Caching Mechanism Research Articles

CTJIF-ICN: A Coadjuvant Trust Joint Interest Forwarding Mechanism in Information Centric Networks

The Information centric networks (ICN) transforms the focal point of current Internet paradigm to data centric approach from host centric approach by allowing content driven forwarding and in-network caching mechanisms. Though NDN (Named data networking) paradigm of ICN assures a secure content communication, it is vulnerable to different attacks by the malicious nodes. To minimize the hazards from compromised nodes and to improve the network security, the remaining nodes should transparently receive information about such nodes. This will restrict the forwarding strategy to exploit these malicious nodes for forwarding interest and content as well. Our protocol introduces a dynamic model for prediction of trust in order to evaluate the node trust. Proposed approach observes the historical behaviors of node and uses extended fuzzy logic rules for the prediction of future behaviors to evaluate the node’s trust value. This prediction model is incorporated within the trust based forwarding mechanism that aims to forward interest through secure and shortest path. The extensive simulation study has been carried out to analyze the protocol performance in ns-3 driven ndnSIM-2.0 simulator for performance metrics such as data discovery latency, packet delivery ratio, network overhead, detection ratio and cache hit ratio. When we integrate our trust joint forwarding strategy to state-of-the-art protocols, their performance is significantly improved up to approximately 10-35% against stated performance measures for realistic network topology.

Efficient asynchronous federated learning with sparsification and quantization

SummaryWhile data is distributed in multiple edge devices, federated learning (FL) is attracting more and more attention to collaboratively train a machine learning model without transferring raw data. FL generally exploits a parameter server and a large number of edge devices during the whole process of the model training, while several devices are selected in each round. However, straggler devices may slow down the training process or even make the system crash during training. Meanwhile, other idle edge devices remain unused. As the bandwidth between the devices and the server is relatively low, the communication of intermediate data becomes a bottleneck. In this article, we propose time‐efficient asynchronous federated learning with sparsification and quantization, that is, TEASQ‐Fed. TEASQ‐Fed can fully exploit edge devices to asynchronously participate in the training process by actively applying for tasks. We utilize control parameters to choose an appropriate number of parallel edge devices, which simultaneously execute the training tasks. In addition, we introduce a caching mechanism and weighted averaging with respect to model staleness to further improve the accuracy. Furthermore, we propose a sparsification and quantitation approach to compress the intermediate data to accelerate the training. The experimental results reveal that TEASQ‐Fed improves the accuracy (up to 16.67% higher) while accelerating the convergence of model training (up to twice faster).

Real-time processing and optimization strategies for IoT data streams

Abstract With the development of industrial IoT and the arrival of smart manufacturing, the field of edge computing has gained more and more attention. However, traditional industrial computing scenarios relying on industrial clouds make data latency a greater challenge. In this paper, for the contradiction between edge devices and task resource allocation encountered in edge computing scenarios in smart manufacturing, we propose an industrial internet task scheduling model for smart manufacturing and introduce a scheduling node state matrix to realize the state management of each scheduling subtask. Aiming at the problem of multiple tasks seizing resources in a complex, intelligent manufacturing environment, the study combines the caching mechanism to realize the task offloading computational processing of order scheduling, in which the caching mechanism is used to solve the problem of computational resource limitations at the edge. It is found through simulation that when the computational task factor ξk =2 is larger, more offloading power is allowed to be transmitted to the edge ser ver for computation. For computational tasks with smaller task factor ξk , the device tends to allocate more computational rate to that computational task. Eventually the data queue length will be continuously reduced and the data queue is concentrated in the interval of very small values, this result verifies that the task scheduling algorithm is able to perform task scheduling efficiently and reduce the latency.

