Traditional Cache Research Articles

On the Use of Embedding Techniques for Modeling User Navigational Behavior in Intelligent Prefetching Strategies

ABSTRACTIn today's data‐intensive client‐server systems, traditional caching methods often fail to meet the demands of modern applications, especially in mobile environments with unstable network conditions. This research addresses the challenge of improving data delivery by proposing an advanced prefetching framework that utilizes various embedding techniques. We explore how to model user navigation using graph‐based, autoencoder‐based, and sequence‐to‐sequence‐based embedding methods and assess their impact on prefetching accuracy and efficiency. Our study shows that utilizing these embedding techniques with supervised learning models improves prefetching performance. We also present a software architecture that blends supervised and unsupervised learning approaches, along with user‐specific and collective learning models, to create a robust prefetching mechanism. The contributions of this study include developing a scalable prefetching solution using machine learning/deep learning algorithms and providing an open‐source prototype of the proposed architecture. This paper offers a significant improvement over previous research and provides valuable insights for enhancing the performance of data‐intensive applications.

Optimistic online caching for batched requests

In this paper, we investigate ‘optimistic’ online caching policies, distinguished by their use of future request predictions derived, for example, from machine learning models. Traditional online optimistic policies, grounded in the Follow-The-Regularized-Leader (FTRL) algorithm, incur a higher computational cost compared to classic policies like Least Frequently Used (LFU) and Least Recently Used (LRU). This is due to each cache state update necessitating the resolution of a constrained optimization problem. To address this problem, we introduce and analyze the ‘batched’ version of two distinct FTRL-based optimistic policies. In this approach, the cache updates occur less frequently, thereby amortizing the update cost over time or over multiple requests. Rather than updating the cache with each new request, the system accumulates a batch of requests before modifying the cache content. First, we present a batched version of the Optimistic Bipartite Caching (OBC) algorithm, that works for single requests, then we introduce a new optimistic batched caching policy, the Per-Coordinate Optimistic Caching (PCOC) algorithm, derived from the per-coordinate-based FTRL. We demonstrate that these online algorithms maintain ‘vanishing regret’ in the batched case, meaning their average performance approaches over time that of an optimal static file allocation, regardless of the sequence of file requests. We then compare the performance of these two strategies with each other and against optimistic versions of LFU and LRU. Our experimental results indicate that this batched optimistic approach outperforms traditional caching policies on both stationary and real-world file request traces.

Edge collaborative caching solution based on improved NSGA II algorithm in Internet of Vehicles

With the rapid development of the Internet of Vehicles technology and the continuous advancement of intelligent transportation, vehicle network traffic is growing exponentially. Traditional caching solutions transmit content through the core network, causing significant latency. A vehicle-cloud collaborative edge caching network architecture was proposed to address this type of problem. Using this architecture, both the on-board terminals and cloud/edge servers could provide computing services. A communication model, caching model, latency model, energy consumption model, load model and multi-objective optimization problem model were designed. The edge collaborative caching solution based on the improved NSGA II algorithm was proposed. By caching some services to edge nodes and nearby vehicles, reducing content access latency and improving resource utilization. The results indicate that the caching solution outperforms the comparison scheme in terms of the comprehensive costs of latency, energy consumption, and load balancing in simulation experiments. It can meet the caching requirements of low latency and low energy consumption for in-vehicle applications, laying a solid foundation for achieving more efficient and reliable connected vehicles and autonomous driving.

Transient Data Caching Based on Maximum Entropy Actor–Critic in Internet-of-Things Networks

AbstractWith the rapid development of the Internet-of-Things (IoT), a massive amount of transient data is transmitted in edge networks. Transient data are highly time-sensitive, such as monitoring data generated by industrial devices. Due to their inefficiency, traditional caching strategies in edge networks are inadequate for handling transient data. Thus, to improve the efficiency of transient data caching, we construct a freshness model of transient data and propose a maximum entropy Actor–Critic-based caching strategy, TD-MEAC-which can improve the freshness of cached data and reduce the long-term caching cost. Simulation results show that the proposed TD-MEAC achieves a higher cache hit rate and maintains a higher average freshness of cached transient data compared with the existing DRL and baseline caching strategies.

