Distributed edge cache acceleration method and device based on PCDN

Abstract

The invention relates to the technical field of Internet content distribution, and discloses a distributed edge cache acceleration method based on PCDN, comprising the following steps: step I: terminal node access and resource initialization, which includes node registration and authentication, resource information reporting and node classification and label management; step II, hierarchical caching strategy and content storage, wherein the hierarchical caching strategy and content storage comprises content popularity grading, cache writing strategy and cache replacement and elimination; and step III, dynamic content scheduling and distribution. According to the distributed edge cache acceleration method and device based on the PCDN, the content popularity and regional distribution are pre-judged in advance through a deep learning model, and bandwidths and resources of different networks are integrated through a network fusion technology, so that the purposes of reducing the number of times of source returning requests, breaking through the transmission limitation of a traditional single network (for example, only depending on a cellular network) and improving the transmission efficiency can be achieved. The effects of node contribution traceability and automatic execution excitation are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet content distribution, and in particular to a PCDN-based distributed edge cache acceleration method and device. Background Art

[0002] With the continuous innovation of network technology, PCDN technology has emerged. PCDN, or P2P content distribution network, integrates P2P technology into the traditional CDN architecture. P2P technology has a decentralized nature, allowing nodes (user devices) in the network to communicate directly with each other and share resources. PCDN leverages this feature to tap into the large amount of idle, fragmented resources in the edge network, attempting to reduce distribution costs while ensuring service quality. This brings new ideas and solutions to the content distribution field, showing application potential in business scenarios such as video on demand, live streaming, and large file downloads.

[0003] Traditional CDNs rely on manual configuration or simple rules (such as content popularity thresholds) for caching. They cannot dynamically adapt to real-time traffic fluctuations, only support single network access (such as fixed broadband), and cannot utilize idle resources of mobile networks (such as 4G / 5G) or short-distance communications (such as Bluetooth). Distribution efficiency is low in weak network environments (such as subways and mountainous areas). Summary of the Invention

[0004] (1) Technical problems solved

[0005] In response to the shortcomings of the existing technology, the present invention provides a distributed edge cache acceleration method and device based on PCDN, which has the advantages of reducing the number of back-to-source requests, breaking through the limitations of traditional single network transmission (such as relying only on cellular networks), achieving traceability of node contributions, and automatic execution of incentives, thereby solving the problems mentioned in the above background technology.

[0006] (2) Technical solution

[0007] To achieve the above objectives, the present invention provides the following technical solution: a PCDN-based distributed edge cache acceleration method, comprising the following steps:

[0008] Step I: Terminal node access and resource initialization, which includes node registration and authentication, resource information reporting, and node classification and label management;

[0009] Step II: Layered caching strategy and content storage, including content popularity grading, cache write strategy, and cache replacement and elimination;

[0010] Step III: Dynamic content scheduling and distribution, which includes request parsing and node discovery, optimal node selection algorithm, content distribution execution, transmission optimization and flow control;

[0011] Step IV: Node incentives and network maintenance, including contribution value measurement, point exchange and penalty mechanism, and network topology optimization;

[0012] Step V: Data security and privacy protection, including content encryption transmission, privacy protection, and access control;

[0013] Step VI: Intelligent content prediction and pre-caching, including prediction model construction, pre-caching strategy formulation, and real-time adjustment and feedback;

[0014] Step VII: cross-network convergence acceleration, including multi-network access and convergence and edge network collaboration;

[0015] Step VIII: Blockchain-driven trusted network, which includes blockchain ledger records, smart contract incentives, and trust assessment and consensus mechanisms.

[0016] Preferably, in the node registration and authentication terminal step, the device applies to the central node to join the PCDN network using a unique identifier (such as a MAC address). After security verification (such as whitelisting and encrypted handshake), a node ID is assigned to ensure the legitimacy of the device. In the resource information reporting step, the node actively submits static resources such as storage / bandwidth / location, as well as dynamic status such as real-time load and cache content list, for the central node to build a global resource map. In the node classification and label management step, nodes are divided into super nodes (high bandwidth / storage), ordinary nodes, and lightweight nodes according to resource capabilities, and are marked with labels such as geographic location and network type to facilitate accurate scheduling.