Cost-Optimized Dynamic Access Control Policy Using Blockchain and Machine Learning for Enhanced Security in IoT Smart Homes

The rapid adoption of Internet of Things (IoT) devices in smart homes has led to growing security vulnerabilities, primarily due to the limitations of traditional, static access control mechanisms. This paper presents a novel, dynamic access control policy that leverages the immutable and transparent nature of Blockchain technology, specifically Ethereum, along with machine learning algorithms to enhance security measures. By integrating machine learning algorithms like Support Vector Machines (SVM) and Neural Networks, the proposed system can adapt and respond to changing behavioural patterns and potential threats in real time. Additionally, a caching mechanism implemented on the Ethereum Blockchain is introduced to optimize system performance and reduce latency. Experimental results demonstrate significant improvements in access control security, system efficiency, and adaptability. The findings of this paper not only contribute to the advancement of secure access control policies for IoT smart homes but pave the way for future research in integrating Blockchain and machine learning for robust and scalable IoT security solutions.

Performance Optimization of Machine Learning Algorithms Based on Spark

Abstract This paper proposes a performance optimization strategy for Spark-based machine learning algorithms in Shuffle and memory data management modules. The Shuffle module is optimized by introducing Observer monitoring module in Spark cluster to achieve task status monitoring and dynamic ShuffleWrite task generation. Meanwhile, an adaptive caching mechanism for RDD data addresses the lack of in-memory data caching. The performance-optimized algorithm performs well in the experiments, with a clustering accuracy of 89% and a response time that is 5% faster than the Random Forest algorithm. In road network traffic state discrimination, the optimized algorithm’s classification decision F-measure value is as high as 99.53%, which is 5.32% higher than that before unoptimization, and the running time is 767 seconds less than that of the unoptimized algorithm when dealing with about 6,880,000 pieces of data, which significantly improves the efficiency and accuracy.

Large-scale Data Mining Method based on Clustering Algorithm Combined with MAPREDUCE

With the continuous deepening and development of information technology, the diversity and amount of information in data continue to grow. Effectively mining these text data to extract valuable content has become an urgent task in the field of data research. This study combines the MapReduce distributed system with the K-means clustering algorithm to meet the challenges of large-scale data mining. At the same time, the paper use a distributed caching mechanism to solve the problem of repeated application of resources for multiple MapReduce collaborative operations and improve data mining efficiency. The combination of MapReduce's distributed computing and the advantages of K-means clustering algorithm provides an efficient and scalable method for large-scale data mining. Experimental results combining internal and external indicators show that the advantage of combining K-means with MapReduce is to fully utilize the distributed and parallel computing characteristics of MapReduce, providing users with an efficient and scalable data mining tool. Through this research, the paper provide new methods and insights for large-scale data mining, improving the efficiency and accuracy of data mining.

Improving Cache Hits On Replacment Blocks Using Weighted LRU-LFU Combinations

Block replacement refers to the process of selecting a block of data or a cache line to be evicted or replaced when a new block needs to be brought into a cache or a memory hierarchy. In computer systems, block replacement policies are used in caching mechanisms, such as in CPU caches or disk caches, to determine which blocks are evicted when the cache is full and new data needs to be fetched. The combination of LRU (Least Recently Used) and LFU (Least Frequently Used) in a weighted manner is known as the "LFU2" algorithm. LFU2 is an enhanced caching algorithm that aims to leverage the benefits of both LRU and LFU by considering both recency and frequency of item access. In LFU2, each item in the cache is associated with two counters: the usage counter and the recency counter. The usage counter tracks the frequency of item access, while the recency counter tracks the recency of item access. These counters are used to calculate a combined weight for each item in the cache. Based on the experimental results, the LRU-LFU combination method succeeded in increasing cache hits from 94.8% on LFU and 95.5% on LFU to 96.6%.

