A cache management method based on message forwarding times

Abstract

The application discloses a kind of based on message forwarding times of cache management method, comprising: S1, the message is forwarded in current network times;S2, according to message remaining lifetime and message in network is forwarded times to define message utility value, according to message utility value design message forwarding and discard strategy;S3, first judge whether node is destination node when forwarding, if it is directly transmission carries message and completes this message transmission, if not, enter S4;S4, if the cache occupancy of node is less than the threshold value set, it indicates that there is larger space in the current cache of node, message is sorted in ascending order according to size and then is transmitted.Conversely, message is sorted in ascending order according to utility value and then is transmitted.S5, after message successfully reaches destination node, execute ACK strategy.The application can effectively improve single node cache utilization, so that message transmission and discard strategy are more reasonable, while improving message delivery rate, can effectively reduce network overhead and average delay.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cache management of opportunistic networks, and in particular to a cache management method based on message forwarding times. BACKGROUND

[0002] Due to the characteristics of frequent interruption of communication links and short-distance communication of nodes of the opportunistic network, in order to complete communication, the opportunistic network adopts a "store-carry-forward" routing mode. On the one hand, the node can store and carry the message for a relatively long period of time until a suitable forwarding opportunity appears; in the process of forwarding the message, the opportunistic network usually adopts a multi-copy routing mechanism to improve the performance of the network. The combination of long-term storage and message replication (message multi-copy transmission) is easy to exhaust the limited storage resources of the node. At the same time, the limited transmission bandwidth of the node and the strong mobility of the node make the transmission data limited when the communication opportunity appears. In order to improve the performance of the routing algorithm, therefore, the design of the cache management becomes a key field worth exploring in the opportunistic network.

[0003] In the flooding routing algorithm, by designing the message forwarding and discarding mechanism, a relatively balanced forwarding opportunity is obtained for each message, which can effectively improve the delivery rate of the message and the fairness of the message distribution.

[0004] The cache management technology includes two aspects: message forwarding mechanism and discarding mechanism. The combination of long-term storage and message replication of the opportunistic network is easy to exhaust the limited storage resources of the node, and the limited transmission bandwidth and the strong mobility of the node make the data limited when the communication connection is established. Therefore, the design of an efficient cache management mechanism becomes the key to the performance of the routing algorithm. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a cache management method based on message forwarding times in view of the defects in the prior art.

[0006] The technical scheme adopted by the present application to solve the technical problem is:

[0007] The present application provides a cache management method based on message forwarding times, in which the propagation range of the message in the network is dynamically estimated by calculating the message utility value, and the message forwarding and discarding strategy is designed according to the message utility value, and the higher the message utility value, the smaller the message diffusion range. The method comprises the following steps:

[0008] S1, counting the number of times the message is forwarded in the opportunistic network;

[0009] S2, defining the message utility value according to the remaining survival time of the message and the number of times the message is forwarded in the network, and designing the message forwarding and discarding strategy according to the message utility value;

[0010] S3, judging whether the node is the destination node before the message is forwarded, if yes, executing step S5, if not, executing step S4;

[0011] S4, judging whether the cache occupancy rate of the current node is higher than the set threshold before the message is forwarded, if yes, the node preferentially transmits the message with high utility value to expand its diffusion range in the network, thereby improving the message delivery rate in the network, otherwise, the messages are sorted in ascending order according to the message size and then forwarded;

[0012] S5, executing the ACK strategy, when the message reaches the destination node, the destination node will create a confirmation message corresponding to the message, and the confirmation message is diffused to the entire network, so that the node carrying the message deletes the cached message in the local after receiving the confirmation message;

[0013] S6, the message forwarding is ended.

