Heterogeneous network convergence processing method and device based on edge computing and storage medium

Abstract

The invention provides a heterogeneous network convergence processing method and device based on edge computing and a storage medium, and the method comprises the steps: fusing each target heterogeneous network with an internet protocol version 6 (IPv6) network, and constructing a plurality of target protocol conversion gateways corresponding to the target heterogeneous networks and the IPv6 network; according to the plurality of target protocol conversion gateways, establishing an edge cache heterogeneous network convergence system and determining a corresponding edge cache strategy and an edge cache operation mechanism; according to an edge caching strategy, obtaining a trained edge calculation model; and in each preset time slot, according to the edge cache heterogeneous network convergence system and the edge cache operation mechanism, inputting the request content of the target user in each target heterogeneous network to the edge computing model, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content. The method is used for achieving the technical effect of improving the heterogeneous network convergence and management efficiency.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a heterogeneous network fusion processing method, device and storage medium based on edge computing. Background Art

[0002] In recent years, with the annual growth of civil aviation passenger transport volume and the continuous expansion of civil aviation scale, the intelligent control of airport surface operations has become the core of improving airport operation safety and efficiency. However, the perception and analysis of aircraft, vehicles, personnel and business in airport surface operations are still not in-depth enough, there are information islands between operating entities, the active coordination ability between activity targets is weak, and the intelligence level of global decision-making is insufficient. The interoperability and high-speed transmission of airport multi-subject heterogeneous networks, as an important basis for the transmission of massive monitoring information and aviation business operation control at airport surfaces, heterogeneous network fusion processing has become a direction with application prospects.

[0003] In the prior art, the heterogeneous network fusion processing method mainly stores data information in a dedicated server on the ground, and uses the dedicated server to exchange and process information for multiple heterogeneous networks.

[0004] In the prior art, the path between data information and the dedicated server is long, resulting in a delay in accessing information, which reduces the communication efficiency between networks and presents a technical problem of low efficiency in heterogeneous network fusion processing. Summary of the invention

[0005] The present application provides a heterogeneous network fusion processing method, device and storage medium based on edge computing, so as to achieve the technical effect of improving the efficiency of heterogeneous network fusion processing.

[0006] In a first aspect, the present application provides a heterogeneous network fusion processing method based on edge computing, comprising:

[0007] Each target heterogeneous network is integrated with the Internet Protocol Version 6 (IPv6) network to construct multiple target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network;

[0008] According to multiple target protocol conversion gateways, an edge cache heterogeneous network fusion system is established and the corresponding edge cache strategy and edge cache operation mechanism are determined;

[0009] According to the edge cache strategy, obtain the trained edge computing model;

[0010] At each preset time slot, according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, the request content of the target user in each target heterogeneous network is input into the edge computing model, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content.

[0011] Optionally, the target heterogeneous network includes an aeronautical telecommunication network network based on the Internet Protocol Suite, an aeronautical telecommunication network network based on the Open System Interconnection Reference Model, and an airport surface broadband mobile communication system network, and the target protocol conversion gateway includes a first protocol conversion gateway, a second protocol conversion gateway, and a third protocol conversion gateway, and each target heterogeneous network is respectively integrated with an Internet Protocol version 6 (IPv6) network to construct a plurality of target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network, including:

[0012] The aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network are set as the first target subnet, and the corresponding target security protocol and target routing protocol are configured according to the aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network; wherein the first target subnet is the IPv6 subnet;

[0013] Constructing a first protocol conversion gateway according to the first target subnet, the target security protocol and the target routing protocol;

[0014] According to the requirements between the aviation telecommunication network and the IPv6 network based on the open system interconnection reference model, multiple target function modules are established, and according to the target function modules, a second protocol conversion gateway is constructed, wherein the target function modules include a packet receiving and sending module, a network layer address mapping module, a network layer message header conversion module and a transport layer message conversion module;

[0015] The airport broadband mobile communication system network and the IPv6 network are set as the second target subnet; wherein the second target subnet is the IPv6 subnet;

[0016] According to the target protocol authentication method, user authentication is performed, and after the user authentication is passed, business resources are accessed according to the Internet address of the business server, and IPv6 data transmission is performed to build a third protocol conversion gateway.

[0017] Optionally, according to the requirements between the aeronautical telecommunication network and the IPv6 network based on the open system interconnection reference model, a plurality of target function modules are established, and according to the target function modules, a second protocol conversion gateway is constructed, including:

[0018] According to the requirements of processing network packets between the aeronautical telecommunication network and the IPv6 network based on the open system interconnection reference model, a packet receiving and sending module is established;

[0019] Establish a network layer address mapping module based on the network layer address mapping requirements between the network service access point address in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 address in the IPv6 network;

[0020] According to the information conversion requirements between the connectionless network protocol message header in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network in the IPv6 message header, a network layer message header conversion module is established;

[0021] According to the conversion requirements between the error detection and recovery protocols of the transport layer in the aeronautical telecommunication network based on the open system interconnection reference model and the transmission control protocol of the transport layer in the IPv6 network, a transport layer message conversion module is established;

[0022] A second protocol conversion gateway is constructed according to the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module.

[0023] Optionally, an edge cache heterogeneous network fusion system is established based on multiple target protocol conversion gateways, including:

[0024] Determine the target router corresponding to each target protocol gateway and the target user corresponding to each target heterogeneous network;

[0025] For each target heterogeneous network, according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, a data connection is established between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, wherein the data connection includes a physical connection or a logical connection;

[0026] According to the information transmission requirements between each target protocol conversion gateway, a corresponding physical connection is established between any two target protocol conversion gateways;

[0027] According to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways and the target router, an edge cache heterogeneous network fusion system is constructed.

[0028] Optionally, determining a corresponding edge cache strategy includes:

[0029] A reinforcement learning agent corresponding to each target protocol conversion gateway;

[0030] According to the edge cache heterogeneous network fusion system, determine the request content of the target user, the entire cache content of the target protocol conversion gateway, and the number of times the request content hits the target protocol conversion gateway;

[0031] Determine the content popularity of the current preset time slot based on the ratio of the requested content to the total cached content of the target protocol conversion gateway;

[0032] Based on the content popularity, all cached contents of the target protocol conversion gateway, and the number of times the requested content hits the target protocol conversion gateway, each reinforcement learning agent is modeled, and the corresponding model state, action, and reward are determined to obtain the edge computing model corresponding to each reinforcement learning agent. The model state, action, and reward are determined in the following way:

[0033] The environment status observed by the target protocol conversion gateway is:

[0034]

[0035] in, represents the environmental state observed by the target protocol conversion gateway g at the preset time slot t, represents the content popularity of the target protocol conversion gateway g in time slot t, represents the cache content of the target protocol conversion gateway g in time slot t;

[0036] After determining the state of the environment, the reinforcement learning agent in the gateway is converted according to the target protocol to determine the model action in the cache update phase Model actions include updating the cache content to obtain the cache content for the next preset time slot Each content in the target protocol conversion gateway has a corresponding index {1, 2, 3, ..., M}, and the content includes cached content and uncached content;

[0037] Before the cache is updated, the popularity of cached content and non-cached content is calculated based on the frequency of requests. Sort in descending order, This means deleting cached content. The lowest-ranked k contents in the cached content are obtained, and the highest-ranked k contents in the uncached content are obtained for supplementation;

[0038] The model reward corresponding to the edge computing model is defined as:

[0039]

[0040] in, Indicates that the edge cache gateway g is in state Take Action The rewards you get, represents the number of times the content requested by the target user through the i-th path hits the gateway cache in the t+1 time slot, w i is the weight related to the delay reduction on the i-th path, w i Obtained by the following formula:

[0041]

[0042] Among them, wi represents the weight related to the delay reduction on the i-th path, Represents the delay reduction caused by hitting the gateway cache on the i-th path;

[0043] The model state, model action, and model reward determined in the edge computing model are used as the edge caching strategy.