Blockchain-Enabled On-Path Caching for Efficient and Reliable Content Delivery in Information-Centric Networks

As the demand for online content continues to grow, traditional Content Distribution Networks (CDNs) are facing significant challenges in terms of scalability and performance. Information-Centric Networking (ICN) is a promising new approach to content delivery that aims to address these issues by placing content at the center of the network architecture. One of the key features of ICNs is on-path caching, which allows content to be cached at intermediate routers along the path from the source to the destination. On-path caching in ICNs still faces some challenges, such as the scalability of the cache and the management of cache consistency. To address these challenges, this paper proposes several alternative caching schemes that can be integrated into ICNs using blockchain technology. These schemes include Bloom filters, content-based routing, and hybrid caching, which combine the advantages of off-path and on-path cachings. The proposed blockchain-enabled on-path caching mechanism ensures the integrity and authenticity of cached content, and smart contracts automate the caching process and incentivize caching nodes. To evaluate the performance of these caching alternatives, the authors conduct experiments using real-world datasets. The results show that on-path caching can significantly reduce network congestion and improve content delivery efficiency. The Bloom filter caching scheme achieved a cache hit rate of over 90% while reducing the cache size by up to 80% compared to traditional caching. The content-based routing scheme also achieved high cache hit rates while maintaining low latency.

EMS-i : An Efficient Memory System Design with Specialized Caching Mechanism for Recommendation Inference

Recommendation systems have been widely embedded into many Internet services. For example, Meta’s deep learning recommendation model (DLRM) shows high prefictive accuracy of click-through rate in processing large-scale embedding tables. The SparseLengthSum (SLS) kernel of the DLRM dominates the inference time of the DLRM due to intensive irregular memory accesses to the embedding vectors. Some prior works directly adopt near data processing (NDP) solutions to obtain higher memory bandwidth to accelerate SLS. However, their inferior memory hierarchy induces low performance-cost ratio and fails to fully exploit the data locality. Although some software-managed cache policies were proposed to improve the cache hit rate, the incurred cache miss penalty is unacceptable considering the high overheads of executing the corresponding programs and the communication between the host and the accelerator. To address the issues aforementioned, we propose EMS-i , an efficient memory system design that integrates Solide State Drive (SSD) into the memory hierarchy using Compute Express Link (CXL) for recommendation system inference. We specialize the caching mechanism according to the characteristics of various DLRM workloads and propose a novel prefetching mechanism to further improve the performance. In addition, we delicately design the inference kernel and develop a customized mapping scheme for SLS operation, considering the multi-level parallelism in SLS and the data locality within a batch of queries. Compared to the state-of-the-art NDP solutions, EMS-i achieves up to 10.9× speedup over RecSSD and the performance comparable to RecNMP with 72% energy savings. EMS-i also saves up to 8.7× and 6.6 × memory cost w.r.t. RecSSD and RecNMP, respectively.

An Optimized AMS Based Cloud Downloading Service with Advanced Caching and Intelligent Data Distribution Mechanism

The popularity of peer-to-peer video content downloading has surged due to diverse content availability and convenient sharing among users. However, scaling systems to accommodate the growing number of users and content items poses a challenge. This research aims to optimize video content downloading in peer-to-peer systems. The objective is to improve performance by developing advanced caching mechanisms, an intelligent data distribution algorithm, and efficient bandwidth resource management. The proposed approach involves implementing innovative caching mechanisms that store frequently accessed content closer to users, reducing download time. An intelligent data distribution algorithm minimizes bottlenecks and maximizes download speeds. Efficient bandwidth resource management ensures fair allocation. Results demonstrate significant enhancements in download time and overall system performance, leading to improved user experience. This research addresses the need for an optimized video content downloading system to handle increasing user and content volumes. The findings hold the potential to enhance user experiences, facilitate seamless video sharing, and advance peer-to-peer video content downloading.

