JP6283618B2 - Packet control apparatus, method and program - Google Patents

Description

  The present invention relates to control of communication amount in, for example, M2M (Machine to Machine) communication.

  In recent years, with the expansion of the M2M platform, it has become easier to place a control server that was previously placed on-premises in a LAN (Local Area Network) in the cloud and transmitted between the control server and the terminal. It is expected that packets will be transmitted via a WAN (Wide Area Network). For this reason, the packet is likely to be delayed through the network, and the number of packets discarded in the control server or terminal on the receiving side may increase due to the influence (see, for example, Non-Patent Document 1). This causes an increase in the processing load on the control server or terminal on the receiving side and leads to one of the causes of congestion on the network, which is very undesirable. Further, in order to reduce transmission delay, it is necessary to enhance network facilities.

  On the other hand, as related art, there is a packet shaping method in which statistical processing is performed based on information such as jitter from a received packet sequence, and output control is performed based on the result of the statistical processing (see, for example, Patent Document 1). .

JP 2002-264583 A

J. Xiong, and J. Lam, “Stabilization of networked control systems with a logic ZOH,” IEEE Transactions on Automatic Control, 2009, 54 (2): 358-363.

  However, in the technique described in Patent Document 1, when a packet is discarded, the packet that cannot be discarded so as to reduce the fluctuation in the time direction of the arrival time of the control signal and should not be discarded. May be destroyed.

  The present invention has been made by paying attention to the above circumstances, and the object of the present invention is to preliminarily discard a packet to be discarded by the communication device on the receiving side, and to prevent fluctuations in the time direction of the arrival time of the packet. Another object of the present invention is to provide a packet control apparatus, method, and program that reduce the amount of communication in the network and thereby reduce the processing load on the communication apparatus and the equipment cost in the network.

To achieve the above object, according to a first aspect of the present invention, there is provided a first communication device that transmits a packet at a constant transmission cycle, and a packet transmitted from the first communication device is received via a network. A packet control device arranged on a communication path with a second communication device, the theoretical arrival time calculated based on a packet transmission period by the first communication device, and the first A determination unit that compares the actual arrival time of the packet transmitted from the communication device and determines whether or not the actual arrival time is delayed by a predetermined allowable value or more from the theoretical arrival time; and the determination unit The packet is discarded when it is determined that the actual arrival time of the packet is delayed more than the allowable value from the theoretical arrival time, and the packet is A packet output control means for transferring to said second communication device, wherein the packet output control means calculates a discard frequency of the packets by, determines whether the range of the threshold value set the calculated discarded frequency pre And a determination control means for variably controlling at least one of the theoretical arrival time and the allowable value so that the discard frequency falls within the threshold range when it is determined that the discard frequency exceeds the threshold range. It is what I did.

The second aspect of the present invention further comprises classification means for classifying the arriving packet into a plurality of series based on additional information of the packet arriving at the apparatus prior to the determination processing by the determining means, The determination processing is performed for each series classified by the classification means.

According to the first aspect of the present invention, in the packet control device arranged on the transmission line, when it is determined that the actual arrival time of the packet is delayed from the theoretical arrival time by a predetermined allowable value or more, The packet is discarded. Therefore, the delayed packet is discarded on the transmission line before reaching the second communication device, thereby reducing fluctuations in the time direction of the arrival time of the packet in the second communication device and packet control. The amount of communication in the network from the device to the second communication device can be reduced, and the occurrence of congestion can be reduced.
Further, at least one of the theoretical arrival time and the allowable value is variably controlled so that the packet discard frequency falls within the threshold range. For this reason, it is possible to perform packet thinning-out control under always optimal conditions.

  According to the second aspect of the present invention, at least one of the theoretical arrival time and the allowable value is variably controlled so that the packet discard frequency falls within the threshold range. For this reason, it is possible to perform packet thinning-out control under always optimal conditions.

According to the second aspect of the present invention, it is possible to perform optimum thinning control according to the characteristics of each packet sequence by classifying the arrived packet into a plurality of sequences based on the additional information of the packet. It becomes.

  That is, according to the present invention, a packet to be discarded is discarded in advance by the communication device on the receiving side, thereby reducing fluctuations in the time direction of the arrival time of the packet and reducing the amount of communication in the network. It is possible to provide a packet control apparatus, method, and program capable of reducing the processing load of the apparatus and the equipment cost in the network.