Research on high performance web service construction method based on JavaScript asynchronous programming technique

Abstract In the face of massive concurrent user access in the era of big data, how to build high-performance web services has become one of the difficulties to be solved by network applications. This paper utilizes Node.js architecture to fully utilize the event-based programming and non-blocking I/O characteristics of JavaScript language to improve the CPU utilization of the web service side. It also proposes the strategy of predicting user-accessed resources for cache replacement through a Gaussian mixture model to further optimize the performance of high-concurrency web services. The performance of the high concurrency web framework constructed using Node.js technology plays excellent, the performance of multi-threaded Node.js is better than single-threaded, and compared with Apache architecture, the average response time, request-response rate, and data throughput test of Node.js architecture in high concurrency scenarios is more advantageous than Apache architecture. The Gaussian mixture clustering model is effective in dealing with data with high-dimensional features, and the average accuracy of the four clustering processes of each algorithm is, in descending order, GMM (79.67%), BIRCH (78.09%), and K-means (77.52%). In addition, the cache replacement strategy based on Gaussian Mixture Model is more effective, with an accuracy rate of nearly 80% and a byte hit rate of nearly 45% when the cache capacity reaches 214KB, which are both higher than the traditional cache replacement strategy with the same cache capacity.

Cache allocation policy based on user preference using reinforcement learning in mobile edge computing

SummaryIn mobile edge computing (MEC), due to the limited computing resources and power of mobile augmented reality (MAR) devices, cache identification results which can reduce power consumption and executing time of mobile devices are the solution to process MAR tasks. In this paper, we study an allocation cache problem in MAR systems. The allocation cache problem is formulated as maximizing the cache utility of cache hit ratio and user preference factor. To solve this problem, a cache resource allocation and cache space adjustment policy for edge computing systems is proposed. We also propose an improved double deep Q‐network (DDQN) algorithm to learn this policy. Simulation results show that the policy greatly improves the cache hit ratio compared with the traditional caching policy.

An Efficient Gustavson-Based Sparse Matrix–Matrix Multiplication Accelerator on Embedded FPGAs

Sparse-matrix sparse-matrix multiplication (SpMM) is an important kernel in multiple areas, e.g., data analytics and machine learning. Due to the low on-chip memory requirement, the consistent data format, and the simplified control logic, the Gustavson’s algorithm is a promising backbone algorithm for SpMM on hardware accelerators. However, the off-chip memory traffic still limits the performance of the algorithm, especially on embedded FPGAs. Previous researchers optimize the Gustavson’s algorithm targeting high bandwidth memory-based architectures and their solutions cannot be directly applied to embedded FPGAs with traditional DDRs. In this work, we propose an efficient Gustavson-based sparse matrix-matrix multiplication accelerator on embedded FPGAs. The proposed design fully considers the feature of off-chip memory access on embedded FPGAs and the dataflow of the Gustavson’s algorithm. At first, we analyze the parallelism of the algorithm and propose to perform the algorithm with element-wise parallelism, which reduces the idle time of processing elements caused by synchronization. Further, we show a counter-intuitive example that the traditional cache leads to worse performance. Then, we propose a novel access pattern-aware cache scheme called SpCache, which provides quick responses to reduce bank conflicts caused by irregular memory accesses and combines streaming and caching to handle requests that access ordered elements of unpredictable length. Moreover, we propose to perform the merge on part of partial results, which removes some redundant merges in the naive implementation and has little postprocessing overhead. Finally, we conduct experiments on the Xilinx Zynq-UltraScale ZCU106 platform with a set of benchmarks from the SuiteSparse matrix collection. The experimental results show that the proposed design achieves an average 1.75x performance speedup compared to the baseline.

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.

Social-Aware Decentralized Cooperative Caching for Internet of Vehicles

As the industry related to intelligent vehicles becomes increasingly mature, emerging in-vehicle applications and services are mushrooming. They have strict requirements on service response delay and network bandwidth. In the face of dynamic environment in the Internet of Vehicles (IoV), how to design an adaptive edge caching strategy becomes a challenge. To cope with this challenge, some researchers have introduced optimization methods based on learning algorithms in cooperative caching. However, general learning algorithms tend to waste bandwidth and computing resources on repetitive task. To this end, we make full use of the idle resources in the road to build a cooperative caching system and propose a Social-Aware Decentralized Cooperative caching (SADC) for IoV. This strategy uses the federated learning framework to train the collaborative caching algorithm based on Deep Reinforcement Learning (DRL). Among them, the Road Side Unit (RSU) is responsible for the training and updating of the global model, and vehicles use local data to provide local updates to the RSU, which then averages the updates provided by all vehicles to improve the shared model. In addition, we use the social network of vehicle users to obtain vehicle contact rates in different areas. The SADC strategy can reduce the content transmission latency and response time, thereby improving the experience quality of vehicle users. Compared with traditional caching strategies, this strategy reduces the average content access delay by about 20%. We also demonstrate the effectiveness of this strategy using an extensive set of experiments.