Hybrid Internet Architecture and Protocol (HIAP): A Self-Evolving and Transformative Framework for Enabling Seamless Real-Time Applications and Secure Peer-to-Peer File Sharing in the Internet of Everything (IoE)

The Internet of Everything (IoE) is rapidly growing and utilized in various applications. However, the growing mobile traffic and services pose significant challenges in flexibility, movement, accessibility, and safety. The existing internet architecture and protocol must be improved to address the challenges. The article examines the attributes and prerequisites of forthcoming networking applications while emphasizing the constraints of conventional network architecture and protocols in fulfilling these needs. The research presents a new Hybrid Internet Architecture and Protocol (HIAP), the Self-Evolving and Transformative (SET) architecture. This architecture is designed to provide diverse control works and smart configuration options for different applications and networking conditions. This HIAP framework's primary emphasis lies in transport protocols and mechanisms for peer-to-peer file sharing. This study proposes implementing a deadline-aware multipath transport protocol within the framework of the Internet architecture with seamless support of real-time applications requiring strict adherence to latency demands. The HIAP framework incorporates evolvability, adaptable routing, and in-network cache mechanisms to enhance contentdelivery. To conduct a more comprehensive examination and conceptualize the peer-to-peer file-sharing system, the research created a pragmatic blueprint for setting the file transfer protocol on the HIAP. The study presented the groundwork for an Internet architecture better suited to the changing demands of the IoE and its various applications. This HIAP architecture aims to be more adaptable, efficient, and transformative. The present study involves a comparative analysis between the proposed research and current architecture. The findings provide evidence of the progress made by the suggested research in addressing the challenges related to multipath and safety in future architectural designs.

A Smart Control System for the Oil Industry Using Text-to-Speech Synthesis Based on IIoT

Oil refineries have high operating expenses and are often exposed to increased asset integrity risks and functional failure. Real-time monitoring of their operations has always been critical to ensuring safety and efficiency. We proposed a novel Industrial Internet of Things (IIoT) design that employs a text-to-speech synthesizer (TTS) based on neural networks to build an intelligent extension control system. We enhanced a TTS model to achieve high inference speed by employing HiFi-GAN V3 vocoder in the acoustic model FastSpeech 2. We experimented with our system on a low resources-embedded system in a real-time environment. Moreover, we customized the TTS model to generate two target speakers (female and male) using a small dataset. We performed an ablation analysis by conducting experiments to evaluate the performance of our design (IoT connectivity, memory usage, inference speed, and output speech quality). The results demonstrated that our system Real-Time Factor (RTF) is 6.4 (without deploying the cache mechanism, which is a technique to call the previously synthesized speech sentences in our system memory). Using the cache mechanism, our proposed model successfully runs on a low-resource computational device with real-time speed (RTF equals 0.16, 0.19, and 0.29 when the memory has 250, 500, and 1000 WAV files, respectively). Additionally, applying the cache mechanism has reduced memory usage percentage from 16.3% (for synthesizing a sentence of ten seconds) to 6.3%. Furthermore, according to the objective speech quality evaluation, our TTS model is superior to the baseline TTS model.

A DDPG Hybrid of Graph Attention Network and Action Branching for Multi-Scale End-Edge-Cloud Vehicular Orchestrated Task Offloading

With the development of 5G technologies and the wide application of artificial intelligence (AI), the mobile intelligent equipment and the providing services have both seen a significant rise in number and type. Some services such as vehicular tasks may go beyond the capability of the mobile equipment so that task offloading is required to help deliver the services. However, the graph structure of task offloading data which can be a key to further improve algorithm performance is seldom considered in future 6G-AI combined communication systems. In this paper, we propose an efficient end-edge-cloud orchestration system that combines storage-partition and computation-shared in cloud cluster, cache mechanism and cybertwin components. At the same time, we model this dynamic system as a graph structure composed of nodes and edges and propose a novel task offloading algorithm that incorporates graph attention network (GAT) and action branching into deep deterministic policy gradient (DDPG) framework. Numerical results show that our offloading scheme achieves a good performance boost compared with other baselines.