[0017] Preferably, the content heat grading step divides the content into hot spots / regional hot spots / low frequencies according to the request frequency, and matches different node hierarchical storage (such as super nodes storing the full set of hot spots, and ordinary nodes storing regional fragments). The cache write strategy step actively pre-stores hot content (super nodes pull the full set, and ordinary nodes cache regional fragments) or passively caches requested content, and lightweight nodes only store fragments (such as short video key frames). The cache replacement and elimination step is based on the improved LRU-K algorithm, combining access frequency, geographic location, and node type to calculate elimination priority, thereby releasing low-value cache space.

[0018] Preferably, in the request parsing and node discovery step, the user request triggers the central node to retrieve the online node holding the content, and the filtering conditions include distance, content integrity, and node status (such as bandwidth utilization <70%). The optimal node selection algorithm step quantifies the node priority through multi-dimensional scoring (distance 40% + bandwidth 30% + cache freshness 20% + reputation value 10%), and gives priority to neighboring, low-load, and high-reputation nodes. The content distribution execution step gives priority to local response or P2P segment transmission (such as video split into 10 segments for parallel download), returns to the source central server when there is no cache, and triggers cache preheating. The transmission optimization and flow control step dynamically adjusts the rate based on the BBR / QUIC protocol. When the node load exceeds 80%, new requests are suspended to avoid congestion and ensure stable transmission.

[0019] Preferably, the contribution value measurement step quantifies node contributions according to storage capacity (1GB / day = 1 point), transmission traffic (1GB = 2 points), and relay forwarding (1GB = 1.5 points). The hot content coefficient can be increased. The point exchange and penalty mechanism step exchanges points for traffic / storage and other rights and interests. Points are deducted from nodes with high offline rates or providing erroneous content. In serious cases, they are marked as untrusted nodes. The network topology optimization step regularly cleans up zombie nodes, dynamically adjusts the super node coverage according to the distribution of new nodes, and ensures regional redundancy (at least 3 super nodes / region).

[0020] Preferably, the content encryption transmission step uses AES-256 to encrypt cache content and transmission data to prevent man-in-the-middle attacks and ensure data transmission security. The privacy protection step only stores coarse-grained locations at the provincial / municipal levels, strips away sensitive device information (such as IMEI), and obfuscates resource reporting data through differential privacy technology. The access control step limits content access rights through a Token mechanism. Users need to carry valid credentials to obtain decryption keys to prevent unauthorized access.

[0021] Preferably, the prediction model construction step analyzes historical requests, time, location and other data based on algorithms such as LSTM, predicts future popular content, and plans the cache layout in advance. The pre-cache strategy formulation step stores the content in the corresponding node in advance according to the prediction results (such as super nodes storing global hotspots, and regional nodes storing local predicted content) to improve the cold start hit rate. The real-time adjustment and feedback step dynamically corrects the prediction model according to the actual request data. If the prediction deviation exceeds the threshold (such as 20%), the cache strategy adaptive adjustment is triggered.

[0022] Preferably, the nodes in the multi-network access and fusion step simultaneously access the cellular / Wi-Fi / Bluetooth Mesh network, the central node dynamically allocates the traffic ratio, aggregates multi-link bandwidth to improve transmission efficiency, and the edge network collaboration step is linked with the edge computing node and the Internet of Things gateway to offload some computing tasks (such as transcoding) to the edge layer, reducing the pressure on the central node and shortening the response chain.

