TWI804337B - Method to reduce packet transmission delay - Google Patents

Abstract

A method for reducing packet transmission delay, which includes: a preparation step, in a multi-path network, a source device transmits at least one packet to a destination device through multiple paths, and determines a packet according to the quantity of the at least one packet Threshold value, and calculate an error rate according to the packet error rate of the plurality of paths; a judgment step, judge whether the error rate is greater than the threshold value; a first transmission step, when the result of the judgment step is yes, the The packet is copied into multiple copies, and the multiple packets are sent to the destination device through the plurality of paths; and a second transmission step, when the result of the judging step is no, the packet is sent to the destination device through one of the paths destination device. The result of the judgment step is used to dynamically switch the transmission steps of the packet. When interference occurs, the packet is copied into multiple copies for transmission, so as to improve the transmission success rate, reduce the average delay time, and further reduce the transmission delay.

Description

Method to reduce packet transmission delay

The invention relates to a packet transmission method, especially a method capable of reducing packet transmission delay.

Packet is the smallest data unit that can be transmitted in the network. The data on the computer is usually continuous and has a certain capacity. For example, the capacity is 100KiloByte (KB). In the packet switching technology, the data will be divided into multiple data blocks before transmission, and the data blocks will form a packet with the header. , the header (Header) includes the total length of the packet (Total Length), source device IP address (Source IP Address), destination device IP address (Destination IP Address), identification code (Identification), check code (Checksum) and other information . These packets go their own way in the network, and after arriving at the destination device, they are reassembled into the original data according to the identification code. Because the order in which each packet arrives at the destination device may be different from the order at which it departs, the original order of the packets must be judged by the identification code when the packets are reassembled.

In the online world, data is converted into electricity or transmitted as radio waves. Taking radio waves as an example, during the transmission of packets from the source device to the destination device, if a sudden thunderstorm interferes with the transmission medium (air), the data being transmitted will definitely be destroyed, making it impossible for the destination device to receive correct packets. The so-called packet error rate means the number of incorrect packets received divided by the total number of packets received, that is to say, the total number of packets received by the destination device The proportion of incorrect packets in the traffic.

Because the wireless network (Wireless Network) has the characteristics of sharing the transmission medium, when the packets are transmitted in the wireless network, it is easy to interfere with each other. This is because the transmission medium of the wireless network is air, so in theory, as long as the You can connect to the wireless network within the range covered by the wireless network. Take Wireless Fidelity (Wi-Fi) 2.4Ghz as an example, the 2.4GHz frequency band operates in the frequency range of 2.412GHz~2.484GHz. It may operate in the 2.4GHz frequency band, so using the Wi-Fi 2.4GHz frequency band to transmit packets may overlap with other signals (such as Bluetooth) and cause interference, so that the destination device cannot receive correct packets.

After the destination device receives the packet, it will judge whether the packet is correct according to the check code in the header. Once the destination device finds that the check code does not match, it means that the packet has been damaged and will discard the packet, and force the source device to resend the packet to maintain data correctness and integrity, but retransmitting packets will cause transmission delays.

Although, the packet error rate of the wireless network will often change. For example, near the microwave oven, the user uses the Wi-Fi 2.4GHz frequency band of the mobile phone to transmit the packet. Because of the electromagnetic wave interference of the microwave oven, the packet error rate will increase, and it is not near the microwave oven. , the packet error rate will decrease.

However, if only a single path is used to transmit packets, as long as there is interference, there will be a delay in transmission. Once in a state of interference for a long time, it will take a long time to wait. Telephone or video conferencing, etc., greatly affect the user experience. For example, when using the Internet phone, the call is dropped for a long time (Dropped Call) or the sound is bad during the call, that is, there is noise during the call, or the sound is intermittent.

Since not all sources of interference can be easily removed, multipath (Referring to a plurality of paths) to transmit the packet can more effectively improve the situation that the transmission delay will be caused by interference. That is, when one of the paths is disturbed, packets can also be transmitted through the other paths. For example, some mobile phones support the function of dual-band Wi-Fi, which can use the 2.4GHz frequency band and the 5GHz frequency band simultaneously. When the 2.4GHz frequency band is interfered, the packets can still be transmitted through the 5GHz frequency band.

Generally speaking, when using multi-path to transmit packets, different paths are alternately selected to transmit packets to avoid congestion.