Joint Content Caching and Recommendation in Opportunistic Mobile Networks Through Deep Reinforcement Learning and Broad Learning

Edge caching has been a research hotspot of the Mobile Edge Computing (MEC) in recent years, which is an effective way to ease the burden of traffic in cellular networks. It places contents to the edge of the networks and assists contents transmission via Device-to-Device (D2D) links. Traditional caching strategies strictly depend on the personal preferences of users, and they are scarcely possible to reduce the transmission cost while ensuring the high effectiveness. Fortunately, recent studies have found that the combination of caching and recommendation can effectively improve the efficiency of edge caching and reduce the transmission cost. In this paper, we jointly consider the content caching and recommendation through Opportunistic Mobile Networks (OMNs) to reduce the cost of Content Service Center (CSC). In order to obtain the optimal caching and recommendation solutions with sparse rating matrix, we propose a Joint Content Caching and Recommender System (JCCRS). In JCCRS, a Broad Incremental Learning based Collaborative Filtering algorithm, named BILCF, is first proposed to predict the missing ratings. Afterwards, we quantify the relationship between each pair of Mobile Users (MUs) according to their mobility, similarity and preference. The content caching and recommendation problem is then modeled as a Non-Linear Integer Programming (NLIP) problem and we prove that it belongs to NP-hard. To solve this problem, a Deep Deterministic Policy Gradient (DDPG) based Content Caching and Recommendation method, named DCRM, is further proposed to obtain the approximate optimal solutions. Extensive experiments on both a realistic dataset and a synthetic dataset validated by the realistic data show that the proposed algorithms outperform other baseline methods under different scenarios.

A delayed eviction caching replacement strategy with unified standard for edge servers

The current explosion of user traffic compels Internet service providers to cache contents at edge servers, so as to reduce the response time of user requests. To deal with such vast amounts of content in edge servers with limited storage, an efficient caching replacement strategy is necessary, consisting of admission and eviction processes. Facing such a large number of user requests and limited storage of edge servers, traditional caching replacement strategies have encountered two major problems. Firstly, they conduct eviction process immediately after cache miss, which ignores the fetching time, during which there may still be multiple access requests to the evicted contents. Secondly, most existing solutions treat admission process and eviction process separately, with different standards, some contents might be fetched and evicted back and forth, seriously affecting user experience. To solve these two problems and improve caching performance, we design two caching modules, Delayed-Eviction and Unified-Standard, and integrate them together into Adele framework. By conducting extensive simulations using real traces, we demonstrate that Adele can improve the hit rate by 31% compared with the state-of-the-art solutions.

A GNN-based proactive caching strategy in NDN networks

As people spend more time watching and sharing videos online, it is critical to provide users with a satisfactory quality of experience (QoE). Leveraging the in-network caching and named-based routing features in Named Data Networks (NDNs), our paper aims to improve user experience through caching. We propose a graph neural network-gain maximization (GNN-GM) cache placement algorithm. First, we use a GNN model to predict users’ ratings of unviewed videos. Second, we consider the total predicted rating of a video as the gain of the cached video. Third, we propose a cache placement algorithm to maximize the caching gain and actively cache videos. Cache replacement is implemented based on the cache gain ranking of videos, with higher cache gain videos replacing lower cache gain videos. We compare GNN-GM with two state-of-the-art caching strategies, namely the NMF-based caching strategy and GNN-CPP. GNN-GM is also compared with two traditional caching strategies, LCE and LRU, LCE and FIFO. We evaluate the five caching strategies using real-world datasets in a tree network topology, a real-world network topology GEANT, and various random topologies. The experimental results show that our caching policy significantly improves cache hit ratio, latency and server load. Notably, GNN-GM achieves a 25% higher cache hit rate, 5% lower latency and 7% lower server load than GNN-CPP in GEANT.

