Pluto and Charon: A Time and Memory Efficient Collaborative Edge AI Framework for Personal LLMs Fine-Tuning

Large language models (LLMs) have unlocked a plethora of powerful applications at the network edge, such as intelligent personal assistants. Data privacy and security concerns have prompted a shift towards edge-based fine-tuning of personal LLMs, away from cloud reliance. However, this raises issues of computational intensity and resource scarcity, hindering training efficiency and feasibility. While current studies investigate parameter-efficient fine-tuning (PEFT) techniques to mitigate resource constraints, our analysis indicates that these techniques are not sufficiently resource-efficient for edge devices. To tackle these challenges, we propose Pluto and Charon (PAC), a time and memory efficient collaborative edge AI framework for personal LLMs fine-tuning. PAC breaks the resource wall of personal LLMs fine-tuning with a sophisticated algorithm-system co-design. (1) Algorithmically, PAC implements a personal LLMs fine-tuning technique that is efficient in terms of parameters, time, and memory. It utilizes Parallel Adapters to circumvent the need for a full backward pass through the LLM backbone. Additionally, an activation cache mechanism further streamlining the process by negating the necessity for repeated forward passes across multiple epochs. (2) Systematically, PAC leverages edge devices in close proximity, pooling them as a collective resource for in-situ personal LLMs fine-tuning, utilizing a hybrid data and pipeline parallelism to orchestrate distributed training. The use of the activation cache eliminates the need for forward pass through the LLM backbone,enabling exclusive fine-tuning of the Parallel Adapters using data parallelism. Extensive evaluation based on prototype implementation demonstrates that PAC remarkably outperforms state-of-the-art approaches, achieving up to 8.64x end-to-end speedup and up to 88.16% reduction in memory footprint.

Asteroid: Resource-Efficient Hybrid Pipeline Parallelism for Collaborative DNN Training on Heterogeneous Edge Devices

On-device Deep Neural Network (DNN) training has been recognized as crucial for privacy-preserving machine learning at the edge. However, the intensive training workload and limited onboard computing resources pose significant challenges to the availability and efficiency of model training. While existing works address these challenges through native resource management optimization, we instead leverage our observation that edge environments usually comprise a rich set of accompanying trusted edge devices with idle resources beyond a single terminal. We propose Asteroid, a distributed edge training system that breaks the resource walls across heterogeneous edge devices for efficient model training acceleration. Asteroid adopts a hybrid pipeline parallelism to orchestrate distributed training, along with a judicious parallelism planning for maximizing throughput under certain resource constraints. Furthermore, a fault-tolerant yet lightweight pipeline replay mechanism is developed to tame the device-level dynamics for training robustness and performance stability. We implement Asteroid on heterogeneous edge devices with both vision and language models, demonstrating up to 12.2x faster training than conventional parallelism methods and 2.1x faster than state-of-the-art hybrid parallelism methods through evaluations. Furthermore, Asteroid can recover training pipeline 14x faster than baseline methods while preserving comparable throughput despite unexpected device exiting and failure.

FlocOff: Data Heterogeneity Resilient Federated Learning with Communication-Efficient Edge Offloading

Federated Learning (FL) has emerged as a fundamental learning paradigm to harness massive data scattered at geo-distributed edge devices in a privacy-preserving way. Given the heterogeneous deployment of edge devices, however, their data are usually Non-IID, introducing significant challenges to FL including degraded training accuracy, intensive communication costs, and high computing complexity. Towards that, traditional approaches typically utilize adaptive mechanisms, which may suffer from scalability issues, increased computational overhead, and limited adaptability to diverse edge environments. To address that, this paper instead leverages the observation that the computation offloading involves inherent functionalities such as node matching and service correlation to achieve data reshaping and proposes Federated learning based on computing Offloading (FlocOff) framework, to address data heterogeneity and resource-constrained challenges. Specifically, FlocOff formulates the FL process with Non-IID data in edge scenarios and derives rigorous analysis on the impact of imbalanced data distribution. Based on this, FlocOff decouples the optimization in two steps, namely : (1) Minimizes the Kullback-Leibler (KL) divergence via Computation Offloading scheduling (MKL-CO); (2) Minimizes the Communication Cost through Resource Allocation (MCC-RA). Extensive experimental results demonstrate that the proposed FlocOff effectively improves model convergence and accuracy by 14.3\%-32.7\% while reducing data heterogeneity under various data distributions.