[0014] Further, the method for counting the number of times of message forwarding in the opportunistic network in step S1 of the present application is:

[0015] The number of generated copies is increased by a field FCm in the header of the message k , and the initial value is 0, indicating that the message is created but has not been successfully forwarded; when the message is successfully forwarded once, a message copy is generated, and the FCm k value is increased by 1; the number of times of message forwarding is represented by the FCm k value; k k

[0016] At t1, node S creates message m k , and the FCm k value of the message is initialized to 0; at this time, node A and node C enter the communication range of node S in succession, node S first transmits the message m k to node A, and the FCm k value of node A is set to 1; when node S transmits the message m k to node C, the FCm k value of node C is set to 2; at t2, node A and node C establish a communication connection, and node A selects node C as the relay node, since the FCm k value of node A is less than the FCm k value of node C, the FCm k value of node A is set to 2; at t3, when node A encounters the relay node F, the FCm k value of the relay node F and the FCm k value of node A are set to 3; at this moment, node C encounters the destination node D, and the message m​​k The message was successfully transmitted to the destination node D, completing the message transmission.

[0017] Furthermore, in step S2 of the present invention, the message utility value Qm k The calculation formula is:

[0018]

[0019] Among them, TTLm k It is message m k Remaining survival time, FCm k This indicates the number of times a message has been forwarded on the network.

[0020] Furthermore, the message forwarding and discarding strategy in step S2 of the present invention is as follows:

[0021] The message utility value reflects the degree of message diffusion in the network. The smaller the diffusion range of a message, the higher its message utility value. When a forwarding opportunity occurs or a node's buffer overflows, the node will calculate the message utility value of each message in the buffer and sort them. The message transmission and discarding order will be designed based on the message utility value.

[0022] Furthermore, in step S4 of the present invention, the node cache occupancy rate R c The calculation formula is:

[0023]

[0024] Among them, M i C represents the size of the buffer occupied by message i. total This represents the total cache capacity of the current node, where n represents the total number of messages in the current node's local cache. The node cache utilization ratio R is used to... c This is used to measure the current cache status of a node. When the node's cache occupancy ratio is close to or equal to 1, it means that when other nodes send new messages, the node will discard the messages in order to receive them.

[0025] The beneficial effects of this invention are: the cache management method based on message forwarding count of this invention designs the message forwarding and discarding mechanism by defining message utility value. The higher the utility value of a message with a small diffusion range in the network, the more balanced the forwarding opportunities of each message are, thus improving the overall performance of the routing algorithm. Attached Figure Description

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0027] Figure 1 This is a flowchart illustrating the message forwarding strategy process of a cache management strategy based on the number of message forwardings in an embodiment of the present invention.

[0028] Figure 2 is a schematic diagram of message forwarding at t1-t3 of an embodiment of the present application. DETAILED DESCRIPTION

[0029] In order to make the objects, technical solutions and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0030] As shown in Figure 1 , the present application is based on a cache management method based on message forwarding times, defines a message utility value in combination with the number of times a message is forwarded in a network and the remaining survival time of the message, dynamically estimates the propagation range of the message in the network by calculating the message utility value, designs a message forwarding and discarding strategy according to the message utility value, and the higher the message utility value of a message with a small message diffusion range. When forwarding a message, the message with a high utility value is forwarded first; when the cache occupancy rate R c of a node is greater than a set threshold R threshold , the messages are forwarded in order of utility value Qm k ; if not, the messages are forwarded in order of message size.

[0031] The method comprises the following steps:

[0032] S1, counting the number of times a message is forwarded in a network;

[0033] S2, defining a message utility value according to the remaining survival time of the message and the number of times the message is forwarded in the network, and designing a message forwarding and discarding strategy according to the message utility value.

[0034] S3, judging whether the node is a destination node before forwarding the message, if yes, executing S5, if not, executing S4.

[0035] S4, judging whether the cache occupancy rate of the current node is higher than a set threshold before forwarding the message, if yes, the node preferentially transmits the message with a high utility value to expand its diffusion range in the network, thereby improving the message delivery rate in the network; otherwise, the messages are sorted in ascending order according to the message size and then forwarded.

[0036] S5, executing an ACK strategy, when the message reaches the destination node, the destination node creates a confirmation message corresponding to the message, diffuses the confirmation message to the entire DTN, so that the node carrying the message m deletes the cached message m in the local after receiving the confirmation message.

[0037] S6, ending the message forwarding.

