arXiv subjects

Marco Ajmone Marsan

Publications and source records attributed to Marco Ajmone Marsan.

12 recordsLinked to original sources

arXiv2026

The Multiserver-Job Stochastic Recurrence Equation for Cloud Computing Performance Evaluation

We study the Multiserver-Job Queuing Model (MJQM) with general independent arrivals and service times under FCFS scheduling, using stochastic recurrence equations (SREs) and ergodic theory. We prove the monotonicity and separability properties of the MJQM SRE, enabling the application of the monotone-separable extension of Loynes' theorem and the formal definition of the MJQM stability condition. Based on these results, we introduce and implement two algorithms: one for drawing sub-perfect samples (SPS) of the system's workload and the second one to estimate the system's stability condition given the statistics of the jobs' input stream. The SPS algorithm allows for a massive GPU parallelization, greatly improving the efficiency of performance metrics evaluation. We also show that this approach extends to more complex systems, including MJQMs with typed resources.

cs.PF↗

arXiv2025

Improving Nonpreemptive Multiserver Job Scheduling with Quickswap

Modern data center workloads are composed of multiserver jobs, computational jobs that require multiple servers in order to run. A data center server can run many multiserver jobs in parallel, as long as it has sufficient resources to meet their demands. However, multiserver jobs are generally stateful, meaning that job preemptions incur significant overhead from saving and reloading the state associated with running jobs. Hence, most systems try to avoid these costly job preemptions altogether. Given these constraints, a scheduling policy must determine what set of jobs to run in parallel at each moment in time to minimize the mean response time across a stream of arriving jobs. Unfortunately, simple non-preemptive policies such as FCFS may leave many servers idle, resulting in high mean response times or even system instability. Our goal is to design and analyze non-preemptive scheduling policies for multiserver jobs that maintain high system utilization to achieve low mean response time. One well-known non-preemptive policy, Most Servers First (MSF), prioritizes jobs with higher server requirements and achieves high resource utilization. However, MSF causes extreme variability in job waiting times, and can perform significantly worse than FCFS in practice. To address this, we propose and analyze a class of scheduling policies called MSF-Quick Swap (MSFQ) that performs well. MSFQ reduces the variability of job waiting times by periodically granting priority to other jobs in the system. We provide both stability results and an analysis of mean response time under MSFQ to prove that our policy dramatically outperforms MSF in the case where jobs request one or all the servers. In more complex cases, we evaluate MSFQ in simulation. We show that, with some additional optimization, variants of the MSFQ policy can greatly outperform MSF and FCFS on real-world multiserver job workloads.

cs.PF↗

arXiv2025

Assessing the Benefits of Ground Vehicles as Moving Urban Base Stations

In the evolution towards 6G user-centric networking, the moving network (MN) paradigm can play an important role. In a MN, some small cell base stations (BS) are installed on top of vehicles, and enable a more dynamic, flexible and sustainable, network operation. By "following" the users movements and adapting dynamically to their requests, the MN paradigm enables a more efficient utilization of network resources, mitigating the need for dense small cell BS deployments at the cost of an increase in resource utilization due to wireless backhauling. This aspect is at least partly compensated by the shorter distance between users and BS, which allows for lower power and Line-of-Sight communications. While the MN paradigm has been investigated for some time, to date, it is still unclear in which conditions the advantages of MN outweigh the additional resource costs. In this paper, we propose a stochastic geometry framework for the characterization of the potential benefits of the MN paradigm as part of an HetNet in urban settings. Our approach allows the estimation of user-perceived performance, accounting for wireless backhaul connectivity as well as base station resource scheduling. We formulate an optimization problem for determining the resource-optimal network configurations and BS scheduling which minimize the overall amount of deployed BSs in a QoS-aware manner, and the minimum vehicular flow between different urban districts required to support them, and we propose an efficient stochastic heuristic to solve it. Our numerical assessment suggests that the MN paradigm, coupled with appropriate dynamic network management strategies, significantly reduces the amount of deployed network infrastructure while guaranteeing the target QoS perceived by users.

