arXiv subjects

Erik Elmroth

Publications and source records attributed to Erik Elmroth.

At least 19 recordsLinked to original sources

arXiv2026

Performance Drift Detection in Machine Learning as a Service (MLaaS) for IoT Environments

Machine Learning as a Service (MLaaS) is a powerful cloud paradigm enabling data-driven intelligent applications in Internet of Things (IoT) environments, widely adopted across healthcare, smart homes, and industry due to its cost-effectiveness. However, the dynamic nature of IoT frequently alters data distributions, affecting MLaaS stability, while periodic MLaaS updates further introduce performance drift. Unlike traditional ML systems, MLaaS clients operate as black-box users without access to internal data or parameters, making drift detection particularly challenging. To address this, we propose a novel MLaaS Performance Drift Detection framework for IoT environments. The framework first employs an MLaaS extraction model that learns service behavior from input-output pairs and identifies prediction-influenced features. Building on this, the proposed MLaaS Performance Drift Detection (MPDD) model jointly captures variations in input data and MLaaS behavior. We further design an Adaptive-Temporal Performance Drift Detection Mechanism (APDDM) that dynamically adjusts monitoring frequency based on behavioral and data variations, enabling timely drift detection for effective service management. Extensive experiments on real-world datasets demonstrate that MPDD achieves up to 22-25% accuracy improvement over baseline drift detection methods. APDDM provides an average accuracy gain of approximately 4% and reduces the miss detection rate by around 9% compared to fixed-interval monitoring.

cs.LG↗

arXiv2026

Workload Buoyancy: Keeping Apps Afloat by Identifying Shared Resource Bottlenecks

Modern multi-tenant, hardware-heterogeneous computing environments pose significant challenges for effective workload orchestration. Simple heuristics for assessing workload performance, such as CPU utilization or application-level metrics, are often insufficient to capture the complex performance dynamics arising from resource contention and noisy-neighbor effects. In such environments, performance bottlenecks may emerge in any shared system resource, leading to unexpected and difficult-to-diagnose degradation. This paper introduces buoyancy, a novel abstraction for characterizing workload performance in multi-tenant systems. Unlike traditional approaches, buoyancy integrates application-level metrics with system-level insights of shared resource contention to provide a holistic view of performance dynamics. By explicitly capturing bottlenecks and headroom across multiple resources, buoyancy facilitates resource-aware and application-aware orchestration in a manner that is intuitive, extensible, and generalizable across heterogeneous platforms. We evaluate buoyancy using representative multi-tenant workloads to illustrate its ability to expose performance-limiting resource interactions. Buoyancy provides a 19.3% better indication of bottlenecks compared to traditional heuristics on average. We additionally show how buoyancy can act as a drop-in replacement for conventional performance metrics, enabling improved observability and more informed scheduling and optimization decisions.

cs.DC↗

arXiv2026

Beyond the Limits: Flexible and Congestion-Aware Cluster Scheduling for the Cloud

Workload scheduling in cloud environments often relies on simplistic assumptions about application resource needs and hardware utilization. Overlooking application-level performance objectives and hardware resource contention that leads to inefficient resource usage and degraded performance. This paper addresses two key limitations of current approaches. First, unnecessarily strict enforcement of service level objectives (SLOs) often leads to resource underutilization and poor energy efficiency. Second, lack of congestion awareness in shared resources such as last-level cache (LLC) and memory bandwidth. In this paper, we propose two complementary strategies to address these limitations: (i) integrating soft SLO limits that allow controlled overcommitment and tolerate minor, transient violations to improve cluster efficiency, and (ii) introducing resource-aware scheduling and rescheduling based on real-time congestion insights for shared resources such as last-level cache (LLC) and memory bandwidth. Our results show that soft SLO limits reduce corrective rescheduling actions by 49% compared to hard-limit approaches while maintaining acceptable performance guarantees. Additionally, resource-aware scheduling decreases node-level congestion by 8% and further mitigates SLO violations, demonstrating the effectiveness of incorporating application-level flexibility and hardware-level insights into scheduling and rescheduling decisions.

cs.DC↗

arXiv2026

Survival of the Cheapest: Cost-Aware Hardware Adaptation for Adversarial Robustness