The figure which shows the whole structure of the M2M network system provided with the packet control apparatus which concerns on one Embodiment of this invention. The figure which shows the deployment location of a packet control apparatus. The block diagram which shows the function structure of the packet control apparatus shown in FIG. The flowchart which shows the procedure of the packet control processing by the packet control apparatus shown in FIG. 3, and a processing content. The figure for demonstrating the thinning-out process by the packet control apparatus shown in FIG. The figure which shows the arrangement | positioning place of the packet control apparatus which concerns on other embodiment of this invention.

Embodiments according to the present invention will be described below with reference to the drawings.
[One Embodiment]
(Constitution)
FIG. 1 is a diagram showing an overall configuration of an M2M network system including a packet control device according to an embodiment of the present invention. In this network system, a user LAN NW3 is connected to the Internet NW1 via a carrier network NW2, and a user terminal 3 as a second communication device is accommodated in the user LAN NW3.

  The user terminal 3 consists of portable terminals, such as a smart phone and a tablet-type terminal, for example, and transmits / receives a signal between household appliances, vital devices, measuring devices, and robot devices. A function of receiving measurement data transmitted from each device, packetizing it, and transmitting it to the control server 2 described later, and receiving a control signal transmitted by the packet from the control server 2 and transmitting it to each device Has a function of controlling the operation.

  A control server 2 as a first communication device is connected to the Internet NW1. The control server 2 is composed of a cloud computer operated by a service provider. Then, various services are executed by receiving and analyzing measurement data and the like sent from the user terminal 3 in packets. Examples of services executed with respect to the measuring device 5 include a user health care service based on vital data, a remote control service for home appliances based on measurement data such as temperature and humidity, and the measurement data on the position and movement of robot devices. Robot equipment operation services.

  Incidentally, the packet control device 1 is arranged in the inter-network connection device 4 connected to the carrier network NW2. FIG. 2 exemplifies candidates for the location of the packet control device 1, and the packet control device 1 is arranged in an inter-network connection device arranged at the connection end of the carrier network NW2 with the Internet NW1 and the user LAN NW3. . For example, an inter-network connection device 41 arranged at a relay point equivalent to an ISP (Internet Service Provider) connection point or a PGW (Packet Data Network GateWay), or an inter-network connection device 42 arranged at a relay point equivalent to an edge router or a base station. Alternatively, it is arranged at any of the connection terminals 40 of the user LAN NW3 connected to the carrier network NW2, for example, an inter-network connection device 40 arranged at a relay point corresponding to an HGW (Home GateWay) or M2MGW (Machine to Machine GateWay). As the inter-network connection devices 40, 41, 42, for example, a router, a switch, or a gateway is used.

  In this embodiment, the case of controlling transmission of a packet transmitted from the control server 2 to the user terminal 3 is taken as an example. In this case, the network having the highest effect of reducing fluctuation in the time direction of the arrival time of the packet at the user terminal 3 The case where it arrange | positions in the inter-connection apparatus 40 (P1 in FIG. 2) is demonstrated.

  Next, the configuration of the packet control device 1 will be described. FIG. 3 is a block diagram showing the functional configuration.

  The packet control device 1 includes a central processing unit (CPU) and a memory, and includes a packet receiving unit 11, a packet identifying unit 12, a packet receiving unit 13, a packet arrival time recording unit 14, and a storage unit 15. A packet buffer 16, a thinning-out processing unit 17, a packet transfer unit 18, a discard packet analysis unit 19, and a parameter setting unit 20. Among these, the packet receiving unit 11, the packet identifying unit 12, the packet receiving unit 13, the packet arrival time recording unit 14, the thinning processing unit 17, the packet transfer unit 18, the discarded packet analysis unit 19, and the parameter setting unit 20 are all illustrated. This is realized by causing the CPU to execute a program stored in the program memory not to be executed.

  The packet receiving unit 11 receives a packet transmitted from the carrier network NW2. The packet identification unit 12 identifies a sequence of packets based on the received additional information of the packet, and performs a process of distributing each sequence. This packet identification and distribution is performed based on, for example, 5-tuple inserted in the packet header, that is, the transmission source IP address, transmission destination IP address, transmission source port number, transmission destination port number, or protocol number. The packet receiving unit 13 receives the packets distributed by the packet identification unit 12 by series and stores them in the packet buffer 16. In addition, as a packet sequence identification method, another method (Deep Packet Inspection) that uses packet payload information as an identification item is also conceivable.

