Wi-Fi channel frequency response database for contactless human activity recognition

This database collects the channel frequency response (CFR) vectors captured through the Nexmon CSI extraction tool from an Asus RT-AC86U IEEE 802.11ac Wi-Fi router working with a total bandwidth of 80 MHz. The dataset is collected in three different environments, i.e., a bedroom, a living room and a University laboratory, while one person performs one among seven activities of interest within the room. The CFR data for the empty room (E) is also provided. We obtained data from three volunteers (a male, and two females) while they were walking (W) or running (R) around, jumping (J) in place, sitting (L) or standing (S) somewhere in the room, sitting down and standing up (C) continuously, and doing arm gym (H). Each CFR sample results in complex-valued channel information from 242 data sub-channels for each transmit-receive antennas pair. In our experiments, with one transmitter antenna and four at the monitoring device, each sample corresponds to four vectors of 242 complex values. Although the total number of sub-channels at 80 MHz is 256, each antenna vector has 242 components as the CFR is only provided for data sub-channels, namely sub-channels whose indexes are {-122, ..., -2} and {2, ..., 122}, i.e., no CFR value is provided for the control sub-channels. For more information about the setup, please, refer to the related publication. This dataset was used to design and assess the performance of SHARP presented in the article ''SHARP: Environment and Person Independent Activity Recognition with Commodity IEEE 802.11 Access Points'' by Francesca Meneghello, Domenico Garlisi, Nicolò Dal Fabbro, Ilenia Tinnirello, Michele Rossi. The Python source code is available at https://github.com/signetlabdei/SHARP. If you use this dataset, please cite our paper: @misc{meneghello2022SHARP, url = {https://arxiv.org/abs/2103.09924}, author = {Meneghello, Francesca and Garlisi, Domenico and Fabbro, Nicolò Dal and Tinnirello, Ilenia and Rossi, Michele}, title = {Environment and Person Independent Activity Recognition with a Commodity IEEE 802.11ac Access Point}, publisher = {arXiv}, year = {2021} }本数据集收录了通过Nexmon信道状态信息（Channel State Information，CSI）提取工具，从总带宽为80 MHz的华硕（Asus）RT-AC86U IEEE 802.11ac Wi-Fi路由器中采集得到的信道频域响应（Channel Frequency Response，CFR）向量。本数据集采集于三种不同场景：卧室、客厅与大学实验室，实验期间单名受试者在室内完成7种预设目标动作之一。同时还提供了空房间（E）场景下的信道频域响应数据。本次数据采集招募了三名受试者（1名男性、2名女性），分别完成以下动作：室内行走（W）、室内跑步（R）、原地跳跃（J）、静坐（L）、站立（S）、连续坐站交替（C）以及手臂健身训练（H）。针对每一组收发天线对，每个信道频域响应样本均包含来自242个数据子信道的复数值信道信息。本实验中，发射端配备1根天线，监测端配备4根天线，因此每个样本对应4组各含242个复数值的向量。尽管80 MHz带宽下总共有256个子信道，但由于信道频域响应仅提供数据子信道的取值（即索引范围为{-122, …, -2}与{2, …, 122}的子信道），因此每个天线向量仅包含242个分量，未提供控制子信道的信道频域响应数值。若需了解实验部署的更多细节，请参阅相关学术论文。本数据集被用于设计并评估SHARP模型的性能，该模型出自论文《SHARP：基于商用IEEE 802.11接入点的跨环境与跨受试者动作识别》，作者为Francesca Meneghello、Domenico Garlisi、Nicolò Dal Fabbro、Ilenia Tinnirello以及Michele Rossi。该模型的Python源代码可从以下网址获取：https://github.com/signetlabdei/SHARP。若您使用本数据集，请引用以下论文： @misc{meneghello2022SHARP, url = {https://arxiv.org/abs/2103.09924}, author = {Meneghello, Francesca and Garlisi, Domenico and Fabbro, Nicolò Dal and Tinnirello, Ilenia and Rossi, Michele}, title = {Environment and Person Independent Activity Recognition with a Commodity IEEE 802.11ac Access Point}, publisher = {arXiv}, year = {2021} }