Deploying adversarially robust machine learning systems requires continuous trade-offs between robustness, cost, and latency. We present an autonomic decision-support framework providing a quantitative foundation for adaptive hardware selection and hyper-parameter tuning in cloud-native deep learning. The framework applies accelerated failure time (AFT) models to quantify the effect of hardware choice, batch size, epochs, and validation accuracy on model survival time. This framework can be naturally integrated into an autonomic control loop (monitor--analyse--plan--execute, MAPE-K), where system metrics such as cost, robustness, and latency are continuously evaluated and used to adapt model configurations and hardware selection. Experiments across three GPU architectures confirm the framework is both sound and cost-effective: the Nvidia L4 yields a 20% increase in adversarial survival time while costing 75% less than the V100, demonstrating that expensive hardware does not necessarily improve robustness. The analysis further reveals that model inference latency is a stronger predictor of adversarial robustness than training time or hardware configuration.

cs.CR↗

arXiv2026

LA-IMR: Latency-Aware, Predictive In-Memory Routing and Proactive Autoscaling for Tail-Latency-Sensitive Cloud Robotics

Hybrid cloud-edge infrastructures now support latency-critical workloads ranging from autonomous vehicles and surgical robotics to immersive AR/VR. However, they continue to experience crippling long-tail latency spikes whenever bursty request streams exceed the capacity of heterogeneous edge and cloud tiers. To address these long-tail latency issues, we present Latency-Aware, Predictive In-Memory Routing and Proactive Autoscaling (LA-IMR). This control layer integrates a closed-form, utilization-driven latency model with event-driven scheduling, replica autoscaling, and edge-to-cloud offloading to mitigate 99th-percentile (P99) delays. Our analytic model decomposes end-to-end latency into processing, network, and queuing components, expressing inference latency as an affine power-law function of instance utilization. Once calibrated, it produces two complementary functions that drive: (i) millisecond-scale routing decisions for traffic offloading, and (ii) capacity planning that jointly determines replica pool sizes. LA-IMR enacts these decisions through a quality-differentiated, multi-queue scheduler and a custom-metric Kubernetes autoscaler that scales replicas proactively -- before queues build up -- rather than reactively based on lagging CPU metrics. Across representative vision workloads (YOLOv5m and EfficientDet) and bursty arrival traces, LA-IMR reduces P99 latency by up to 20.7 percent compared to traditional latency-only autoscaling, laying a principled foundation for next-generation, tail-tolerant cloud-edge inference services.

cs.DC↗

arXiv2026

Trust-Aware Routing for Distributed Generative AI Inference at the Edge

Emerging deployments of Generative AI increasingly execute inference across decentralized and heterogeneous edge devices rather than on a single trusted server. In such environments, a single device failure or misbehavior can disrupt the entire inference process, making traditional best-effort peer-to-peer routing insufficient. Coordinating distributed generative inference therefore requires mechanisms that explicitly account for reliability, performance variability, and trust among participating peers. In this paper, we present G-TRAC, a trust-aware coordination framework that integrates algorithmic path selection with system-level protocol design to ensure robust distributed inference. First, we formulate the routing problem as a \textit{Risk-Bounded Shortest Path} computation and introduce a polynomial-time solution that combines trust-floor pruning with Dijkstra's search, achieving sub-millisecond median routing latency at practical edge scales, and remaining below 10 ms at larger scales. Second, to operationally support the routing logic in dynamic environments, the framework employs a \textit{Hybrid Trust Architecture} that maintains global reputation state at stable anchors while disseminating lightweight updates to edge peers via background synchronization. Experimental evaluation on a heterogeneous testbed of commodity devices demonstrates that G-TRAC significantly improves inference completion rates, effectively isolates unreliable peers, and sustains robust execution even under node failures and network partitions.

cs.DC↗

arXiv2026

Speculative Policy Orchestration: A Latency-Resilient Framework for Cloud-Robotic Manipulation

Cloud robotics enables robots to offload high-dimensional motion planning and reasoning to remote servers. However, for continuous manipulation tasks requiring high-frequency control, network latency and jitter can severely destabilize the system, causing command starvation and unsafe physical execution. To address this, we propose Speculative Policy Orchestration (SPO), a latency-resilient cloud-edge framework. SPO utilizes a cloud-hosted world model to pre-compute and stream future kinematic waypoints to a local edge buffer, decoupling execution frequency from network round-trip time. To mitigate unsafe execution caused by predictive drift, the edge node employs an $ε$-tube verifier that strictly bounds kinematic execution errors. The framework is coupled with an Adaptive Horizon Scaling mechanism that dynamically expands or shrinks the speculative pre-fetch depth based on real-time tracking error. We evaluate SPO on continuous RLBench manipulation tasks under emulated network delays. Results show that even when deployed with learned models of modest accuracy, SPO reduces network-induced idle time by over 60% compared to blocking remote inference. Furthermore, SPO discards approximately 60% fewer cloud predictions than static caching baselines. Ultimately, SPO enables fluid, real-time cloud-robotic control while maintaining bounded physical safety.

