Related papers: ODMA: On-Demand Memory Allocation Strategy for LLM…

We consider the allocation of spectral and power resources to the mobiles (i.e., user equipment (UE)) in a cell every subframe (1 ms) for the Long Term Evolution (LTE) orthogonal frequency division multiple access (OFDMA) cellular network.…

Low-Rank Adaptation (LoRA) fine-tunes large models by learning low-rank updates on top of frozen weights, dramatically reducing trainable parameters and memory. However, there is still a gap between full training with low-rank projections…

The paradigm of large language model (LLM) reasoning is shifting from parameter scaling to test-time compute scaling, yet many existing approaches still rely on uniform brute-force sampling (for example, fixed best-of-N or self-consistency)…

We propose MatchRDMA, a proactive, segmented, and rate-matched long-haul RDMA scheme for geo-distributed LLM training over OTN. By coordinating source and destination OTN rates, it improves inter-DC throughput by up to 20x compared with…

Online Continual Learning (OCL) involves sequentially arriving data and is particularly challenged by catastrophic forgetting, which significantly impairs model performance. To address this issue, we introduce a novel framework, Online…

This work considers the orthogonal frequency division multiple access (OFDMA) technology that enables multiple unmanned aerial vehicles (multi-UAV) communication systems to provide on-demand services. The main aim of this work is to derive…

In cloud services, virtual machine (VM) scheduling is a typical Online Dynamic Multidimensional Bin Packing (ODMBP) problem, characterized by large-scale complexity and fluctuating demands. Traditional optimization methods struggle to adapt…

The memory capacity in edge devices is often limited due to constraints on cost, size, and power. Consequently, memory competition leads to inevitable page swapping in memory-constrained mixed-criticality edge devices, causing slow storage…

Cache-assisted ultra-dense mobile edge computing (MEC) networks are a promising solution for meeting the increasing demands of numerous Internet-of-Things mobile devices (IMDs). To address the complex interferences caused by small base…

Today's high-speed switches employ an on-chip shared packet buffer. The buffer is becoming increasingly insufficient as it cannot scale with the growing switching capacity. Nonetheless, the buffer needs to face highly intense bursts and…

Large Language Models (LLMs) face limitations due to the high demand on GPU memory and computational resources when handling long contexts. While sparsify the Key-Value (KV) cache of transformer model is a typical strategy to alleviate…

Large language models (LLMs) often suffer from catastrophic forgetting in continual learning (CL) scenarios, where performance on previously learned tasks degrades severely while training on sequentially arriving tasks. Although pioneering…

This paper considers the resource allocation algorithm design for downlink multiple-input single-output (MISO) orthogonal frequency division multiple access (OFDMA) ultra-reliable low latency communication (URLLC) systems. To meet the…

Multiple access techniques are vital for 5G and beyond. While Orthogonal Frequency Division Multiple Access (OFDMA) is standard, its high peak-to-average power ratio (PAPR) reduces energy efficiency in uplink transmissions. This paper…

A large language model (LLM) is one of the most important emerging machine learning applications nowadays. However, due to its huge model size and runtime increase of the memory footprint, LLM inferences suffer from the lack of memory…

In this work, we present the design for both pilot-uncoupled and pilot-free on-off multiple access (ODMA) receivers in unsourced random access (URA) for multiple-input multiple-output (MIMO) systems. Unlike pilot-coupled ODMA, where on-off…

Memory bandwidth is known to be a performance bottleneck for FPGA accelerators, especially when they deal with large multi-dimensional data-sets. A large body of work focuses on reducing of off-chip transfers, but few authors try to improve…

Deep Reinforcement Learning (DRL) has demonstrated strong performance in robotic control but remains susceptible to out-of-distribution (OOD) states, often resulting in unreliable actions and task failure. While previous methods have…

RDMA is vital for efficient distributed training across datacenters, but millisecond-scale latencies complicate the design of its reliability layer. We show that depending on long-haul link characteristics, such as drop rate, distance and…

Multi-Head Latent Attention (MLA), introduced in DeepSeek-V2, improves the efficiency of large language models by projecting query, key, and value tensors into a compact latent space. This architectural change reduces the KV-cache size and…