Pub Date : 2025-11-01DOI: 10.1016/j.peva.2025.102530
Xianyu Zhu , Ruohan Wu , Junshi Chen , Hong An
Given the complexity of heterogeneous architectures and multi-node collaboration, large-scale HPC (High-Performance Computing) clusters pose challenges in resource utilization and performance optimization during distributed data parallelism (DDP) training. Performance modeling aims to identify application bottlenecks and guide algorithm design, but existing performance models rarely consider the impact of system architecture on communication performance or provide a systematic analysis of distributed training. To address these issues, this paper proposes swPredictor, a data-driven performance model devised for accurately predicting the performance of DDP training. First, an original performance dataset is developed based on various communication patterns at runtime to avoid systematic errors. Subsequently, a novel multi-branch module FNO-Inception is proposed, combining FNO (Fourier Neural Operator) layer with Inception structure to simultaneously utilize various frequency features. Finally, by introducing the FNO-Inception module, a novel regression model FI-Net is constructed to fit complex nonlinear relationships. The experimental results demonstrate that FI-Net can accurately predict the performance of DDP training on the Sunway OceanLight supercomputer with an overall MAPE of 0.93%, which outperforms the other baseline models.
{"title":"swPredictor: A data-driven performance model for distributed data parallelism training on large-scale HPC clusters","authors":"Xianyu Zhu , Ruohan Wu , Junshi Chen , Hong An","doi":"10.1016/j.peva.2025.102530","DOIUrl":"10.1016/j.peva.2025.102530","url":null,"abstract":"<div><div>Given the complexity of heterogeneous architectures and multi-node collaboration, large-scale HPC (High-Performance Computing) clusters pose challenges in resource utilization and performance optimization during distributed data parallelism (DDP) training. Performance modeling aims to identify application bottlenecks and guide algorithm design, but existing performance models rarely consider the impact of system architecture on communication performance or provide a systematic analysis of distributed training. To address these issues, this paper proposes swPredictor, a data-driven performance model devised for accurately predicting the performance of DDP training. First, an original performance dataset is developed based on various communication patterns at runtime to avoid systematic errors. Subsequently, a novel multi-branch module FNO-Inception is proposed, combining FNO (Fourier Neural Operator) layer with Inception structure to simultaneously utilize various frequency features. Finally, by introducing the FNO-Inception module, a novel regression model FI-Net is constructed to fit complex nonlinear relationships. The experimental results demonstrate that FI-Net can accurately predict the performance of DDP training on the Sunway OceanLight supercomputer with an overall MAPE of 0.93%, which outperforms the other baseline models.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102530"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145516999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.peva.2025.102512
Jacob Bergquist , Erol Gelenbe , Mohammed Nasereddin , Karl Sigman
The Massive Access Problem arises due to devices that forward packets simultaneously to servers in rapid succession, or by malevolent software in devices that flood network nodes with high-intensity traffic. To protect servers from such events, attack detection (AD) software is installed on servers, and the Quasi-Deterministic Transmission Policy (QDTP) has been proposed to “shape traffic” and protect servers, allowing attack detection to proceed in a timely fashion by delaying some of the incoming packets individually based on their arrival times. QDTP does not cause packet loss, and can be designed so that it does not increase end-to-end packet delay. Starting with measurements taken on an experimental test-bed where the QDPT algorithm is installed on a dedicated processor, which precedes the server itself, we show that QDPT protects the server from attacks by accumulating arriving packets at the input of the QDTP processor, then forwarding them at regular intervals to the server. We compare the behaviour of the server, with and without the use of QDTP, showing the improvement it achieves, provided that its “delay” parameter is correctly selected. We analyze the sample paths associated with QDTP and prove that when its delay parameter is chosen in a specific manner, the end-to-end delay of each packet remains unchanged as compared to an ordinary First-In-First-Out system. An approach based on stationary ergodic processes is developed for the stability conditions. Assuming mutually independent and identically distributed inter-arrival times, service times and QDTP delays, we exhibit the positive recurrent structure of a two-dimensional Markov process and its regeneration points.