[0023] Preferably, the blockchain ledger recording step utilizes blockchain to store data such as node registration, contribution, and content transmission, and ensures that it cannot be tampered with through a hash chain, thereby achieving full process traceability. The smart contract incentive step automatically executes point issuance and redemption through a blockchain smart contract without manual intervention, thereby ensuring that the incentive rules are transparent and the execution is efficient. The trust evaluation and consensus mechanism step calculates the node reputation value based on blockchain records, adopts a PoS / DPoS consensus mechanism to elect trusted nodes to participate in management, and punishes violations (such as deducting points and isolating nodes).

[0024] A PCDN-based distributed edge cache acceleration device includes a terminal node device and a central management node device. The terminal node device includes a resource management and reporting module, an intelligent cache engine module, a P2P transmission module, and a security and privacy module. The central management node device includes a global resource scheduling center, an intelligent routing and prediction engine, an incentive and trust management module, and a security supervision and compliance module.

[0025] Compared with the existing technology, the present invention provides a distributed edge cache acceleration method and device based on PCDN, which has the following beneficial effects:

[0026] 1. This PCDN-based distributed edge cache acceleration method and device uses a deep learning model to predict content popularity and regional distribution in advance, and integrates the bandwidth and resources of different networks through network fusion technology, thereby reducing the number of back-to-source requests, breaking through the limitations of traditional single network transmission (such as relying solely on cellular networks), and achieving traceability of node contributions and incentives for automatic execution.

[0027] 2. This PCDN-based distributed edge caching acceleration method and device integrates idle storage, bandwidth, and other resources on the user side to achieve edge caching and local distribution of content. Through global resource scheduling, intelligent routing, and incentive mechanisms, it improves node resource utilization, dynamically optimizes network topology, and builds a trusted ecosystem, thereby ensuring data security and privacy and enhancing the efficiency, stability, and sustainability of the PCDN network. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a schematic diagram of the architecture of the distributed edge cache acceleration method for PCDN of the present invention;

[0029] Figure 2 This is a schematic diagram of the architecture of the distributed edge cache acceleration device of PCDN of the present invention. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] See also Figure 1-2 A distributed edge cache acceleration method based on PCDN includes the following steps:

[0032] Step I: Terminal node access and resource initialization, which includes node registration and authentication, resource information reporting, and node classification and label management.

[0033] Step II: Layered caching strategy and content storage, which includes content popularity grading, cache write strategy, and cache replacement and elimination;

[0034] Step III: Dynamic content scheduling and distribution, which includes request parsing and node discovery, optimal node selection algorithm, content distribution execution, transmission optimization and flow control;

[0035] Step IV: Node incentives and network maintenance, which include contribution value measurement, point exchange and penalty mechanisms, and network topology optimization.

[0036] Step V: Data security and privacy protection, which includes content encryption transmission, privacy protection, and access control;

[0037] Step VI: Intelligent content prediction and pre-caching, which includes prediction model construction, pre-caching strategy formulation, and real-time adjustment and feedback;

[0038] Step VII: Cross-network convergence acceleration, which includes multi-network access and convergence and edge network collaboration;

[0039] Step VIII: Blockchain-driven trusted network. The blockchain-driven trusted network includes blockchain ledger records, smart contract incentives, and trust assessment and consensus mechanisms.

[0040] In the node registration and authentication step, a terminal device applies to the central node to join the PCDN network using a unique identifier (such as a MAC address). After security verification (such as whitelisting and encrypted handshake), a node ID is assigned to ensure the legitimacy of the device. In the resource information reporting step, the node actively submits static resources such as storage / bandwidth / location, as well as dynamic status such as real-time load and cache content list, so that the central node can build a global resource map. In the node classification and label management step, nodes are divided into super nodes (high bandwidth / storage), ordinary nodes, and lightweight nodes according to resource capabilities, and are marked with tags such as geographic location and network type to facilitate accurate scheduling.

