100 milliseconds, respectively, they are used for parallel transmission. Here, it is assumed that frame # 1 is transmitted from communication channel # 1, and frame # 2 is transmitted from communication channel # 2 200 milliseconds later. At this time, frame # 2 arrives at the transmission destination earlier than frame # 1. If the frame # 2 is being processed or there is no frame waiting for processing, the frame # 2 is processed as it is without using the guard time, and the order is reversed with respect to the later-arrived frame # 1. . When the guard time is used, the propagation delay difference is adjusted. Therefore, even if the frame # 2 arrives early, the processing is performed from the frame # 1, and the order is not reversed.

  Next, the operation of the parallel transmission method and system of the present invention will be described. By using time information for frame processing, the reference for determining the order in which frames are processed becomes a single reference regardless of the flow or the transmission device. Therefore, the degree of freedom of transmission in the transmission device is increased, and the rearrangement processing of frames arriving from a plurality of communication paths is facilitated in the reception device. Furthermore, not only P2P transmission but also data transmission using a plurality of communication paths in various forms such as P2M, M2P, and M2M transmission becomes possible.

  In addition, since the propagation delay time can be measured for each frame, the frame rearrangement process can be reliably performed even when the propagation delay time frequently changes. At this time, by performing frame processing according to the calculated transmission interval, fluctuation at the time of data output can be suppressed.

  That is, the present invention is a parallel transmission method in which data transmission is performed via a plurality of communication paths. The feature of the present invention is that the transmission device and the reception device are time-synchronized with each other. Giving first time information that is the frame generation time or the frame transmission time to a transmission frame, selecting one of the plurality of communication paths, transmitting the frame from the selected communication path, The receiving apparatus determines the processing order of the frames based on the first time information given to the frames, and executes processing from the first frame in the processing order.

  Further, the receiving device calculates the second time information obtained by adding a guard time to the first time information or the time when the frame is received, and associates the second time information with the frame, and the processing order. The processing can be executed from the first frame whose current time has passed the second time information.

  Alternatively, the receiving device calculates a transmission interval of the frame based on the first time information given to the frame, and calculates second time information obtained by adding a guard time to the time when the frame is received. Then, the second time information can be associated with the frame, and the process can be executed according to the transmission interval for the frame in which the processing order is first and the current time has passed the second time information.

  Alternatively, the receiving device measures the propagation delay time of the plurality of communication paths by subtracting the first time information given to the frame from the reception time of the frame, and the maximum propagation delay of the plurality of communication paths The guard time of the communication channel is calculated by subtracting the propagation delay time of the communication channel from the time, or a fixed value is given as the guard time regardless of the propagation delay time, or the maximum propagation delay time A value obtained by adding or multiplying fixed values can be given as the guard time.

  Alternatively, the receiving device can receive and process the frames transmitted from a plurality of the transmitting devices at the same time.

  Further, the present invention can be viewed from the viewpoint of a parallel transmission system. That is, the present invention is a parallel transmission system that performs data transmission via a plurality of communication channels, and the feature of the present invention is that the transmission device and the reception device are time-synchronized with each other. Giving first time information that is the frame generation time or the frame transmission time to a transmission frame, selecting one of the plurality of communication paths, transmitting the frame from the selected communication path, The receiving apparatus determines the processing order of the frames based on the first time information given to the frames, and executes processing from the first frame in the processing order.

  Further, the receiving device calculates the second time information obtained by adding a guard time to the first time information or the time when the frame is received, and associates the second time information with the frame, and the processing order. The processing can be executed from the first frame whose current time has passed the second time information.

  Alternatively, the receiving device calculates a transmission interval of the frame based on the first time information given to the frame, and calculates second time information obtained by adding a guard time to the time when the frame is received. Then, the second time information can be associated with the frame, and the process can be executed according to the transmission interval for the frame in which the processing order is first and the current time has passed the second time information.

  Further, the present invention can be viewed from the viewpoint of a receiving apparatus. That is, the present invention is a receiving apparatus applied to a parallel transmission system that performs data transmission via a plurality of communication paths, and the feature of the present invention is that the transmitting apparatus facing the own receiving apparatus is time-synchronized with each other. And the first time information given to the frame received from one of the plurality of communication paths, wherein the first time information that is the frame generation time or the frame transmission time is given to the data transmission frame. The processing order of the frames is determined based on the information, and the processing is executed from the first frame in the processing order.

  In addition, the first time information or the second time information obtained by adding a guard time to the time when the frame is received is calculated, the second time information is associated with the frame, and the processing order is first and present The processing can be executed from the frame whose time has passed the second time information.

  Alternatively, the transmission interval of the frame is calculated based on the first time information given to the frame, and second time information obtained by adding a guard time to the time when the frame is received is calculated. The second time information is associated, and the process can be executed in accordance with the transmission interval for the frame in which the processing order is first and the current time has passed the second time information.

  In addition, the present invention can be viewed from the viewpoint of a transmission apparatus. That is, the present invention is a transmitting device applied to a parallel transmission system that performs data transmission via a plurality of communication paths, and the feature of the present invention is that a receiving device opposite to its own transmitting device is time-synchronized with each other. The first time information that is the frame generation time or the frame transmission time is assigned to the data transmission frame, one of the plurality of communication paths is selected, and the frame is transmitted from the selected communication path. There is a place to do.

  Further, the present invention can be viewed from the viewpoint of a program. That is, the present invention is a program that, when installed in a general-purpose information processing apparatus, causes the general-purpose information processing apparatus to realize a function corresponding to the reception apparatus or transmission apparatus of the present invention.

  By recording the program of the present invention on a recording medium, the general-purpose information processing apparatus can install the program of the present invention using this recording medium. Alternatively, the program of the present invention can be directly installed on the general-purpose information processing apparatus via a network from a server that holds the program of the present invention.