{"title":"Mitigating massive access with Quasi-Deterministic Transmission: Experiments and stationary analysis","authors":"Jacob Bergquist , Erol Gelenbe , Mohammed Nasereddin , Karl Sigman","doi":"10.1016/j.peva.2025.102512","DOIUrl":"10.1016/j.peva.2025.102512","url":null,"abstract":"<div><div>The Massive Access Problem arises due to devices that forward packets simultaneously to servers in rapid succession, or by malevolent software in devices that flood network nodes with high-intensity traffic. To protect servers from such events, attack detection (AD) software is installed on servers, and the Quasi-Deterministic Transmission Policy (QDTP) has been proposed to “shape traffic” and protect servers, allowing attack detection to proceed in a timely fashion by delaying some of the incoming packets individually based on their arrival times. QDTP does not cause packet loss, and can be designed so that it does not increase end-to-end packet delay. Starting with measurements taken on an experimental test-bed where the QDPT algorithm is installed on a dedicated processor, which precedes the server itself, we show that QDPT protects the server from attacks by accumulating arriving packets at the input of the QDTP processor, then forwarding them at regular intervals to the server. We compare the behaviour of the server, with and without the use of QDTP, showing the improvement it achieves, provided that its “delay” parameter is correctly selected. We analyze the sample paths associated with QDTP and prove that when its delay parameter is chosen in a specific manner, the end-to-end delay of each packet remains unchanged as compared to an ordinary First-In-First-Out system. An approach based on stationary ergodic processes is developed for the stability conditions. Assuming mutually independent and identically distributed inter-arrival times, service times and QDTP delays, we exhibit the positive recurrent structure of a two-dimensional Markov process and its regeneration points.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102512"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.peva.2025.102520
Jingfan Meng, Tianji Yang, Jun Xu
Set reconciliation (SetR) is an important research problem that has been studied for over two decades. In this problem, two large sets and of objects (tokens, files, records, etc.) are stored respectively at two different network-connected hosts, which we name Alice and Bob respectively. Alice and Bob need to communicate with each other to learn the set union (which then becomes their reconciled state), at low communication and computation costs. In this work, we study a different problem intricately related to SetR: Alice and Bob collaboratively compute . We call this problem SetX (set intersection). Although SetX is just as important as SetR, it has never been properly studied in its own right. Rather, there is an unspoken perception by the research community that SetR and SetX are equally difficult (in costs), and hence “roughly equivalent.” Our first contribution is to show that SetX is fundamentally a much “cheaper” problem than SetR, debunking this long-standing perception. Our second contribution is to develop a novel SetX solution, the communication cost of which handily beats the information-theoretic lower bound of SetR. This protocol is based on the idea of compressed sensing (CS), which we describe here only for the special case of (We do have a more sophisticated protocol for the general case). Our protocol is for Alice to encode into a CS sketch and send it to Bob, where is a CS matrix with rows and is the binary vector representation of . Our key innovation here is to make (the sketch size) just large enough (for the sketch) to summarize (what Alice misses). In contrast, in existing protocols needs to be large enough to summarize (what Alice knows), which is typically much larger in cardinality. Our third contribution is to design a CS matrix that is both “friendly” to (the performance of) applications and “compliant” with CS theory.
