Jiaxin's Homepage

An OFDMA-based system designed for Ultra-Reliable Low-Latency Communications (URLLC) in Industrial Internet of Things (IIoT).
Achieves 99.995% reliability with 0.8ms average latency through centralized scheduling and robust modulation.
Implemented on USRP SDR platform with GNURadio, supporting up to 32 simultaneous users in the 2.4GHz band.

Software-define Radio Wireless System Design Signal Processing

A real-time wireless system that is implemented on PC-USRP using GNURadio, targeting to provide TSN service to the upper layer with high accuracy synchronization.
Achieving synchronization among nodes to within 100ns, and the end-to-end latency can be down to 3.75ms.
Easy-deployable and reconfigurable for many other time-critical wireless applications.

Software-define Radio Signal Processing Synchornization

A wireless system implemented on real UAVs using Embedded USRP E310 for signal transmission and reception.
Implements Physical-layer Network Coding (PNC) with compute-and-forward protocol for improved spectral efficiency and reliability.
Demonstrates successful network coding operations on airborne platforms with improved throughput over traditional relaying.
Features Doppler shift compensation and adaptive power control for stable performance under UAV mobility.

Software-define Radio Wireless System Design Signal Processing

It was the first-of-its-kind collaborative machine-learning competition to overcome scarcity in the radio frequency (RF) spectrum.
Built a real-time wireless system that leverage Artificial Intelligence to access the spectrum without interfering the other signals.
Cooperated with the research group from the North-Western University
Qualified for Preliminary Event 1 & 2 (PE1 & 2) and presented in the events hosted in Johns Hopkins University Applied Physics Laboratory.

Software-define Radio Wireless System Design Signal Processing

A full-stack time-slotted system built on USRP and GNURadio with flexible frame structure and microsecond-level timing precision.
Supports multiple MAC protocols: TDMA, CSMA, Coded Slotted Aloha, and Physical-layer Network Coding for rapid prototyping.
Features modular design for easy integration of new MAC schemes and real-world performance evaluation.
Serves as the implementation platform for multiple research projects enabling cross-layer optimization studies.

Software-define Radio Wireless System Design

Implemented a reduced-complexity Viterbi decoder for Physical-layer Network Coding (PNC) with soft-decision decoding and SIMD acceleration.
Built the first single-GPP based PNC development environment in GNU Radio, reducing transmission slots from 4 to 2 for doubled throughput.
Demonstrated real-time video streaming with improved quality and reduced latency compared to traditional relaying.
Open-source contribution serving as the foundation for subsequent PNC research projects.

Software-define Radio Wireless System Design Signal Processing

Investigates optimal broadcast modes for offline finding networks (e.g., Apple AirTag, Samsung SmartTag) through large-scale measurement with 50 real-world deployed tags.
Proposes an adaptive broadcast mode selection algorithm that optimizes energy efficiency and finding success rate based on environmental conditions.
Implements a prototype system and demonstrates significant performance improvements over existing solutions.
Published in IEEE Transactions on Mobile Computing (TMC), 2025.

IoT Bluetooth Low Energy Wireless Networks Network Protocols

First practical SDR implementation of Age-of-Information (AoI) oriented random access using USRP N210 and GNURadio.
Proposes and implements a two-level updating mechanism that reduces peak AoI by 30% and average AoI by 20% compared to baseline.
Validates theoretical models through real-world experiments, demonstrating the effectiveness of simple schemes in practical scenarios.
Published in IEEE IECON 2020.

Software-defined Radio Age of Information Random Access USRP GNURadio

An open-source simulation framework for BLE neighbor discovery based on NS-3, accurately modeling BLE 5.0 protocol stack with real-world factors.
Validated against real-world experiments with < 5% error, accounting for duty cycles, scan intervals, and broadcast intervals.
Enables rapid prototyping and evaluation of BLE neighbor discovery schemes without costly hardware deployments.
Published in ACM MSWiM 2022.

BLE 5.0 Network Simulation NS-3 Wireless Protocols

Research Projects