Conventional technologies such as the "route selection method for mixed plural heterogeneous networks" of the Republic of China Patent No. I492584 or the "connection method for analyzing data packets to select the connection path" of the Republic of China Patent No. I481230 are all based on the selection of the most Packets are transmitted over the best path.

In conventional technologies, when it is difficult to select an optimal path, a preset path is usually used to transmit packets. However, when multiple paths are interfered, if only one path is used to transmit the packet, it is still very easy for the destination device to fail to receive the correct packet, resulting in the need to resend the packet, resulting in transmission delay, so the optimal path is selected The way to transmit packets has its limitations.

In view of this, how to get rid of the above deficiency is the technical difficulty that the inventor of this case wants to solve.

In view of the above, the purpose of the present invention is to solve and improve the existing problems and deficiencies in the conventional practice.

To achieve the above object, the present invention provides a method for reducing packet transmission delay, which includes: a preparation step, in a multi-path network, at least one packet is sent to a destination device by a source device through multiple paths, and based on the number of at least one packet A threshold value, and calculate an error rate according to the packet error rate of the plurality of paths; a judgment step, judge whether the error rate is greater than the threshold value; a first transmission step, when the result of the judgment step is yes, send The packet is copied into multiple copies, and the multiple packets are sent to the destination device through the plurality of paths; and a second transmission step, when the result of the judging step is no, the packet is sent to the destination device through one of the paths the target device.

Wherein, after the first transmission step, a confirmation step is included, and the confirmation step is to confirm whether a copy of the packet is successfully transmitted, and after the confirmation step, a deletion step is included, and the deletion step is when the result of the confirmation step is If yes, delete redundant packets of the complex path.

Wherein, after the first transmission step, a confirmation step is included, the confirmation step is to confirm whether a copy of the packet is successfully transmitted, and after the confirmation step, a retransmission step is included, and the retransmission step is the result of the confirmation step If no, retransmit the plurality of packets to the destination device through the plurality of paths.

Wherein, the multipath network is a homogeneous network or a heterogeneous network.

Wherein, the multipath network is a heterogeneous network, and the heterogeneous network is a combination of a wired network and a wireless network.

Wherein, in the preparation step, the packet is an Internet Protocol version 4 packet or an Internet Protocol version 6 packet.

Wherein, in the preparation step, the threshold value is different according to the quantity of the packets.

Wherein, in the preparation step, the plurality of paths are detected at any time, and when the packet error rate of the plurality of paths changes, the error rate is recalculated.

，其中N為該封包的數量。 Among them, the calculation formula of the threshold value is

, where N is the number of packets.

Wherein, in the preparation step, the packet error rate of the plurality of paths is weighted to generate the error rate.

The result of the judgment step is used to dynamically switch the transmission steps of the packet. When interference occurs, the packet is copied into multiple copies for transmission, so that the destination device has a higher chance of receiving the correct packet to improve transmission. The success rate reduces the average delay time, thereby reducing the transmission delay.

S11: Preparatory steps

S12: Judgment step

S13: first transfer step

S14: second transmission step

S15: confirmation step

S16: delete step

S17: retransmission step

21: Source device

22: Purpose device

23: The first packet

24: Second packet

25: Packet error

L: path

[FIG. 1] is a flow chart of the method for reducing packet transmission delay in the present invention.

[Fig. 2] is a flowchart of the improved method of the present invention.

[FIG. 3] is a schematic diagram of the present invention when no packet error occurs and the error rate is greater than the threshold.

[FIG. 4] is a schematic diagram of the present invention when no packet error occurs and the error rate is less than the threshold value.

[FIG. 5] is a schematic diagram of the present invention when a packet error occurs and the error rate is greater than the threshold.

[FIG. 6] is a schematic diagram of the improved present invention when a packet error occurs and the error rate is greater than the threshold value.

[FIG. 7] is a schematic diagram of deleting redundant packets after improvement of the present invention.

In order for your examiner to easily and concisely understand the other features and advantages of the present invention and the effects achieved, the present invention is described in detail as follows with accompanying drawings:

1 and 3, the present invention provides a method for reducing packet transmission delay, which includes: a preparation step S11, a judgment step S12, a first transmission step S13 and a second transmission step S14.