cs.RO↗

arXiv2026

Tiny, Hardware-Independent, Compression-based Classification

The recent developments in machine learning have highlighted a conflict between online platforms and their users in terms of privacy. The importance of user privacy and the struggle for power over user data has been intensified as regulators and operators attempt to police online platforms. As users have become increasingly aware of privacy issues, client-side data storage, management, and analysis have become a favoured approach to large-scale centralised machine learning. However, state-of-the-art machine learning methods require vast amounts of labelled user data, making them unsuitable for models that reside client-side and only have access to a single user's data. State-of-the-art methods are also computationally expensive, which degrades the user experience on compute-limited hardware and also reduces battery life. A recent alternative approach has proven remarkably successful in classification tasks across a wide variety of data -- using a compression-based distance measure (called normalised compression distance) to measure the distance between generic objects in classical distance-based machine learning methods. In this work, we demonstrate that the normalised compression distance is actually not a metric; develop it for the wider context of kernel methods to allow modelling of complex data; and present techniques to improve the training time of models that use this distance measure. We demonstrate that the normalised compression distance works as well as and sometimes better than other metrics and kernels -- while requiring only marginally more computational costs and in spite of the lack of formal metric properties. The end results is a simple model with remarkable accuracy even when trained on a very small number of samples allowing for models that are small and effective enough to run entirely on a client device using only user-supplied data.

cs.LG↗

arXiv2026

Taming Cold Starts: Proactive Serverless Scheduling with Model Predictive Control

Serverless computing has transformed cloud application deployment by introducing a fine-grained, event-driven execution model that abstracts away infrastructure management. Its on-demand nature makes it especially appealing for latency-sensitive and bursty workloads. However, the cold start problem, i.e., where the platform incurs significant delay when provisioning new containers, remains the Achilles' heel of such platforms. This paper presents a predictive serverless scheduling framework based on Model Predictive Control to proactively mitigate cold starts, thereby improving end-to-end response time. By forecasting future invocations, the controller jointly optimizes container prewarming and request dispatching, improving latency while minimizing resource overhead. We implement our approach on Apache OpenWhisk, deployed on a Kubernetes-based testbed. Experimental results using real-world function traces and synthetic workloads demonstrate that our method significantly outperforms state-of-the-art baselines, achieving up to 85% lower tail latency and a 34% reduction in resource usage.

cs.DC↗

arXiv2025

Hardware-Level QoS Enforcement Features: Technologies, Use Cases, and Research Challenges

Recent advancements in commodity server processors have enabled dynamic hardware-based quality-of-service (QoS) enforcement. These features have gathered increasing interest in research communities due to their versatility and wide range of applications. Thus, there exists a need to understand how scholars leverage hardware QoS enforcement in research, understand strengths and shortcomings, and identify gaps in current state-of-the-art research. This paper observes relevant publications, presents a novel taxonomy, discusses the approaches used, and identifies trends. Furthermore, an opportunity is recognized for QoS enforcement utilization in service-based cloud computing environments, and open challenges are presented.

cs.DC↗

arXiv2025

Silent Failures in Stateless Systems: Rethinking Anomaly Detection for Serverless Computing

Serverless computing has redefined cloud application deployment by abstracting infrastructure and enabling on-demand, event-driven execution, thereby enhancing developer agility and scalability. However, maintaining consistent application performance in serverless environments remains a significant challenge. The dynamic and transient nature of serverless functions makes it difficult to distinguish between benign and anomalous behavior, which in turn undermines the effectiveness of traditional anomaly detection methods. These conventional approaches, designed for stateful and long-running services, struggle in serverless settings where executions are short-lived, functions are isolated, and observability is limited. In this first comprehensive vision paper on anomaly detection for serverless systems, we systematically explore the unique challenges posed by this paradigm, including the absence of persistent state, inconsistent monitoring granularity, and the difficulty of correlating behaviors across distributed functions. We further examine a range of threats that manifest as anomalies, from classical Denial-of-Service (DoS) attacks to serverless-specific threats such as Denial-of-Wallet (DoW) and cold start amplification. Building on these observations, we articulate a research agenda for next-generation detection frameworks that address the need for context-aware, multi-source data fusion, real-time, lightweight, privacy-preserving, and edge-cloud adaptive capabilities. Through the identification of key research directions and design principles, we aim to lay the foundation for the next generation of anomaly detection in cloud-native, serverless ecosystems.