  Thereby, the receiving apparatus or transmitting apparatus of this invention is realizable using a general purpose information processing apparatus.

  According to the present invention, it is possible to perform data transmission using a plurality of highly flexible communication channels, and it is possible to use an application that requires large-capacity data transfer via a network and to provide a service thereof.

  In the embodiment shown below, the communication direction is mainly described using only one direction in order to simplify the description (although some examples may be described using bidirectional), the communication direction is It is not limited. Similarly, as for the number of communication channels, many are described as two, but the number of communication channels is not limited. Further, the number of the plurality of apparatuses in the embodiments described below is not limited. The frame may be variable length or fixed length.

(First Example)
A first embodiment will be described with reference to FIGS. 1 to 5 (an example of “parallel transmission”). FIG. 1 shows a block configuration of the receiving device 3 used in the first embodiment. In FIG. 1, the signal flow is from left to right. The receiving device 3 used in the first embodiment includes a processing order determination unit 10 that determines a processing order of frames based on time information given to a frame received from one of a plurality of communication paths, and a first processing order. A scheduler 11 for executing processing from the frames and a buffer unit 12 for temporarily storing the arrived frames.

  FIG. 2A shows a state in which the transmission device 2 and the reception device 3 perform data transmission via the network 1 using the two communication paths p1 and p2. The time c2 of the transmission device 2 and the time c3 of the reception device 3 are synchronized with the reference time c1 via the network 1 or a network independent of the network 1. However, when the reference time c1 does not exist, one device time (for example, time c2) is used as a reference, and the other device time (for example, time c3) is determined via the network 1 or the network 1 independent of the network 1. You may synchronize with.

  FIG. 2B shows an example in which data is processed and transmitted. In the transmission device 2, the data d1, d2, and d3 delivered from the higher layer side are each framed (frames f1, f2, and f3), distributed to the two communication paths p1 and p2, and transmitted. At this time, time information t1, t2, and t3 (t1 <t2 <t3) are attached to the frames f1, f2, and f3, respectively.

  The frames received by the receiving device 3 are processed with the processing order determined according to the time information regardless of the arrival order. For example, even if the arrival order at the receiving device 3 is frames f2, f3, and f1, the frames are deframed in accordance with the time information in the order of f1, f2, and f3 (processing to remove frame overhead such as time information), Data is also passed to the upper layer in the order of d1, d2, and d3. Alternatively, the deframing process is completed in the arrival order (f2, f3, f1), and the data is passed to the upper layer in the order of d1, d2, d3 according to the time information associated with the data.

  The “communication path selection” portion of the first embodiment will be described with reference to FIG. 3A to 3F conceptually show processing in which the frame distribution function FS distributes a frame to two communication paths p1 and p2. FIG. 4 shows a principal block configuration of the transmission apparatus 2. FIG. 5 shows a processing flow in the transmission apparatus 2. The transmission device 2 gives time information that is a frame generation time or a frame transmission time to the data transmission frame by the framing processing unit 22, selects one of a plurality of communication paths by the frame distribution function FS21, and A frame is transmitted from the selected communication path. In addition, the link control unit 20 performs association control between the connection destination and the communication path.

  As shown in FIG. 3 (a), the utilization rates of the communication channels p1 and p2 ((the amount of bits flowing in per unit time) / (the transmission rate of the communication channel (bits / second))) are ρ1 and ρ2, respectively. Suppose. The frame distribution function FS distributes frames so that ρ1 = ρ2. For example, when the utilization rate becomes ρ1> ρ2 at a certain time, the next frame is transmitted from the communication path p2. As a result, if the utilization rate is ρ1 <ρ2, the next frame is transmitted from the communication path p1. Conversely, if ρ1> ρ2, the frame is continuously transmitted from the communication path p2. If ρ1 = ρ2, the frame is transmitted from an arbitrary communication path.

Note that the amount of bits flowing in during a unit time can be measured by recording the cumulative length of frames transmitted from the communication path in the unit time. For example, if 100 frames of 1500 bytes are transmitted from a communication channel in one second,
1500 × 8 × 100 = 1.2 megabits / second.

The transmission speed of the communication path is basically the physical rate of the communication path. For example, when a certain communication path is 1 gigabit Ethernet,
The transmission speed of the communication path is 1 gigabit / second. If the same physical link is shared fairly among a plurality of communication paths, the value obtained by dividing the physical rate by the number of shares is the transmission speed of the communication path. For example, if 4 channels share 1 Gigabit Ethernet fairly,
The transmission speed of the communication path = 250 megabits / second.

  As shown in FIG. 3B, it is assumed that the transmission rates of the communication channels p1 and p2 are β1 and β2, respectively. The frame distribution function FS probabilistically selects a communication path according to the ratio of each transmission rate, and transmits the frame. That is, in the example of FIG. 3B, the probability that the communication path p1 is selected is β1 / (β1 + β2), and the probability that the communication path p2 is selected is β2 / (β1 + β2). If the transmission rates of the communication paths p1 and p2 are 1 gigabit / second and 2 gigabit / second, respectively, the columns in which the communication paths p1 and p2 are selected are 1/3 and 2/3, respectively.

  As shown in FIG. 3C, it is assumed that the frame loss rates of the communication paths p1 and p2 are ε1 and ε2, respectively. The frame distribution function FS probabilistically selects a communication path according to the frame loss rate and transmits the frame. That is, in the example of FIG. 3C, when the relationship between the frame loss rates is ε1> ε2, the probability of selecting the communication channel p2 is increased, and more communication channels p2 are selected. If the frame loss rate exceeds the reference value, the communication path is also excluded from the options. For example, when ε1 ≧ ε0 under the reference value ε0, the frame is transmitted only from the communication path p2 (ε2 <ε0).