The preparation step S11 is in a multi-path network, a source device 21 transmits at least one packet to a destination device 22 through a plurality of paths L, and a threshold value is determined according to the quantity of the at least one packet, and according to the plurality A packet error rate (Packet Error Rate, PER) of the path L is used to calculate an error rate.

Specifically, the multipath network can be a homogeneous network or a heterogeneous network. A homogeneous network refers to a network composed of the same communication protocol, and a heterogeneous network refers to a network composed of different communication protocols. A network with two different communication protocols can be called a heterogeneous network, such as Bluetooth ( IEEE802.15) and Wi-Fi (IEEE802.11) are two different communication protocols. The heterogeneous network can be a combination of a wired network and a wireless network (Wireless Network), and the wired network can be Ethernet, Coaxial Cable, Power Line Communication, PLC), the wireless network can be Wireless-Fidelity (Wi-Fi) or fourth-generation mobile communication (4GLTE). The source device 21 and the destination device 22 can be hardware devices such as computers, smart phones, tablet computers or notebook computers that use packet switching technology to transmit data. The packet can be an Internet Protocol Version 4 (IPV4) packet or an Internet Protocol Version 6 (IPV6) packet. Taking the wireless network (Wi-Fi) as an example, the complex path L can be 2.4GHz frequency band (2.412GHz~2.484GHz) of WI-FI and 5GHz frequency band (4.915GHz~5.865GMz) of WI-FI.

When the path is disturbed, the packet will be damaged or lost, making the destination device unable to receive the correct packet. Therefore, the destination device will require the source device to resend the packet in order to maintain the correctness and integrity of the data. The packet error rate means that the received packets The incorrect count is divided by the total count of packets received. That is to say, the higher the packet error rate of the path L, the higher the chance that the packet is transmitted to the destination device 22 through the path L, the destination device 22 cannot receive the correct packet, and the packet must be retransmitted, because Retransmission will increase the transmission time, thereby increasing the transmission delay. Therefore, the higher the packet error rate of the path L, the easier it is to cause transmission delay, which will greatly increase the transmission time.

Since the interference source may change (increase, decrease or disappear) with time, the packet error rate of the path L may also change with time. In the preparation step S11, the multiple paths L can be detected at any time, and when the packet error rate of the multiple paths L changes, the error rate is recalculated.

The threshold value is used as a basis for dynamically switching transmission steps, and the threshold value varies according to the number of packets. Please refer to Table 1, the threshold value can be preset according to the experimental results, the threshold value is:

[Table 1]

Because when the path L has a higher packet error rate, it will have a greater impact on the overall efficiency of the complex path L, so in the preparation step S11, the packet error rate of the complex path L is weighted to generate The error rate (the weighted average is used as the error rate). For example, in the complex path L, the highest packet error rate has a weight of 3, and the other lower packet error rates The weighted value of the error rate is 1, and the weighted value can be determined according to the requirement. The calculation example of the error rate is as follows:

Example 1: The packet error rates of the two paths L are 50% and 80% respectively, the error rates are: ((50%×1)+(80%×3))/(1+3)=72.5%.

Example 2: The packet error rates of the two paths L are 20% and 30% respectively, and the error rates are: ((20%×1)+(30%×3))/(1+3)=27.5%.

Example 3: The packet error rates of the three paths L are 50%, 60% and 80% respectively. The error rates are: ((50%×1)+(60%×1)+(80%×3))/ (1+1+3)=70%.

The judging step S12 is judging whether the error rate is greater than the threshold.

The first transmitting step S13 is to copy the packet into multiple copies when the result of the judging step S12 is yes, and transmit the multiple packets to the destination device 22 through the plurality of paths L.

The second transmitting step S14 is to transmit the packet to the destination device 22 through one of the paths L when the result of the judging step is negative.

Please refer to Fig. 1 and shown in Fig. 3, be two (the first packet 23, the second packet 24) and path quantity be two (the packet error rates of these two paths L are respectively 50% and 80% with the number of packets) ) as an example, the preparation step S11 is that the source device 21 transmits the first packet 23 and the second packet 24 to the destination device 22 through the two paths L. As shown in Table 1, the number of packets is two, and the threshold value is 66%. As shown in Example 1, the packet error rates of the two paths L are 50% and 80%, respectively, and the error rate is 72.5%. The judging step S12 is judging whether the error rate (72.5%) is greater than the threshold (66%). The first transmission step S13 is when the result of the judging step S12 is yes (72.5% is greater than 66%), the first packet 23 is copied into multiple copies, and the multiple second packets are passed through the two paths L A packet 23 is sent to the destination device 22 . Afterwards, the packet error rates of the two paths L do not change, or Although the changed error rate is still greater than the threshold, the second packet 24 is copied into multiple copies, and the multiple copies of the second packet 24 are sent to the destination device 22 through the two paths L.