[0038] The calculation method of the number of times the message is forwarded in the network in step S1 is as follows: for the convenience of calculation, a field FCm k is added in the header of the message m k , and the initial value of FCm k is 0, indicating that the message is created but has not been successfully forwarded. When the message is successfully forwarded once, a message copy is generated, and the FCm k value is increased by 1; when both of the two nodes that meet carry the message m k , the FCm k value of the message m k between the two nodes is updated to the larger one.

[0039] As shown in FIG. 1, at time t1, the node S creates the message m k , and the FCm k value of the message is initialized to 0. At this time, the node A and the node C enter the communication range of the node S successively, the node S first transmits the message m k to the node A, and the FCm k value of the node A is set to 1; when the node S transmits the message m k to the node C, the FCm k value of the node C is set to 2; at time t2, the node A and the node C establish a communication connection, and the node A selects the node C as a relay node. Since the FCm k value of the node A is less than the FCm k value of the node C, the FCm k value of the node A is set to 2; at time t3, when the node A meets the relay node F, the FCm k value of the relay node F and the FCm k value of the node A are set to 3; at this time, the node C meets the destination node D, and the message m k is successfully transmitted to the destination node D, and the message transmission is completed.

[0040] Since each message in the DTN has a life cycle, generally, TTL (Time to Live) is used to represent the remaining survival time of the message, so that the message can be eliminated within a limited time, and network storm is avoided.

[0041] The calculation formula of the message utility value Qm k in step S2 is as follows:

[0042]

[0043] wherein TTLm k is the remaining survival time of the message m k , and FCm c represents the number of times the message is forwarded in the network.

[0044] The message utility value reflects the diffusion degree of the message in the network, and the message utility value of the message with small diffusion range is higher. When the forwarding opportunity occurs or the node buffer overflows, the node calculates the utility value of each message in the buffer and sorts them, and the message transmission and discarding order is designed according to the message utility value.

[0045] The node buffer occupation ratio R in step S4 c measures the current buffer state of the node, and when the node buffer occupation ratio is close to 1 or equal to 1, it means that when the new message is sent by other nodes, the node will generate the discarding of the message for receiving the new message. The formula of the node buffer occupation ratio is:

[0046]

[0047] Wherein, M i represents the size of the occupied buffer of the message i, C total represents the total buffer capacity of the current node, and n represents the total number of messages in the local buffer of the current node.

[0048] In summary, the application discloses a buffer management method based on message utility value, which dynamically estimates the propagation range of the message in the network by calculating the message utility value, and designs the message forwarding and discarding strategy according to the message utility value. The higher the message utility value is, the smaller the message diffusion range is. When forwarding the message, the message with high utility value is forwarded first; when the buffer occupation ratio R c of the node is greater than the set threshold R threshold , the forwarding is sorted according to the utility value Qm k , and if it is less than the message size, the forwarding is sorted according to the message size. The method can improve the message delivery rate, effectively reduce the network overhead and average delay. Compared with the buffer management algorithm (SHLI) of discarding the message with the shortest survival time, the buffer management algorithm (MOFO) of discarding the message with the most forwarding times, the message management mechanism (MPBBM) based on the message survival attribute and the Epidemic algorithm, the method proposed in the application is optimal in terms of message delivery rate, average delay and network overhead.

[0049] The simulation experiment uses the ONE simulation platform, combines the typical flooding algorithm Epidemic, and simulates and compares the method proposed in the application with Epidemic, MPBBM and SHLI. The forwarding order of Epidemic, MPBBM and SHLI all follows the FIFO principle. The simulation platform simulates the Helsinki city map as the node movement range, and the specific parameter settings are shown in Table 1.

[0050] Table 1 Simulation parameters

[0051]

[0052]

[0053] (1)Routing performance under the message lifetime change

[0054] Under the parameter setting of Table 1, the routing performance of each algorithm under the message lifetime of 50min, 100min, 150min, 200min, 250min and 300min is compared.

[0055] With the increase of the message lifetime, the message can exist in the DTN for a longer time, and the Epidemic routing algorithm lacks a cache management strategy, so when the message lifetime increases, the delivery rate of the Epidemic routing algorithm presents a downward trend. The delivery rates of the other three algorithms tend to be stable when the message lifetime is 150min, because the reasonable selection of the forwarding and discarding strategies increases the probability of the message reaching the destination node.