  The frame loss rate is information given from the receiving device, and the receiving device indicates the ratio of the number of frames in which an error is detected with respect to the total number of received frames or the number of frames that have been correctly received with respect to the total number of received frames. The frame loss rate is calculated as a ratio of the number subtracted from the total number of frames.

As shown in FIG. 3D, it is assumed that the propagation delay times of the communication paths p1 and p2 are Td ^(p1) and Td ^(p2) , respectively. The frame distribution function FS removes the communication path from the options when the propagation delay time difference exceeds the reference value. For example, when the reference value is (Td ^(p1) −Td ^(p2) ) ≧ ΔTd under ΔTd, the frame is transmitted only from the communication path p2.

  As shown in FIG. 3E, the frame distribution function FS randomly selects a communication path and transmits a frame. In other words, the probabilities of selecting each communication path are all equal regardless of the state of the communication path (utilization rate, bandwidth, frame loss rate). In the example of FIG. 3E, since there are two communication paths, the probability that the communication paths p1 and p2 are selected is both ½.

  As shown in FIG. 3F, the frame distribution function FS selects a communication path in a predetermined order and transmits a frame. The simplest method is to select all the communication paths once, and in the example of FIG. 3 (e), p1 → p2 → p1 →. As another method, there is a method of giving a rule for determining a selection order in advance and selecting a communication path according to this rule. For example, with the rule of selecting all the communication paths three times, in the example of FIG. 3 (f), p 1 → p 1 → p 1 → p 2 → p 2 → p 2 → p 1 →. The method shown previously can be considered as one of these rules.

  As another method, there is a method in which a table in which the selection order is determined in advance is given and a communication path is selected according to this table. For example, in the case of FIG. 3 (f), assuming that the frame distribution conforms to the table FST, the communication path is p2-> p2-> p1-> p2-> p2-> p2-> p1-> p1->.

  As shown in FIG. 3G, the frame distribution function FS selects a communication path that is not transmitting a frame (non-busy state) and transmits the frame. For example, when the frame f1 is being transmitted from the communication path p1 (the communication path p1 is busy), the frame distribution function FS transmits the frame f2 to be subsequently transmitted from the communication path p2 in the non-busy state.

  As shown in FIG. 3 (h), the frame distribution function FS selects a communication path with a short transmission waiting queue and transmits the frame. For example, when the transmission queues of the communication paths p1 and p2 are q1 and q2 (q1> q2), respectively, the frame distribution function FS distributes the next frame to the communication path p2.

(Second embodiment)
A second embodiment will be described with reference to FIG. 6 (an example of “use of guard time”). FIG. 6A shows, on the time axis, events (processing) for frames arriving from the two communication paths p1 and p2 in the receiving apparatus 3 (the left in the figure is the past and the right in the figure is the future). Corresponding). At time T1, the frame f2 (time information t2) is received from the communication path p2 (I2-1). The guard time for the communication path p2 is GT2, and the processing of the frame f2 is not started until time T3 when the guard time GT2 ends. During this time, the frame f1 (time information t1) arrives from the communication path p1 at time T2 (I1-1).

  The guard time for the communication path p1 is GT1, and the processing of the frame f1 is not started until time T4 when the guard time GT1 ends. The guard time GT2 ends at time T3 during this period, but the processing of the frame f2 is not started because the relationship between the time information t1 and t2 is t1 <t2. At time T4, the guard time GT1 ends, and processing of the frame f1 is started (I1-2). At time T5, the processing of the frame f1 ends (I1-3), and then the processing of the frame f2 starts (I2-2). At time T6, the process for frame f2 ends.

  FIG. 6B shows, on the time axis, events (processing) for frames arriving from the two communication paths p1 and p2 in the receiving apparatus 3 (the left in the figure is the past and the right in the figure is the future). Corresponding). At time T1, the frame f2 (time information t2) is received from the communication path p2 (I2-1). The guard time for the communication path p2 is GT2, and the processing of the frame f2 is not started until time T2 when the guard time GT2 ends. The guard time GT2 ends at time T2, and processing of the frame f2 is started (I2-2). The frame f1 (time information t1) is received from the communication path p1 at time T3 when the frame f2 is being processed (I1-1). Since the relationship between the time information t1 of the frame f1 and the time information t2 of the frame f2 is t1 <t2, the frame f1 is immediately discarded (I1-L). At time T4, the processing for frame f2 ends (I2-3).

(Third embodiment)
A third embodiment will be described with reference to FIG. FIG. 7A shows a state in which the transmission apparatus 2 and the reception apparatus 3 perform data transmission using the communication paths p1 and p2 via the network 1. The time c2 of the transmission device 2 and the time c3 of the reception device 3 are synchronized with the reference time c1 via the network 1 or a network independent of the network 1. However, when the reference time c1 does not exist, one device time (for example, time c2) is used as a reference, and the time of another device (for example, time c3) is synchronized with this via the network 1 or a network independent of the network 1. You may do that.

  FIG. 7B shows an example in which data is processed and transmitted. A processing flow of the receiving device 3 of the third embodiment is shown in FIG. The block configuration of the receiving device 3 is the same as that in FIG. In the transmission device 2, the data d1, d2, and d3 transferred from the upper layer side at the time intervals τ12 and τ23 are respectively framed (frames f1, f2, and f3) and distributed to the two communication paths p1 and p2 for transmission. Is done. At this time, time information t1, t2, and t3 (t1 <t2 <t3: τ12 = t2-t1, τ23 = t3-t2) are attached to the frames f1, f2, and f3, respectively. The frames received by the receiving device 3 are processed with the processing order and the processing time interval determined according to the time information regardless of the arrival order.

