Federated Learning (FL) is an emerging distributed machine learning (ML) technique that enables in-situ model training and inference on decentralized edge devices. We propose Totoro$^+$, a novel scalable FL system that enables massive FL applications to run simultaneously on edge networks. The key insight is to explore a distributed hash table (DHT)-based peer-to-peer (P2P) model to re-architect the centralized FL system design into a fully decentralized one. In contrast to previous studies where many FL applications shared one centralized parameter server, Totoro$^+$assigns a dedicated parameter server to each application. Any edge node can act as any application's coordinator, aggregator, client selector, worker (participant device), or any combination of the above, thereby radically improving scalability and adaptivity. Totoro$^+$introduces three innovations to realize its design: a locality-aware P2P multi-ring structure, a publish/subscribe-based forest abstraction, and a game-theoretic path planning model with a guarantee of an$\epsilon$-approximate Nash equilibrium. Real-world experiments on 500 Amazon EC2 servers show that Totoro$^+$scales gracefully with the number of FL applications and$N$edge nodes speeds up the total training time by$1.2\times -14.0\times$, achieves$\mathcal {O}(\log N)$hops for model dissemination and gradient aggregation with millions of nodes, and efficiently adapts to the practical edge networks and churns.

While originally designed for the cloud, the benefits of the serverless paradigm are vital in Edge/Fog computing environments. In this paper, we propose Ekko, a novel decentralized edge serverless scheduling system, which enables a large number of serverless applications to run simultaneously at the edge through the Functionas-a-Service (FaaS) model. The key insight is to re-architect the common centralized or hierarchical scheduling systems into a fully decentralized one by using the distributed hash table (DHT) based peer-to-peer (P2P) model, in which many distributed schedulers operate autonomously without any centralized state. In sharp contrast to existing studies, any edge node in our system can act as a scheduler, a function worker, a query forwarder, or a storage node, and flexibly switch between these roles, thereby significantly improving scalability and adaptivity. Ekko introduces three design innovations: a boundary-aware P2P organization, distributed shadow schedulers with a keychain scheduling algorithm, and a distributed locality-aware bucket image store. Our evaluation on 500 Amazon EC2 nodes shows that, compared to the state-of-the-art, Ekko reduces the 90-th percentile tail queue wait time by up to 96.6 %, the scheduling time by up to 38.5 %, and the total deployment time by up to 89.5 %, while efficiently scaling to millions of function invocation requests on thousands of edge nodes.

While originally designed for the cloud, the benefits of the serverless paradigm are also vital in Edge/Fog computing environments. This paper presents Capybara, a new scalable, programmable distributed object store for storing and sharing serverless function data objects (state) on edge infrastructures. The key innovations here are (1) achieving scalability and avoiding the significant DRAM cost of indexing metadata servers through a "game-theoretic" DHT-based P2P architecture; (2) providing edge users with a "programmable" handler abstraction to customize data management policies, such as different function image caching policies, warm container "keep-alive" durations, data access control methods, and data replication policies.

Edge applications generate a large influx of sensor data on massive scales, and these massive data streams must be processed shortly to derive actionable intelligence. However, traditional data processing systems are not well-suited for these edge applications as they often do not scale well with a large number of concurrent stream queries, do not support low-latency processing under limited edge computing resources, and do not adapt to the level of heterogeneity and dynamicity commonly present in edge computing environments. As such, we present AgileDart, an agile and scalable edge stream processing engine that enables fast stream processing of many concurrently running low-latency edge applications' queries at scale in dynamic, heterogeneous edge environments. The novelty of our work lies in a dynamic dataflow abstraction that leverages distributed hash table-based peer-to-peer overlay networks to autonomously place, chain, and scale stream operators to reduce query latencies, adapt to workload variations, and recover from failures and a bandit-based path planning model that re-plans the data shuffling paths to adapt to unreliable and heterogeneous edge networks. We show that AgileDart outperforms Storm and EdgeWise on query latency and significantly improves scalability and adaptability when processing many real-world edge stream applications' queries.

Graph Neural Networks (GNNs) have emerged as powerful tools for supervised machine learning over graph-structured data, while sampling-based node representation learning is widely utilized in unsupervised learning. However, scalability remains a major challenge in both supervised and unsupervised learning for large graphs (e.g., those with over 1 billion nodes). The scalability bottleneck largely stems from the mini-batch sampling phase in GNNs and the random walk sampling phase in unsupervised methods. These processes often require storing features or embeddings in memory. In the context of distributed training, they require frequent, inefficient random access to data stored across different workers. Such repeated inter-worker communication for each mini-batch leads to high communication overhead and computational inefficiency.

Federated Learning (FL) is an emerging distributed machine learning (ML) technique that enables in-situ model training and inference on decentralized edge devices. We propose Totoro, a novel scalable FL engine, that enables massive FL applications to run simultaneously on edge networks. The key insight is to explore a distributed hash table (DHT)-based peer-to-peer (P2P) model to re-architect the centralized FL system design into a fully decentralized one. In contrast to previous studies where many FL applications shared one centralized parameter server, Totoro assigns a dedicated parameter server to each individual application. Any edge node can act as any application's coordinator, aggregator, client selector, worker (participant device), or any combination of the above, thereby radically improving scalability and adaptivity. Totoro introduces three innovations to realize its design: a locality-aware P2P multi-ring structure, a publish/subscribe-based forest abstraction, and a bandit-based exploitation-exploration path planning model. Real-world experiments on 500 Amazon EC2 servers show that Totoro scales gracefully with the number of FL applications and N edge nodes, speeds up the total training time by 1.2 × -14.0×, achieves O (logN) hops for model dissemination and gradient aggregation with millions of nodes, and efficiently adapts to the practical edge networks and churns.