3 March 2022 in IEEE Transactions on Mobile Computing
The controller provisioning, which adjusts the number, locations, and members of satellite controllers adaptive to the dynamic network load and topology, fundamentally impacts the performance of software-defined satellite networks (SDSNs). An ideal provisioning strategy should achieve a low total control overhead throughout the entire satellite operation period, which is extremely challenging since the network load can only be predicted in a short time scale . Existing methods can hardly achieve this goal for they greedily configure controllers in each time slot, where switches have to frequently migrate from one controller to another. In this paper, we focus on achieving globally optimized strategies with only current network load information . We first propose a comprehensive control overhead model and formulate the Controller Provisioning Problem (CPP) in SDSNs as a non-convex integer programming problem. To solve the problem, we propose an approximate algorithm named AROA by introducing a regularization framework and based on randomized rounding. We theoretically derive its competitive ratio. To produce strategies in time for future large satellite constellations, we further propose a more efficient heuristic algorithm HROA. Evaluations on our built simulation system show that our proposed methods significantly outperform related schemes in control overhead, latency, and scalability.
9 March 2026 in ACM Computing Surveys
The emergence of large-scale Mixture of Experts (MoE) models represents a significant advancement in artificial intelligence, offering larger model capacity and computational efficiency through conditional computation. However, deploying and running inference on these models presents significant challenges in computational resources, latency, and energy efficiency. This comprehensive survey analyzes optimization techniques for MoE models across the entire system stack. We first establish a taxonomical framework that categorizes optimization approaches into model-level, system-level, and hardware-level optimizations. At the model level, we examine architectural innovations including efficient expert design, attention mechanisms, various compression techniques such as pruning, quantization, and knowledge distillation, as well as algorithm improvement including dynamic routing strategies and expert merging methods. At the system level, we investigate distributed computing approaches, load balancing mechanisms, and efficient scheduling algorithms that enable scalable deployment. Furthermore, we delve into hardware-specific optimizations and co-design strategies that maximize throughput and energy efficiency. This survey provides both a structured overview of existing solutions and identifies key challenges and promising research directions in MoE inference optimization. To facilitate ongoing updates and the sharing of cutting-edge advances in MoE inference optimization research, we have established a repository accessible at https://github.com/MoE-Inf/awesome-moe-inference/ .
8 November 2024 in Communications Biology
The precise classification of cell types from single-cell RNA sequencing (scRNA-seq) data is pivotal for dissecting cellular heterogeneity in biological research. Traditional graph neural network (GNN) models are constrained by reliance on predefined graphs, limiting the exploration of complex cell-to-cell relationships. We introduce scGraphformer, a transformer-based GNN that transcends these limitations by learning an all-encompassing cell-cell relational network directly from scRNA-seq data. Through an iterative refinement process, scGraphformer constructs a dense graph structure that captures the full spectrum of cellular interactions. This comprehensive approach enables the identification of subtle and previously obscured cellular patterns and relationships. Evaluated on multiple datasets, scGraphformer demonstrates superior performance in cell type identification compared to existing methods and showcases its scalability with large-scale datasets. Our method not only provides enhanced cell type classification ability but also reveals the underlying cell interactions, offering deeper insights into functional cellular relationships. The scGraphformer thus holds the potential to significantly advance the field of single-cell analysis and contribute to a more nuanced understanding of cellular behavior.
7 August 2024 in IEEE Transactions on Services Computing
As the most important incremental component in the advent of the 6G era, Low-Earth-Orbit (LEO) satellite networks are becoming increasingly instrumental, and their integration with Software-Defined Networking (SDN) is progressively recognized as a potent strategy for evolving toward truly service-centric networks, where networks are flexiblely reconstructed based on the service demands. Within such networks, the SDN controllers are responsible for network management by making service-aware resource orchestration. Hence, the placement and assignment of controllers emerge as one of the most critical aspects of the network management service, which becomes particularly challenging when confronted with the unique complexities posed by LEO satellite networks, characterized by their highly dynamic topology and unpredictable load fluctuations. In this paper, for the first time, we tackle the issue of controller placement and assignment with a focus on delivering network management services. Firstly, we formulate the adaptive controller placement and assignment problem. Then, we propose the control relation graph (CRG) to capture the control overhead. Next, we present the CRG-based controller placement and assignment algorithm and the sliding window based traffic prediction method. The lookahead-based improvement algorithm is designed to further decrease management costs. Finally, we conduct a series of theoretical analyses including time complexities. Extensive emulation results demonstrate that our algorithms outperform related schemes in terms of response time and load balancing.
19 October 2022 in IEEE Computer Architecture Letters
Multi-modal neural networks have become increasingly pervasive in many machine learning application domains due to their superior accuracy by fusing various modalities. However, they present many unique characteristics such as multi-stage execution, frequent synchronization and high heterogeneity, which are not well understood in the system and architecture community. In this article, we first present and characterize a set of multi-modal neural network workloads of different sizes at inference stage. We then explore their important implications from system and architecture aspects. We hope that our work can help guide future software/hardware design and optimization for efficient inference of multi-modal DNN applications.