[0041] Through the above technical solutions, the uniqueness of device identity is established, and illegal device access is prevented through security verification (such as TPM chip signature), the legitimacy of network nodes is guaranteed, and malicious devices are prevented from abusing resources. The overall security is improved, and dynamic data such as node storage / bandwidth / location is collected to build a global resource map. The central node has real-time knowledge of network resource distribution, providing a data basis for intelligent scheduling, and resource utilization is improved by 30%+. Resource capacity is graded (super node / ordinary node) and marked with geographic tags to achieve differentiated scheduling (such as super nodes storing the entire hotspot, regional nodes storing fragments), shortening the content distribution path, and reducing latency by 40%+.

[0042] The content heat classification step divides the content into hot spots / regional hot spots / low frequencies according to the request frequency, matching different node levels for storage (such as super nodes storing the entire set of hot spots, and ordinary nodes storing regional fragments). The cache write strategy step actively pre-stores hot content (super nodes pull the entire set, and ordinary nodes cache regional fragments) or passively caches requested content. Lightweight nodes only store fragments (such as short video key frames). The cache replacement and elimination step is based on the improved LRU-K algorithm, combining access frequency, geographic location, and node type to calculate elimination priority and release low-value cache space.

[0043] Through the above technical solution, content levels are divided according to request frequency (hotspot / regional hotspot / low frequency), matching node storage capacity, hot content hit rate is increased by 50%, central server load is reduced by 35%, hot content is actively pre-stored and requested content is passively cached, high-frequency content is arranged in advance, cold start response speed is increased by 60%, user waiting time is reduced, low-value cache is released based on the improved LRU algorithm, cache space utilization is increased to 85%, and the proportion of obsolete content is less than 5%, ensuring efficient use of storage resources.

[0044] In the request parsing and node discovery steps, user requests trigger the central node to retrieve online nodes holding content. The filtering conditions include distance, content integrity, and node status (such as bandwidth utilization <70%). The optimal node selection algorithm step quantifies node priority through multi-dimensional scoring (distance 40% + bandwidth 30% + cache freshness 20% + reputation value 10%), and gives priority to neighboring, low-load, and high-reputation nodes. The content distribution execution step gives priority to local response or P2P fragment transmission (such as splitting the video into 10 segments for parallel download). When there is no cache, it returns to the source central server and triggers cache preheating. The transmission optimization and flow control step dynamically adjusts the rate based on the BBR / QUIC protocol. When the node load exceeds 80%, new requests are suspended to avoid congestion and ensure stable transmission.

[0045] Through the above technical solution, online nodes holding content are retrieved and filtered (such as distance < 50 kilometers, bandwidth utilization < 70%), the range of candidate nodes is narrowed, scheduling efficiency is improved by 70%, invalid queries are avoided, multi-dimensional scoring (distance / bandwidth / reputation value) quantifies node priority, and the accuracy of selecting the optimal node is > 92%. The end-to-end delay is stable at 10-20ms, P2P fragment transmission is prioritized, and when there is no cache, the source is returned and preheating is triggered. The content distribution efficiency is improved by 200%, the return rate is reduced to less than 15%, saving central bandwidth costs, and the rate is dynamically adjusted based on the QUIC protocol to prevent node overload. The transmission stability is improved by 90%, and the node bandwidth utilization is balanced and controlled at 70%-80%, ensuring that users can access the Internet normally.

[0046] The contribution value measurement step quantifies node contributions based on storage capacity (1GB / day = 1 point), transmission traffic (1GB = 2 points), and relay forwarding (1GB = 1.5 points). The hot content coefficient can be increased. In the point redemption and penalty mechanism step, points are redeemed for traffic / storage and other rights. Points are deducted from nodes with high offline rates or providing erroneous content. In serious cases, they are marked as untrusted nodes. In the network topology optimization step, zombie nodes are regularly cleaned up, and the super node coverage is dynamically adjusted according to the distribution of new nodes to ensure regional redundancy (at least 3 super nodes / region).