[0044] Optionally, determining a corresponding edge cache operation mechanism includes:

[0045] According to each preset time slot, a content distribution phase and a cache update phase are set in the edge cache heterogeneous network fusion system;

[0046] In the content distribution stage, according to the edge cache heterogeneous network fusion system, the request content of the target user in each target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network are determined, and it is judged whether the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user;

[0047] If yes, then obtain the target cache content from the target protocol conversion gateway; if no, then obtain the target cache content from other protocol conversion gateways, wherein the target cache content is the cache content corresponding to the request content in the target protocol conversion gateway, and the other protocol conversion gateway is any target protocol conversion gateway except the target protocol conversion gateway corresponding to the target heterogeneous network;

[0048] In the cache update phase, the cache content in each target protocol conversion gateway is updated at each preset time slot according to the edge cache strategy;

[0049] Determine the edge cache operation mechanism based on the content distribution phase and cache update phase.

[0050] Optionally, according to the edge cache strategy, a trained edge computing model is obtained, including:

[0051] After the edge computing model is established, the corresponding evaluation network loss function is determined according to the DQN algorithm. The evaluation network loss function is obtained by the following formula:

[0052]

[0053] TargetQ=r i +γmax a Q target (s i+1 ,a;θ - )

[0054] Among them, Loss(θ) represents the evaluation network loss function, E represents the mean, and s irepresents the model environment in time slot i, a i represents the model action in time slot i, γ i represents the discount rate in time slot i, s i+1 represents the model environment in the (i+1) time slot, B represents the number of small batch samples in the experience replay pool, and θ represents the evaluation network Q eval Parameter, θ - Represents the target network Q target Parameters, evaluation network Q eval Update the parameters by the following formula:

[0055] θ t+1 ←θ t -l*▽ θ Loss(θ)

[0056] Among them, θ t+1 Represents the evaluation network Q at time slot t+1 eval The parameter θ t Represents the evaluation network Q in time slot t eval Parameters, l represents the learning rate;

[0057] According to the evaluation network loss function, the edge computing model corresponding to the edge caching strategy is trained in each preset time slot to obtain the trained edge computing model.

[0058] In a second aspect, the present application provides a heterogeneous network fusion processing device based on edge computing, including:

[0059] A construction module is used to merge each target heterogeneous network with an Internet Protocol version 6 (IPv6) network, and to construct a plurality of target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network;

[0060] A first processing module is used to establish an edge cache heterogeneous network fusion system and determine a corresponding edge cache strategy and edge cache operation mechanism according to multiple target protocol conversion gateways;

[0061] The second processing module is used to obtain the trained edge computing model according to the edge cache strategy;

[0062] The control module is used to input the request content of the target user in each target heterogeneous network into the edge computing model at each preset time slot according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content.

[0063] Optionally, the building blocks are also used to:

[0064] The aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network are set as the first target subnet, and the corresponding target security protocol and target routing protocol are configured according to the aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network; wherein the first target subnet is the IPv6 subnet, and each network corresponds to an IPv6 subnet;

[0065] Constructing a first protocol conversion gateway according to the first target subnet, the target security protocol and the target routing protocol;

[0066] According to the requirements between the aviation telecommunication network and the IPv6 network based on the open system interconnection reference model, multiple target function modules are established, and according to the target function modules, a second protocol conversion gateway is constructed, wherein the target function modules include a packet receiving and sending module, a network layer address mapping module, a network layer message header conversion module and a transport layer message conversion module;

[0067] The airport broadband mobile communication system network and the IPv6 network are set as the second target subnet; wherein the second target subnet is an IPv6 subnet, and each network corresponds to an IPv6 subnet;

[0068] According to the target protocol authentication method, user authentication is performed, and after the user authentication is passed, business resources are accessed according to the Internet address of the business server, and IPv6 data transmission is performed to build a third protocol conversion gateway.

[0069] Optionally, the building blocks are also used to:

[0070] According to the requirements of processing network packets between the aeronautical telecommunication network and the IPv6 network based on the open system interconnection reference model, a packet receiving and sending module is established;

[0071] Establish a network layer address mapping module based on the network layer address mapping requirements between the network service access point address in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 address in the IPv6 network;

[0072] According to the information conversion requirements between the connectionless network protocol message header in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network in the IPv6 message header, a network layer message header conversion module is established;

[0073] According to the conversion requirements between the error detection and recovery protocols of the transport layer in the aeronautical telecommunication network based on the open system interconnection reference model and the transmission control protocol of the transport layer in the IPv6 network, a transport layer message conversion module is established;

[0074] A second protocol conversion gateway is constructed according to the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module.

[0075] Optionally, the first processing module is further used for:

[0076] Determine the target router corresponding to each target protocol gateway and the target user corresponding to each target heterogeneous network;

[0077] For each target heterogeneous network, according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, a data connection is established between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, wherein the data connection includes a physical connection or a logical connection;

[0078] According to the information transmission requirements between each target protocol conversion gateway, a corresponding physical connection is established between any two target protocol conversion gateways;

[0079] According to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways and the target router, an edge cache heterogeneous network fusion system is constructed.

[0080] Optionally, the first processing module is further used for:

[0081] A reinforcement learning agent corresponding to each target protocol conversion gateway;

[0082] According to the edge cache heterogeneous network fusion system, determine the request content of the target user, the entire cache content of the target protocol conversion gateway, and the number of times the request content hits the target protocol conversion gateway;

[0083] Determine the content popularity of the current preset time slot based on the ratio of the requested content to the total cached content of the target protocol conversion gateway;

[0084] Based on the content popularity, all cached contents of the target protocol conversion gateway, and the number of times the requested content hits the target protocol conversion gateway, each reinforcement learning agent is modeled, and the corresponding model state, action, and reward are determined to obtain the edge computing model corresponding to each reinforcement learning agent. The model state, action, and reward are determined in the following way:

[0085] The environment status observed by the target protocol conversion gateway is:

[0086]

[0087] in, represents the environmental state observed by the target protocol conversion gateway g at the preset time slot t, represents the content popularity of the target protocol conversion gateway g in time slot t, represents the cache content of the target protocol conversion gateway g in time slot t;

[0088] After determining the state of the environment, the reinforcement learning agent in the gateway is converted according to the target protocol to determine the model action in the cache update phase Model actions include updating the cache content to obtain the cache content for the next preset time slot Each content in the target protocol conversion gateway has a corresponding index {1, 2, 3, ..., M}, and the content includes cached content and uncached content;

[0089] Before the cache is updated, the popularity of cached content and non-cached content is calculated based on the frequency of requests. Sort in descending order, This means deleting cached content. The lowest-ranked k contents in the cached content are obtained, and the highest-ranked k contents in the uncached content are obtained for supplementation;

[0090] The model reward corresponding to the edge computing model is defined as:

[0091]

[0092] in, Indicates that the edge cache gateway g is in state Take Action The rewards you get, represents the number of times the content requested by the target user through the i-th path hits the gateway cache in the t+1 time slot, w i is the weight related to the delay reduction on the i-th path, w i Obtained by the following formula:

[0093]

[0094] Among them, w i represents the weight related to the delay reduction on the i-th path, Represents the delay reduction caused by hitting the gateway cache on the i-th path;

[0095] The model state, model action, and model reward determined in the edge computing model are used as the edge caching strategy.

[0096] Optionally, the first processing module is further used for:

[0097] According to each preset time slot, a content distribution phase and a cache update phase are set in the edge cache heterogeneous network fusion system;

[0098] In the content distribution stage, according to the edge cache heterogeneous network fusion system, the request content of the target user in each target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network are determined, and it is judged whether the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user;

[0099] If yes, then obtain the target cache content from the target protocol conversion gateway; if no, then obtain the target cache content from other protocol conversion gateways, wherein the target cache content is the cache content corresponding to the request content in the target protocol conversion gateway, and the other protocol conversion gateway is any target protocol conversion gateway except the target protocol conversion gateway corresponding to the target heterogeneous network;

[0100] In the cache update phase, the cache content in each target protocol conversion gateway is updated at each preset time slot according to the edge cache strategy;

[0101] Determine the edge cache operation mechanism based on the content distribution phase and cache update phase.

