LifeHD: Lifelong Intelligence Beyond the Edge Using Hyperdimensional Computing

LifeHD: Lifelong Intelligence Beyond the Edge Using Hyperdimensional Computing On-device learning has emerged as a prevailing trend that avoids the slow response time and costly communication of cloud-based learning. The ability to learn continuously and indefinitely in a changing environment, and with resource constraints, is critical for real sensor deployments. However, existing designs are inadequate […]

Read More

Collaborative Networked Estimation and Inference through Graph Neural Networks

Collaborative Networked Estimation and Inference through Graph Neural Networks Graph neural networks (GNNs) are deep learning architectures that learn powerful representations of graphs and graph signals. GNNs have shown remarkable performance across various domains, such as biology and drug discovery, quantum chemistry, robotics, social networks, and recommender systems. Their success can be attributed to their […]

Read More

Rethinking Deep Learning Compression Using Information Theoretic Structures

Rethinking Deep Learning Compression Using Information Theoretic Structures Neural compression has brought tremendous progress in designing lossy compressors with good rate-distortion (RD) performance at low complexity. Thus far, neural compression design involves transforming the source to a latent vector, which is then rounded to integers and entropy coded. While this approach has been shown to […]

Read More

Data-driven Adaptive Network Models: Quantitative Group Testing

  < BACK TO ALL TILOS NETWORKS PROJECTS Data-driven Adaptive Network Models: Quantitative Group Testing The quantitative group testing (QGT) problem aims at learning an underlying binary vector x of length n with a sparsity parameter k. The information acquisition process about x involves conducting pooled measurements, also known as group tests, where the outcome […]

Read More

Empirical Network Optimization via Non-uniform Sampling

Empirical Network Optimization via Non-uniform Sampling Large-scale distributed optimization is a key component of the design of multi-scale network protocols. Traditionally, architecture-level hyperparameters and protocol parameters are optimized manually at run-time, which requires a human fine-tuning step to produce an optimized network performance metric. We will capitalize on the abundance of data and past design […]

Read More

Empirical Network Optimization via Distributed Zeroth-order Optimization

Empirical Network Optimization via Distributed Zeroth-order Optimization Distributed zeroth-order optimization forms the backbone of empirical network optimization, and has wide applications in federated learning, where multiple agents use their local data and computation to collectively train a model. There are two main challenges in practical network optimization tasks. The first is the unavailable gradient of […]

Read More

Learning to Slice Wi-Fi Networks: A State-Augmented Primal-Dual Approach

Learning to Slice Wi-Fi Networks: A State-Augmented Primal-Dual Approach In enterprise settings, it is vital to manage network operations to support multiple use cases with different requirements. Additionally, 3GPP includes architectures to integrate Wi-Fi in converged connectivity (5G + Wi-Fi) in enterprise. Network slicing allows an access point (AP) to allocate the network resources across […]

Read More