cs.DC↗

arXiv2025

MTF-Grasp: A Multi-tier Federated Learning Approach for Robotic Grasping

Federated Learning (FL) is a promising machine learning paradigm that enables participating devices to train privacy-preserved and collaborative models. FL has proven its benefits for robotic manipulation tasks. However, grasping tasks lack exploration in such settings where robots train a global model without moving data and ensuring data privacy. The main challenge is that each robot learns from data that is nonindependent and identically distributed (non-IID) and of low quantity. This exhibits performance degradation, particularly in robotic grasping. Thus, in this work, we propose MTF-Grasp, a multi-tier FL approach for robotic grasping, acknowledging the unique challenges posed by the non-IID data distribution across robots, including quantitative skewness. MTF-Grasp harnesses data quality and quantity across robots to select a set of "top-level" robots with better data distribution and higher sample count. It then utilizes top-level robots to train initial seed models and distribute them to the remaining "low-level" robots, reducing the risk of model performance degradation in low-level robots. Our approach outperforms the conventional FL setup by up to 8% on the quantity-skewed Cornell and Jacquard grasping datasets.

cs.LG↗

arXiv2025

Optimising Virtual Resource Mapping in Multi-Level NUMA Disaggregated Systems

Disaggregated systems have a novel architecture motivated by the requirements of resource intensive applications such as social networking, search, and in-memory databases. The total amount of resources such as memory and CPU cores is very large in such systems. However, the distributed topology of disaggregated server systems result in non-uniform access latency and performance, with both NUMA aspects inside each box, as well as additional access latency for remote resources. In this work, we study the effects complex NUMA topologies on application performance and propose a method for improved, NUMA-aware, mapping for virtualized environments running on disaggregated systems. Our mapping algorithm is based on pinning of virtual cores and/or migration of memory across a disaggregated system and takes into account application performance, resource contention, and utilization. The proposed method is evaluated on a 288 cores and around 1TB memory system, composed of six disaggregated commodity servers, through a combination of benchmarks and real applications such as memory intensive graph databases. Our evaluation demonstrates significant improvement over the vanilla resource mapping methods. Overall, the mapping algorithm is able to improve performance by significant magnitude compared the default Linux scheduler used in system.

cs.DC↗

arXiv2024

FloRa: Flow Table Low-Rate Overflow Reconnaissance and Detection in SDN

Software Defined Networking (SDN) has evolved to revolutionize next-generation networks, offering programmability for on-the-fly service provisioning, primarily supported by the OpenFlow (OF) protocol. The limited storage capacity of Ternary Content Addressable Memory (TCAM) for storing flow tables in OF switches introduces vulnerabilities, notably the Low-Rate Flow Table Overflow (LOFT) attacks. LOFT exploits the flow table's storage capacity by occupying a substantial amount of space with malicious flow, leading to a gradual degradation in the flow-forwarding performance of OF switches. To mitigate this threat, we propose FloRa, a machine learning-based solution designed for monitoring and detecting LOFT attacks in SDN. FloRa continuously examines and determines the status of the flow table by closely examining the features of the flow table entries. Upon detecting an attack FloRa promptly activates the detection module. The module monitors flow properties, identifies malicious flows, and blacklists them, facilitating their eviction from the flow table. Incorporating novel features such as Packet Arrival Frequency, Content Relevance Score, and Possible Spoofed IP along with Cat Boost employed as the attack detection method. The proposed method reduces CPU overhead, memory overhead, and classification latency significantly and achieves a detection accuracy of 99.49%, which is more than the state-of-the-art methods to the best of our knowledge. This approach not only protects the integrity of the flow tables but also guarantees the uninterrupted flow of legitimate traffic. Experimental results indicate the effectiveness of FloRa in LOFT attack detection, ensuring uninterrupted data forwarding and continuous availability of flow table resources in SDN.