[0102] Optionally, the second processing module is further used for:

[0103] After the edge computing model is established, the corresponding evaluation network loss function is determined according to the DQN algorithm. The evaluation network loss function is obtained by the following formula:

[0104]

[0105] TargetQ=r i +γmax a Q target (s i+1 ,a;θ - )

[0106] Among them, Loss(θ) represents the evaluation network loss function, E represents the mean, and s i represents the model environment in time slot i, a i represents the model action in time slot i, γ i represents the discount rate in time slot i, s i+1 represents the model environment in the (i+1) time slot, B represents the number of small batch samples in the experience replay pool, and θ represents the evaluation network Q eval Parameter, θ - Represents the target network Q target Parameters, evaluation network Q eval Update the parameters by the following formula:

[0107] θ t+1 ←θ t -l*▽ θ Loss(θ)

[0108] Among them, θ t+1 Represents the evaluation network Q at time slot t+1 eval The parameter θ t Represents the evaluation network Q in time slot t eval Parameters, l represents the learning rate;

[0109] According to the evaluation network loss function, the edge computing model corresponding to the edge caching strategy is trained in each preset time slot to obtain the trained edge computing model.

[0110] In a third aspect, the present application provides a heterogeneous network fusion processing device based on edge computing, including:

[0111] Processor and memory;

[0112] Memory stores computer-executable instructions;

[0113] The processor executes the computer-executable instructions stored in the memory, so that the processor executes various possible implementations of the first aspect as described above.

[0114] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement various possible implementation methods as in the first aspect.

[0115] In a fifth aspect, the present application provides a computer program product, which, when executed by a processor, is used to implement various possible implementations as in the first aspect.

[0116] The present application provides a heterogeneous network fusion processing method, device and storage medium based on edge computing. By fusing each target heterogeneous network with the Internet Protocol Version 6 (IPv6) network respectively, a plurality of target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network are constructed; according to the plurality of target protocol conversion gateways, an edge cache heterogeneous network fusion system is established and the corresponding edge cache strategy and edge cache operation mechanism are determined; according to the edge cache strategy, a trained edge computing model is obtained; at each preset time slot, the request content of the target user in each target heterogeneous network is input into the edge computing model according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, so that the edge The computing model updates the cache content in the target protocol conversion gateway according to the request content, thereby integrating the heterogeneous networks by building the target protocol conversion gateway, improving network interoperability, and realizing the interconnection between the heterogeneous networks by establishing an edge cache heterogeneous network fusion system. At the same time, according to the edge cache operation mechanism and the edge computing model corresponding to the edge cache strategy, the request content of the target user is analyzed and cached in the target protocol conversion gateway, so that the cache content changes with the change of user needs, shortening the network request path, reducing the average delay, solving the technical problem of low efficiency of heterogeneous network fusion processing, and achieving the technical effect of improving the efficiency of heterogeneous network fusion processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0118] Figure 1 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 1 ;

[0119] Figure 2 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 2 ;

[0120] Figure 3 A schematic diagram of the protocol conversion structure between an aviation telecommunication network based on the Internet Protocol Suite and IPv6 provided in an embodiment of the present application;

[0121] Figure 4 A schematic diagram of a mapping method between a network service access point address and an IPv6 address provided in an embodiment of the present application;

[0122] Figure 5 A schematic diagram of a process for converting transport layer messages between a transport control protocol and an error detection and recovery protocol of a transport layer provided in an embodiment of the present application;

[0123] Figure 6 A schematic diagram of the protocol conversion structure between the aviation telecommunication network and the open system network and IPv6 provided in the embodiment of the present application;

[0124] Figure 7 A schematic diagram of the user authentication phase of the airport broadband mobile communication system network and IPv6 protocol conversion provided in an embodiment of the present application;

[0125] Figure 8 A network topology diagram of the protocol conversion between the airport broadband mobile communication system network and IPv6 provided in the embodiment of the present application;

[0126] Fig. 9 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 3 ;

[0127] Fig.10 A heterogeneous network fusion topology diagram provided in an embodiment of the present application;

[0128] Fig.11 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 4 ;

[0129] Fig.12 A schematic diagram of an edge caching strategy for a third protocol conversion gateway provided in an embodiment of the present application;

[0130] Fig.13 A schematic diagram of the structure of a heterogeneous network fusion processing device based on edge computing provided in an embodiment of the present application;

[0131] Fig.14 A hardware structure diagram of a heterogeneous network fusion processing device based on edge computing provided in an embodiment of the present application.

[0132] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0133] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0134] In the prior art, the path between data information and the dedicated server is long, resulting in a delay in accessing information, which reduces the communication efficiency between networks and presents a technical problem of low efficiency in heterogeneous network fusion processing.

[0135] The present application provides a heterogeneous network fusion processing method, device and storage medium based on edge computing. By constructing a target protocol conversion gateway, the heterogeneous networks are integrated to improve network interoperability. By establishing an edge cache heterogeneous network fusion system, the interconnection between the heterogeneous networks is realized. At the same time, according to the edge cache operation mechanism and the edge computing model corresponding to the edge cache strategy, the request content of the target user is analyzed and cached in the target protocol conversion gateway, so that the cache content changes with the change of user needs, shortening the network request path, reducing the average delay, solving the technical problem of low efficiency of heterogeneous network fusion processing, and achieving the technical effect of improving the efficiency of heterogeneous network fusion processing.

[0136] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0137] Figure 1 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 1 .like Figure 1 As shown, the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application includes:

[0138] S101, integrating each target heterogeneous network with the Internet Protocol version 6 (IPv6) network, and constructing multiple target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network;

[0139] In this embodiment, the target heterogeneous networks include an Aeronautical Telecommunication Network / Internet Protocol Suite (ATN / IPS) network based on the Internet Protocol Suite, an Aeronautical Telecommunication Network / Open System Interconnection Reference Model (ATN / OSI) network based on the Open System Interconnection Reference Model, and an Aeronautical Mobile Airport Communications System (AeroMACS) network.

[0140] The ATN / IPS network, the ATN / OSI network and the AeroMACS network are respectively integrated with the IPv6 network, and a target protocol conversion gateway corresponding to each target heterogeneous network is constructed according to the following methods S201 to S209.

[0141] S102, establishing an edge cache heterogeneous network fusion system according to multiple target protocol conversion gateways and determining corresponding edge cache strategies and edge cache operation mechanisms;

[0142] According to the constructed target protocol conversion gateway, combined with the following methods S901 to S906, a corresponding edge cache heterogeneous network fusion system is established, and according to the following methods S1101 to S1111, the corresponding edge cache strategy and edge cache operation mechanism of the system are determined.

[0143] S103. Obtain a trained edge computing model according to an edge cache strategy.

[0144] According to the determined edge caching strategy, combined with the following methods S1111 to S1112, the trained edge computing model is obtained.

[0145] S104. In each preset time slot, according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, the request content of the target user in each target heterogeneous network is input into the edge computing model, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content.

[0146] In this embodiment, the preset time slot refers to a preset time interval, and the interval period may be but is not limited to one minute or one hour, and the cache content refers to the network data packets cached in the target protocol conversion gateway.

[0147] In each preset time slot, according to the architecture of the edge cache heterogeneous network fusion system and its corresponding edge cache operation mechanism, the request content submitted by the target user in each target heterogeneous network is imported into the edge computing model for processing, so that the model updates the cache content in the corresponding target protocol conversion gateway in the edge cache heterogeneous network fusion system according to the request content of the target user.

[0148] The present application provides a heterogeneous network fusion processing method based on edge computing, which integrates each target heterogeneous network with the Internet Protocol Version 6 (IPv6) network respectively, and constructs multiple target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network; according to the multiple target protocol conversion gateways, an edge cache heterogeneous network fusion system is established and the corresponding edge cache strategy and edge cache operation mechanism are determined; according to the edge cache strategy, a trained edge computing model is obtained; at each preset time slot, according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, the request content of the target user in each target heterogeneous network is input into the edge computing model, so that the edge computing model The cache content in the target protocol conversion gateway is updated according to the request content, so that by building the target protocol conversion gateway, the heterogeneous networks are integrated to improve network interoperability, and by establishing an edge cache heterogeneous network fusion system, the interconnection between the heterogeneous networks is achieved. At the same time, according to the edge cache operation mechanism and the edge computing model corresponding to the edge cache strategy, the request content of the target user is analyzed and cached in the target protocol conversion gateway, so that the cache content changes with the change of user needs, shortening the network request path, reducing the average delay, solving the technical problem of low efficiency of heterogeneous network fusion processing, and achieving the technical effect of improving the efficiency of heterogeneous network fusion processing.

