Position fingerprinting (FP), in which a common position signature is based on the received signal strength (RSS), is one of the most efficient indoor positioning methods. Another position signature, known as the channel impulse response (CIR), is regarded as a linear temporal filter, which characterizes the multipath channel of the operating environment. We implement a channel sounder based on an orthogonal frequency-division multiplexing system to collect off-line/online CIR measurements and develop a ray-tracing (RT) channel predictor to capture the main characteristics of the channel for the off-line predicted database. We are the first to utilize RT as a channel predictor to assist indoor FP using CIR measurements. We utilize coarse localization to classify the reference points (RPs) based on the access point with the strongest RSS. We propose an RT-assisted FP (RAFP) method, in which we estimate a position by fusing the measured and predicted signatures to find the RPs with the highest correlation values between the online measurement and the off-line measured and simulated CIR databases. Experimental results show that the RAFP-positioning with a hybrid of the predicted and measured CIR-reduces the FP localization error by 25%. By incorporating simulated CIRs, the RAFP has the advantages in reducing human labor for off-line measurement collection and in using less number of CIR measurements to maintain a satisfactory performance. The results encourage a further development to reduce the cost by replacing the sounding system with the wireless network interface cards for a scalable deployment.

Radar sensing technology offers an innovative approach to human-computer interaction, distinguished by its robust sensing capabilities impervious to acoustic and optical disturbances, thus presenting a superior alternative for reliable user engagement. Gesture recognition with mmWave frequency-modulated continuous wave radar extracts the range and velocity from the raw data, such as range-doppler image (RDI). Besides, the angle estimated using a multiple-input-multiple-output (MIMO) radar also contains rich gesture information. Thus, to leverage the MIMO radar technique, we design the usage of the azimuth/elevation-based range-angle images (RAIs) with the average over slow time with the RDI as the spectrum map input for gesture recognition. Since gesture motion is related to the track of position and velocity, we utilize feature extraction on convolutional neural networks of two different spectrum maps, each time sequence learning by cascading long short-term memory, and then fuse the two networks in the end to recognize hand gestures. We validate the proposed scheme based on hand gestures collected by several subjects in different rooms using the 77 GHz mmWave radar from Texas Instrument. With the various combinations of antennas, we observed that the higher angular resolution provided by azimuth and elevation angle could help to enhance the machine-learning model’s ability. By utilizing angle alongside velocity information, the late fusion network provides a classification accuracy of 94.67% and 97.43% among twelve gestures in terms of per frame and sequence accuracy.

Location estimation and tracking for the mobile devices have attracted a significant amount of attention in recent years. The network-based location estimation schemes have been widely adopted based on the radio signals between the mobile device and the base stations. The location estimators associated with the Kalman filtering techniques are exploited to both acquire location estimation and trajectory tracking for the mobile devices. However, most of the existing schemes become inapplicable for location tracking due to the deficiency of signal sources. In this paper, two predictive location tracking algorithms are proposed to alleviate this problem. The predictive location tracking (PLT) scheme utilizes the predictive information obtained from the Kalman filter in order to provide the additional signal inputs for the location estimator. Furthermore, the geometric-assisted PLT (GPLT) scheme incorporates the geometric dilution of precision (GDOP) information into the algorithm design. Persistent accuracy for location tracking can be achieved by adopting the proposed GPLT scheme, especially with inadequate signal sources. Numerical results demonstrate that the GPLT algorithm can achieve better precision in comparison with other network-based location tracking schemes.

Near-ground localization is a special type of localization scenario that is widely considered in both academic and industry. In this communication, with considering the multipath condition in near-ground environments, we conducted field measurements and proposed an antenna height-dependent time of arrival (TOA) ranging error model. Experiment results show that TOA distance measurement errors accord with Gaussian distribution, and its mean and variance are directly related with antenna height and communication distance. Based on the proposed TOA ranging error model, Cramer-Rao lower bound under typical scenarios have been calculated, and localization performance simulation has been studied. The experiment results indicate that the proposed TOA ranging error model has significant theoretical and practical application values.

To acquire higher network capacity, the cell selection problem is formulated based on the partially observable Markov decision process (POMDP). It observes/predicts unavailable cell loading information from those non-serving base stations and takes actions for maintaining active base station set and handover target selection. Various utility functions are designed to consider different factors in the proposed POMDP-based cell selection (POCS) schemes, including system capacity, handover time, and mobility of mobile station. With the considerations of cell load as a hidden Markov model and predicted mobility in the reward function, simulation results show that the proposed POCS schemes can outperform conventional received signal strength based and load balancing methods.