This is because when the error rate is greater than the threshold value, it means that the overall packet error rate of the two paths L is high, and the destination device 22 is less likely to receive the correct packet, so the first packet 23 and the second packet The second packet 24 is copied into multiple copies, and the copied packets are sent to the destination device 22 through the two paths L, so that the destination device 22 has a higher chance to receive the correct packet, so as to improve transmission Success rate, avoiding the need to retransmit due to packet errors, reducing the average delay time, thereby reducing transmission delay.

Please refer to Fig. 1 and shown in Fig. 4, be two (the first packet 23, the second packet 24) and path quantity be two (the packet error rates of these two paths L are respectively 20% and 30% with the number of packets) ) as an example, the preparation step S11 is that the source device 21 transmits the first packet 23 and the second packet 24 to the destination device 22 through the two paths L. As shown in Table 1, the number of packets is two, and the threshold value is 66%. As shown in Example 2, the packet error rates of the two paths L are 20% and 30%, respectively, and the error rate is 27.5%. The judging step S12 is judging whether the error rate (27.5%) is greater than the threshold (66%). The second transmitting step S14 is to transmit the first packet 23 to the destination device 22 through one of the paths L when the result of the judging step S12 is negative (27.5% is not greater than 66%). Afterwards, the packet error rate of the two paths L does not change, or the error rate after the change is still not greater than the threshold value, and then the first packet 24 is sent to the destination device through one of the paths L twenty two. Generally, the first packet 24 and the second packet 24 are transmitted to the destination device 22 through different paths L to avoid congestion.

This is because when the error rate is less than the threshold value, it means that the complex number path L The lower the overall packet error rate, the easier it is for the destination device 22 to receive the correct packet, so the packet can be directly transmitted through one of the paths L without duplicating the packet.

Please refer to Fig. 1 and shown in Fig. 5, be two (the first packet 23, the second packet 24) and path quantity be two (the packet error rates of these two paths L are respectively 50% and 80% with the number of packets) ) as an example, the preparation step S11 is that the source device 21 transmits the first packet 23 and the second packet 24 to the destination device 22 through the two paths L. As shown in Table 1, the number of packets is two, and the threshold value is 66%. As shown in Example 1, the packet error rates of the two paths L are 50% and 80%, respectively, and the error rate is 72.5%. The judging step S12 is judging whether the error rate (72.5%) is greater than the threshold (66%). The first transmission step S13 is when the result of the judging step S12 is yes (72.5% is greater than 66%), the first packet 23 is copied into multiple copies, and the multiple second packets are passed through the two paths L A packet 23 is sent to the destination device 22 . However, during the process of transmitting the multiple copies of the first packet 23 through the two paths L, a packet error 25 occurs in one path L of the two paths L and must be retransmitted, which increases the transmission time. Ultimately, the overall transmission time depends on the path L that completes the slowest, because a packet error 25 occurs on one of the paths L, so the overall transmission time is three unit times.

Please refer to Fig. 2 and Fig. 6, the method of this case can be further improved as, after the first transmission step S13, include a confirmation step S15, the confirmation step S15 is to confirm whether there is a successful transmission of the packet, and the confirmation After the step S15, a deleting step S16 is included. The deleting step S16 is to delete redundant packets of the plurality of paths L when the result of the confirming step S15 is yes. Furthermore, after the confirmation step S15, a retransmission step S17 is also included. The retransmission step S17 is to retransmit the multiple packets to the destination device 22 through the plurality of paths L when the result of the confirmation step S15 is negative. .

，其中N為該封包的數量。 After the method in this case has been improved, the calculation formula of the threshold value is adjusted to

, where N is the number of packets.

Example 4: The number of packets N is 2, the threshold=-0.0036×1+0.0921×1+0.4083=0.4968=49.68%).

Example 5: The number of packets N is 4, the threshold=-0.0036×4+0.0921×2+0.4083=0.5781=57.81%).