[0149] Figure 2 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 2 .like Figure 2 As shown, this embodiment, based on the above embodiment, provides a supplementary description of the construction process of the target protocol conversion gateway, including:

[0150] S201, setting the aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network as the first target subnet, and configuring the corresponding target security protocol and target routing protocol according to the aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network;

[0151] In this embodiment, the target security protocol refers to the IP security protocol (IPsec) and the Internet Key Exchange Protocol Version 2 (IKEv2) security protocol, the target routing protocol refers to the Border Gateway Protocol Four (BGP-4) routing protocol, and the first target subnet refers to the IPv6 subnet.

[0152] Figure 3 A schematic diagram of the protocol conversion structure between the aviation telecommunication network based on the Internet protocol suite and IPv6 provided in the embodiment of the present application is shown in FIG. Figure 3 As shown in the figure, both ATN / IPS and TCP / IPv6 are based on the TCP / IP five-layer model. Each model includes the physical layer (PHY), the data link layer (MAC), the network layer (Network Layer), the transport layer (Transport Layer) and the application layer (APP) from bottom to top. Among them, the network layer is configured with the IPv6 network, and the transport layer is configured with the transmission configuration protocol (Transmission Control Protocol, TCP protocol). The difference between ATN / IPS and TCP / IPv6 protocols is in the application layer. The application layer of ATN / IPS consists of the ATN application layer interface IPS Data System (IPSDS) and the ATN Application Aviation Service Environment (ATN-APP ASE);

[0153] To integrate the ATN / IPS network and the IPv6 network and obtain the ATN / IPS-IPv6 gateway, you need to: first set the ATN / IPS network and the IPv6 network into two IPv6 subnets, and configure the corresponding subnet addresses on the corresponding gateway interfaces; configure IPsec and IKEv2 protocols between the ATN / IPS router and the IPv6 network router, and finally configure BGP-4 protocol between the two autonomous systems (AS) of the ATN / IPS network and the IPv6 network.

[0154] S202, constructing a first protocol conversion gateway according to the first target subnet, the target security protocol and the target routing protocol;

[0155] In this embodiment, the first protocol conversion gateway refers to a target protocol conversion gateway between an aeronautical telecommunication network based on the Internet Protocol Suite and an IPv6 network.

[0156] According to the configured IPv6 subnet, IPsec and IKEv2 protocols and BGP-4 protocol, the first protocol conversion gateway between the aviation telecommunication network based on the Internet Protocol Suite and the IPv6 network is constructed.

[0157] S203, establishing a packet receiving and sending module according to the requirement of processing network packets between the aeronautical telecommunication network and the IPv6 network based on the open system interconnection reference model;

[0158] In this embodiment, the packet receiving and sending module is used to receive or send network data packets from an aeronautical telecommunication network based on the open system interconnection reference model and an IPv6 network.

[0159] According to the demand for receiving or sending network data packets between the ATN / OSI network and the IPv6 network, the processing function of the network data packets between the ATN / OSI network and the IPv6 network is determined, and based on this function, corresponding packet receiving and sending modules are established for the TN / OSI network and the IPv6 network.

[0160] S204, establishing a network layer address mapping module according to the network layer address mapping requirements between the network service access point address in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 address in the IPv6 network;

[0161] In this embodiment, the network layer address mapping module is used to map the network service access point (NSAP) address in the aeronautical telecommunication network based on the open system interconnection reference model with the IPv6 address of the IPv6 protocol network layer.

[0162] According to the requirement of one-to-one mapping between the Network Service Access Point (NSAP) address in the ATN / OSI network and the IPv6 address of the network layer of the IPv6 protocol, the mapping function between the NSAP address and the IPv6 address between the ATN / OSI network and the IPv6 network is determined, and based on this function, a network layer address mapping module is constructed so that each network can access the other network through the correct address information.

[0163] S205, establishing a network layer header conversion module according to the information conversion requirements between the connectionless network protocol header in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network in the IPv6 header;

[0164] In this embodiment, the network layer header conversion module is used to convert the connectionless network protocol (CLNP) header information in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 header information in the IPv6 network.

[0165] According to the requirement of information conversion between CLNP message header information in ATN / OSI network and IPv6 message header information in IPv6 network, the information conversion function of message header between ATN / OSI network and IPv6 network is determined, and according to this function, the corresponding network layer message header conversion module is established.

[0166] Figure 4 A schematic diagram of a mapping method between a network service access point address and an IPv6 address provided in an embodiment of the present application, such as Figure 4As shown, when the CLNP address needs to be converted to an IPv6 address, the 20-byte message header information (470027 81 43484E 00 ARS LOC SYS(MAC)SEL) corresponding to the CLNP address needs to be converted into the 16-byte message header information (0240 278143484E00 LOC SYS(MAC)) corresponding to the IPv6 address; similarly, when the IPv6 address needs to be converted to a CLNP address, the message header information (IPv6) of the IPv6 address needs to be converted into the 20-byte message header information (350000IPv6 00) corresponding to the CLNP address.

[0167] S206, establishing a transport layer message conversion module according to the conversion requirements between the error detection and recovery class protocol of the transport layer in the aeronautical telecommunication network based on the open system interconnection reference model and the transmission control protocol of the transport layer in the IPv6 network;

[0168] In this embodiment, the transport layer message conversion module is used to convert the error detection and recovery class (TP4) protocol of the transport layer in the aviation telecommunication network based on the open system interconnection reference model with the transmission control protocol (TCP) of the transport layer in the IPv6 network.

[0169] According to the protocol conversion requirements between the TP4 protocol of the transport layer in the ATN / OSI network and the TCP protocol of the transport layer in the IPv6 network, the protocol conversion function of the transport layer between the ATN / OSI network and the IPv6 network is determined, and based on this function, the corresponding transport layer message conversion module is established.

[0170] Figure 5 A schematic diagram of a flow chart of a transport layer message conversion between a transport control protocol and an error detection and recovery protocol of a transport layer provided in an embodiment of the present application, such as Figure 5 As shown, according to the conversion requirements between the TP4 protocol of the transport layer in the ATN / OSI network and the TCP protocol of the transport layer in the IPv6 network, the following is determined: Figure 5The conversion process of the transport layer message shown in the figure, wherein the "three-way handshake (establishing a connection)" of the TCP host corresponds to the "three-way handshake" of the TP4 host. During the handshake process, the TCP host sends a TCP message segment with a synchronization sequence number (Synchronize, SYN) flag to the ATN / OSI-IPv6 gateway to establish a connection. The ATN / OSI-IPv6 gateway establishes a connection with the TP4 host by sending a message with a CR flag. After receiving the message, the TP4 host sends a message segment with a CC flag to the gateway to establish a connection between the TP4 host and the ATN / OSI-IPv6 gateway. At the same time, the gateway feeds back a TCP message segment with SYN and confirmation character (Acknowledge, ACK) flags to the TCP host after receiving the SYN message, so as to establish a data connection between the TCP host and the ATN / OSI-IPv6 gateway.

[0171] During the transmission process, the TCP host will send a TCP message with an ACK flag to the ATN / OSI-IPv6 gateway. After receiving the message, the gateway will send a message with a DT flag to the TP4 host. After the host receives the message, it will convert it into a message with an AK flag and feed it back to the ATN / OSI-IPv6 gateway, so that the gateway will send the message segment with ACK data (DATA) to the TCP host for data transmission;

[0172] The "four waves" of the TCP host correspond to the "two waves (disconnection)" of the TP4 host. During the wave process, the TCP host sends a message segment with a Finish (FIN) mark to the ATN / OSI-IPv6 gateway. After receiving the message segment, the ATN / OSI-IPv6 gateway feeds back a message segment with an ACK mark to TCP for confirmation. The TP4 host sends a message segment with a DR mark to the ATN / OSI-IPv6 gateway, so that the gateway sends a message segment with a FIN ACK mark to the TCP host to disconnect from the TCP host. After receiving the message segment, TCP feeds back a message segment with an ACK mark to the ATN / OSI-IPv6 gateway for confirmation. After confirmation, the ATN / OSI-IPv6 gateway sends a message segment with a DC mark to the TP4 host to disconnect from the TP4 host.

[0173] S207, constructing a second protocol conversion gateway according to the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module;

[0174] In this embodiment, the second protocol conversion gateway refers to a target protocol conversion gateway between an aeronautical telecommunication network based on the open system interconnection reference model and an IPv6 network.