{"title":"CommonSense: Efficient Set Intersection (SetX) protocol based on compressed sensing","authors":"Jingfan Meng, Tianji Yang, Jun Xu","doi":"10.1016/j.peva.2025.102520","DOIUrl":"10.1016/j.peva.2025.102520","url":null,"abstract":"<div><div>Set reconciliation (SetR) is an important research problem that has been studied for over two decades. In this problem, two large sets <span><math><mi>A</mi></math></span> and <span><math><mi>B</mi></math></span> of objects (tokens, files, records, etc.) are stored respectively at two different network-connected hosts, which we name Alice and Bob respectively. Alice and Bob need to communicate with each other to learn the set union <span><math><mrow><mi>A</mi><mo>∪</mo><mi>B</mi></mrow></math></span> (which then becomes their reconciled state), at low communication and computation costs. In this work, we study a different problem intricately related to SetR: Alice and Bob collaboratively compute <span><math><mrow><mi>A</mi><mo>∩</mo><mi>B</mi></mrow></math></span>. We call this problem SetX (set intersection). Although SetX is just as important as SetR, it has never been properly studied in its own right. Rather, there is an unspoken perception by the research community that SetR and SetX are equally difficult (in costs), and hence “roughly equivalent.” Our first contribution is to show that SetX is fundamentally a much “cheaper” problem than SetR, debunking this long-standing perception. Our second contribution is to develop a novel SetX solution, the communication cost of which handily beats the information-theoretic lower bound of SetR. This protocol is based on the idea of compressed sensing (CS), which we describe here only for the special case of <span><math><mrow><mi>A</mi><mo>⊆</mo><mi>B</mi></mrow></math></span> (We do have a more sophisticated protocol for the general case). Our protocol is for Alice to encode <span><math><mi>A</mi></math></span> into a CS sketch <span><math><mrow><mi>M</mi><msub><mrow><mi>1</mi></mrow><mrow><mi>A</mi></mrow></msub></mrow></math></span> and send it to Bob, where <span><math><mi>M</mi></math></span> is a CS matrix with <span><math><mi>l</mi></math></span> rows and <span><math><msub><mrow><mi>1</mi></mrow><mrow><mi>A</mi></mrow></msub></math></span> is the binary vector representation of <span><math><mi>A</mi></math></span>. Our key innovation here is to make <span><math><mi>l</mi></math></span> (the sketch size) just large enough (for the sketch) to summarize <span><math><mrow><mi>B</mi><mo>∖</mo><mi>A</mi></mrow></math></span> (what Alice misses). In contrast, in existing protocols <span><math><mi>l</mi></math></span> needs to be large enough to summarize <span><math><mi>A</mi></math></span> (what Alice knows), which is typically much larger in cardinality. Our third contribution is to design a CS matrix <span><math><mi>M</mi></math></span> that is both “friendly” to (the performance of) applications and “compliant” with CS theory.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102520"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145517051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-24DOI: 10.1016/j.peva.2025.102510
Dieter Fiems , Balakrishna J. Prabhu
We consider a strategic routing game for a two-class discriminatory processor-sharing queue with an additional cost for joining the premium class. We show that, depending on the specific parameters of the system, various equilibria can coexist, including equilibria where the queueing system is not ergodic for the equilibrium traffic split. We also investigate how the server can select the priority of the classes and the fees charged to the customers to maximise its revenue. We then investigate learning strategies that converge to particular equilibria. Finally, we study how the elasticity of the traffic demand affects the equilibrium solutions.
{"title":"User equilibria in heterogeneous discriminatory processor sharing queues","authors":"Dieter Fiems , Balakrishna J. Prabhu","doi":"10.1016/j.peva.2025.102510","DOIUrl":"10.1016/j.peva.2025.102510","url":null,"abstract":"<div><div>We consider a strategic routing game for a two-class discriminatory processor-sharing queue with an additional cost for joining the premium class. We show that, depending on the specific parameters of the system, various equilibria can coexist, including equilibria where the queueing system is not ergodic for the equilibrium traffic split. We also investigate how the server can select the priority of the classes and the fees charged to the customers to maximise its revenue. We then investigate learning strategies that converge to particular equilibria. Finally, we study how the elasticity of the traffic demand affects the equilibrium solutions.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102510"},"PeriodicalIF":0.8,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-02DOI: 10.1016/j.peva.2025.102509
Julianna Bor, Peter G. Harrison
Join-the-Shortest-Queue (JSQ) is the scheduling policy of choice for many network providers, cloud servers, and traffic management systems, where individual queues are served under the processor sharing (PS) queueing discipline. A numerical solution for the response time distribution in two parallel PS queues with JSQ scheduling is derived for the first time. Using the generating function method, two partial differential equations (PDEs) are obtained corresponding to conditional response times, where the conditioning is on a particular traced task joining the first or the second queue. These PDEs are functional equations that contain partial generating functions and their partial derivatives, and therefore cannot be solved by commonly used techniques. We are able to solve these PDEs numerically with good accuracy and perform the deconditioning with respect to the queue-length probabilities by evaluating a certain complex integral. Numerical results for the density and the first four moments compare well against regenerative simulation.