  For example, even if the arrival order at the receiving device 3 is the frames f2, f3, and f1, the frames are deframed according to the time information in the order of f1, f2, and f3 and the processing time intervals τ12 and τ23 (frames such as time information) Data is also transferred to the upper layer side in the order of d1, d2, and d3 and at time intervals τ12 and τ23. Alternatively, the deframing process is completed in the arrival order (f2, f3, f1), and the data is passed to the upper layer in the order of d1, d2, d3 and the time intervals τ12, τ23 according to the time information associated with the data.

(Fourth embodiment)
A fourth embodiment will be described with reference to FIGS. 9 to 13 ("Propagation delay time measurement" portion). FIG. 9A shows a state in which data transmission is performed between the transmission device 2 and the reception device 3 using the two communication paths p1 and p2. FIG. 9B shows the relationship of events (processing) between the transmission device 2 and the reception device 3 using the time axis (the upper side corresponds to the past and the lower side corresponds to the future. ). For example, the transmission device 2 transmits the frame f1 at time Ts (= t1), and the reception device 3 receives the frame f1 at time Tr. The time difference corresponds to the propagation delay time and is expressed as an expression shown in FIG. The receiving device 3 can perform measurement every time a frame is received, or can perform measurement using an arbitrary frame with a fixed time interval. Furthermore, as a value for calculating the guard time, an individual measurement value may be used, or a value obtained by statistically processing the measurement value may be used.

  In FIG. 9 (d), data transmission is performed between the transmission device 2 and the reception device 3 using the two communication paths p1 and p2, and the propagation delay time measurement frames fm1 and fm2 are further transmitted. It shows the state. The propagation delay time measurement frame uses a header having the same configuration as the data transmission frame. The two are distinguished by using identification information in the header or identification information stored in a payload portion into which data is stored.

  Further, the propagation delay time measurement frame can store the management information of the communication path, the transmission device 2 and the reception device 3 in the payload portion, and the transmission device 2 and the reception device 3 use these information to change the device state. Variable.

  FIG. 9E shows the relationship of events (processing) between the transmission device 2 and the reception device 3 using a time axis (the upper side corresponds to the past and the lower side corresponds to the future). For example, the transmission device 2 transmits the frame fm1 at time Ts (= tm1). The transmission device 2 can transmit the propagation delay measurement frame at an arbitrary timing or periodically. The receiving device 3 receives the frame fm1 at time Tr. The time difference corresponds to the propagation delay time and is expressed as an expression shown in FIG. The receiving device 3 can perform measurement every time a propagation delay time measurement frame is received, or can perform measurement using an arbitrary propagation delay extension time measurement frame with a fixed time interval. Furthermore, as a value for calculating the guard time, an individual measurement value may be used, or a value obtained by statistically processing the measurement value may be used.

Next, the “guard time calculation” portion in the fourth embodiment will be described with reference to FIGS. FIG. 10A shows a state in which data transmission is performed between the transmission device 2 and the reception device 3 using the two communication paths p1 and p2. Td ^(p1) and Td ^(p2) indicate the propagation delay times of the communication paths p1 and p2, respectively. Since Td ^(p1) > Td ^(p2) , the propagation delay time Td ^{(p1 ) of} the communication path ^{p1. )} Is the maximum propagation delay time Td ^(max) . At this time, with reference to the maximum propagation delay time Td ^(max) (= Td ^(p1) ), a value obtained by subtracting the propagation delay time Td ^(j) (j = 1, 2) of the communication path to be calculated from this value. A value obtained by adding the constant value α is the guard time of the communication channel, and is expressed as an expression shown in FIG. Alternatively, the maximum propagation delay time Td ^(max) (= Td ^(p1) ) is used as a reference, and the value obtained by subtracting the propagation delay time Td ^(j) (j = 1, 2) of the communication path to be calculated from this value is constant. A value obtained by multiplying the value α is the guard time of the communication channel, and is expressed as an equation shown in FIG. Further, with reference to the maximum propagation delay time Td ^(max) (= Td ^(p1) ), a value obtained by adding or multiplying this value by a constant value α is the guard time for all the communication paths, and FIG. 10 (d) or (e ). Furthermore, the fixed value is the guard time for all communication paths, and is expressed as an expression shown in FIG.

  Next, the α portion in FIG. 10 will be described with reference to FIG. FIG. 11A shows a state in which data is transmitted between the transmission device 2 and the reception device 3 using the two communication paths p1 and p2. The communication path p1 passes through the link L11, the relay node R2, and the link L12, and the communication path p2 passes through the link L21, the relay node R1, the link L22, the relay node R2, the link L23, the relay node R3, and the link L24. Here, the link indicates a physical communication medium such as an optical fiber or an electric cable.

  The time (= propagation delay time) required from the start of transmission by the transmission device 2 to the completion of reception by the reception device 3 can be broadly divided into two elements. One is the time to propagate through the link and is related to the length of the link. For example, if the link is an optical fiber, the length of the link is LT, the speed of light is C, and the refractive index of the optical fiber is n, the propagation time in the link is expressed by LT / (C / n). The other is the transmission time associated with transmission / reception, which varies depending on the frame length, the transmission speed of the apparatus, and the number of relay nodes.

  For example, if the frame length is D (bits), the transmission rate of the device is β (bits / second), and the transit time of the relay node (number of devices N) is only related to the transmission rate, the transmission time associated with transmission / reception Is represented by (N + 1) × D / β (when the propagation speeds of all devices are the same, and the relay node starts transmission after receiving all frames). The former (the time for propagation through the link) varies depending on external environmental conditions such as temperature, but the variation is small.

  In FIG. 11 (a), the total length of the fiber (the total length of L11 and L12 in the communication path p1 and the total length of L21, L22, and L23 in the communication path p2) is 100 kilometers, and is 1 meter longer due to temperature changes. In other words (if it becomes 100.001 kilometers), the increase is 5 nanoseconds. On the other hand, the latter (transmission time associated with transmission / reception) varies greatly depending on the frame length.