[0175] Figure 6 The schematic diagram of the protocol conversion structure between the aviation telecommunication network and the open system network and IPv6 provided in the embodiment of the present application is as follows: Figure 6 As shown in the figure, the ATN / OSI network is a seven-layer model based on OSI, which includes the physical layer, data link layer, network layer, transport layer, session layer, presentation layer and application layer from bottom to top. The network layer is configured with CLNP network, the transport layer is configured with TP4 protocol, the session layer is configured with session control protocol (Session Control Protocols, COSP), and the presentation layer is configured with control plane policing (Control Plane Policing, COPP). After the ATN / OSI network and the IPv6 network are integrated, the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module are sequentially placed in the second protocol conversion gateway composed of the ATN / OSI network and the IPv6 network, and finally the following is formed. Figure 6 ATN / IPS-IPv6 gateway shown.

[0176] S208, setting the airport broadband mobile communication system network and IPv6 network as the second target subnet;

[0177] In this embodiment, the second target subnet refers to an IPv6 subnet.

[0178] To integrate the airport broadband mobile communication system network and the IPv6 network, it is first necessary to set the airport broadband mobile communication system network and the IPv6 network into two IPv6 subnets and configure the corresponding subnet addresses at the corresponding gateway interfaces.

[0179] S209, performing user authentication according to the target protocol authentication method, and after the user authentication is passed, accessing the service resources according to the Internet address of the service server, performing IPv6 data transmission, so as to build a third protocol conversion gateway.

[0180] In this embodiment, the target protocol authentication method refers to the TLS (Transport Layer Security Protocol) + EAP (Extensible Authentication Protocol) + RADIUS (Remote User Dial-In Authentication System) protocol authentication method, TLS is responsible for establishing an encrypted transmission channel, EAP is used for the authentication process, RADIUS is an AAA (Authentication, Authorization, Accounting) authentication protocol used to manage user identity authentication and access control, and the third protocol conversion gateway refers to the target protocol conversion gateway between the airport broadband mobile communication system network and the IPv6 network.

[0181] User authentication is performed according to the TLS+EAP+RADIUS protocol. After the user authentication is passed, the user accesses the business resources according to the IP address of the business server and performs IPv6 data transmission to build a third protocol conversion gateway between the airport broadband mobile communication system network and the IPv6 network.

[0182] Figure 7 A schematic diagram of the airport broadband mobile communication system network and IPv6 protocol conversion user authentication phase provided by the embodiment of the present application, such as Figure 7 As shown, during the authentication process, the IPv6 terminal first sends a packet with a specific flow label identifier to request the service corresponding to the flow label. At this time, the destination address of the packet sent is the AeroMACS-IPv6 gateway; the gateway initiates an EAP request to the ASN-GW (access service network gateway) to establish an EAP connection with the ASN-GW; the gateway needs to configure a mapping table of flow labels and destination ASN-GW addresses at the initial moment, and after receiving the packet sent by the IPv6 terminal, it modifies the destination address in the IPv6 message header to the corresponding ASN-GW address in the mapping table according to the flow label identifier in the message; the ASN-GW is also an AAA client, and after receiving the data packets sent by the gateway and various transportation tools through the base station (Base Station, BS), it diverts them to the corresponding CSN (content service network) according to the flow label in the message header, and the CSN then sends the service request packet to the AAA server (AAAserver); the AAA server stores the user identity information of this service, which is used to verify whether the received packet has the authority to access this service. If there is no authority, an access rejection response is returned, otherwise, the IP address of the service server is returned to the user for the user to request the corresponding service.

[0183] Figure 8 The network topology diagram of the protocol conversion between the airport broadband mobile communication system network and IPv6 provided in the embodiment of the present application is as follows: Figure 8As shown, the AeroMACS network is based on an IP model and includes an AAA server and an ASN-GW, wherein the AAA server, the ASN-GW and the TCP / IPv6 protocol stack are configured with a Wide Area Network Interface (WAN I / F), an IPv6 protocol, a User Datagram Protocol (UDP), a TLS protocol and an EAP protocol. Different from the TCP / IPv6 protocol stack, the AAA server and the ASN-GW are also configured with a Remote Authentication Dial-in User System (RADIUS Server) protocol. The AAA server, the ASN-GW and the TCP / IPv6 protocol stack in the AeroMACS network are integrated to obtain an AeroMACS-IPv6 gateway including the WAN I / F protocol, the EAP protocol and the IPv6 subnets of the AeroMACS network and the IPv6 network.

[0184] The present application provides a heterogeneous network fusion processing method based on edge computing, which builds a target protocol conversion gateway by integrating various heterogeneous networks with the IPv6 network; thereby achieving interconnection between multiple heterogeneous network protocols, improving network interoperability, solving the technical problem of low efficiency in heterogeneous network fusion processing, and achieving the technical effect of improving the efficiency of heterogeneous network fusion processing.

[0185] Fig. 9 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 3 .like Fig. 9 As shown, this embodiment, based on the above embodiment, provides a supplementary description of the establishment process of the edge cache heterogeneous network fusion system, including:

[0186] S901, determining a target router corresponding to each target protocol gateway and a target user corresponding to each target heterogeneous network;

[0187] In this embodiment, each target protocol gateway corresponds to a number of target routers, and each target heterogeneous network includes a number of target users.

[0188] According to each target protocol gateway, a corresponding number of target routers and a corresponding target user in each target heterogeneous network are determined.

[0189] S902, for each target heterogeneous network, according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, establish a data connection between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network;

[0190] In this embodiment, the data connection includes a physical connection or a logical connection.

[0191] For each target heterogeneous network, a physical connection or a logical connection is established between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network.

[0192] S903, establishing a corresponding physical connection between any two target protocol conversion gateways according to the information transmission requirements between each target protocol conversion gateway;

[0193] S904, constructing an edge cache heterogeneous network fusion system according to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways, and the target router;

[0194] According to the data connection between the target protocol conversion gateway and the target user, the physical connection between any two target protocol conversion gateways and the target router, a corresponding edge cache heterogeneous network fusion system is constructed.

[0195] Fig.10 The heterogeneous network fusion topology diagram provided in the embodiment of the present application is as follows: Fig.10 As shown in the figure, the edge cache heterogeneous network fusion system is mainly composed of AeroMACS network, ATN / IPS network, and ATN / OSI network. Under each target heterogeneous network, a corresponding target protocol conversion gateway is established with the IPv6 network, namely, AeroMACS-IPv6 gateway, ATN / IPS-IPv6 gateway, and ATN / OSI-IPv6 gateway. Each target protocol conversion gateway is connected to each other through the IPv6 network, forming an edge cache heterogeneous network fusion system together.

[0196] It should be noted that for a user in a heterogeneous network, the storage space of the protocol conversion gateway in the network is limited, and at most M contents can be stored. When the content requested by the user is not cached in the target protocol conversion gateway corresponding to the user, the corresponding content is requested from users in other heterogeneous networks. The number of contents that users in the entire network may request is at most N.

[0197] The present application provides a heterogeneous network fusion processing method based on edge computing. By constructing an edge cache heterogeneous network fusion system and determining the corresponding edge cache operation mechanism according to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways and the target router; thereby, through the edge cache heterogeneous network fusion system, the network request path between the target user and the target protocol conversion gateway is shortened, the average delay of information access is reduced, the technical problem of low efficiency of heterogeneous network fusion processing is solved, and the technical effect of improving the efficiency of heterogeneous network fusion processing is achieved.

[0198] Fig.11 The process of the heterogeneous network fusion processing method based on edge computing provided in the embodiment of the present application Figure 4 ,like Fig.11 As shown, based on the above embodiment, this embodiment provides additional explanations on the edge cache strategy, the process of determining the edge cache operation mechanism, and the training process of the edge computing model, including:

[0199] S1101, providing a reinforcement learning agent corresponding to each target protocol conversion gateway;

[0200] S1102, determining the request content of the target user, all cache contents of the target protocol conversion gateway, and the number of times the request content hits the target protocol conversion gateway according to the edge cache heterogeneous network fusion system;

[0201] S1103, according to the ratio of the request content to the total cache content of the target protocol conversion gateway, determine the content popularity of the current preset time slot, and according to the content popularity, the total cache content of the target protocol conversion gateway and the number of times the request content hits the target protocol conversion gateway, model each reinforcement learning agent, and determine the corresponding model state, action and reward;

[0202] In this embodiment, the content popularity is used to indicate the popularity of the content within a certain area over a period of time.