A Content Poisoning Attack Detection and Prevention System in Vehicular Named Data Networking

Named data networking (NDN) is gaining momentum in vehicular ad hoc networks (VANETs) thanks to its robust network architecture. However, vehicular NDN (VNDN) faces numerous challenges, including security, privacy, routing, and caching. Specifically, the attackers can jeopardize vehicles’ cache memory with a Content Poisoning Attack (CPA). The CPA is the most difficult to identify because the attacker disseminates malicious content with a valid name. In addition, NDN employs request–response-based content dissemination, which is inefficient in supporting push-based content forwarding in VANET. Meanwhile, VNDN lacks a secure reputation management system. To this end, our contribution is three-fold. We initially propose a threshold-based content caching mechanism for CPA detection and prevention. This mechanism allows or rejects host vehicles to serve content based on their reputation. Secondly, we incorporate a blockchain system that ensures the privacy of every vehicle at roadside units (RSUs). Finally, we extend the scope of NDN from pull-based content retrieval to push-based content dissemination. The experimental evaluation results reveal that our proposed CPA detection mechanism achieves a 100% accuracy in identifying and preventing attackers. The attacker vehicles achieved a 0% cache hit ratio in our proposed mechanism. On the other hand, our blockchain results identified tempered blocks with 100% accuracy and prevented them from storing in the blockchain network. Thus, our proposed solution can identify and prevent CPA with 100% accuracy and effectively filters out tempered blocks. Our proposed research contribution enables the vehicles to store and serve trusted content in VNDN.

Two-Hop Cooperative Caching and UAVs Deployment Based on Potential Game

This paper explores the joint cache placement and 3D deployment of Unmanned Aerial Vehicle (UAV) groups, utilizing potential game theory and a two-hop UAV cooperative caching mechanism, which could create a tradeoff between latency and coverage. The proposed scheme consists of three parts: first, the initial 2D location of UAV groups is determined through K-means, with the optimal altitude based on the UAV coverage radius. Second, to balance the transmission delay and coverage, the MOS (Mean Opinion Score) and coverage are designed to evaluate the performance of UAV-assisted networks. Then, the potential game is modeled, which transfers the optimization problem into the maximization of the whole network utility. The locally coupling effect resulting from action changes among UAVs is considered in the design of the potential game utility function. Moreover, a log-linear learning scheme is applied to solve the problem. Finally, the simulation results verify the superiority of the proposed scheme in terms of the achievable transmission delay and coverage performance compared with two other tested schemes. The coverage ratio is close to 100% when the UAV number is 25, and the user number is 150; in addition, this game outperforms the benchmarks when it comes to maximizing MOS of users.

Detection of Cache Pollution Attack Based on Ensemble Learning in ICN-Based VANET

Content Centric Network (CCN) can be extended to efficiently and reliably support content delivery and solve the network performance degradation caused by dynamic topology and intermittent connectivity of Vehicle Ad hoc NETwork (VANET). However, the in-network caching mechanism of Vehicular Content Centric Network (VCCN) is vulnerable against Cache Pollution Attack (CPA), where attackers aim to fill the buffer space with non-popular contents by releasing fake requests. Unavoidably, the cache hit ratio of content requests from legal users is degraded and the content retrieval latency is increased under CPA. Hence, it is critical to detect and mitigate CPA. The current solutions for static CCN cannot be directly applied into dynamic VCCN. In this paper, we propose a detection scheme based on hybrid heterogeneous multi-classifier ensemble learning, where CPA is determined by the cooperation of multiple vehicles. In our scheme, each vehicle can build or join a cluster whose head possesses more common moving attributes of position, speed and direction with other members. Besides, the cluster head as a base learner is responsible for training its own classifier by making some relevant statistics on requests and hit ratio. Specifically, the problem of ensemble classifier making from the individual classifiers is formulated as a linear optimization problem, with the goal of minimizing the false ratio of detecting CPA. The generalization ability of ensemble learning can make very accurate predictions on CPA. By comparison, our detection scheme outperforms the existing schemes in terms of detection ratio, hit ratio, retrieval delay. Besides, simulations have proved that the overfitting problem of adopting a singe base learning algorithm can be alleviated in our scheme.