[0047] Through the above technical solution, node contributions are quantified according to storage / transmission / relay, node participation enthusiasm is increased by 80%, resource contribution increases by 250% year-on-year, positive incentives are combined with violation penalties (such as point deductions and node isolation), node online stability is improved to 95%, content accuracy is greater than 99%, forming a virtuous ecological cycle, zombie nodes are regularly cleaned up, super node coverage is dynamically adjusted, network connectivity is improved to 99.9%, regional redundant nodes ensure disaster recovery capabilities, and fault recovery time is less than 5 minutes.

[0048] The content encryption transmission step uses AES-256 to encrypt cached content and transmitted data to prevent man-in-the-middle attacks and ensure data transmission security. The privacy protection step only stores coarse-grained locations at the provincial / municipal level, strips sensitive device information (such as IMEI), and uses differential privacy technology to obfuscate resource reporting data. The access control step uses a token mechanism to limit content access rights. Users must carry valid credentials to obtain decryption keys to prevent unauthorized access.

[0049] Through the above technical solution, AES-256 encryption prevents data theft, the transmission link security level reaches financial-grade standards, the man-in-the-middle attack interception rate is 100%, the geographic location is blurred, sensitive information (such as IMEI) is stripped, and the risk of user privacy leakage is reduced to 0.01%, complying with compliance requirements such as GDPR. Token tokens verify permissions and bind decryption keys, with a 100% unauthorized access interception rate, ensuring the controllability of content distribution.

[0050] The prediction model construction step analyzes historical requests, time, location and other data based on algorithms such as LSTM, predicts future popular content, and plans the cache layout in advance. The pre-cache strategy formulation step stores the content in the corresponding node in advance according to the prediction results (such as super nodes store global hotspots, and regional nodes store local predicted content) to improve the cold start hit rate. The real-time adjustment and feedback step dynamically corrects the prediction model according to the actual request data. If the prediction deviation exceeds the threshold (such as 20%), the cache strategy adaptive adjustment is triggered.

[0051] Through the above technical solution, LSTM predicts future popular content and regional distribution, the pre-cache hit rate is increased by 40%, the efficiency of dealing with sudden traffic is improved by 70%, the cost of manual intervention is reduced, the predicted content is stored in the corresponding node in advance, the response time of cold start requests is shortened to less than 50ms, the user freeze rate is reduced by 65%, the prediction model is corrected according to actual requests, the prediction deviation is controlled within 10%, and the model adaptation cycle is shortened to 1 hour.

[0052] In the multi-network access and fusion step, nodes simultaneously access the cellular / Wi-Fi / Bluetooth Mesh network. The central node dynamically allocates traffic proportions and aggregates multi-link bandwidth to improve transmission efficiency. The edge network collaboration step is linked with edge computing nodes and IoT gateways to offload some computing tasks (such as transcoding) to the edge layer, reducing the pressure on the central node and shortening the response chain.

[0053] Through the above technical solution, cellular / Wi-Fi / Bluetooth Mesh bandwidth is aggregated, the transmission rate is increased by 150%-200%, the stability in weak network environments is improved by 40%, the coverage blind spots are reduced by 50%, and by linking with edge computing nodes to offload tasks such as transcoding, the computing load of central nodes is reduced by 45%, the dynamic content processing delay is less than 10ms, and real-time interactive scenarios are supported.

[0054] The blockchain ledger recording step uses the blockchain to store data such as node registration, contribution, and content transmission, and ensures that it cannot be tampered with through the hash chain, achieving traceability of the entire process. The smart contract incentive step automatically executes the issuance and redemption of points through the blockchain smart contract without human intervention, ensuring that the incentive rules are transparent and the execution is efficient. The trust assessment and consensus mechanism step calculates the node reputation value based on blockchain records, adopts the PoS / DPoS consensus mechanism to elect trusted nodes to participate in management, and punishes violations (such as deducting points and isolating nodes).