  In FIG. 11A, when the transmission rate β is 1 gigabit / second, the transmission time on the communication path p1 is about 1.02 microseconds @ 64 bytes and 24 microseconds @ 1500 bytes. The transmission time on the communication path p2 is about 23 microseconds and about 2.05 microseconds @ 64 bytes and 48 microseconds @ 1500 bytes, and the difference is about 46 microseconds. If a frame with a larger size is used, the difference becomes larger. For this reason, it is important to consider the frame length range, transmission rate, and number of relay nodes when providing a margin (a constant value α) for guard time calculation.

  FIG. 11B is a table that gives a constant value α for calculating the gart time according to the difference between the maximum frame length and the minimum frame length, and the constant value α is determined with reference to this table. For example, when ΔD1 is less than 100 bytes, ΔD2 is 100 bytes or more and less than 500 bytes, ΔD3 is 500 bytes or more and less than 1000 bytes, and so on, the minimum and maximum frame lengths used for transmission are 1200 bytes and 1500 bytes, respectively. The constant value α = α3.

  FIG. 11C is a table that gives a constant value α for guard time calculation according to the difference between the maximum frame length and the minimum frame length and the number of relay nodes, and the constant value α is determined with reference to this table. . For example, ΔD1 is less than 100 bytes, ΔD2 is from 100 bytes to less than 500 bytes, ΔD3 is from 500 bytes to less than 1000 bytes, N1 is less than 5, N2 is from 5 to less than 10, N3 is from 10 to less than 15, In this case, if the minimum and maximum frame lengths used for transmission are 1200 bytes, 1500 bytes, and the number of relay nodes is 7, the constant value α = α23.

  FIG. 11D is a table that gives a constant value α for guard time calculation according to the difference between the maximum frame length and the minimum frame length and the transmission rate, and the constant value α is determined with reference to this table. For example, ΔD1 is less than 100 bytes, ΔD2 is 100 bytes or more and less than 500 bytes, ΔD3 is 500 bytes or more and less than 1000 bytes, B1 is less than 100 gigabits / second, B2 is 100 megabits / second or more and less than 1 gigabit / second, B3 Is 1 gigabit / second or more and less than 10 gigabit / second,..., If the minimum and maximum frame lengths used for transmission are 1200 bytes, 1500 bytes, and the transmission speed is 10 megabits / second, a constant value α = α13 It becomes.

  FIG. 11 (e) is a table that gives a constant value α for guard time calculation according to the difference between the maximum frame length and the minimum frame length, the number of relay nodes, and the transmission rate. To decide. For example, ΔD1 is less than 100 bytes, ΔD2 is from 100 bytes to less than 500 bytes, ΔD3 is from 500 bytes to less than 1000 bytes, N1 is less than 5, N2 is from 5 to less than 10, N3 is from 10 to less than 15, If B1 is less than 100 gigabits / second, B2 is 100 gigabits / second or more and less than 1 gigabit / second, B3 is 1 gigabit / second or more and less than 10 gigabit / second, ... When 1200 bytes, 1500 bytes, the number of relay nodes is 7, and the transmission speed is 10 megabits / second, respectively, the constant value α = α123.

  In addition, as shown in FIG. 11 (f), there is a method in which a fixed value α0 is given as a constant value α and used for guard time calculation regardless of conditions.

  FIG. 12 shows a block configuration of the receiving device 3 of the fourth embodiment. FIG. 13 shows a processing flow of the receiving apparatus 3 of FIG. The processing order determination unit 30 acquires time information and performs delay time calculation and GT calculation processing. The scheduler 31 calls the frame (data) having the earliest time information and outputs data if the GT has passed. The buffer unit 32 performs a process of giving GT.

(Fifth embodiment)
A fifth embodiment will be described with reference to FIG. 14 (example of “M2P type”). FIG. 14A shows a state in which the transmitting apparatuses 2A and 2B are performing data transmission via the network 1 using the two communication paths {pA1, pA2} and {pB1, pB2} with the receiving apparatus 3, respectively. Show. The times c2A and c2B of the transmission devices 2A and 2B and the time c3 of the reception device c3 are synchronized with the reference time c1 via the network 1 or a network independent of the network 1. However, when the reference time c1 does not exist, one device time (for example, time c2A) is used as a reference, and the time of other devices (for example, time c2B, c3) is obtained via the network 1 or the network 1 independent of the network 1. You may synchronize with.

  FIG. 14B shows an example in which data is processed and transmitted. In the transmission device 2A, the data d10, d13, d14 delivered from the upper layer side are each framed (frames f10, f13, f14), distributed to the two communication paths pA1, pA2, and transmitted. At this time, time information t10, t13, and t14 are attached to the frames f10, f13, and f14, respectively. Similarly, in the transmission device 2B, the data d11, d12, and d15 are each framed (frames f11, f12, and f15) and distributed to the two communication paths pB1 and pB2 and transmitted. At this time, time information t11, t12, and t15 are attached to the frames f11, f12, and f15, respectively.

  Here, the relationship of time information is t10 <t11 <t12 <t13 <t14 <t15. The frames received by the receiving device 3 are processed with the processing order determined according to the time information regardless of the arrival order. For example, even if the arrival order at the receiving device 3 is frames f10, f12, f15, f13, f11, f14, the frames are deframed in the order of f10, f11, f12, f13, f14, f15 (time information etc. The data is also transferred to the upper layer in the order of d10, d11, d12, d13, d14, and d15. Alternatively, the deframing process is finished in the arrival order (f10, f12, f15, f13, f11, f14), and the data is higher in the order of d10, d11, d12, d13, d14, d15 according to the time information associated with the data. Passed to the layer side.