[0203] At each protocol conversion gateway, according to different request preferences of users in the current period, and according to the ratio of the target cache content to the total cache content of the target protocol conversion gateway, a content popularity is set for the user's requested content;

[0204] Specifically, the content popularity of the target protocol conversion gateway's content f at time slot t is equal to the ratio of the local protocol conversion gateway's content f requests at time slot t to all content requests at time slot t. Content popularity is positively correlated with the frequency with which the content is accessed.

[0205] Based on the basic model parameters, each reinforcement learning agent is modeled and the corresponding model state, action, and reward are determined in the following way:

[0206] The environment status observed by the target protocol conversion gateway is:

[0207]

[0208] in, represents the environmental state observed by the target protocol conversion gateway g at the preset time slot t, represents the content popularity of the target protocol conversion gateway g in time slot t, Represents the cache content of the target protocol conversion gateway g in time slot t.

[0209] After determining the state of the environment, the reinforcement learning agent in the gateway is converted according to the target protocol to determine the model action in the cache update phase Model actions include updating the cache content to obtain the cache content for the next preset time slot Among them, each content in the target protocol conversion gateway has a corresponding index {1,2,3,…,M}, and the content includes cached content and uncached content;

[0210] Before the cache is updated, the popularity of cached content and non-cached content is calculated based on the frequency of requests. Sort descending, This means deleting cached content. The lowest-ranked k contents in the cache are retrieved, and the highest-ranked k contents in the uncached contents are obtained for replenishment.

[0211] The model reward corresponding to the edge computing model is defined as:

[0212]

[0213] in, Indicates that the edge cache gateway g is in state Take Action The rewards you get, represents the number of times the content requested by the target user through the i-th path hits the gateway cache in the t+1 time slot, w i is the weight related to the delay reduction on the i-th path, w i Obtained by the following formula:

[0214]

[0215] Among them, w i represents the weight related to the delay reduction on the i-th path, Represents the delay reduction caused by hitting the gateway cache on the i-th path;

[0216] Fig.12 A schematic diagram of the edge caching strategy of the third protocol conversion gateway provided in the embodiment of the present application is as follows: Fig.12 As shown in the figure, taking the edge cache strategy deployed at the third protocol conversion gateway between the airport broadband mobile communication system network and the IPv6 network as an example, d ug is the latency from the local user to the AeroMACS-IPv6 gateway, is the delay from ATN / OSI-IPv6 gateway or ATN / IPS-IPv6 to other network users. The average access delay of user content is the weighted delay of the number of user content requests N, defined as:

[0217]

[0218] Among them, d i is the request delay of the i-th content, is the number of requests for the i-th content in time slot t. It can be seen that maximizing the long-term reward can minimize the average access delay in the network.

[0219] S1104, using the model state, model action, and model reward determined in the edge computing model as an edge cache strategy;

[0220] S1105: after determining that the edge cache heterogeneous network fusion system is established, setting a content distribution phase and a cache update phase in the edge cache heterogeneous network fusion system according to each preset time slot;

[0221] In this embodiment, the content distribution phase refers to the phase in which each protocol conversion gateway receives and processes user requests in the network, and the cache update phase refers to the phase in which each protocol conversion gateway updates the cache content in its storage module.

[0222] After the establishment of the edge cache heterogeneous network fusion system is completed, according to each preset time slot, the edge cache operation mechanism corresponding to the edge cache heterogeneous network fusion system is divided into a content distribution phase and a cache update phase.

[0223] S1106. In the content distribution phase, according to the edge cache heterogeneous network fusion system, the request content of the target user in each target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network are determined;

[0224] In the content distribution phase, the request content corresponding to the target user of each target heterogeneous network in the edge cache heterogeneous network fusion system and the target protocol conversion gateway corresponding to the target heterogeneous network are determined.

[0225] S1107, determining whether the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user;

[0226] S1108: If yes, obtain the target cache content from the target protocol conversion gateway;

[0227] In this embodiment, the target cache content refers to the cache content corresponding to the request content in the target protocol conversion gateway.

[0228] If the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user, the target cache content is directly obtained from the target protocol conversion gateway.

[0229] S1109, if not, obtaining the target cache content from other protocol conversion gateways;

[0230] In this embodiment, other protocol conversion gateways refer to any target protocol conversion gateway except the target protocol conversion gateway corresponding to the target heterogeneous network.

[0231] If the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network does not contain the request content of the target user, the target cache content is directly obtained from other protocol conversion gateways different from the target protocol conversion gateway.

[0232] S1110 . In the cache update phase, according to the edge cache strategy, the cache content in each target protocol conversion gateway is updated in each preset time slot.

[0233] In the cache update phase, the cache content in each target protocol conversion gateway is updated in each preset time slot according to the determined edge cache strategy.

[0234] S1111. Determine an edge cache operation mechanism according to a content distribution phase and a cache update phase;

[0235] The edge cache operation mechanism is determined according to the execution actions of the content distribution phase and the cache update phase determined in each time slot.

[0236] S1112. After determining that the edge computing model is established, determine the corresponding evaluation network loss function according to the DQN algorithm;

[0237] In this embodiment, the evaluation network loss function is obtained by the following formula:

[0238]

[0239] TargetQ=r i +γmax a Q target (s i+1 ,a;θ - )

[0240] Among them, Loss(θ) represents the evaluation network loss function, E represents the mean, and s i represents the model environment in time slot i, a i represents the model action in time slot i, γ i represents the discount rate in time slot i, s i+1represents the model environment in the (i+1) time slot, B represents the number of small batch samples in the experience replay pool, and θ represents the evaluation network Q eval Parameter, θ - Represents the target network Q target Parameters, evaluation network Q eval Update the parameters by the following formula:

[0241] θ t+1 ←θ t -l*▽ θ Loss(θ)

[0242] Among them, θ t+1 Represents the evaluation network Q at time slot t+1 eval The parameter θ t Represents the evaluation network Q in time slot t eval Parameters, l represents the learning rate;

[0243] S1113. According to the evaluation network loss function, the edge computing model corresponding to the edge caching strategy is trained in each preset time slot to obtain the trained edge computing model.

[0244] The present application provides a heterogeneous network fusion processing method based on edge computing. By determining an edge caching strategy and establishing an edge computing model, the edge computing technology is utilized to analyze the content frequently requested by users, and the content is cached in a target conversion protocol gateway using a reinforcement learning strategy, so that the cached content of the target conversion protocol in the gateway changes with the change of user needs. The effective control of the request content of the target users in each heterogeneous network by the edge caching heterogeneous network fusion system is strengthened, the active coordination capability between the target users and the target protocol conversion gateway in each heterogeneous network and the intelligence level of the global decision-making are improved, the technical problem of low efficiency of heterogeneous network fusion processing is solved, and the technical effect of improving the efficiency of heterogeneous network fusion processing is achieved.

[0245] Fig.13 This is a schematic diagram of the structure of a heterogeneous network fusion processing device based on edge computing provided in an embodiment of the present application. The device of this embodiment can be in the form of software and / or hardware. Fig.13 As shown, the heterogeneous network fusion processing device 1300 based on edge computing provided in the embodiment of the present application includes: a construction module 1301, a first processing module 1302, a second processing module 1303, and a control module 1304:

[0246] A construction module 1301 is used to merge each target heterogeneous network with an Internet Protocol version 6 (IPv6) network, and construct multiple target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network;

[0247] The first processing module 1302 is used to establish an edge cache heterogeneous network fusion system according to multiple target protocol conversion gateways and determine the corresponding edge cache strategy and edge cache operation mechanism;

[0248] The second processing module 1303 is used to obtain the trained edge computing model according to the edge cache strategy;

[0249] The control module 1304 is used to input the request content of the target user in each target heterogeneous network into the edge computing model at each preset time slot according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content.