Galaxy: A Resource-Efficient Collaborative Edge AI System for In-situ Transformer Inference

Transformer-based models have unlocked a plethora of powerful intelligent applications at the edge, such as voice assistant in smart home. Traditional deployment approaches offload the inference workloads to the remote cloud server, which would induce substantial pressure on the backbone network as well as raise users' privacy concerns. To address that, in-situ inference has been recently recognized for edge intelligence, but it still confronts significant challenges stemming from the conflict between intensive workloads and limited on-device computing resources. In this paper, we leverage our observation that many edge environments usually comprise a rich set of accompanying trusted edge devices with idle resources and propose Galaxy, a collaborative edge AI system that breaks the resource walls across heterogeneous edge devices for efficient Transformer inference acceleration. Galaxy introduces a novel hybrid model parallelism to orchestrate collaborative inference, along with a heterogeneity-aware parallelism planning for fully exploiting the resource potential. Furthermore, Galaxy devises a tile-based fine-grained overlapping of communication and computation to mitigate the impact of tensor synchronizations on inference latency under bandwidth-constrained edge environments. Extensive evaluation based on prototype implementation demonstrates that Galaxy remarkably outperforms state-of-the-art approaches under various edge environment setups, achieving up to 2.5x end-to-end latency reduction.

Implementation of Big AI Models for Wireless Networks with Collaborative Edge Computing

Big Artificial Intelligence (AI) models have emerged as a crucial element in various intelligent applications at the edge, such as voice assistants in smart homes and autonomous robotics in smart factories. Training big AI models, e.g., for personalized fine-tuning and continual model refinement, poses significant challenges to edge devices due to the inherent conflict between limited computing resources and intensive workload associated with training. Despite the constraints of on-device training, traditional approaches usually resort to aggregating training data and sending it to a remote cloud for centralized training. Nevertheless, this approach is neither sustainable, which strains long-range backhaul transmission and energy-consuming datacenters, nor safely private, which shares users' raw data with remote infrastructures. To address these challenges, we alternatively observe that prevalent edge environments usually contain a diverse collection of trusted edge devices with untapped idle resources, which can be leveraged for edge training acceleration. Motivated by this, in this article, we propose collaborative edge training, a novel training mechanism that orchestrates a group of trusted edge devices as a resource pool for expedited, sustainable big AI model training at the edge. As an initial step, we present a comprehensive framework for building collaborative edge training systems and analyze in-depth its merits and sustainable scheduling choices following its workflow. To further investigate the impact of its parallelism design, we empirically study a case of four typical parallelisms from the perspective of energy demand with realistic testbeds. Finally, we discuss open challenges for sustainable collaborative edge training to point to future directions of edge-centric big AI model training.

A3D: Adaptive, Accurate, and Autonomous Navigation for Edge-Assisted Drones

Accurate navigation is of paramount importance to ensure flight safety and efficiency for autonomous drones. Recent research starts to use Deep Neural Networks to enhance drone navigation given their remarkable predictive capability for visual perception. However, existing solutions either run DNN inference tasks on drones in situ, impeded by the limited onboard resource, or offload the computation to external servers which may incur large network latency. Few works consider jointly optimizing the offloading decisions along with image transmission configurations and adapting them on the fly. In this paper, we propose A3D, an edge server assisted drone navigation framework that can dynamically adjust task execution location, input resolution, and image compression ratio in order to achieve low inference latency, high prediction accuracy, and long flight distances. Specifically, we first augment state-of-the-art convolutional neural networks for drone navigation and define a novel metric called Quality of Navigation as our optimization objective which can effectively capture the above goals. We then design a deep reinforcement learning based neural scheduler at the drone side for which an information encoder is devised to reshape the state features and thus improve its learning ability. To further support simultaneous multi-drone serving, we extend the edge server design by developing a network-aware resource allocation algorithm, which allows provisioning containerized resources aligned with drones' demand. We finally implement a proof-of-concept prototype with realistic devices and validate its performance in a real-world campus scene, as well as a simulation environment for thorough evaluation upon AirSim. Extensive experimental results show that A3D can reduce end-to-end latency by 28.06% and extend the flight distance by up to 27.28% compared with non-adaptive solutions.