(Sixth embodiment)
A sixth embodiment will be described with reference to FIG. 15 (an example related to “change in communication path state”).

  FIG. 15 shows the state of communication paths (p1, p2, p3) between the transmission device 2 and the reception device 3 and the relationship between events (processing) and time (the left is the past, the right is the diagram). Corresponding to the future). A state where the communication path is used is indicated by a thick solid line, and a state where the communication path is not used is indicated by a broken line.

  As shown in FIG. 15A, the use of the communication path p2 that has been used together with the communication path p1 at time Tc is stopped (Ip2-SP), and the use of the communication path p3 is started simultaneously (Ip3-ST). There is no change in the number of communication paths used between the transmission device 2 and the reception device 3. The transmission device 2 selects an available communication path and transmits a frame. The receiving device 3 selects a usable communication path and transmits a frame. The receiving device 3 processes a frame that arrives from an available communication path.

  As shown in FIG. 15B, the use of the communication path p3 is started at time Tc1 (Ip3-ST). The number of communication channels used between the transmission device 2 and the reception device 3 is two to three, including the communication channels p1 and p2. Use of the communication paths p1 and p3 is stopped at time Tc2 (Ip1-ST, Ip3-SP). The number of communication paths used between the transmission device 2 and the reception device 3 is reduced from three to one. The transmission device 2 selects an available communication path and transmits a frame. The receiving device 3 processes a frame that arrives from an available communication path.

(Seventh embodiment)
A seventh embodiment will be described with reference to FIG. 16 (example of “P2M type”).

  FIG. 16A shows a state in which the transmitting apparatus 2 performs data transmission via the network 1 using the two communication paths {pA1, pA2} and {pB1, pB2} respectively with the receiving apparatuses 3A and 3B. Show. The time c2 of the transmission device 2 and the times c3A and c3B of the reception devices c3A and c3B are synchronized with the reference time c1 via the network 1 or a network independent of the network 1. However, when the reference time c1 does not exist, one device time (for example, time c2) is used as a reference, and the other device time (for example, time c3A, c3B) is transmitted via the network 1 or the network 1 independent of the network 1. You may synchronize with.

  FIG. 16B shows an example in which data is processed and transmitted. In the transmission device 2, the data d20, d23, d22, d23, d24, d25 passed from the upper layer side is framed (frames f20, f21, f22, f23, f24, f25), and four communication paths pA1, It is distributed to pA2, pB1, and pB2 and transmitted. At this time, time information t20, t21, t22, t23, t24, and t25 (t20 <t21 <t22 <t23 <t24 <t25) are attached to the frames f20, f21, f22, f23, f24, and f25, respectively. The frames received by the receiving device 3A are processed according to the time information, regardless of the arrival order.

For example, even if the arrival order at the receiving device 3A is the frames f21, f24, and f23, the frames are deframed in the order of f21, f23, and f24 (processing that removes frame overhead such as time information), and the data is also d21. , D23, and d24 in this order. (Or, the deframing process is completed in the arrival order (f21, f24, f23), and the data is passed to the upper layer side in the order of d21, d23, d24 according to the time information associated with the data.)
Similarly, the frames received by the receiving device 3B are processed with the processing order determined according to the time information regardless of the arrival order. For example, even if the arrival order at the receiving device 3B is the frames f22, f20, and f25, the frames are deframed in the order of f20, f22, and f25, and the data is also transferred to the upper layer in the order of d20, d22, and d25. Passed. Alternatively, the deframing process is completed in the arrival order (f22, f20, f25), and the data is passed to the upper layer side in the order of d20, d22, d25 according to the time information associated with the data.

(Eighth Example)
An eighth embodiment will be described with reference to FIG. 17 (example of “M2M type”).

  FIG. 17A shows a state in which the transmission / reception devices 23A, 23B, and 23C perform data transmission through the network 1 using a total of six communication paths (pAB, pAC, pBC, pBA, pCA, and pCA). Show. The times c23A, c23B, c23C of the transmission / reception devices 23A, 23B, 23C are synchronized with the reference time c1 via the network 1 or a network independent of the network 1. However, if the reference time c1 does not exist, one device time (for example, time c23A) is used as a reference, and the other device time (for example, time c23B, c23C) is transmitted via the network 1 or a network independent of the network 1. You may synchronize with this.

  FIG. 17B shows an example in which data is processed and transmitted. In the transmission / reception device 23A, the data dA1 and dA2 delivered from the higher layer side are respectively framed (fA1 and fB2) and transmitted from the two communication paths pAB and pAC. At this time, time information tA1 and tA2 are attached to the frames fA1 and fA2, respectively. Similarly, in the transmission / reception device 23B, the data dB1, dB2, and dB3 are framed (fB1, fB2, and fB3) and transmitted from the two communication paths pBA and pBC. At this time, time information tB1, tB2, and tB3 are attached to the frames fB1, fB2, and fB3, respectively.

  Similarly, in the transmission / reception device 23C, the data dC1, dC2, and dC3 are each framed (fC1, fC2, and fC3) and transmitted from the two communication paths pCA and pCB. At this time, time information tC1, tC2, and tC3 are attached to the frames fC1, fC2, and fC3, respectively. Here, the relationship of time information is tB1 <tA1 <tB2 <tC1 <tC2 <tA2 <tC3 <tB3. The frames received by the transmission / reception device 23A are processed with the processing order determined according to the time information regardless of the arrival order.

  For example, even if the arrival order at the transmitting / receiving apparatus 23A is the frames fC1, fB2, and fC2, the frames are deframed in the order of fB2, fC1, and fC2 (processing that removes frame overhead such as time information), and the data is also dB2. , DC1, and dC2 in this order. Alternatively, the deframing process is completed in the arrival order (fC1, fB2, fC2), and the data can be passed to the upper layer in the order of dB2, dC1, and dC2 according to the time information associated with the data.