加入最短队列（join -the- short - queue， JSQ）是许多网络提供商、云服务器和流量管理系统所选择的调度策略，其中单个队列在处理器共享（processor sharing， PS）队列规则下提供服务。本文首次给出了采用JSQ调度的两个并行PS队列响应时间分布的数值解。利用生成函数方法，得到了与条件响应时间相对应的两个偏微分方程（PDEs），其中条件响应是针对加入第一个或第二个队列的特定跟踪任务。这些偏微分方程是包含偏生成函数及其偏导数的函数方程，因此不能用常用的技术来求解。我们能够以较好的精度对这些偏微分方程进行数值求解，并通过求某个复积分对队列长度概率进行去条件化。密度和前4阶矩的数值模拟结果与再生模拟结果比较吻合。
{"title":"Response time in a pair of processor sharing queues with Join-the-Shortest-Queue scheduling","authors":"Julianna Bor, Peter G. Harrison","doi":"10.1016/j.peva.2025.102509","DOIUrl":"10.1016/j.peva.2025.102509","url":null,"abstract":"<div><div>Join-the-Shortest-Queue (JSQ) is the scheduling policy of choice for many network providers, cloud servers, and traffic management systems, where individual queues are served under the processor sharing (PS) queueing discipline. A numerical solution for the response time distribution in two parallel PS queues with JSQ scheduling is derived for the first time. Using the generating function method, two partial differential equations (PDEs) are obtained corresponding to conditional response times, where the conditioning is on a particular traced task joining the first or the second queue. These PDEs are functional equations that contain partial generating functions and their partial derivatives, and therefore cannot be solved by commonly used techniques. We are able to solve these PDEs numerically with good accuracy and perform the deconditioning with respect to the queue-length probabilities by evaluating a certain complex integral. Numerical results for the density and the first four moments compare well against regenerative simulation.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102509"},"PeriodicalIF":0.8,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145099156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study a model of auction representative of the 5G auction in France. We determine the optimal strategy of a bidder, assuming that the valuations of competitors are unknown to this bidder and that competitors adopt the straightforward bidding strategy. Our model is based on a Partially Observable Markov Decision Process (POMDP). This POMDP admits a concise statistics, avoiding the solution of a dynamic programming equation in the space of beliefs. In addition, under this optimal strategy, the expected gain of the bidder does not decrease if competitors deviate from straightforward bidding. We illustrate our results by numerical experiments, comparing the value of the bidder with the value of a perfectly informed one.