cs.NI↗

arXiv2024

A Training Rate and Survival Heuristic for Inference and Robustness Evaluation (TRASHFIRE)

Machine learning models -- deep neural networks in particular -- have performed remarkably well on benchmark datasets across a wide variety of domains. However, the ease of finding adversarial counter-examples remains a persistent problem when training times are measured in hours or days and the time needed to find a successful adversarial counter-example is measured in seconds. Much work has gone into generating and defending against these adversarial counter-examples, however the relative costs of attacks and defences are rarely discussed. Additionally, machine learning research is almost entirely guided by test/train metrics, but these would require billions of samples to meet industry standards. The present work addresses the problem of understanding and predicting how particular model hyper-parameters influence the performance of a model in the presence of an adversary. The proposed approach uses survival models, worst-case examples, and a cost-aware analysis to precisely and accurately reject a particular model change during routine model training procedures rather than relying on real-world deployment, expensive formal verification methods, or accurate simulations of very complicated systems (\textit{e.g.}, digitally recreating every part of a car or a plane). Through an evaluation of many pre-processing techniques, adversarial counter-examples, and neural network configurations, the conclusion is that deeper models do offer marginal gains in survival times compared to more shallow counterparts. However, we show that those gains are driven more by the model inference time than inherent robustness properties. Using the proposed methodology, we show that ResNet is hopelessly insecure against even the simplest of white box attacks.

cs.LG↗

arXiv2022

A Qualitative Evaluation of Service Mesh-based Traffic Management for Mobile Edge Cloud

Service mesh is getting widely adopted as the cloud-native mechanism for traffic management in microservice-based applications, in particular for generic IT workloads hosted in more centralized cloud environments. Performance-demanding applications continue to drive the decentralization of modern application execution environments, as in the case of mobile edge cloud. This paper presents a systematic and qualitative analysis of state-of-the-art service mesh to evaluate how suitable its design is for addressing the traffic management needs of performance-demanding application workloads hosted in a mobile edge cloud environment. With this analysis, we argue that today's dependability-centric service mesh design fails at addressing the needs of the different types of emerging mobile edge cloud workloads and motivate further research in the directions of performance-efficient architectures, stronger QoS guarantees and higher complexity abstractions of cloud-native traffic management frameworks.

cs.DC↗

arXiv2021

Towards Soft Circuit Breaking in Service Meshes via Application-agnostic Caching

Service meshes factor out code dealing with inter-micro-service communication, such as circuit breaking. Circuit breaking actuation is currently limited to an "on/off" switch, i.e., a tripped circuit breaker will return an application-level error indicating service unavailability to the calling micro-service. This paper proposes a soft circuit breaker actuator, which returns cached data instead of an error. The overall resilience of a cloud application is improved if constituent micro-services return stale data, instead of no data at all. While caching is widely employed for serving web service traffic, its usage in inter-micro-service communication is lacking. Micro-services responses are highly dynamic, which requires carefully choosing adaptive time-to-life caching algorithms. We evaluate our approach through two experiments. First, we quantify the trade-off between traffic reduction and data staleness using a purpose-build service, thereby identifying algorithm configurations that keep data staleness at about 3% or less while reducing network load by up to 30%. Second, we quantify the network load reduction with the micro-service benchmark by Google Cloud called Hipster Shop. Our approach results in caching of about 80% of requests. Results show the feasibility and efficiency of our approach, which encourages implementing caching as a circuit breaking actuator in service meshes.

cs.NI↗

arXiv2020

Resource Management for Blockchain-enabled Federated Learning: A Deep Reinforcement Learning Approach

Blockchain-enabled Federated Learning (BFL) enables mobile devices to collaboratively train neural network models required by a Machine Learning Model Owner (MLMO) while keeping data on the mobile devices. Then, the model updates are stored in the blockchain in a decentralized and reliable manner. However, the issue of BFL is that the mobile devices have energy and CPU constraints that may reduce the system lifetime and training efficiency. The other issue is that the training latency may increase due to the blockchain mining process. To address these issues, the MLMO needs to (i) decide how much data and energy that the mobile devices use for the training and (ii) determine the block generation rate to minimize the system latency, energy consumption, and incentive cost while achieving the target accuracy for the model. Under the uncertainty of the BFL environment, it is challenging for the MLMO to determine the optimal decisions. We propose to use the Deep Reinforcement Learning (DRL) to derive the optimal decisions for the MLMO.

cs.LG↗