  Similarly, the frames received by the transmission / reception device 23B are processed with the processing order determined according to the time information regardless of the arrival order. For example, even if the arrival order at the transmission / reception device 23B is the frames fC3 and fA2, the frames are deframed in the order of fA2 and fC3, and the data is also passed to the upper layer in the order of dA2 and dC3. Alternatively, the deframing process is completed in the arrival order (fC3, fA2), and the data is also passed to the upper layer in the order of dA2, dC3 in accordance with the time information associated with the data.

  Similarly, the frames received by the transmission / reception device 23C are processed with the processing order determined according to the time information regardless of the arrival order. For example, even if the arrival order at the transmitter / receiver 23C is the frames fA1, fB1, and fB3, the frames are deframed in the order of fB1, fA1, and fB3, and the data is also moved to the upper layer side in the order of dB1, dA1, and dB3. Passed. Alternatively, the deframing process is completed in the arrival order (fA1, fB1, fB3), and the data is passed to the upper layer in the order of dB1, dA1, and dB3 according to the time information associated with the data.

(Ninth Example)
A ninth embodiment will be described with reference to FIG. 18 (example of “application uses frame time information”).

  In FIG. 18, a terminal device (hereinafter referred to as a transmission terminal) Term2 having a transmission device 2 and a terminal device (hereinafter referred to as a reception terminal) Term3 having a reception device 3 perform data transmission using two communication paths p1 and p2. Indicates the state of going. A common application is used between the transmitting terminal Term2 and the receiving terminal Term3, and application entities (hereinafter simply referred to as applications) are 2aF and 3aF, respectively.

  Transmission data a30 is passed from the application of the transmission terminal Term2 to the lower function 2iF. Here, an intermediate function 2iF is considered between the application 2aF and the transmission apparatus 2, but the application 2aF and the transmission apparatus 2 may share an interface. In that case, a30 = d30. Conversely, a plurality of functions may exist. The same applies to the interface between the application 3aF of the receiving terminal Term3 and the receiving function 3.

  The data a30 passed to the function 2iF is passed to the transmitting apparatus 2 as data d30 with a header for the function 2iF attached. The frame generation unit 2mF of the transmission device 2 generates a frame f30 by adding a header including time information to the received data d30, and transmits the frame f30 from the selected communication path (2pF and 3pF are physical interfaces). The receiving device 3 receives the frame f30, determines the processing order using the frame processing unit 3mF time information, and then processes the frame f30.

  The data d30 is passed to the function 3iF, the header is processed by the function 3iF, and then passed to the application 3aF as data a30. In parallel with this processing, time information t30 is passed from the receiving device 3 to the application 3aF. The application uses the time information t30 as auxiliary information for quality assurance and improvement.

(Tenth embodiment)
A tenth embodiment will be described with reference to FIG. 19 (an example of “frame processing of receiving apparatus”). The processing flow of FIG. 19 is shown in FIG. FIG. 19A shows, on the time axis, events (processing) for frames transmitted from the two communication paths p1 and p2 in the transmission device 2 (the left in the figure is the past, the right in the figure is the future). Corresponding). At time T1, the frame f1 (time information T1) is transmitted to the communication path p1 (Is1-1). Further, the frame f2 (time information T2) is transmitted to the communication path p2 at time T2 (Is2-1).

  FIG. 19B shows, on the time axis, events (processing) for frames arriving from the two communication paths p1 and p2 in the receiving apparatus 3 (the left in the figure is the past, the right in the figure is the future). Corresponding). At time T3, the frame f2 (time information T2) is received from the communication path p2 (Ir2-1). The value of the time information acquired from the frame f2 is T2, and the processing of the frame f2 is not started until time T7 obtained by adding the fixed time Tadd to T2. During this time, the frame f1 (time information T1) arrives from the communication path p1 at time T4 (Ir1-1). The value of the time information acquired from the frame f1 is T1, and the processing of the frame f1 is not started until time T5 obtained by adding the fixed time Tadd to T1. At time T5, processing of the frame f1 is started (Ir1-2), and processing of the frame f1 is completed at time T6 (Ir1-3). Processing at frame f2 is started at time T7 when time T2-T1 has elapsed from time T6 (Ir2-2), and processing at frame f2 ends at time T8 (Ir2-3).

  Instead of the fixed time Tadd, a maximum propagation delay time Td (max) or a value obtained by adding (multiplying) a fixed value to the maximum propagation delay time Td (max) may be used.

(Eleventh Example)
The eleventh embodiment is a program that, when installed in a general-purpose information processing device, causes the general-purpose information processing device to realize a function corresponding to the transmission device or the reception device of the present embodiment. The program is recorded on a recording medium and installed in a general-purpose information processing apparatus, or installed in a general-purpose information processing apparatus via a communication line, so that the transmission apparatus of the present embodiment is installed in the general-purpose information processing apparatus. Alternatively, a function corresponding to the receiving device can be realized. The general-purpose information processing apparatus is, for example, a general-purpose personal computer.

  According to the present invention, it is possible to perform data transmission using a plurality of highly flexible communication paths, and it is possible to use an application that requires large-capacity data transfer via a network and to provide the service. And a network with high added value for both users.