Example 6: The number of packets N is 8, the threshold=-0.0036×9+0.0921×3+0.4083=0.6522=65.22%).

Please refer to Fig. 2, Fig. 6 and shown in Fig. 7 again, be two (the first packet 23, the second packet 24) and the number of paths with the number of packets (the packet error rate of these two paths L is respectively 50 % and 80%) as an example, the preparation step S11 is that the source device 21 transmits the first packet 23 and the second packet 24 to the destination device 22 through the two paths L. As shown in Example 4, the number of packets is two, and the threshold value is 49.68%. As shown in Example 1, the packet error rates of the two paths L are 50% and 80%, respectively, and the error rate is 72.5%. The judging step S12 is judging whether the error rate (72.5%) is greater than the threshold (49.68%). The first transmission step S13 is when the result of the judging step S12 is yes (72.5% is greater than 49.68%), the first packet 23 is copied into multiple copies, and the multiple second packets are passed through the two paths L A packet 23 is sent to the destination device 22 . Although a packet error 25 occurs in one path L of the two paths L during the process of transmitting the multiple copies of the first packet 23 through the two paths L, it is confirmed through the confirming step S15 that the first packet 23 has a copy After the transmission is successful, the redundant (not yet delivered) first packet 23 of the plurality of paths L is deleted through the deleting step S16. Afterwards, the packet error rate of the two paths L does not change, or the error rate after the change is still not greater than the threshold value despite the change, then the second packet 24 is copied into multiple copies, and passed through the two The path L takes the multipart Two packets 24 are sent to the destination device 22 .

This is because it is confirmed that one copy of the packet has been successfully transmitted through the confirming step S15, so there is no need to retransmit, and because the packet has been successfully transmitted, the multiple copies of the packet that are copied are deleted through the deleting step S16, so the overall The transmission time is two unit times, which can effectively further reduce the transmission delay.

In this case, the following methods can be used to further optimize:

1. If the plurality of paths L is two paths, and in the preparation step S11, the packet error rates of the two paths are higher or lower than the threshold value at the same time, the error rate does not need to be calculated, and the judgment can be accelerated.

Example 7: The packet error rates of the two paths L are 60% and 40% respectively, and the threshold value is 66%, because the error rate must be between 40% and 60%, and the result of the error rate must be low is lower than the threshold, so the calculation of the error rate can be omitted.

2. Since the bandwidths of the multiple paths L are not necessarily the same, when considering the amount of packet allocation, the bandwidths of the multiple paths L can be unified through a standardized formula before further evaluation. For the convenience of description, the packet success rate (Packet Success Rate, PSR) is introduced, and PSR=1-PER.

Example 8: The first path: bandwidth (Bandwidth, BW) = 100Mbps, packet error rate (Packet Error Rate, PER) = 20%, effective bandwidth = bandwidth × packet success rate = 100Mbps × (1-20%) =100Mbps×0.8=80Mbps. The second path: BW=50Mbps, PER=60%, effective bandwidth=50Mbps×(1-60%)=50Mbps×0.4=20Mbps. Since the bandwidth of the first path (100Mbps) is different from that of the second path (50Mbps), it is necessary to unify the bandwidths of the first path and the second path through a standardized formula before further evaluation.

In essence, the effective bandwidth will not change, that is, new packet success rate (New_PSR) × new bandwidth (New_BW) = old bandwidth (Old_BW) × old packet success rate (Old_PSR), so it can be standardized Formula: New_PSR=Old_BW×Old_PSR/New_BW, New_PER=1-New_PSR. That is, the bandwidth of the second path can be standardized to 100Mbps using the standardized formula, and New_PSR=50Mbps×(1-60%)/100Mbps=50Mbps×0.4/100Mbps=20Mbps/100Mbps=20%, New_PER=1-20% can be obtained =80%. It means to standardize the bandwidth (BW) of the second path from 50Mbps to 100Mbps, then the packet error rate (PER)=80%. Standardized results: the first path: BW=100Mbps, PER=20%. The second path: BW=100Mbps, PER=80%.

The packet allocation amount of the plurality of paths L is proportional to the effective bandwidth, that is, the number of allocated packets=total packets×effective bandwidth of the path/effective bandwidth of the total path.