{"title":"Optimal strategy against straightforward bidding in clock auctions","authors":"Jad Zeroual , Marianne Akian , Aurélien Bechler , Matthieu Chardy , Stéphane Gaubert","doi":"10.1016/j.peva.2025.102502","DOIUrl":"10.1016/j.peva.2025.102502","url":null,"abstract":"<div><div>We study a model of auction representative of the 5G auction in France. We determine the optimal strategy of a bidder, assuming that the valuations of competitors are unknown to this bidder and that competitors adopt the straightforward bidding strategy. Our model is based on a Partially Observable Markov Decision Process (POMDP). This POMDP admits a concise statistics, avoiding the solution of a dynamic programming equation in the space of beliefs. In addition, under this optimal strategy, the expected gain of the bidder does not decrease if competitors deviate from straightforward bidding. We illustrate our results by numerical experiments, comparing the value of the bidder with the value of a perfectly informed one.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"169 ","pages":"Article 102502"},"PeriodicalIF":0.8,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144724472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-20DOI: 10.1016/j.peva.2025.102501
Ilias Iliadis, Mark Lantz
Magnetic tape is a digital data storage technology that has evolved continuously over the last seven decades. It provides a cost-effective way to retain the rapidly increasing volumes of data being created in recent years. The low cost per terabyte combined with tape’s low energy consumption make it an appealing option for storing infrequently accessed data and has resulted in a resurgence in use of the technology. Power and operational failures may damage tapes and lead to data loss. To protect stored data against loss and achieve high data reliability, an erasure coding scheme is employed. A theoretical model capturing the effect of tape failures and latent errors on system reliability is developed. Closed-form expressions are derived for the Mean Time to Data Loss () and the Expected Annual Fraction of Effective Data Loss () reliability metric, which assesses losses at the file, object, or block, level. The results obtained demonstrate that, for realistic values of bit error rates, reliability is affected by the presence of latent errors. The effect of system parameters on reliability is assessed by conducting a sensitivity evaluation. The reliability improvement achieved by employing erasure coding schemes with increased capability is demonstrated. The theoretical results derived can be used to dimension and provision tape libraries to provide desired levels of data durability.
{"title":"Reliability evaluation of tape library systems","authors":"Ilias Iliadis, Mark Lantz","doi":"10.1016/j.peva.2025.102501","DOIUrl":"10.1016/j.peva.2025.102501","url":null,"abstract":"<div><div>Magnetic tape is a digital data storage technology that has evolved continuously over the last seven decades. It provides a cost-effective way to retain the rapidly increasing volumes of data being created in recent years. The low cost per terabyte combined with tape’s low energy consumption make it an appealing option for storing infrequently accessed data and has resulted in a resurgence in use of the technology. Power and operational failures may damage tapes and lead to data loss. To protect stored data against loss and achieve high data reliability, an erasure coding scheme is employed. A theoretical model capturing the effect of tape failures and latent errors on system reliability is developed. Closed-form expressions are derived for the Mean Time to Data Loss (<span><math><mtext>MTTDL</mtext></math></span>) and the Expected Annual Fraction of Effective Data Loss (<span><math><mtext>EAFEDL</mtext></math></span>) reliability metric, which assesses losses at the file, object, or block, level. The results obtained demonstrate that, for realistic values of bit error rates, reliability is affected by the presence of latent errors. The effect of system parameters on reliability is assessed by conducting a sensitivity evaluation. The reliability improvement achieved by employing erasure coding schemes with increased capability is demonstrated. The theoretical results derived can be used to dimension and provision tape libraries to provide desired levels of data durability.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"169 ","pages":"Article 102501"},"PeriodicalIF":1.0,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144338331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-14DOI: 10.1016/j.peva.2025.102500
Nader Alfares, George Kesidis
We consider a content-caching system for Virtual or Augmented Reality (VR/AR) that is shared by a number of user groups. The cache could be located in an edge–cloud datacenter and the users could be mobile. Each user group operates its own LRU-list of a certain capacity in the shared cache. The length of objects simultaneously appearing in plural LRU-lists is equally divided (apportioned) among them, i.e., object sharing among the LRUs. We provide a working-set approximation for this system to quickly estimate the cache-hit probabilities for objects apportioned across user groups. We also prove that a solution to the working-set approximation exists. A way to reduce ripple evictions is discussed and some numerical performance results are provided based on Zipf-distributed object popularities. To evaluate our apportioned object sharing system in a VR context, we introduce a benchmark specifically designed for VR streaming, addressing the latency and Quality of Experience (QoE) challenges that arise due to VR’s dynamic fields of view (FoVs) and user synchronization requirements. Using the VR benchmark, we present additional numerical results for cache-hit rates and users’ QoE to illustrate the system’s effectiveness under VR streaming conditions.