Stream Aggregation with Compressed Sliding Windows

High performance stream aggregation is critical for many emerging applications that analyze massive volumes of data. Incoming data needs to be stored in a sliding window during processing, in case the aggregation functions cannot be computed incrementally. Updating the window with new incoming values and reading it to feed the aggregation functions are the two primary steps in stream aggregation. Although window updates can be supported efficiently using multi-level queues, frequent window aggregations remain a performance bottleneck as they put tremendous pressure on the memory bandwidth and capacity. This article addresses this problem by enhancing StreamZip, a dataflow stream aggregation engine that is able to compress the sliding windows. StreamZip deals with a number of data and control dependency challenges to integrate a compressor in the stream aggregation pipeline and alleviate the memory pressure posed by frequent aggregations. In addition, StreamZip incorporates a caching mechanism for dealing with skewed-key distributions in the incoming data stream. In doing so, StreamZip offers higher throughput as well as larger effective window capacity to support larger problems. StreamZip supports diverse compression algorithms offering both lossless and lossy compression to integers as well as floating-point numbers. Compared to designs without compression, StreamZip lossless and lossy designs achieve up to 7.5× and 22× higher throughput, while improving the effective memory capacity by up to 5× and 23×, respectively.

µFuncCache: A User-Side Lightweight Cache System for Public FaaS Platforms

Building cloud-native applications based on public “Function as a Service” (FaaS) platforms has become an attractive way to improve business roll-out speed and elasticity, as well as reduce cloud usage costs. Applications based on FaaS are usually designed with multiple different cloud functions based on their functionality, and there will be call relationships between cloud functions. At the same time, each cloud function may depend on other services provided by cloud providers, such as object storage services, database services, and file storage services. When there is a call relationship between cloud functions, or between cloud functions and other services, a certain delay will occur, and the delay will increase with the length of the call chain, thereby affecting the quality of application services and user experience. Therefore, we introduce μFuncCache, a user-side lightweight caching mechanism to speed up data access for public FaaS services, fully utilizing the container delay destruction mechanism and over-booked memory commonly found in public FaaS platforms, to reduce function call latency without the need to perceive and modify the internal architecture of public clouds. Experiments in different application scenarios have shown that μFuncCache can effectively improve the performance of FaaS applications by consuming only a small amount of additional resources, while achieving a maximum reduction of 97% in latency.

An Improved Shortest Path Campus Navigation Algorithm

Navigation within college campuses can be a challenging task for students, staff, and visitors. With the increasing reliance on smartphones as a common utility, it is possible to develop a dedicated navigation system to aid in navigating these complex environments. In an effort to improve upon traditional navigation methods, we explored the use of the A* algorithm and identified an opportunity to enhance its performance by implementing bidirectional search. This allows the algorithm to search simultaneously from both the start and end points, reducing the search space and improving efficiency. Additionally, we utilized a caching mechanism to store previously calculated paths, which allows the algorithm to reuse these values and further improve its performance. Our approach offers a unique solution to the problem of navigating on college campuses and has the potential to significantly improve the user experience.

AFDX network equipment fault diagnosis technology

This paper focuses on the network equipment fault monitoring and diagnosis software, and studies the fault diagnosis of the monitored AFDX network based on the network algorithm. Firstly, the range fault characteristic parameters are designed to identify the fault type, and the correlation between the detection results and the fault characteristic parameters at each location can be obtained. Secondly, the data storage management scheme of the first level filtering and the second level caching mechanism is designed for the data collected in the detection. Then, according to the designed fault classification, fault diagnosis methods are given respectively, and the occasional anomalies are identified and suppressed. Finally, the network fault diagnosis verification module is designed, and the experimental verification is carried out from the perspectives of real-time and concurrency. The verification results prove the effectiveness of the method.