  The packet arrival time recording unit 14 detects a packet arrival time ta (k) that is a time at which each received packet k arrives for each of the sorted sequences, and stores the detected packet arrival time ta (k). 15 is stored. The storage unit 15 stores in advance a theoretical arrival time TAT calculated based on the packet transmission period of the control server 2, an estimated value I of the packet arrival interval, and an allowable delay amount D. The initial value D0 of the allowable delay amount D is set to D0 = 10 msec by the service provider, for example. The initial value I0 of the estimated value I of the packet arrival interval can be obtained as I0 = T0 / N0 by sampling the number N0 of packets that have arrived during a certain time T0.

The thinning processing unit 17 includes a determination unit 171, a parameter update unit 172, and an output control unit 173. The determination unit 171 has the following processing functions.
(1) Each time the received packet k is stored in the packet buffer 16, the arrival time ta (k), the theoretical arrival time TAT, and the allowable delay amount D of the received packet are read from the storage unit 15. Then, it is determined whether or not the arrival time ta (k) of the received packet is delayed by the allowable delay amount D or more with respect to the theoretical arrival time TAT, and the determination result is given to the output control unit 173.
(2) If the arrival time ta (k) of the received packet is less than or equal to the theoretical arrival time TAT, the theoretical arrival time TAT is updated to the received packet arrival time ta (k), and the updated theoretical arrival time TAT Is stored in the storage unit 15.

  Based on the determination result given from the determination unit 171, the output control unit 173 receives the received packet when the arrival time ta (k) of the received packet is not delayed by the allowable delay amount D or more than the theoretical arrival time TAT. The packet k is output from the packet buffer 16 to the packet transfer unit 18, and when the arrival time ta (k) of the received packet is delayed by the allowable delay amount D or more than the theoretical arrival time TAT, the received packet k is packetized. Discard from the buffer 16. The packet transfer unit 18 transmits the received packet k 1 read from the packet buffer 16 to the user terminal 3.

  The discarded packet analysis unit 19 reads a packet to be discarded from the packet buffer 16 and measures the number of discarded packets. Then, it is determined whether or not there are consecutive discard packets exceeding a certain threshold value, and the determination result is given to the parameter setting unit 20. The parameter setting unit 20 sets parameters of the allowable delay amount D or the estimated value I of the packet arrival interval, or both parameters, when the discard packet analysis unit 19 determines that the discard packets are continuously generated for a certain threshold value or more. The changed parameter is stored in the storage unit 15. The change in the estimated value I of the packet arrival interval is reflected in the theoretical arrival time TAT.

(Operation)
Next, the packet control operation by the apparatus configured as described above will be described with reference to the flowchart shown in FIG.
The storage unit 15 stores in advance the theoretical arrival time TAT based on the packet transmission period by the control server 2, the estimated value I of the packet arrival interval, and the initial values of the allowable delay amount D in advance. To do.

(1) Packet distribution reception When the packet receiving unit 11 receives the packet k 1, first, in step S11, the packet identification unit 12 identifies the sequence of the packet k 1 based on the received header information of the packet k 1. Is done. For example, at least one of a source IP address, a destination IP address, a source port number, a destination port number, and a protocol number is extracted from the header information. Based on the extracted 5-tuple, which source and which Whether the packet sequence is transferred to or from the transmission destination is identified. Then, the received packet is sorted and stored in the packet buffer 16 for each identified sequence by the packet receiving unit 13. This distribution makes it possible to perform a thinning process for each received packet series.
In step S 12, the packet reception time ta (k) is measured by the packet arrival time recording unit 14 at the timing when the packet reception unit 13 receives the packet, and stored in the storage unit 15.

(2) Packet Discard or Transfer When the received packet k is stored in the packet buffer 16, the thinning-out processing unit 17 determines whether to transfer or discard the received packet k as follows. .

That is, the determination unit 171 first reads the arrival time ta (k), the theoretical arrival time TAT, and the allowable delay amount D 1 of the received packet k 1 from the storage unit 15 in step 13 and arrives at the reception time ta (k) of the received packet. Is compared with the theoretical arrival time TAT. As a result of this comparison,
TAT ≥ ta (k)
If the received packet arrival time ta (k) is within the theoretical arrival time TAT, the parameter update unit 172 sets the theoretical arrival time TAT to the received packet arrival time ta (k) in step S15. (TAT = ta (k)), and the theoretical arrival time TAT stored in the storage unit 15 is updated to the changed value. In step S16, under the control of the output control unit 173, the received packet k is read from the packet buffer 16, supplied to the packet transfer unit 18, and transmitted from the packet transfer unit 18 to the user terminal 3. .