{"title":"Virtual caching with apportioned objects for mobile virtual reality","authors":"Nader Alfares, George Kesidis","doi":"10.1016/j.peva.2025.102500","DOIUrl":"10.1016/j.peva.2025.102500","url":null,"abstract":"<div><div>We consider a content-caching system for Virtual or Augmented Reality (VR/AR) that is shared by a number of user groups. The cache could be located in an edge–cloud datacenter and the users could be mobile. Each user group operates its own LRU-list of a certain capacity in the shared cache. The length of objects simultaneously appearing in plural LRU-lists is equally divided (apportioned) among them, i.e., object sharing among the LRUs. We provide a working-set approximation for this system to quickly estimate the cache-hit probabilities for objects apportioned across user groups. We also prove that a solution to the working-set approximation exists. A way to reduce ripple evictions is discussed and some numerical performance results are provided based on Zipf-distributed object popularities. To evaluate our apportioned object sharing system in a VR context, we introduce a benchmark specifically designed for VR streaming, addressing the latency and Quality of Experience (QoE) challenges that arise due to VR’s dynamic fields of view (FoVs) and user synchronization requirements. Using the VR benchmark, we present additional numerical results for cache-hit rates and users’ QoE to illustrate the system’s effectiveness under VR streaming conditions.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"169 ","pages":"Article 102500"},"PeriodicalIF":1.0,"publicationDate":"2025-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-12DOI: 10.1016/j.peva.2025.102499
Michel Mandjes , Jiesen Wang
This paper examines a model involving two dynamic Erdős–Rényi random graphs that evolve in parallel, with edges in each graph alternating between being present and absent according to specified on- and off-time distributions. A key feature of our setup is regime switching: the graph that is observed at any given moment depends on the state of an underlying background process, which is modeled as an alternating renewal process. This modeling framework captures a common situation in various real-world applications, where the observed network is influenced by a (typically unobservable) background process. Such scenarios arise, for example, in economics, communication networks, and biological systems.
In our setup we only have access to aggregate quantities such as the number of active edges or the counts of specific subgraphs (such as stars or complete graphs) in the observed graph; importantly, we do not observe the mode. The objective is to estimate the on- and off-time distributions of the edges in each of the two dynamic Erdős–Rényi random graphs, as well as the distribution of time spent in each of the two modes. By employing parametric models for the on- and off-times and the background process, we develop a method of moments approach to estimate the relevant parameters. Experimental evaluations are conducted to demonstrate the effectiveness of the proposed method in recovering these parameters.
{"title":"Inference for dynamic Erdős–Rényi random graphs under regime switching","authors":"Michel Mandjes , Jiesen Wang","doi":"10.1016/j.peva.2025.102499","DOIUrl":"10.1016/j.peva.2025.102499","url":null,"abstract":"<div><div>This paper examines a model involving two dynamic Erdős–Rényi random graphs that evolve in parallel, with edges in each graph alternating between being present and absent according to specified on- and off-time distributions. A key feature of our setup is regime switching: the graph that is observed at any given moment depends on the state of an underlying background process, which is modeled as an alternating renewal process. This modeling framework captures a common situation in various real-world applications, where the observed network is influenced by a (typically unobservable) background process. Such scenarios arise, for example, in economics, communication networks, and biological systems.