[0056] Because the strategy of the application can reasonably spread the message to the entire network, with the increase of the message lifetime, the number of messages that can be delivered also increases greatly, so that the network overhead is reduced. The MPBBM and SHLI strategies select reasonable discarding strategies, which increases the possibility of high-quality messages reaching the destination node, and also reduces the number of forwarding times, thereby reducing the network overhead of the overall strategy. Overall, the MF-BMS strategy reduces by 65.65% compared with the Epidemic routing algorithm, reduces by 10.04% compared with the MPBBM strategy, and reduces by 17.97% compared with the SHLI strategy.

[0057] The average delay of the strategy of the application is higher than that of the other three strategies, and the reason is not repeated here. Overall, the MF-BMS improves by 10.73% compared with the Epidemic routing algorithm, improves by 10.6% compared with the MPBBM strategy, and improves by 23.14% compared with the SHLI strategy.

[0058] The average number of hops of the Epidemic routing strategy is the highest, and the other three are relatively close. Through analysis, it can be known that the MF-BMS reduces by 67.23% compared with the Epidemic routing algorithm, improves by 11.41% compared with the MPBBM strategy, and improves by 8.84% compared with the SHLI strategy.

[0059] The average number of hops of the strategy of the application is relatively high, because the purpose of the strategy of the application is to make each message spread as evenly as possible, avoid excessive forwarding of some messages to reduce the transmission opportunity of other messages, and does not have the purpose of selecting high-quality relay nodes for transmission, but tries to expand the message diffusion range, so that the number of hops required for the message to reach the target node is relatively large.

[0060] Summarizing the above experiments, compared with the other four algorithms, the strategy of the application exchanges the performance advantages in message delivery rate, average delay, network overhead for the cost of more hop count.

[0061] (2) Routing performance under the change of simulation time

[0062] Under the parameter setting of Table 1, the routing performance of each algorithm under the simulation time of 2h, 6h, 10h, 14h, 18h and 22h is compared.

[0063] The message delivery rate of the strategy of the application is the highest, and the delivery rates of the four algorithms tend to be stable after the simulation is carried out to 10h with the increase of simulation time, because the message arrival at the destination node and message generation in the DTN basically maintain a stable state. In addition, the average delivery rate of the strategy of the application is improved by 50.20% compared with Epidemic routing algorithm, improved by 12.16% compared with MPBBM strategy, and improved by 23.10% compared with SHLI strategy. This is mainly because reasonable forwarding and ACK confirmation mechanism can provide more effective cache space for a single node in the DTN, so that the effective messages carried by the single node increase, while the SHLI and MPBBM strategies only consider how to discard unsuitable messages, and do not fully utilize the cache, so that the delivery rate of the strategy is limited.

[0064] The network load rate of the strategy of the application is the lowest, which is reduced by 109.13% compared with Epidemic routing algorithm, reduced by 11.01% compared with MPBMM strategy, and reduced by 26.74% compared with the average network load rate of SHLI strategy.

[0065] The average delay of the strategy of the application is the highest, which is improved by 8.58% compared with Epidemic routing algorithm, improved by 9.04% compared with MPBBM strategy, and improved by 22.93% compared with SHLI strategy.

[0066] The average hop count of Epidemic strategy is the highest, which is reduced by 70.24% compared with Epidemic routing algorithm, improved by 8.21% compared with MPBBM strategy, and improved by 0.97% compared with SHLI strategy. This phenomenon is similar to the above, and the reason is not described again.

[0067] (3) Routing performance under the change of message generation interval

[0068] Under the parameter setting of Table 1, the routing performance of each algorithm under the message generation interval of 5s, 15s, 25s, 35s and 55s is compared. The message generation interval is inversely proportional to the message generation rate.

[0069] The message delivery rate increases positively with the increase of message generation interval, because with the increase of message generation interval, the message generation rate gradually becomes small, the messages in the network gradually become less, and the occurrence of message discarding due to frequent congestion of the node under the condition of cache limitation is reduced, so the delivery rate is increasing. Among them, the delivery rate of the strategy of the present application is the highest.