QoE-Aware Efficient Content Distribution Scheme For Satellite-Terrestrial Networks

The satellite-terrestrial networks (STN) utilize the spacious coverage and low transmission latency of the Low Earth Orbit (LEO) constellation to transfer requested content for subscribers especially in remote areas. With the development of storage and computing capacity of satellite onboard equipment, it is considered promising to leverage in-network caching technology on STN to improve content distribution efficiency. However, traditional caching and distribution schemes are not suitable in STN, considering dynamic satellite propagation links and time-varying topology. More specifically, the unevenness of user distribution heightens difficulties for assurance of user quality of experience. To address these problems, we first propose a density-based network division algorithm. The STN is divided into a series of blocks with different sizes to amortize the data delivery costs. To deploy the caching satellites, we analyze the link connectivity and propose an approximate minimum coverage vertex set algorithm. Then, a novel cache node selection algorithm is designed for optimal subscriber matching. On the basis of time-varying network model, the STN cache content updating mechanism is derived to enable a stable and sustainable quality of user experience. The simulation results demonstrate that the proposed user-oriented STN content distribution scheme can obviously reduce the average propagation delay and network load under different network conditions and has better stability and self-adaptability under continuous time variation.

A hybrid caching strategy for information-centric satellite networks based on node classification and popular content awareness

Low Earth Orbit (LEO) satellite networks are inexpensive to deploy and have wide coverage, and are widely available for differentiated content distribution services. With the development of on-board storage and computing capabilities, in-network caching technology in information centric networking (ICN) has proven to be an effective way to increase the throughput and content distribution efficiency of satellite networks. However, traditional caching and distribution schemes, which do not take into account the high-speed motion of satellite nodes and dynamic changes in topology, are not applicable to satellite networks. To address these issues, a hybrid caching strategy for satellite networks based on node classification and popular content awareness (NCPCA) is proposed. Firstly, the time slot partitioning method based on interlayer similarity is proposed to transform the dynamically changing process into a set of time slots with stable topology. Next, the satellite nodes are dynamically divided into two categories by fully considering the changes in connection relationships and interaction order during the spatial–temporal evolution of the satellite nodes. Nodes with satellite topological and functional characteristics are screened out as core nodes with TOPSIS algorithm, and the remaining nodes are regarded as edge nodes. Finally, a probabilistic caching scheme based on content popularity is adopted with core nodes as caching nodes to ensure caching performance and promote the diversity of cached content. Simulation results show that this strategy can effectively improve cache hit rate, reduce user request delay and content fetching hops, and promote the stable operation of satellite networks compared with other caching strategies.

NOMA-Based Integrated Satellite-Terrestrial Networks with Wireless Caching

To decrease the transmission delay and alleviate communication congestion, caching has been applied in the integrated satellite-terrestrial networks (ISTNs). In traditional caching schemes, the cache-enabled relays update files during off-peak hours and push them during on-peak hours. However, for the enormous number of end devices in ISTNs, the long-time waiting for off/on-peak hours would decrease the quality of service. To address this issue, this paper proposes a two-tier nonorthogonal multiple access- (NOMA-) based ISTN with wireless caching. Specifically, data transmission of this model consists of two phases. For the first phase, named the file-push-and-placement (FPAP), the satellite employs the NOMA protocol to send information to both the relay and first-tier user. While for the second phase, which is called the file-push-and-delivery (FPAD), second-tier users are served by the relay employing NOMA. Performance analysis of the proposed configuration is carried out, focusing on the exact and asymptotic outage probability (OP), diversity order, and hit probability. Moreover, the influence of key factors on the system performance is also investigated. Compared with the traditional configuration, it is shown analytically and numerically that the proposed scheme achieves lower OP and higher diversity order.

Age-constrained dynamic content replacing and delivering for UAV-assisted context awareness

In this work, we employ the cache-enabled UAV to provide context information delivery to end devices that make timely and intelligent decisions. Different from the traditional network traffic, context information varies with time and brings in the age-constrained requirement. The cached content items should be refreshed timely based on the age status to guarantee the freshness of user-received contents, which however consumes additional transmission resources. The traditional cache methods separate the caching and the transmitting, which are not suitable for the dynamic context information. We jointly design the cache replacing and content delivery based on both the user requests and the content dynamics to maximize the offloaded traffic from the ground network. The problem is formulated based on the Markov Decision Process (MDP). A sufficient condition of cache replacing is found in closed form, whereby a dynamic cache replacing and content delivery scheme is proposed based on the Deep Q-Network (DQN). Extensive simulations have been conducted. Compared with the conventional popularity-based and the modified Least Frequently Used (i.e., LFU-dynamic) schemes, the UAV can offload around 30 % traffic from the ground network by utilizing the proposed scheme in the urban scenario, according to the simulation results.