[0250] In a possible implementation, the construction module 1301 is further used to:

[0251] The aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network are set as the first target subnet, and the corresponding target security protocol and target routing protocol are configured according to the aeronautical telecommunication network based on the Internet protocol suite and the IPv6 network; wherein the first target subnet is the IPv6 subnet, and each network corresponds to an IPv6 subnet;

[0252] Constructing a first protocol conversion gateway according to the first target subnet, the target security protocol and the target routing protocol;

[0253] According to the requirements between the aviation telecommunication network and the IPv6 network based on the open system interconnection reference model, multiple target function modules are established, and according to the target function modules, a second protocol conversion gateway is constructed, wherein the target function modules include a packet receiving and sending module, a network layer address mapping module, a network layer message header conversion module and a transport layer message conversion module;

[0254] The airport broadband mobile communication system network and the IPv6 network are set as the second target subnet; wherein the second target subnet is an IPv6 subnet, and each network corresponds to an IPv6 subnet;

[0255] According to the target protocol authentication method, user authentication is performed, and after the user authentication is passed, business resources are accessed according to the Internet address of the business server, and IPv6 data transmission is performed to build a third protocol conversion gateway.

[0256] In a possible implementation, the construction module 1301 is further used to:

[0257] According to the requirements of processing network packets between the aeronautical telecommunication network and the IPv6 network based on the open system interconnection reference model, a packet receiving and sending module is established;

[0258] Establish a network layer address mapping module based on the network layer address mapping requirements between the network service access point address in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 address in the IPv6 network;

[0259] According to the information conversion requirements between the connectionless network protocol message header in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network in the IPv6 message header, a network layer message header conversion module is established;

[0260] According to the conversion requirements between the error detection and recovery protocols of the transport layer in the aeronautical telecommunication network based on the open system interconnection reference model and the transmission control protocol of the transport layer in the IPv6 network, a transport layer message conversion module is established;

[0261] A second protocol conversion gateway is constructed according to the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module.

[0262] In a possible implementation, the first processing module 1302 is further configured to:

[0263] Determine the target router corresponding to each target protocol gateway and the target user corresponding to each target heterogeneous network;

[0264] For each target heterogeneous network, according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, a data connection is established between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, wherein the data connection includes a physical connection or a logical connection;

[0265] According to the information transmission requirements between each target protocol conversion gateway, a corresponding physical connection is established between any two target protocol conversion gateways;

[0266] According to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways and the target router, an edge cache heterogeneous network fusion system is constructed.

[0267] In a possible implementation, the first processing module 1302 is further configured to:

[0268] A reinforcement learning agent corresponding to each target protocol conversion gateway;

[0269] According to the edge cache heterogeneous network fusion system, determine the request content of the target user, the entire cache content of the target protocol conversion gateway, and the number of times the request content hits the target protocol conversion gateway;

[0270] Determine the content popularity of the current preset time slot based on the ratio of the requested content to the total cached content of the target protocol conversion gateway;

[0271] Based on the content popularity, all cached contents of the target protocol conversion gateway, and the number of times the requested content hits the target protocol conversion gateway, each reinforcement learning agent is modeled, and the corresponding model state, action, and reward are determined to obtain the edge computing model corresponding to each reinforcement learning agent. The model state, action, and reward are determined in the following way:

[0272] The environment status observed by the target protocol conversion gateway is:

[0273]

[0274] in, represents the environmental state observed by the target protocol conversion gateway g at the preset time slot t, represents the content popularity of the target protocol conversion gateway g in time slot t, represents the cache content of the target protocol conversion gateway g in time slot t;

[0275] After determining the state of the environment, the reinforcement learning agent in the gateway is converted according to the target protocol to determine the model action in the cache update phase Model actions include updating the cache content to obtain the cache content for the next preset time slot Among them, each content in the target protocol conversion gateway has a corresponding index {1,2,3,…,M}, and the content includes cached content and uncached content;

[0276] Before the cache is updated, the popularity of cached content and non-cached content is calculated based on the frequency of requests. Sort descending, This means deleting cached content. The lowest-ranked k contents in the cached content are obtained, and the highest-ranked k contents in the uncached content are obtained for supplementation;

[0277] The model reward corresponding to the edge computing model is defined as:

[0278]

[0279] in, Indicates that the edge cache gateway g is in state Take Action The rewards you get, represents the number of times the content requested by the target user through the i-th path hits the gateway cache in the t+1 time slot, w i is the weight related to the delay reduction on the i-th path, w i Obtained by the following formula:

[0280]

[0281] Among them, w i represents the weight related to the delay reduction on the i-th path, Represents the delay reduction caused by hitting the gateway cache on the i-th path;

[0282] The model state, model action, and model reward determined in the edge computing model are used as the edge caching strategy.

[0283] In a possible implementation, the first processing module 1302 is further configured to:

[0284] According to each preset time slot, a content distribution phase and a cache update phase are set in the edge cache heterogeneous network fusion system;

[0285] In the content distribution stage, according to the edge cache heterogeneous network fusion system, the request content of the target user in each target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network are determined, and it is judged whether the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user;

[0286] If yes, then obtain the target cache content from the target protocol conversion gateway; if no, then obtain the target cache content from other protocol conversion gateways, wherein the target cache content is the cache content corresponding to the request content in the target protocol conversion gateway, and the other protocol conversion gateway is any target protocol conversion gateway except the target protocol conversion gateway corresponding to the target heterogeneous network;

[0287] In the cache update phase, the cache content in each target protocol conversion gateway is updated at each preset time slot according to the edge cache strategy;

[0288] Determine the edge cache operation mechanism based on the content distribution phase and cache update phase.

[0289] In a possible implementation, the second processing module 1303 is further configured to:

[0290] After the edge computing model is established, the corresponding evaluation network loss function is determined according to the DQN algorithm. The evaluation network loss function is obtained by the following formula:

[0291]

[0292] TargetQ=r i +γmax a Q target (s i+1 ,a;θ - )

[0293] Among them, Loss(θ) represents the evaluation network loss function, E represents the mean, and s i represents the model environment in time slot i, a i represents the model action in time slot i, γ i represents the discount rate in time slot i, s i+1 represents the model environment in the (i+1) time slot, B represents the number of small batch samples in the experience replay pool, and θ represents the evaluation network Q eval Parameter, θ - Represents the target network Q target Parameters, evaluation network Q eval Update the parameters by the following formula:

[0294] θ t+1 ←θ t -l*▽ θ Loss(θ)

[0295] Among them, θ t+1 Represents the evaluation network Q at time slot t+1 eval The parameter θ t Represents the evaluation network Q in time slot t eval Parameters, l represents the learning rate;

[0296] According to the evaluation network loss function, the edge computing model corresponding to the edge caching strategy is trained in each preset time slot to obtain the trained edge computing model.

[0297] The present application provides a heterogeneous network fusion processing device based on edge computing, which can implement the above method embodiment. Its implementation principle and technical effects are similar, and this embodiment will not be repeated here.

[0298] Fig.14 This is a hardware structure diagram of heterogeneous network fusion processing based on edge computing provided in the embodiment of the present application. Fig.14 As shown, the heterogeneous network fusion processing device 1400 based on edge computing includes:

[0299] Processor 1401 and memory 1402;

[0300] Memory stores computer-executable instructions;

[0301] The processor executes the computer execution instructions stored in the memory 1402, so that the heterogeneous network fusion processing device based on edge computing executes the heterogeneous network fusion processing method based on edge computing as described above.

[0302] It should be understood that the processor 1401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), or application-specific integrated circuits (ASIC). A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor.

[0303] The memory 1402 may include a high-speed random access memory (RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a disk or an optical disk, etc.

[0304] An embodiment of the present application correspondingly provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the heterogeneous network fusion processing method based on edge computing as described above.

[0305] The embodiment of the present application also provides a computer program product accordingly. When the computer program is executed by a processor, it is used to implement the heterogeneous network fusion processing method based on edge computing as described above.

[0306] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0307] It should be further noted that, although the various steps in the flowchart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0308] It should be understood that the above-mentioned device embodiments are only illustrative, and the device of the present application can also be implemented in other ways. For example, the division of units / modules in the above-mentioned embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0309] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present application may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.

[0310] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. If not specifically stated, the processor may be any appropriate hardware processor, such as a CPU, a GPU, an FPGA, a DSP, an ASIC, etc. If not specifically stated, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc.