[0055] Through the above technical solution, node behavior data is stored, cannot be tampered with and is traceable, the credibility of the data is improved to 99.99%, the audit efficiency is improved by 80%, and regulatory compliance requirements are met. Points are automatically issued and redeemed without human intervention, the incentive execution efficiency is improved by 95%, the node income error is less than 0.1%, and the possibility of human cheating is eliminated. The reputation value is calculated based on blockchain records, and PoS elects trusted nodes. The node cheating rate is reduced to below 0.1%, the trust establishment efficiency is improved by 90%, and the network autonomy is significantly enhanced.

[0056] A PCDN-based distributed edge cache acceleration device includes a terminal node device and a central management node device. The terminal node device includes a resource management and reporting module, an intelligent cache engine module, a P2P transmission module and a security and privacy module. The central management node device includes a global resource scheduling center, an intelligent routing and prediction engine, an incentive and trust management module and a security supervision and compliance module. The resource management and reporting module collects node resource data in real time, provides a basis for scheduling, and improves resource utilization; the intelligent cache engine module optimizes content storage and improves hit rate through layered and semantic caching; the P2P transmission module accelerates content distribution and ensures stable transmission with the help of protocol optimization and penetration technology; the security and privacy module strengthens data encryption and access control to protect security and compliance; the global resource scheduling center integrates resources to achieve load balancing; the intelligent routing and prediction engine accurately matches nodes and pre-caches content to reduce latency; the incentive and trust management module stimulates node vitality and credibility through points and blockchain mechanisms; the security supervision and compliance module ensures content legality and privacy protection, and builds a solid network security defense line. Each module works together to drive an efficient, reliable and secure distributed acceleration network.

[0057] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A distributed edge cache acceleration method based on PCDN, characterized in that: The following steps are involved: Step I: Terminal node access and resource initialization, which includes node registration and authentication, resource information reporting, and node classification and label management; Step II: Layered caching strategy and content storage, including content popularity grading, cache write strategy, and cache replacement and elimination; Step III: Dynamic content scheduling and distribution, which includes request parsing and node discovery, optimal node selection algorithm, content distribution execution, transmission optimization and flow control; Step IV: Node incentives and network maintenance, including contribution value measurement, point exchange and penalty mechanism, and network topology optimization; Step V: Data security and privacy protection, including content encryption transmission, privacy protection, and access control; Step VI: Intelligent content prediction and pre-caching, including prediction model construction, pre-caching strategy formulation, and real-time adjustment and feedback; Step VII: cross-network convergence acceleration, including multi-network access and convergence and edge network collaboration; Step VIII: Blockchain-driven trusted network, which includes blockchain ledger records, smart contract incentives, and trust assessment and consensus mechanisms.

2. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: In the node registration and authentication terminal step, a device applies to the central node to join the PCDN network using a unique identifier (such as a MAC address). After security verification (such as whitelisting and encrypted handshake), a node ID is assigned to ensure the legitimacy of the device. In the resource information reporting step, the node actively submits static resources such as storage / bandwidth / location, as well as dynamic status such as real-time load and cache content list, for the central node to build a global resource map. In the node classification and label management step, nodes are divided into super nodes (high bandwidth / storage), ordinary nodes, and lightweight nodes according to resource capabilities, and are marked with tags such as geographic location and network type to facilitate accurate scheduling.

3. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: The content heat classification step divides the content into hot spots / regional hot spots / low frequencies according to the request frequency, and matches different node hierarchical storage (such as super nodes storing the entire set of hot spots, and ordinary nodes storing regional fragments). The cache write strategy step actively pre-stores hot content (super nodes pull the entire set, and ordinary nodes cache regional fragments) or passively caches requested content. Lightweight nodes only store fragments (such as short video key frames). The cache replacement and elimination step is based on the improved LRU-K algorithm, combining access frequency, geographic location, and node type to calculate elimination priority and release low-value cache space.

4. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: In the request parsing and node discovery step, the user request triggers the central node to retrieve the online node holding the content. The filtering conditions include distance, content integrity, and node status (such as bandwidth utilization <70%). The optimal node selection algorithm step quantifies the node priority through multi-dimensional scoring (distance 40% + bandwidth 30% + cache freshness 20% + reputation value 10%), and gives priority to neighboring, low-load, and high-reputation nodes. The content distribution execution step gives priority to local response or P2P fragment transmission (such as splitting the video into 10 segments for parallel download). When there is no cache, it returns to the source central server and triggers cache preheating. The transmission optimization and flow control step dynamically adjusts the rate based on the BBR / QUIC protocol. When the node load exceeds 80%, new requests are suspended to avoid congestion and ensure stable transmission.

5. The PCDN-based distributed edge cache acceleration method according to claim 1, wherein: The contribution value measurement step quantifies node contributions based on storage capacity (1GB / day = 1 point), transmission traffic (1GB = 2 points), and relay forwarding (1GB = 1.5 points). The hot content coefficient can be increased. The point exchange and penalty mechanism step exchanges points for traffic / storage and other rights. Points are deducted from nodes with high offline rates or providing erroneous content. In serious cases, they are marked as untrusted nodes. The network topology optimization step regularly cleans up zombie nodes and dynamically adjusts the super node coverage according to the distribution of new nodes to ensure regional redundancy (at least 3 super nodes / region).

6. The PCDN-based distributed edge cache acceleration method and device according to claim 1, characterized in that: The content encryption transmission step uses AES-256 to encrypt cached content and transmitted data to prevent man-in-the-middle attacks and ensure data transmission security. The privacy protection step only stores coarse-grained locations at the provincial / municipal levels, strips away sensitive device information (such as IMEI), and obfuscates resource reporting data through differential privacy technology. The access control step limits content access rights through a token mechanism. Users need to carry valid credentials to obtain decryption keys to prevent unauthorized access.

7. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: The prediction model construction step analyzes historical requests, time, location and other data based on algorithms such as LSTM, predicts future popular content, and plans the cache layout in advance. The pre-cache strategy formulation step stores the content in the corresponding node in advance according to the prediction results (such as super nodes store global hotspots, and regional nodes store local predicted content) to improve the cold start hit rate. The real-time adjustment and feedback step dynamically corrects the prediction model according to the actual request data. If the prediction deviation exceeds the threshold (such as 20%), the cache strategy adaptive adjustment is triggered.

8. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: The nodes in the multi-network access and fusion step simultaneously access the cellular / Wi-Fi / Bluetooth Mesh network, the central node dynamically allocates traffic ratios, aggregates multi-link bandwidth to improve transmission efficiency, and the edge network collaboration step is linked with edge computing nodes and IoT gateways to offload some computing tasks (such as transcoding) to the edge layer, reducing the pressure on the central node and shortening the response chain.

9. The PCDN-based distributed edge cache acceleration method according to claim 1, characterized in that: The blockchain ledger recording step utilizes blockchain to store data such as node registration, contribution, and content transmission, and ensures that it cannot be tampered with through a hash chain, thereby achieving full process traceability. The smart contract incentive step automatically executes point issuance and redemption through blockchain smart contracts without manual intervention, ensuring that the incentive rules are transparent and the execution is efficient. The trust assessment and consensus mechanism step calculates the node reputation value based on blockchain records, adopts the PoS / DPoS consensus mechanism to elect trusted nodes to participate in management, and punishes violations (such as deducting points and isolating nodes).

10. The PCDN-based distributed edge cache acceleration device according to claim 1, comprising a terminal node device and a central management node device, characterized in that: The terminal node device includes a resource management and reporting module, an intelligent cache engine module, a P2P transmission module and a security and privacy module, and the central management node device includes a global resource scheduling center, an intelligent routing and prediction engine, an incentive and trust management module and a security supervision and compliance module.

Images