Example 9: The first path: BW=100Mbps, PER=20%, effective bandwidth=100Mbps×PSR=100Mbps×(1-20%)=80Mbps. The second path: BW=100Mbps, PER=80%, effective bandwidth is 100Mbps×PSR=(1-80%)=20Mbps. If there are 5 packets to be transmitted, the number of packets assigned to the first path=5×80/(80+20)=4 packets are assigned, the number of packets assigned to the second path=5×20/(80+20 ) = assigned to 1 packet.

Example 10: The third path: BW=100Mbps, PER=90%, effective bandwidth=100Mbps×PSR=100Mbps×(1-90%)=10Mbps. The fourth path: BW=100Mbps, PER=95%, effective bandwidth=100Mbps×PSR=100×(1-95%)=5Mbps. The fifth path: BW=100Mbps, PER=95%, effective bandwidth=100Mbps×PSR=100×(1-95%)=5Mbps. If there are 4 packets to be transmitted, the number of packets assigned to the third path=4×10/(10+5+5)=2 packets are assigned, the number of packets assigned to the fourth path=4×5/(10 +5+5)=1 packet is allocated, the number of packets allocated to the fifth path=4×5/(10+5+5)=1 packet is allocated.

To sum up, in this way, under the multi-path network, according to the comparison result of the error rate and the threshold value to confirm the interference situation, the transmission steps of the packet are dynamically switched, and when the interference occurs, the packet is copied The transmission is performed in multiple parts, so that the destination device 22 has a higher chance to receive the correct packet, so as to improve the transmission success rate, reduce the average delay time, and further reduce the transmission delay. In addition, this case can be further improved. After the first transmission step S13, a confirmation step S15, the deletion step S16 and the retransmission step S17 are included to confirm whether the copied packet is in one of the paths The transmission on L is successful, so as to delete the redundant (not yet delivered) packet on the other path L, which can effectively further reduce the transmission delay.

The above discussion is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention; therefore, equivalent modifications, combinations, replacements or transfers, etc., are made without departing from the spirit and scope of the present invention. , should be covered within the scope of the patent application of the present invention.

S11: Preparatory steps

S12: Judgment step

S13: first transfer step

S14: second transmission step

Claims (9)

A method for reducing packet transmission delay, which includes: a preparation step, in a multi-path network, a source device transmits at least one packet to a destination device through multiple paths, and determines a packet according to the quantity of the at least one packet threshold value, and calculate an error rate according to the packet error rate of the plurality of paths, wherein the calculation formula of the threshold value is

, where N is the quantity of the packet; a judging step, judging whether the error rate is greater than the threshold; a first transmission step, when the result of the judging step is yes, copying the packet into multiple copies, and passing through the A plurality of paths transmit the plurality of packets to the destination device; and a second transmission step, when the result of the judging step is negative, transmit the packet to the destination device through one of the paths. The method for reducing packet transmission delay as described in claim 1, wherein after the first transmission step, a confirmation step is included, the confirmation step is to confirm whether a copy of the packet is successfully transmitted, and after the confirmation step, a deletion is included step, the deleting step is when the result of the confirming step is yes, deleting redundant packets of the plurality of paths. The method for reducing packet transmission delay as described in Claim 1, wherein after the first transmission step, a confirmation step is included, the confirmation step is to confirm whether a copy of the packet is successfully transmitted, and after the confirmation step, A retransmission step is included, and the retransmission step is to retransmit the plurality of packets to the destination device through the plurality of paths when the result of the confirmation step is negative. The method for reducing packet transmission delay as claimed in claim 1, wherein the multipath network is a homogeneous network or a heterogeneous network. The method for reducing packet transmission delay as claimed in claim 1, wherein the multipath network is a heterogeneous network, and the heterogeneous network is a combination of a wired network and a wireless network. The method for reducing packet transmission delay as claimed in claim 1, wherein in the preparation step, the packet is an Internet Protocol version 4 packet or an Internet Protocol version 6 packet. The method for reducing packet transmission delay as claimed in claim 1, wherein in the preparation step, the threshold value is different according to the number of packets. The method for reducing packet transmission delay as described in Claim 1, wherein in the preparation step, the multiple paths are detected at any time, and when the packet error rate of the multiple paths changes, the error rate is recalculated. The method for reducing packet transmission delay as claimed in claim 1, wherein in the preparation step, the packet error rate of the plurality of paths is weighted to generate the error rate.

Images