cs.NI↗

arXiv2025

Self-organisation of common good usage and an application to Internet services

Natural and human-made common goods present key challenges due to their susceptibility to degradation, overuse, or congestion. We explore the self-organisation of their usage when individuals have access to several available commons but limited information on them. We propose an extension of the Win-Stay, Lose-Shift (WSLS) strategy for such systems, under which individuals use a resource iteratively until they are unsuccessful and then shift randomly. This simple strategy leads to a distribution of the use of commons with an improvement against random shifting. Selective individuals who retain information on their usage and accordingly adapt their tolerance to failure in each common good improve the average experienced quality for an entire population. Hybrid systems of selective and non-selective individuals can lead to an equilibrium with equalised experienced quality akin to the ideal free distribution. We show that these results can be applied to the server selection problem faced by mobile users accessing Internet services and we perform realistic simulations to test their validity. Furthermore, these findings can be used to understand other real systems such as animal dispersal on grazing and foraging land, and to propose solutions to operators of systems of public transport or other technological commons.

cs.MA↗

arXiv2024

Green Operations of SWIPT Networks: The Role of End-User Devices

Internet of Things (IoT) devices often come with batteries of limited capacity that are not easily replaceable or rechargeable, and that constrain significantly the sensing, computing, and communication tasks that they can perform. The Simultaneous Wireless Information and Power Transfer (SWIPT) paradigm addresses this issue by delivering power wirelessly to energy-harvesting IoT devices with the same signal used for information transfer. For their peculiarity, these networks require specific energy-efficient planning and management approaches. However, to date, it is not clear what are the most effective strategies for managing a SWIPT network for energy efficiency. In this paper, we address this issue by developing an analytical model based on stochastic geometry, accounting for the statistics of user-perceived performance and base station scheduling. We formulate an optimization problem for deriving the energy optimal configuration as a function of the main system parameters, and we propose a genetic algorithm approach to solve it. Our results enable a first-order evaluation of the most effective strategies for energy-efficient provisioning of power and communications in a SWIPT network. We show that the service capacity brought about by users brings energy-efficient dynamic network provisioning strategies that radically differ from those of networks with no wireless power transfer.

cs.NI↗

arXiv2024

Context-Aware Orchestration of Energy-Efficient Gossip Learning Schemes

Fully distributed learning schemes such as Gossip Learning (GL) are gaining momentum due to their scalability and effectiveness even in dynamic settings. However, they often imply a high utilization of communication and computing resources, whose energy footprint may jeopardize the learning process, particularly on battery-operated IoT devices. To address this issue, we present Optimized Gossip Learning (OGL)}, a distributed training approach based on the combination of GL with adaptive optimization of the learning process, which allows for achieving a target accuracy while minimizing the energy consumption of the learning process. We propose a data-driven approach to OGL management that relies on optimizing in real-time for each node the number of training epochs and the choice of which model to exchange with neighbors based on patterns of node contacts, models' quality, and available resources at each node. Our approach employs a DNN model for dynamic tuning of the aforementioned parameters, trained by an infrastructure-based orchestrator function. We performed our assessments on two different datasets, leveraging time-varying random graphs and a measurement-based dynamic urban scenario. Results suggest that our approach is highly efficient and effective in a broad spectrum of network scenarios.

cs.NI↗

arXiv2024

Optimal Allocation of Tasks and Price of Anarchy of Distributed Optimization in Networked Computing Facilities

The allocation of computing tasks for networked distributed services poses a question to service providers on whether centralized allocation management be worth its cost. Existing analytical models were conceived for users accessing computing resources with practically indistinguishable (hence irrelevant for the allocation decision) delays, which is typical of services located in the same distant data center. However, with the rise of the edge-cloud continuum, a simple analysis of the sojourn time that computing tasks observe at the server misses the impact of diverse latency values imposed by server locations. We therefore study the optimization of computing task allocation with a new model that considers both distance of servers and sojourn time in servers. We derive exact algorithms to optimize the system and we show, through numerical analysis and real experiments, that differences in server location in the edge-cloud continuum cannot be neglected. By means of algorithmic game theory, we study the price of anarchy of a distributed implementation of the computing task allocation problem and unveil important practical properties such as the fact that the price of anarchy tends to be small -- except when the system is overloaded -- and its maximum can be computed with low complexity.