The principal part block block diagram of the receiver of a 1st Example. The figure which shows the whole structure and the concept of a process of 1st Example. The figure which shows the concept of selection of the several communication path of 1st Example. The principal part block block diagram of the transmitter of a 1st Example. The figure which shows the processing flow of the transmitter of a 1st Example. The figure for demonstrating the example of "use of guard time" of 2nd Example. The figure which shows the whole structure of a 3rd Example, and the concept of a process. The figure which shows the processing flow of the receiver of a 3rd Example. The figure for demonstrating the "propagation delay time measurement" part of 4th Example. The figure for demonstrating the "guard time calculation" part of 4th Example. The figure for demonstrating (alpha) part in FIG. The principal part block block diagram of the receiver of 4th Example. The figure which shows the processing flow of the receiver of 4th Example. The figure which shows the whole structure and the concept of a process of 5th Example. The figure for demonstrating the example in connection with the "change of a communication channel state" of 6th Example. The figure for demonstrating the example in connection with "P2M" of 7th Example. The figure for demonstrating the example in connection with "M2M" of 8th Example. The figure for demonstrating the example in connection with the "time information and application" of 9th Example. The figure for demonstrating the example of the "frame processing of a receiver" of 10th Example. The figure which shows the processing flow of the receiver of 10th Example.

Explanation of symbols

1 Network 2, 2A, 2B Transmitting device 3, 3A, 3B Receiving device 10, 30 Processing order determination unit 11, 31 Scheduler 12, 32 Buffer unit 20 Link control unit 21 FS (Frame distribution function)
22 Framing processor 23A, 23B, 23C Transmitter / receiver c1-c3, c2A, c2B, c3A, c3B, c23A, c23B, c23C, T1-T8 Time d1-d30 Data f1-f30 Frame L11-L24 Link p1, p2, pA1, pA2, pB1, pB2, pAB, pAC, pBA, pBC, pCA, pCB, communication channel t1 to t30 Time information Term2 Transmitting terminal Term3 Receiving terminal 2aF, 3aF Application a30, d30 Transmitting data 2iF, 3iF Function 2mF Frame generation unit 3mF Frame processor 2pF, 3pF Physical interface

Claims (14)

A parallel transmission method for transmitting data via multiple communication paths,
The transmitter and receiver are time synchronized with each other,
The transmission device assigns first time information that is the frame generation time or the frame transmission time to a data transmission frame, selects one of the plurality of communication paths, and selects the frame as the selected communication. Send from the road,
The parallel transmission method characterized in that the receiving device determines the processing order of the frames based on the first time information given to the frames, and executes processing from the first frame in the processing order. .   The reception device calculates second time information obtained by adding a guard time to the first time information or the time when the frame is received, associates the second time information with the frame, The parallel transmission method according to claim 1, wherein the process is executed from the frame whose current time has passed the second time information.   The receiving device calculates a transmission interval of the frame based on the first time information given to the frame, calculates second time information obtained by adding a guard time to the time of receiving the frame, 2. The parallel transmission according to claim 1, wherein the second time information is associated with the frame, and processing is executed according to the transmission interval for the frame in which the processing order is first and the current time has passed the second time information. Method.   The receiving apparatus measures the propagation delay time of the plurality of communication paths by subtracting the first time information given to the frame from the reception time of the frame, and determines the maximum propagation delay time of the plurality of communication paths. The guard time of the communication channel is calculated by subtracting the propagation delay time of the communication channel, or a fixed value is given as a guard time regardless of the propagation delay time, or a fixed value is set for the maximum propagation delay time. 4. The parallel transmission method according to claim 2, wherein a value obtained by adding or multiplying is given as a guard time.   The parallel transmission method according to claim 1, wherein the receiving apparatus receives and processes the frames transmitted from a plurality of the transmitting apparatuses simultaneously. A parallel transmission system that performs data transmission via multiple communication paths,
The transmitter and receiver are time synchronized with each other,
The transmission device assigns first time information that is the frame generation time or the frame transmission time to a data transmission frame, selects one of the plurality of communication paths, and selects the frame as the selected communication. Send from the road,
The parallel transmission system, wherein the receiving device determines the processing order of the frames based on the first time information given to the frames, and executes processing from the first frame in the processing order .   The reception device calculates second time information obtained by adding a guard time to the first time information or the time when the frame is received, associates the second time information with the frame, The parallel transmission system according to claim 6, wherein the process is executed from the frame whose current time has passed the second time information.   The receiving device calculates a transmission interval of the frame based on the first time information given to the frame, calculates second time information obtained by adding a guard time to the time of receiving the frame, The parallel transmission according to claim 6, wherein the second time information is associated with the frame, and the process is executed according to the transmission interval for the frame in which the processing time is first and the current time has passed the second time information. system. A receiving device applied to a parallel transmission system that performs data transmission through a plurality of communication paths,
Synchronize time with the transmitting device opposite to the receiving device,
First time information that is the frame generation time or the frame transmission time is given to the data transmission frame, and the first time information given to the frame received from one of the plurality of communication paths A receiving apparatus characterized in that the processing order of the frames is determined based on the processing, and the processing is executed from the first frame in the processing order.   The second time information obtained by adding a guard time to the time when the first time information or the frame is received is calculated, the second time information is associated with the frame, the processing time is first and the current time is The receiving device according to claim 9, wherein the processing is executed from the frame that has passed the second time information.   A transmission interval of the frame is calculated based on the first time information given to the frame, second time information obtained by adding a guard time to a time when the frame is received is calculated, and the second time information is added to the frame. The receiving apparatus according to claim 9, wherein the second time information is associated, and the process is executed according to the transmission interval for the frame in which the processing time is first and the current time has passed the second time information.   A program that, when installed in a general-purpose information processing device, causes the general-purpose information processing device to realize a function corresponding to the receiving device according to any one of claims 9 to 11. A transmission device applied to a parallel transmission system that performs data transmission via a plurality of communication paths,
Synchronize time with the receiving device opposite to its own transmitting device,
First time information that is the frame generation time or the frame transmission time is assigned to a data transmission frame, one of the plurality of communication paths is selected, and the frame is transmitted from the selected communication path. A transmitter characterized by the above.   14. A program that, when installed in a general-purpose information processing apparatus, causes the general-purpose information processing apparatus to realize a function corresponding to the transmission apparatus according to claim 13.

Images