</div><div>In our setup we only have access to aggregate quantities such as the number of active edges or the counts of specific subgraphs (such as stars or complete graphs) in the observed graph; importantly, we do not observe the mode. The objective is to estimate the on- and off-time distributions of the edges in each of the two dynamic Erdős–Rényi random graphs, as well as the distribution of time spent in each of the two modes. By employing parametric models for the on- and off-times and the background process, we develop a method of moments approach to estimate the relevant parameters. Experimental evaluations are conducted to demonstrate the effectiveness of the proposed method in recovering these parameters.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"169 ","pages":"Article 102499"},"PeriodicalIF":1.0,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-02DOI: 10.1016/j.peva.2025.102498
Josu Doncel , Mohamad Assaad
We study a monitoring system in which a single source sends status updates to a monitor through a communication channel. The communication channel is modeled as a queueing system, and we assume that attacks occur following a random process. When an attack occurs, all packets in the queueing system are discarded. While one might expect attacks to always negatively impact system performance, we demonstrate in this paper that, from the perspective of Age of Information (AoI), attacks can in some cases reduce the AoI. Our objective is to identify the conditions under which AoI is reduced and to determine the attack rate that minimizes or reduces AoI. First, we analyze single and tandem M/M/1/1 queues with preemption and show that attacks cannot reduce AoI in these cases. Next, we examine a single M/M/1/1 queue without preemption and establish necessary and sufficient conditions for the existence of an attack rate that minimizes AoI. For this scenario, we also derive an upper bound for the optimal attack rate and prove that it becomes tight when the arrival rate of updates is very high. Through numerical experiments, we observe that attacks can reduce AoI in tandem M/M/1/1 queues without preemption, as well as in preemptive M/M/1/2 and M/M/1/3 queues. Furthermore, we show that the benefit of attacks on AoI increases with the buffer size.
我们研究了一个监控系统，其中单个源通过通信通道向监视器发送状态更新。通信通道被建模为一个排队系统，我们假设攻击发生在一个随机的过程中。当攻击发生时，队列系统中的所有报文将被丢弃。虽然人们可能期望攻击总是对系统性能产生负面影响，但我们在本文中证明，从信息时代（Age of Information, AoI）的角度来看，攻击在某些情况下可以降低AoI。我们的目标是确定减少AoI的条件，并确定最小化或减少AoI的攻击率。首先，我们分析了具有抢占的单队列和串联M/M/1/1队列，并表明在这些情况下攻击不能降低AoI。接下来，我们研究了一个没有抢占的M/M/1/1队列，并建立了使AoI最小的攻击率存在的充分必要条件。对于这种情况，我们还推导了最优攻击率的上界，并证明当更新到达率非常高时，它会变得很紧。通过数值实验，我们观察到攻击可以减少无抢占的M/M/1/1队列中的AoI，以及抢占的M/M/1/2和M/M/1/3队列中的AoI。此外，我们还表明，攻击AoI的好处随着缓冲区大小的增加而增加。
{"title":"Can attacks reduce Age of Information?","authors":"Josu Doncel , Mohamad Assaad","doi":"10.1016/j.peva.2025.102498","DOIUrl":"10.1016/j.peva.2025.102498","url":null,"abstract":"<div><div>We study a monitoring system in which a single source sends status updates to a monitor through a communication channel. The communication channel is modeled as a queueing system, and we assume that attacks occur following a random process. When an attack occurs, all packets in the queueing system are discarded. While one might expect attacks to always negatively impact system performance, we demonstrate in this paper that, from the perspective of Age of Information (AoI), attacks can in some cases reduce the AoI. Our objective is to identify the conditions under which AoI is reduced and to determine the attack rate that minimizes or reduces AoI. First, we analyze single and tandem M/M/1/1 queues with preemption and show that attacks cannot reduce AoI in these cases. Next, we examine a single M/M/1/1 queue without preemption and establish necessary and sufficient conditions for the existence of an attack rate that minimizes AoI. For this scenario, we also derive an upper bound for the optimal attack rate and prove that it becomes tight when the arrival rate of updates is very high. Through numerical experiments, we observe that attacks can reduce AoI in tandem M/M/1/1 queues without preemption, as well as in preemptive M/M/1/2 and M/M/1/3 queues. Furthermore, we show that the benefit of attacks on AoI increases with the buffer size.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"169 ","pages":"Article 102498"},"PeriodicalIF":1.0,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144196352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号