On the other hand, as a result of comparison in step 13,
TAT <ta (k)
Is determined. In this case, in step S14,
TAT + D> ta (k)
It is determined whether or not. That is, it is determined whether or not the delay amount with respect to the theoretical arrival time TAT of the arrival time ta (k) of the received packet k is smaller than the allowable delay amount D. If TAT + D> ta (k) as a result of the determination, the delay amount of the received packet k is within the allowable range, and the process proceeds to step S16. Under the control of the output control unit 173, the received packet k 1 is read from the packet buffer 16, supplied to the packet transfer unit 18, and transmitted from the packet transfer unit 18 to the user terminal 3.

On the other hand, as a result of the comparison in step S14,
TAT + D ≤ ta (k)
Suppose that. In this case, since the delay amount of the received packet k exceeds the allowable amount D 1, it is determined that the received packet k is unnecessary. In step S17, the packet k is discarded from the packet buffer 16 under the control of the output control unit 173.

  FIG. 5 shows an example of the packet discard and transfer operations described above. In the figure, it is assumed that the packets c and e of the packets a to g transmitted from the control server 2 arrive at the packet control device 1 after being delayed by the allowable delay amount D or more. In this case, the packets c and e are discarded in the packet control device 1 and are not transferred to the user terminal 3.

(3) Discarded packet analysis and parameter update When the packet is discarded, the discarded packet analyzer 19 operates in step S18, and the discarded packet analyzer 19 counts the number of discarded packets. Then, it is determined whether or not there are consecutive discard packets exceeding a certain threshold value, and the determination result is given to the parameter setting unit 20.

The following example is given as a method for analyzing a discarded packet.
(3-1) Count the number of discarded packets, and if a number of discarded packets occur continuously beyond a certain threshold, add a correction term to the parameter of the allowable delay amount D or the estimated value I of the packet arrival interval, or both Then, correction is performed to prevent the packets from being discarded continuously.

An example of the correction method will be described below. When the number of consecutive packet discards is N, the threshold is Nth, the initial packet arrival interval is I0, and the allowable delay is D0.
N> Nth
Then I = 2I0,
N ≤ Nth
Then, I = I0.

  At this time, the allowable delay amount D is constant at D = D0. A method that dynamically changes D is also assumed.

(3-2) If the degree of thinning is reduced after the addition, conversely, a correction that subtracts the correction term is performed.
(3-3) Repeated addition and subtraction corrections to converge to an optimum state.

  Finally, when the parameter setting unit 20 determines in step S19 that the number of discarded packets continuously exceeds a certain threshold, at least one parameter of the allowable delay amount D and the estimated value I of the packet arrival interval is set. Be changed. And the parameter memorize | stored in the memory | storage part 15 is updated by the value after the said change.

  When an example of the process for one received packet k is completed, in step S20, the estimated arrival time I of the packet arrival interval is added to the theoretical arrival time TAT, and the theoretical arrival time TAT stored in the storage unit 15 is obtained. , Updated to the theoretical arrival time TAT after the above addition.

(Effect of the embodiment)
As described above in detail, in this embodiment, the packet control device 1 is arranged in a router or the like on the packet transmission path between the control server 2 and the user terminal 3, and in this packet control device 1, the packet receiving unit 11 The received packet k is sorted according to the series by the packet identification unit 12 based on the header information, and then the packet arrival time ta (k) of the received packet k is compared with the theoretical arrival time TAT by the thinning processing unit 17. It is determined whether the packet arrival time ta (k) is delayed by a predetermined allowable delay amount D or more than the theoretical arrival time TAT. If the packet arrival time ta (k) is delayed, the received packet k is discarded. The received packet k is transferred to the user terminal.

  Therefore, a packet whose delay amount exceeds the allowable amount D 1 is discarded on the packet transmission path by the packet control device 1 arranged in the router or the like, so that the packet to be discarded is not transferred to the user terminal 3. For this reason, the packet discarding process in the user terminal 3 is not required, thereby reducing the processing load on the user terminal. In addition, since the delayed packet is discarded on the packet transmission path, the amount of communication in the downstream network can be suppressed, thereby suppressing the occurrence of congestion.

  In addition, the discard packet analysis unit 19 counts the number of discarded packets and determines whether or not a discard packet is continuously generated for a certain threshold value or more, and it is determined that a discard packet is continuously generated for a certain threshold value or more. In this case, the allowable delay amount D or the estimated value I of the packet arrival interval (the change of the estimated value I is reflected in the theoretical arrival time TAT) is updated. Therefore, it is possible to set the optimum allowable delay amount D or the estimated value I of the packet arrival interval.