Orchestrating data placement and query execution in heterogeneous CPU-GPU DBMS

There has been a growing interest in using GPU to accelerate data analytics due to its massive parallelism and high memory bandwidth. The main constraint of using GPU for data analytics is the limited capacity of GPU memory. Heterogeneous CPU-GPU query execution is a compelling approach to mitigate the limited GPU memory capacity and PCIe bandwidth. However, the design space of heterogeneous CPU-GPU query execution has not been fully explored. We aim to improve state-of-the-art CPU-GPU data analytics engine by optimizing data placement and heterogeneous query execution. First, we introduce a semantic-aware fine-grained caching policy which takes into account various aspects of the workload such as query semantics, data correlation, and query frequency when determining data placement between CPU and GPU. Second, we introduce a heterogeneous query executor which can fully exploit data in both CPU and GPU and coordinate query execution at a fine granularity. We integrate both solutions in Mordred, our novel hybrid CPU-GPU data analytics engine. Evaluation on the Star Schema Benchmark shows that the semantic-aware caching policy can outperform the best traditional caching policy by up to 3x. Compared to existing GPU DBMSs, Mordred can outperform by an order of magnitude.

Dynamic Routing Strategy for Directed Transmissions of High-Valued Contents in NGSO Satellite-Based Internet

Nongeostationary orbit (NGSO) satellite-based Internet is an important part of the future network. One single NGSO satellite has the characteristics of high-speed movement, small coverage area, and short connection time; therefore, the method of multisatellite collaboration is adopted in NGSO satellite-based Internet. Caching is an effective technique to reduce the transmission load and delay, especially for high-value content. However, the traditional caching cannot be applied to the NGSO satellite network directly due to the frequent switching of service satellites. As a result, some high-valued cached data need to be transmitted among satellites, such that an effective routing between the source and destination node needs to be established orderly in the satellite network. This paper proposes a dynamic routing strategy based on multilayer satellite architecture for unicast and multicast scenarios, respectively. For the unicast, a routing strategy based on Q -learning is proposed to optimize bandwidth and delay simultaneously, and the effectiveness is verified by the comparison against the traditional routing strategy. For the multicast, an advanced Q -learning multicast routing algorithm based on spatial directionality is proposed to optimize the overall transmission success rate and resource cost, and the superiority of the proposed routing strategy is verified by simulation results.

CVC: A Collaborative Video Caching Framework Based on Federated Learning at the Edge

With the rapid development of Internet social media platforms and the popularization of smart terminal devices, video services such as short videos have surged. However, the massive amount of smart devices connected to the core network causes the increase of load on the backhaul link, which makes traditional cloud computing unable to meet the low-latency requirement of user for video services fully. To this end, we propose to implement proactive video caching at the edge to improve the quality of user experience. Thus, a novel collaborative video caching framework at the edge, CVC, is proposed. In this framework, we present a video request prediction model based on federated learning, aiming to reduce communication costs. Also, we design two methods to improve the hit rate and reduce latency, including a collaborative caching decision method ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CCD</i> ) and a collaborative service response method ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CSR</i> ). Under the video service scenario, the experimental results show that CVC can improve the cache hit rate and reduce the waiting delay of users compared with the traditional caching algorithms. Simultaneously, CVC can also reduce the overall system’s communication cost and caching cost.

Research on Load Balancing and Caching Strategy for Central Network

The central network architecture takes the information content as the processing object and can better meet the user access needs. However, the traditional network node load balancing and caching strategy has been difficult to adapt to the development of the central network architecture. Therefore, this article proposed a load balancing and cache management strategy based on the central network in order to provide a reference for the architecture and application of the central network. Based on the analysis of the existing network architecture, this article expounded the networking form and characteristics of the central network system. By analyzing the load scheduling requirements of network nodes, combined with the transmission characteristics of resource requests in the central network, a cache node load balancing method suitable for the requirements of the central network was proposed based on the existing load balancing algorithms. In addition, by analyzing the cache management strategy and its impact on the performance of the network system, a cache node information forwarding model with resource requests as the object in the central network environment was given. Finally, through experimental test and comparative analysis, the results showed that the method proposed in this article had certain feasibility and effectiveness. Compared with other typical strategies, the load balancing and caching strategy proposed in this article can better meet the user needs of the central network.