Source author record

Jianhua He

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

ResearcherUnclaimed source record

cond-mat.mes-hall cond-mat.mtrl-sci Networking and Internet Architecture astro-ph.CO astro-ph.GA Computer Vision Information Theory math.IT Multimedia physics.ins-det

Catalog footprint

What is connected

6works

10topics

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

6G for Vehicle-to-Everything (V2X) Communications: Enabling Technologies, Challenges, and Opportunities

We are on the cusp of a new era of connected autonomous vehicles with unprecedented user experiences, tremendously improved road safety and air quality, highly diverse transportation environments and use cases, as well as a plethora of advanced applications. Realizing this grand vision requires a significantly enhanced vehicle-to-everything (V2X) communication network which should be extremely intelligent and capable of concurrently supporting hyper-fast, ultra-reliable, and low-latency massive information exchange. It is anticipated that the sixth-generation (6G) communication systems will fulfill these requirements of the next-generation V2X. In this article, we outline a series of key enabling technologies from a range of domains, such as new materials, algorithms, and system architectures. Aiming for truly intelligent transportation systems, we envision that machine learning will play an instrumental role for advanced vehicular communication and networking. To this end, we provide an overview on the recent advances of machine learning in 6G vehicular networks. To stimulate future research in this area, we discuss the strength, open challenges, maturity, and enhancing areas of these technologies.

preprint2022arXiv

DNN-Driven Compressive Offloading for Edge-Assisted Semantic Video Segmentation

Deep learning has shown impressive performance in semantic segmentation, but it is still unaffordable for resource-constrained mobile devices. While offloading computation tasks is promising, the high traffic demands overwhelm the limited bandwidth. Existing compression algorithms are not fit for semantic segmentation, as the lack of obvious and concentrated regions of interest (RoIs) forces the adoption of uniform compression strategies, leading to low compression ratios or accuracy. This paper introduces STAC, a DNN-driven compression scheme tailored for edge-assisted semantic video segmentation. STAC is the first to exploit DNN's gradients as spatial sensitivity metrics for spatial adaptive compression and achieves superior compression ratio and accuracy. Yet, it is challenging to adapt this content-customized compression to videos. Practical issues include varying spatial sensitivity and huge bandwidth consumption for compression strategy feedback and offloading. We tackle these issues through a spatiotemporal adaptive scheme, which (1) takes partial strategy generation operations offline to reduce communication load, and (2) propagates compression strategies and segmentation results across frames through dense optical flow, and adaptively offloads keyframes to accommodate video content. We implement STAC on a commodity mobile device. Experiments show that STAC can save up to 20.95% of bandwidth without losing accuracy, compared to the state-of-the-art algorithm.

preprint2020arXiv

The parameter-free Finger-Of-God model and its application to 21cm intensity mapping

Using the galaxy catalog built from ELUCID N-body simulation and the semi-analytical galaxy formation model, we have built a mock HI intensity mapping map. We have implemented the Finger-of-God (FoG) effect in the map by considering the galaxy HI gas velocity dispersion. By comparing the HI power spectrum in the redshift space with the measurement from IllustrisTNG simulation, we have found that such FoG effect can explain the discrepancy between current mock map built from N-body simulation and Illustris TNG simulation. Then we built a parameter-free FoG model and a shot-noise model to calculate the HI power spectrum. We found that our model can accurately fit both the monopole and quadrupole moments of the HI matter power spectrum. Our method of building the mock HI intensity map and the parameter-free FoG model will be widely useful for the up-coming 21cm intensity mapping experiments, such as CHIME, Tianlai, BINGO, FAST and SKA. It is also crucial for us to study the non-linear effects in 21cm intensity mapping.

preprint2016arXiv

Optimization of an underwater in-situ LaBr3:Ce spectrometer with energy self-calibration and efficiency calibration

An underwater in situ gamma ray spectrometer based on LaBr3 was developed and optimized to monitor marine radioactivity. The intrinsic background mainly from La138 and Ac227 of LaBr3 was well determined by low background measurement and pulse shape discrimination method. A method of self-calibration using three internal contaminant peaks was proposed to eliminate the peak shift during long term monitoring. With experiments under different temperatures, the method was proved to be helpful for maintaining long term stability. To monitor the marine radioactivity, the spectrometer efficiency was calculated via water tank experiment as well as Monte Carlo simulation.

preprint2011arXiv

Probing Band-Tail States in Silicon MOS Heterostructures with Electron Spin Resonance

We present an electron spin resonance (ESR) approach to characterize shallow electron trapping in band-tail states at Si/SiO2 interfaces in metal-oxide-semiconductor (MOS) devices and demonstrate it on two MOS devices fabricated at different laboratories. Despite displaying similar low temperature (4.2 K) peak mobilities, our ESR data reveal a significant difference in the Si/SiO2 interface quality of these two devices, specifically an order of magnitude difference in the number of shallow trapped charges at the Si/SiO2 interfaces. Thus, our ESR method allows a quantitative evaluation of the Si/SiO2 interface quality at low electron densities, where conventional mobility measurements are not possible.

preprint2010arXiv

Spin relaxation and coherence times for electrons at the Si/SiO2 interface

While electron spins in silicon heterostructures make attractive qubits, little is known about the coherence of electrons at the Si/SiO2 interface. We report spin relaxation (T1) and coherence (T2) times for mobile electrons and natural quantum dots at a 28Si/SiO2 interface. Mobile electrons have short T1 and T2 of 0.3 us at 5 K. In line with predictions, confining electrons and cooling increases T1 to 0.8 ms at 350 mK. In contrast, T2 for quantum dots is around 10 us at 350 mK, increasing to 30 us when the dot density is reduced by a factor of two. The quantum dot T2 is shorter than T1, indicating that T2 is not controlled by T1 at 350 mK but is instead limited by an extrinsic mechanism. The evidence suggests that this extrinsic mechanism is an exchange interaction between electrons in neighboring dots.