[0070] The average delay of the strategy of the present application is also the lowest compared with the other four algorithms, and the reason is not described here.

[0071] The network overhead of the strategy of the present application is the lowest, because when the receiving node buffer overflows, the number of times of receiving the same message and the number of discarded messages by the control node of the present application are controlled, so the network overhead is the smallest.

[0072] The average hop count of the strategy of the present application is the highest. The reason is not described here.

[0073] In summary, under the condition of change of message generation interval, the performance of the strategy of the present application is the best in terms of delivery rate, network overhead and average delay.

[0074] It should be understood that those skilled in the art can make improvements or changes according to the above description, and all these improvements and changes shall belong to the protection scope of the appended claims of the present application.

Claims

1. A cache management method based on message forwarding count, characterized in that, In opportunistic networks, the propagation range of a message is dynamically estimated by calculating its utility value. Message forwarding and discarding strategies are then designed based on this utility value; messages with smaller propagation ranges have higher utility values. This method includes the following steps: S1, the number of times a message is forwarded in an opportunity network; S2. Define the message utility value based on the remaining lifetime of the message and the number of times the message is forwarded in the network, and design message forwarding and discarding strategies based on the message utility value; The message utility value in step S2 Qm k The calculation formula is: in, TTLm k It's a message. m k The remaining survival time, FCm k This indicates the number of times a message has been forwarded on the network; S3. Before forwarding the message, determine whether the node is the destination node. If it is the destination node, proceed to step S5; otherwise, proceed to step S4. S4. Before forwarding a message, determine whether the current node's cache utilization rate is higher than the set threshold. If it is higher, the node prioritizes transmitting messages with higher message utility values ​​to expand their diffusion range in the network, thereby improving the message delivery rate in the network. Otherwise, the messages are forwarded after being sorted in ascending order by message size. S5. Execute the ACK strategy. When the message reaches the destination node, the destination node will create an acknowledgment message corresponding to the message and spread the acknowledgment message throughout the network. This will allow the node carrying the message to delete the cached message in its local memory after receiving the acknowledgment message. S6. Message forwarding ends.

2. The cache management method based on message forwarding count according to claim 1, characterized in that, The method for counting the number of times a message is forwarded in the opportunistic network in step S1 is as follows: To facilitate message calculation m k To record the number of replicas, add a field to the message header to indicate the number of message replicas. FCm k Its initial value is 0, indicating that the message has been created but has not yet been successfully forwarded; once the message is successfully forwarded, a message copy is generated. FCm k Increment the value by 1; pass FCm k Values ​​are used to represent messages m k The number of times it was forwarded; At time t1, node S Create message m k The message FCm k The value is initialized to 0. At this time, node A and node C enter the communication range of node S one after the other. Node S sends the message first. m k When transmitted to A, node A's FCm k The value is set to 1; when node S sends a message m k When transmitted to node C, node C's FCm k The value is set to 2; at time t2, node A and node C Establish a communication connection, and node A selects a node. C As a relay node, due to node A's FCm k Less than node C FCm k The value, at this time, will be the value of node A. FCm k The value is set to 2; at time t3, when node A encounters relay node F, it will transfer the value of relay node F to the relay node. FCm k Value and its own FCm k The value is set to 3; at this moment, node C encounters the destination node D and sends the message. m k The message was successfully transmitted to the destination node D, completing the message transmission.

3. The cache management method based on message forwarding count according to claim 1, characterized in that, The message forwarding and discarding strategy in step S2 is as follows: The message utility value reflects the degree of message diffusion in the network. The smaller the diffusion range of a message, the higher its message utility value. When a forwarding opportunity occurs or a node's buffer overflows, the node will calculate the message utility value of each message in the buffer and sort them. The message transmission and discarding order will be designed based on the message utility value.

4. The cache management method based on message forwarding count according to claim 1, characterized in that, Node cache occupancy rate in step S4 R c The calculation formula is: in, M i This indicates the size of the buffer occupied by message i. C total This indicates the total cache capacity of the current node. n This indicates the total number of messages in the current node's local cache.

Images