Papers for
edge computing developers
Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.
Communication limits shape distributed consensus and learning in networks
Distributed Consensus, Optimization, and Learning under Communication Constraints
Abstract: Many modern engineering applications require distributed problem-solving in which multiple nodes in a network cooperate to achieve a common optimization objective. Numerous distributed optimization and learning algorithms rely on consensus-based distributed coordination to ensure network-wide agreement via iterative local interactions. In these algorithms, the accuracy, frequency, and reliability of information exchange among network nodes directly influence system performance, including convergence behavior, training delay, and model accuracy. This tutorial article adopts a bottom-up approach to present the fundamentals of consensus averaging and distributed optimization methods, and highlights the impact of network connectivity, information flow, and communication uncertainty on network cooperation and optimization performance. Building on these theoretical insights, we discuss key design principles for communication-constrained distributed learning across diverse network topologies and review common approaches to communication-efficient learning, particularly in wireless networks. Despite substantial progress, further efforts are needed to bridge the gap between theory and practice and fully realize the potential of cooperative and networked intelligence.
FedSharp improves decentralized learning accuracy under data noise and attacks
FedSharp: Byzantine Robust Decentralized Federated Learning Under Noisy Channel
Abstract: Decentralized federated learning (DFL) eliminates the central aggregation server but remains vulnerable to data heterogeneity, Byzantine attacks, and channel noise. We show that noise accumulates as models propagate across nodes, becoming increasingly significant as the learning rate decays. Moreover, accumulated noise enables Byzantine attackers to exploit loss-based model selection by mimicking honest model averages. To address these challenges, we propose screened, hop-aware, ranked, proximal averaging, a robust DFL framework (FedSharp). In this approach, each node maintains models from its 2b predecessors, ensuring at least b honest candidates when at most b nodes are Byzantine. FedSharp screens suspicious models using coordinate-wise median deviations, averages the b lowest-loss candidates to reduce fresh channel-noise variance by a factor of b, and applies noise-adaptive gradient-norm regularization and proximal training to mitigate accumulated noise and non-IID drift. Experiments on a 100-node ring with 33 Byzantine nodes show that FedSharp achieves 93.5-94.9% accuracy on MNIST, compared with 29.4-38.1% for a state-of-the-art single-model-selection benchmark, and 32.8-48.8% on CIFAR-10, compared with 10.0-30.5%, across Gaussian and hidden attacks under IID and non-IID data. On MNIST, FedSharp also achieves accuracy within approximately one percentage point of the benchmark operating without channel noise.
Quantized encoders speed up earth observation on common x86 computers
Quantized Edge Inference of a Self-Supervised SAR-Optical Encoder on Commodity x86 Hardware
Abstract: Self-supervised vision transformers are increasingly used as reusable encoders for Earth observation, but their inference cost on the hardware available in the field is rarely characterized. We deploy a pretrained SAR-optical masked-autoencoder encoder on three commodity x86 nodes, from a low-power CPU without AVX2 to an AVX2+VNNI reference, at FP32, FP16 and INT8 on the CPU and, where present, the integrated GPU, and report latency, energy (RAPL and wall plug) and linear-probe accuracy. Quantization is accuracy-neutral: INT8 changes micro-mAP by only +0.0006, within the seed spread, so precision becomes a pure cost decision. INT8 lowers CPU latency on every node, by 1.45x even on the CPU without AVX2, and CPU energy wherever it is measured, and the VNNI reference (1.53x) suggests a further gain beyond AVX2. The Pareto-optimal device shifts from the integrated GPU, which is 4.38x faster than the CPU on the weakest node, to the CPU as the CPU class improves.
Queue aware token communications improve ai and wireless service balance
Queue-Aware Token Communications: Bridging AI Representation and Wireless Service
Abstract: The rapid advancement of foundation models and edge intelligence is transforming wireless networks from data-oriented communication systems into artificial intelligence (AI)-native service platforms. Token communications (TokenCom) has emerged as a promising paradigm that directly transmits task-oriented AI representations. However, existing studies typically assume that generated tokens can be transmitted immediately, overlooking the queueing dynamics arising from the mismatch between token workload generation and time-varying wireless service capability. To address this limitation, we introduce Queue-Aware TokenCom, which explicitly incorporates token queues into the AI service pipeline. We present an end-to-end architecture integrating AI representation generation, token workload management, wireless transmission, and downstream task inference. We then discuss two key design directions: budget-aware token representation optimization and queue-aware token service optimization, with the token budget serving as a critical interface between AI performance and communication efficiency. Moreover, we investigate fixed and adaptive token budget strategies to characterize the fundamental tradeoff between AI performance and communication efficiency in dynamic wireless environments. Finally, we outline research directions toward closed-loop, task-aware, and scalable TokenCom systems.