cs.GT↗

arXiv2023

On the Limit Performance of Floating Gossip

In this paper we investigate the limit performance of Floating Gossip, a new, fully distributed Gossip Learning scheme which relies on Floating Content to implement location-based probabilistic evolution of machine learning models in an infrastructure-less manner. We consider dynamic scenarios where continuous learning is necessary, and we adopt a mean field approach to investigate the limit performance of Floating Gossip in terms of amount of data that users can incorporate into their models, as a function of the main system parameters. Different from existing approaches in which either communication or computing aspects of Gossip Learning are analyzed and optimized, our approach accounts for the compound impact of both aspects. We validate our results through detailed simulations, proving good accuracy. Our model shows that Floating Gossip can be very effective in implementing continuous training and update of machine learning models in a cooperative manner, based on opportunistic exchanges among moving users.

stat.ML↗

arXiv2019

DeepFloat: Resource-Efficient Dynamic Management of Vehicular Floating Content

Opportunistic communications are expected to playa crucial role in enabling context-aware vehicular services. A widely investigated opportunistic communication paradigm for storing a piece of content probabilistically in a geographica larea is Floating Content (FC). A key issue in the practical deployment of FC is how to tune content replication and caching in a way which achieves a target performance (in terms of the mean fraction of users possessing the content in a given region of space) while minimizing the use of bandwidth and host memory. Fully distributed, distance-based approaches prove highly inefficient, and may not meet the performance target,while centralized, model-based approaches do not perform well in realistic, inhomogeneous settings. In this work, we present a data-driven centralized approach to resource-efficient, QoS-aware dynamic management of FC.We propose a Deep Learning strategy, which employs a Convolutional Neural Network (CNN) to capture the relationships between patterns of users mobility, of content diffusion and replication, and FC performance in terms of resource utilization and of content availability within a given area. Numerical evaluations show the effectiveness of our approach in deriving strategies which efficiently modulate the FC operation in space and effectively adapt to mobility pattern changes over time.

cs.NI↗

arXiv2018

A Deep Learning Strategy for Vehicular Floating Content Management

Floating Content (FC) is a communication paradigm for the local dissemination of contextualized information through D2D connectivity, in a way which minimizes the use of resources while achieving some specified performance target. Existing approaches to FC dimensioning are based on unrealistic system assumptions that make them, highly inaccurate and overly conservative when applied in realistic settings. In this paper, we present a first step towards the development of a cognitive approach to efficient dynamic management of FC. We propose a deep learning strategy for FC dimensioning, which exploits a Convolutional Neural Network(CNN) to efficiently modulate over time the resources employed by FC in a QoS-aware manner. Numerical evaluations show that our approach achieves a maximum rejection rate of3%, and resource savings of 37.5% with respect to the benchmark strategy

cs.NI↗

arXiv2017

Selecting the top-quality item through crowd scoring

We investigate crowdsourcing algorithms for finding the top-quality item within a large collection of objects with unknown intrinsic quality values. This is an important problem with many relevant applications, for example in networked recommendation systems. The core of the algorithms is that objects are distributed to crowd workers, who return a noisy and biased evaluation. All received evaluations are then combined, to identify the top-quality object. We first present a simple probabilistic model for the system under investigation. Then, we devise and study a class of efficient adaptive algorithms to assign in an effective way objects to workers. We compare the performance of several algorithms, which correspond to different choices of the design parameters/metrics. In the simulations we show that some of the algorithms achieve near optimal performance for a suitable setting of the system parameters.

cs.HC↗

arXiv2014

The Importance of Being Earnest in Crowdsourcing Systems

This paper presents the first systematic investigation of the potential performance gains for crowdsourcing systems, deriving from available information at the requester about individual worker earnestness (reputation). In particular, we first formalize the optimal task assignment problem when workers' reputation estimates are available, as the maximization of a monotone (submodular) function subject to Matroid constraints. Then, being the optimal problem NP-hard, we propose a simple but efficient greedy heuristic task allocation algorithm. We also propose a simple ``maximum a-posteriori`` decision rule. Finally, we test and compare different solutions, showing that system performance can greatly benefit from information about workers' reputation. Our main findings are that: i) even largely inaccurate estimates of workers' reputation can be effectively exploited in the task assignment to greatly improve system performance; ii) the performance of the maximum a-posteriori decision rule quickly degrades as worker reputation estimates become inaccurate; iii) when workers' reputation estimates are significantly inaccurate, the best performance can be obtained by combining our proposed task assignment algorithm with the LRA decision rule introduced in the literature.

cs.HC↗