Source author record

Hanwen Cao

Hanwen Cao appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Information Theory math.IT Computer Vision eess.SP Machine Learning Artificial Intelligence Networking and Internet Architecture

Catalog footprint

What is connected

5works

7topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Codebook Based Two-Time Scale Resource Allocation Design for IRS-Assisted eMBB-URLLC Systems

This paper investigates the resource allocation algorithm design for wireless systems assisted by large intelligent reflecting surfaces (IRSs) with coexisting enhanced mobile broadband (eMBB) and ultra reliable low-latency communication (URLLC) users. We consider a two-time scale resource allocation scheme, whereby the base station's precoders are optimized in each mini-slot to adapt to newly arriving URLLC traffic, whereas the IRS phase shifts are reconfigured only in each time slot to avoid excessive base station-IRS signaling. To facilitate efficient resource allocation design for large IRSs, we employ a codebook-based optimization framework, where the IRS is divided into several tiles and the phase-shift elements of each tile are selected from a pre-defined codebook. The resource allocation algorithm design is formulated as an optimization problem for the maximization of the average sum data rate of the eMBB users over a time slot while guaranteeing the quality-of-service (QoS) of each URLLC user in each mini-slot. An iterative algorithm based on alternating optimization (AO) is proposed to find a high-quality suboptimal solution. As a case study, the proposed algorithm is applied in an industrial indoor environment modelled via the Quadriga channel simulator. Our simulation results show that the proposed algorithm design enables the coexistence of eMBB and URLLC users and yields large performance gains compared to three baseline schemes. Furthermore, our simulation results reveal that the proposed two-time scale resource allocation design incurs only a small performance loss compared to the case when the IRSs are optimized in each mini-slot.

preprint2022arXiv

Optimization-based Phase-shift Codebook Design for Large IRSs

In this paper, we focus on large intelligent reflecting surfaces (IRSs) and propose a new codebook construction method to obtain a set of pre-designed phase-shift configurations for the IRS unit cells. Since the complexity of online optimization and the overhead for channel estimation scale with the size of the phase-shift codebook, the design of small codebooks is of high importance. We consider both continuous and discrete phase-shift designs and formulate the codebook construction as optimization problems. To solve the optimization problems, we propose an optimal algorithm for the discrete phase-shift design and a low-complexity sub-optimal solution for the continuous design. Simulation results show that the proposed algorithms facilitate the construction of codebooks of different sizes and with different beamwidths. Moreover, the performance of the discrete phaseshift design with 2-bit quantization is shown to approach that of the continuous phase-shift design. Finally, our simulation results show that the proposed designs enable large transmit power savings compared to the existing linear and quadratic codebook designs [1], [2].

preprint2020arXiv

ASAP-Net: Attention and Structure Aware Point Cloud Sequence Segmentation

Recent works of point clouds show that mulit-frame spatio-temporal modeling outperforms single-frame versions by utilizing cross-frame information. In this paper, we further improve spatio-temporal point cloud feature learning with a flexible module called ASAP considering both attention and structure information across frames, which we find as two important factors for successful segmentation in dynamic point clouds. Firstly, our ASAP module contains a novel attentive temporal embedding layer to fuse the relatively informative local features across frames in a recurrent fashion. Secondly, an efficient spatio-temporal correlation method is proposed to exploit more local structure for embedding, meanwhile enforcing temporal consistency and reducing computation complexity. Finally, we show the generalization ability of the proposed ASAP module with different backbone networks for point cloud sequence segmentation. Our ASAP-Net (backbone plus ASAP module) outperforms baselines and previous methods on both Synthia and SemanticKITTI datasets (+3.4 to +15.2 mIoU points with different backbones). Code is availabe at https://github.com/intrepidChw/ASAP-Net

preprint2020arXiv

Complex Sequential Understanding through the Awareness of Spatial and Temporal Concepts

Understanding sequential information is a fundamental task for artificial intelligence. Current neural networks attempt to learn spatial and temporal information as a whole, limited their abilities to represent large scale spatial representations over long-range sequences. Here, we introduce a new modeling strategy called Semi-Coupled Structure (SCS), which consists of deep neural networks that decouple the complex spatial and temporal concepts learning. Semi-Coupled Structure can learn to implicitly separate input information into independent parts and process these parts respectively. Experiments demonstrate that a Semi-Coupled Structure can successfully annotate the outline of an object in images sequentially and perform video action recognition. For sequence-to-sequence problems, a Semi-Coupled Structure can predict future meteorological radar echo images based on observed images. Taken together, our results demonstrate that a Semi-Coupled Structure has the capacity to improve the performance of LSTM-like models on large scale sequential tasks.

preprint2012arXiv

A Robust Signal Classification Scheme for Cognitive Radio

This paper presents a robust signal classification scheme for achieving comprehensive spectrum sensing of multiple coexisting wireless systems. It is built upon a group of feature-based signal detection algorithms enhanced by the proposed dimension cancelation (DIC) method for mitigating the noise uncertainty problem. The classification scheme is implemented on our testbed consisting real-world wireless devices. The simulation and experimental performances agree with each other well and shows the effectiveness and robustness of the proposed scheme.