Monarch MoE reduces forgetting in multimodal AI on limited memory
Monarch-MoE: Structured Experts for Multimodal Continual Instruction Tuning
Abstract: Learning new tasks in multimodal continual instruction tuning can cause catastrophic forgetting of previously acquired knowledge. Although multi-expert LoRA distributes learning across experts, updates to a dense low-rank expert can still affect many output features. We introduce Monarch-MoE, which replaces the dense factors in LoRA experts with block-diagonal factors connected by a fixed permutation. This structure provides blockwise gradient isolation: individual block-pair updates affect only one output block, while connections between all input and output blocks are maintained. To make these compact experts practical on memory-limited hardware, we combine batched computation, fused forward operators, and an optimized backward pass. Across three continual-learning methods evaluated on MLLM-DCL and UCIT with LLaVA-1.5-7B, Monarch-MoE improves first-task retention while reducing adapter parameters by approximately 75% and 50%, respectively. In the main comparison, first-task forgetting is as low as 0.52 percentage points on MLLM-DCL and 0.10 on UCIT, with method-dependent trade-offs in final performance. Our implementation supports fine-tuning and inference on a single 24 GB RTX 4090, with lower measured memory use and competitive runtime in the evaluated settings. Together, these algorithmic and systems improvements make structured experts a practical option for continual adaptation on memory-limited hardware.
Fedoag enables efficient federated learning over varied wireless channels
Over-the-Air Federated Learning in Heterogeneous Mobile Wireless Networks
Abstract: Over-the-air computation has emerged as a scalable and efficient solution for deploying federated learning algorithms in wireless networks by exploiting waveform superposition for simultaneous model aggregation. Most existing work struggles with heterogeneous fading channels. These approaches either enforce unbiased updates from all devices or allow partial device contributions, requiring careful tuning of the convergence bound to mitigate bias under specific fading models. However, the former significantly amplifies receiver noise due to the weakest channel, whereas the latter is sensitive to fading model mismatch and converges only to a biased objective. To tackle these challenges, we propose FedOAG, which employs algorithmic components to automatically satisfy energy constraints via gradient normalization and evenly mix devices' updates through implicit gossiping. Importantly, FedOAG does not require transmission from all devices, nor does it rely on a specific fading model or knowledge of time-varying statistical channel distributions. We show that FedOAG converges to a stationary point of an unbiased non-convex objective at the best possible rate $O(1/\sqrt{T})$ for any stochastic first-order method. We corroborate our analysis with numerical experiments over dynamic wireless conditions on real-world datasets.
FastFederatedLearning speeds up and customizes federated learning
Efficiently Distributed Federated Learning
Abstract: Federated Learning (FL) is experiencing a substantial research interest, with many frameworks being developed to allow practitioners to build federations easily and quickly. Most of these efforts do not consider two main aspects that are key to Machine Learning (ML) software: customizability and performance. This research addresses these issues by implementing an open-source FL framework named FastFederatedLearning (FFL). FFL is implemented in C/C++, focusing on code performance, and allows the user to specify any communication graph between clients and servers involved in the federation, ensuring customizability. FFL is tested against Intel OpenFL, achieving consistent speedups over different computational platforms (x86-64, ARM-v8, RISC-V), ranging from 2.5x and 3.69x. We aim to wrap FFL with a Python interface to ease its use and implement a middleware for different communication backends to be used. We aim to build dynamic federations in which relations between clients and servers are not static, giving life to an environment where federations can be seen as long-time evolving structures and exploited as services.
Open-source tool simulates O-RAN networks for cost-effective testing
Open-source emulation-based test environment to settle O-RAN-compliant trials
Abstract: Experimental tools are a key factor in both academic and industrial research communities to create design evaluations of new networking technologies that involve troubleshooting or changing the planning of deployed networks. Physical Software-Defined Radio (SDR) experimental platforms enable a design solution for the quick prototyping of wireless communication systems. However, SDR-based experimental platforms incur high costs, which leads to scalability limitations in the experimental settings. Having said this, network simulators, emulators, and new testbeds have attracted increasing attention. Emulation-based research prototyping can be distinguished from real communication networks and SDR-based platforms by allowing a tradeoff between cost and flexibility. This paper examines the Mininet-RAN emulation tool, which, as well as Radio Access Network (RAN) modeling, provides a way to test Open RAN Intelligent Controller (RIC) services without the need to deploy an entire RAN infrastructure. The Mininet-RAN creates virtual network elements, such as hosts, L2/L3 devices, controllers, and links, by combining some of the best emulator features, hardware testbeds, and simulators. By running the current code of standard practice Unix/Linux network applications and network stack, the Mininet-RAN enables real-world network data traffic patterns to be delivered to the RIC, regarding the most significant aspect of the dynamic generation of wireless system's KPIs. We provide the basic code of Mininet-RAN for the first two O-RAN Alliance-defined use cases involving V2X and UAV. The xApps are being implemented in O-RAN SC near-RT RIC, with Mininet-RAN which provides a closed-loop validation environment.