Serving Graph Neural Networks With Distributed Fog Servers For Smart IoT Services

Graph Neural Networks (GNNs) have gained growing interest in miscellaneous applications owing to their outstanding ability in extracting latent representation on graph structures. To render GNN-based service for IoT-driven smart applications, traditional model serving paradigms usually resort to the cloud by fully uploading geo-distributed input data to remote datacenters. However, our empirical measurements reveal the significant communication overhead of such cloud-based serving and highlight the profound potential in applying the emerging fog computing. To maximize the architectural benefits brought by fog computing, in this paper, we present Fograph, a novel distributed real-time GNN inference framework that leverages diverse and dynamic resources of multiple fog nodes in proximity to IoT data sources. By introducing heterogeneity-aware execution planning and GNN-specific compression techniques, Fograph tailors its design to well accommodate the unique characteristics of GNN serving in fog environments. Prototype-based evaluation and case study demonstrate that Fograph significantly outperforms the state-of-the-art cloud serving and fog deployment by up to 5.39x execution speedup and 6.84x throughput improvement.

Real-Time High-Resolution Pedestrian Detection in Crowded Scenes via Parallel Edge Offloading

To identify dense and small-size pedestrians in surveillance systems, high-resolution cameras are widely deployed, where high-resolution images are captured and delivered to off-the-shelf pedestrian detection models. However, given the highly computation-intensive workload brought by the high resolution, the resource-constrained cameras fail to afford accurate inference in real time. To address that, we propose Hode, an offloaded video analytic framework that utilizes multiple edge nodes in proximity to expedite pedestrian detection with high-resolution inputs. Specifically, Hode can intelligently split high-resolution images into respective regions and then offload them to distributed edge nodes to perform pedestrian detection in parallel. A spatio-temporal flow filtering method is designed to enable context-aware region partitioning, as well as a DRL-based scheduling algorithm to allow accuracy-aware load balance among heterogeneous edge nodes. Extensive evaluation results using realistic prototypes show that Hode can achieve up to 2.01% speedup with very mild accuracy loss.

GNN at the Edge: Cost-Efficient Graph Neural Network Processing over Distributed Edge Servers

Edge intelligence has arisen as a promising computing paradigm for supporting miscellaneous smart applications that rely on machine learning techniques. While the community has extensively investigated multi-tier edge deployment for traditional deep learning models (e.g. CNNs, RNNs), the emerging Graph Neural Networks (GNNs) are still under exploration, presenting a stark disparity to its broad edge adoptions such as traffic flow forecasting and location-based social recommendation. To bridge this gap, this paper formally studies the cost optimization for distributed GNN processing over a multi-tier heterogeneous edge network. We build a comprehensive modeling framework that can capture a variety of different cost factors, based on which we formulate a cost-efficient graph layout optimization problem that is proved to be NP-hard. Instead of trivially applying traditional data placement wisdom, we theoretically reveal the structural property of quadratic submodularity implicated in GNN's unique computing pattern, which motivates our design of an efficient iterative solution exploiting graph cuts. Rigorous analysis shows that it provides parameterized constant approximation ratio, guaranteed convergence, and exact feasibility. To tackle potential graph topological evolution in GNN processing, we further devise an incremental update strategy and an adaptive scheduling algorithm for lightweight dynamic layout optimization. Evaluations with real-world datasets and various GNN benchmarks demonstrate that our approach achieves superior performance over de facto baselines with more than 95.8% cost eduction in a fast convergence speed.

Edge Robotics: Edge-Computing-Accelerated Multi-Robot Simultaneous Localization and Mapping

With the wide penetration of smart robots in multifarious fields, Simultaneous Localization and Mapping (SLAM) technique in robotics has attracted growing attention in the community. Yet collaborating SLAM over multiple robots still remains challenging due to performance contradiction between the intensive graphics computation of SLAM and the limited computing capability of robots. While traditional solutions resort to the powerful cloud servers acting as an external computation provider, we show by real-world measurements that the significant communication overhead in data offloading prevents its practicability to real deployment. To tackle these challenges, this paper promotes the emerging edge computing paradigm into multi-robot SLAM and proposes RecSLAM, a multi-robot laser SLAM system that focuses on accelerating map construction process under the robot-edge-cloud architecture. In contrast to conventional multi-robot SLAM that generates graphic maps on robots and completely merges them on the cloud, RecSLAM develops a hierarchical map fusion technique that directs robots' raw data to edge servers for real-time fusion and then sends to the cloud for global merging. To optimize the overall pipeline, an efficient multi-robot SLAM collaborative processing framework is introduced to adaptively optimize robot-to-edge offloading tailored to heterogeneous edge resource conditions, meanwhile ensuring the workload balancing among the edge servers. Extensive evaluations show RecSLAM can achieve up to 39% processing latency reduction over the state-of-the-art. Besides, a proof-of-concept prototype is developed and deployed in real scenes to demonstrate its effectiveness.