[Other Embodiments]
The present invention is not limited to the above embodiment. For example, as shown in FIG. 6, the point P2 before entering the carrier network NW2, the point P3 immediately after entering the carrier network NW2, the point P4 before exiting from the carrier network NW2 to the access zone, and the carrier network NW2 from the carrier network NW2 to the access zone The packet control device may be arranged at a location P5 immediately after leaving and a location P6 before entering the user LAN NW3.

  Moreover, although the case where packet control is performed on the packet transmitted from the control server 2 to the user terminal 3 has been described as an example in the embodiment, the packet control is performed on the packet transmitted from the user terminal 3 to the control server 2. You may do it.

  Further, in the above embodiment, the case where packet control is performed on a packet transmitted from the control server 2 to the user terminal 3 has been described as an example. However, packet control is performed on a packet transmitted from the user terminal 3 to the control server 2. You may do it.

  Furthermore, in the above embodiment, the packet control apparatus 1 estimates the packet arrival interval as the estimated value I of the packet arrival interval. However, the value acquired from the control server 2 or the value declared by the service provider is used. The administrator of the packet control apparatus 1 may set the packet control apparatus 1.

  In short, the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

  DESCRIPTION OF SYMBOLS 1 ... Packet control apparatus, 2 ... Control server, 3 ... Terminal, 4 ... Network connection apparatus, 11 ... Packet receiving part, 12 ... Packet identification part, 13 ... Packet receiving part, 14 ... Packet arrival time recording part, 15 ... Storage unit, 16 ... packet buffer, 17 ... decimation processing unit, 18 ... packet transfer unit, 19 ... discard packet analysis unit, 20 ... parameter setting unit, 40, 41, 42 ... network connection device, 171 ... determination unit, 172 ... parameter update unit, 173 ... output control unit.

Claims (5)

Arranged on a communication path between a first communication device that transmits packets at a constant transmission cycle and a second communication device that receives packets transmitted from the first communication device via a network A packet control device,
The theoretical arrival time calculated based on the packet transmission period by the first communication device is compared with the actual arrival time of the packet transmitted from the first communication device. A determination means for determining whether or not a delay greater than a preset allowable value from the theoretical arrival time;
The determination unit discards the packet when it is determined that the actual arrival time of the packet is more than the allowable value from the theoretical arrival time, and discards the packet when it is determined that the packet is not delayed. Packet output control means for transferring to the second communication device ;
When the packet output control means calculates the packet discard frequency, determines whether or not the calculated discard frequency is within a preset threshold range, and when it is determined that the discard frequency exceeds the threshold range A packet control apparatus comprising: a determination control unit that variably controls at least one of the theoretical arrival time and the allowable value so that the discard frequency falls within a threshold range . Prior to the determination processing by the determination means, further comprising a classification means for classifying the arrived packet into a plurality of series based on the additional information of the packet arriving at the own device,
The packet control apparatus according to claim 1 , wherein the determination unit performs the determination process for each series classified by the classification unit. Arranged on a communication path between a first communication device that transmits packets at a constant transmission cycle and a second communication device that receives packets transmitted from the first communication device via a network A packet control method executed by a packet control device,
The theoretical arrival time calculated based on the packet transmission period by the first communication device is compared with the actual arrival time of the packet transmitted from the first communication device. A determination process for determining whether or not the theoretical arrival time is delayed by a preset allowable value or more;
The determination process discards the packet when it is determined that the actual arrival time of the packet is delayed more than the allowable value from the theoretical arrival time, and determines that the packet is not delayed when it is determined that the packet is not delayed. A packet output control process for transferring to the second communication device ;
When calculating the packet discard frequency by the packet output control process, determining whether the calculated discard frequency is within a preset threshold range, and when determining that the discard frequency exceeds the threshold range A packet control method comprising: a determination control step of variably controlling at least one of the theoretical arrival time and the allowable value so that the discard frequency falls within a threshold range . Prior to the determination process by the determination process, further comprising a classification process of classifying the arrived packet into a plurality of series based on the additional information of the packet arriving at the own device,
The packet control method according to claim 3, wherein the determination process performs the determination process for each series of packets classified by the classification process. The program which makes the computer with which the said packet control apparatus performs the process which the said each means with which the packet control apparatus of Claim 1 or 2 comprises comprises .