[0311] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0312] In the above embodiments, the description of each embodiment has its own emphasis. For the part not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0313] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0314] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A heterogeneous network fusion processing method based on edge computing, characterized in that: include: Merging each target heterogeneous network with an Internet Protocol version 6 (IPv6) network, and constructing a plurality of target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network; According to the plurality of target protocol conversion gateways, an edge cache heterogeneous network fusion system is established and corresponding edge cache strategies and edge cache operation mechanisms are determined; Obtaining a trained edge computing model according to the edge caching strategy; In each preset time slot, according to the edge cache heterogeneous network fusion system and the edge cache operation mechanism, the request content of each target user in the target heterogeneous network is input into the edge computing model, so that the edge computing model updates the cache content in the target protocol conversion gateway according to the request content.

2. The method according to claim 1, characterized in that The target heterogeneous networks include an aeronautical telecommunication network based on the Internet Protocol suite, an aeronautical telecommunication network based on the Open Systems Interconnection Reference Model, and an airport surface broadband mobile communication system network; the target protocol conversion gateways include a first protocol conversion gateway, a second protocol conversion gateway, and a third protocol conversion gateway; and the steps of fusing each target heterogeneous network with an Internet Protocol version 6 (IPv6) network to construct multiple target protocol conversion gateways corresponding to the target heterogeneous network and the IPv6 network include: Setting the Internet Protocol Suite-based aeronautical telecommunication network and the IPv6 network as first target subnets, and configuring corresponding target security protocols and target routing protocols based on the Internet Protocol Suite-based aeronautical telecommunication network and the IPv6 network; wherein the first target subnet is an IPv6 subnet; Constructing the first protocol conversion gateway according to the first target subnet, the target security protocol, and the target routing protocol; Establishing a plurality of target functional modules based on requirements between the open systems interconnection reference model-based aeronautical telecommunication network and the IPv6 network, and constructing the second protocol conversion gateway based on the target functional modules, wherein the target functional modules include a packet sending and receiving module, a network layer address mapping module, a network layer message header conversion module, and a transport layer message conversion module; Setting the airport broadband mobile communication system network and the IPv6 network as a second target subnet; wherein the second target subnet is the IPv6 subnet; According to the target protocol authentication method, user authentication is performed, and after the user authentication is passed, service resources are accessed according to the Internet address of the service server and IPv6 data transmission is performed to construct the third protocol conversion gateway.

3. The method according to claim 2, characterized in that The step of establishing a plurality of target functional modules based on requirements between the open systems interconnection reference model-based aeronautical telecommunication network and the IPv6 network, and constructing the second protocol conversion gateway based on the target functional modules includes: Establishing the packet receiving and sending module according to the requirements for processing network packets between the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network; Establishing the network layer address mapping module according to the network layer address mapping requirements between the network service access point address in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 address in the IPv6 network; Establishing the network layer header conversion module according to the information conversion requirements between the connectionless network protocol header in the aeronautical telecommunication network based on the open system interconnection reference model and the IPv6 network and the IPv6 header; Establishing the transport layer message conversion module according to the conversion requirements between the error detection and recovery protocols of the transport layer in the aeronautical telecommunication network based on the open system interconnection reference model and the transmission control protocol of the transport layer in the IPv6 network; The second protocol conversion gateway is constructed according to the packet receiving and sending module, the network layer address mapping module, the network layer message header conversion module and the transport layer message conversion module.

4. The method according to claim 1, wherein The establishing of an edge cache heterogeneous network fusion system based on the plurality of target protocol conversion gateways includes: Determine the target router corresponding to each target protocol gateway and the target user corresponding to each target heterogeneous network; For each target heterogeneous network, according to the information transmission requirements between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, establish a data connection between the target user in the target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network, wherein the data connection includes a physical connection or a logical connection; Establishing a corresponding physical connection between any two target protocol conversion gateways according to information transmission requirements between each two target protocol conversion gateways; The edge cache heterogeneous network fusion system is constructed according to the connection relationship between the target protocol conversion gateway and the target user, the connection relationship between any two target protocol conversion gateways, and the target router.

5. The method according to claim 4, characterized in that Determining the corresponding edge cache strategy includes: A reinforcement learning agent corresponding to each target protocol conversion gateway; Determining, according to the edge cache heterogeneous network fusion system, the request content of the target user, all cache contents of the target protocol conversion gateway, and the number of times the request content hits the target protocol conversion gateway; Determining the content popularity of a current preset time slot based on a ratio of the requested content to all cached content of the target protocol conversion gateway; Model each reinforcement learning agent based on the content popularity, all cached contents of the target protocol conversion gateway, and the number of times the requested content hits the target protocol conversion gateway, and determine the corresponding model state, action, and reward to obtain the edge computing model corresponding to each reinforcement learning agent, wherein the model state, action, and reward are determined in the following manner: The environment state observed by the target protocol conversion gateway is: in, represents the environmental state observed by the target protocol conversion gateway g at the preset time slot t, represents the content popularity of the target protocol conversion gateway g in time slot t, Represents the cache content of the target protocol conversion gateway g in time slot t; After determining the state of the environment, the model action of the cache update phase is determined according to the reinforcement learning agent in the target protocol conversion gateway. The model action includes updating the cache content to obtain the cache content of the next preset time slot Each content in the target protocol conversion gateway has a corresponding index {1, 2, 3, ..., M}, and the content includes cached content and non-cached content; Before the cache is updated, the popularity of the cached content and the uncached content is calculated based on the request frequency. Sort in descending order, This means deleting cached content. The k contents with the lowest ranking in the cached content are obtained, and the k contents with the highest ranking in the uncached content are obtained for supplementation; The model reward corresponding to the edge computing model is defined as: in, Indicates that the edge cache gateway g is in state Take action The rewards received, represents the number of times the target user in time slot t+1 requests the content through path i and hits the gateway cache, w i is the weight related to the delay reduction on the i-th path, w i Obtained by the following formula: Among them, w i represents the weight related to the delay reduction on the i-th path, represents the delay reduction brought by hitting the gateway cache on the i-th path; The model state, the action, and the reward determined in the edge computing model are used as the edge caching strategy.

6. The method according to claim 5, characterized in that Determining the corresponding edge cache operation mechanism includes: According to each preset time slot, setting a content distribution phase and a cache update phase in the edge cache heterogeneous network fusion system; In the content distribution stage, according to the edge cache heterogeneous network fusion system, the request content of the target user in each target heterogeneous network and the target protocol conversion gateway corresponding to the target heterogeneous network are determined, and whether the network data packet of the target protocol conversion gateway corresponding to the target heterogeneous network contains the request content of the target user; If so, obtaining the target cache content from the target protocol conversion gateway; if not, obtaining the target cache content from other protocol conversion gateways, wherein the target cache content is the cache content corresponding to the requested content in the target protocol conversion gateway, and the other protocol conversion gateway is any target protocol conversion gateway other than the target protocol conversion gateway corresponding to the target heterogeneous network; In the cache update phase, according to the edge cache strategy, the cache content in each of the target protocol conversion gateways is updated in each preset time slot; The edge cache operation mechanism is determined according to the content distribution phase and the cache update phase.

7. The method according to claim 6, characterized in that The step of obtaining a trained edge computing model according to the edge caching strategy includes: After the edge computing model is established, the corresponding evaluation network loss function is determined according to the DQN algorithm. The evaluation network loss function is obtained by the following formula: TargetQ=r i +γmax a Q target (s i+1 ,a;θ - ) Among them, Loss(θ) represents the evaluation network loss function, E represents the mean, s i represents the model environment in time slot i, a i represents the model action in the i-th time slot, γ i represents the discount rate in time slot i, s i+1 represents the model environment in the (i+1) time slot, B represents the number of small batch samples in the experience replay pool, and θ represents the evaluation network Q eval Parameter, θ - Represents the target network Q target Parameters, evaluation network Q eval Update the parameters using the following formula: Among them, θ t+1 Represents the evaluation network Q at time slot t+1 eval The parameter θ t Represents the evaluation network Q in time slot t eval Parameters, l represents the learning rate; According to the evaluation network loss function, the edge computing model corresponding to the edge caching strategy is trained in each preset time slot to obtain the trained edge computing model.

8. A heterogeneous network fusion processing device based on edge computing, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

10. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images