Considering the limited spectrum and incomplete coverage of communication infrastructures for mobile edge computing (MEC), we present a cognitive unmanned aerial vehicle (UAV)-assisted MEC system, where UAVs share licensed spectrum with primary user (PU) for computation offloading. A joint trajectory design, power control, and computation task allocation (JTPAC) problem is formulated to minimize the long term system cost under the constraint of PU communication rate. To solve this complicated problem, we transform it into a partially observable Markov decision process (POMDP) based optimization and propose a multi-agent deep deterministic policy gradient (MADDPG)-based JTPAC algorithm. Simulation results show that the proposed algorithm achieves lower system cost compared to the benchmark algorithms and the PU communication rate can be guaranteed under different situations.

The automatic signal modulation recognition of Orthogonal Frequency Division Multiplexing (OFDM) multi-carrier signals holds significant research importance and application value in areas such as Wi-Fi, 4G, and 5G communications. However, traditional automatic modulation recognition methods perform poorly in multi-carrier communication systems, especially in complex real-world channel environments. In this study, we first create an OFDM multi-carrier signal dataset based on real channel data, named OFDM2024, which includes data on various modulation types under multiple signal-to-noise (SNR) ratio conditions. Secondly, we propose a modulation recognition method for OFDM signals based on a temporal-frequency attention mechanism using the Transformer model, called TF-Transformer. First, we design temporal feature extraction and frequency feature extraction modules to extract temporal and frequency domain features from the signal data. Next, we develop a feature fusion module to integrate the obtained features. Finally, we different the modulation types of the signals based on the fused features. Experimental results indicate that TF-Transformer significantly outperforms traditional signal recognition algorithms such as ViT, CNN, and ResNet in terms of performance.

This study focuses on a free space optical (FSO)/radio frequency (RF) satellite-UAV-terrestrial integrated network (SUTIN) to overcome limitations of traditional RF SUTIN on spectrum scarcity and security. Traditional resource allocations for FSO/RF SUTIN often prioritize transmission efficiency while neglecting reliability. Moreover, in this network, atmospheric turbulence, pointing errors, and multipath fading can lead to unreliable data transmission. To resolve these problems, we propose a hybrid FSO/RF SUTIN based on adaptive modulation and coding scheme (MCS) and resource allocation algorithm to ensure both transmission efficiency and reliability with imperfect channel state information (CSI) and random channel fading. A joint optimization problem is formulated, incorporating MCS mode selection, power allocation, and UAV trajectory optimization, with frame error rate (FER) and CSI uncertainty constraints to maximize system throughput. To solve this complicated high dimension optimization problem, we present mathematical mapping models between the optimization variables and communication performance parameters (the average transmission rate and FER) with imperfect CSI and random fading for both FSO and RF link. According to these mapping models, a robust solution based on deep deterministic policy gradient is proposed, integrating decoupling and user clustering techniques to reduce the computational complexity. The simulation results demonstrate that this proposed FSO/RF SUTIN and the corresponding algorithm can improve system throughput and guarantee transmission reliability.

With the increasing complexity of next generation network applications and the coexistence of diverse service requirements, Generative AI (GAI) and Large Models (LMs) based semantic communication are widely regarded as promising solutions to address these challenges. The goal of these systems is not only to reduce system burden by reducing transmission data, but also to adapt to new and complex requirements. In this paper, we propose a semantic communication system designed to meet diverse requirements of speech applications while enabling accurate speech transmission. The semantic encoder comprises an unsupervised model wav2vec 2.0 for learning universal speech representations to enable adaptability across various speech-related tasks. It also includes a prosodic feature encoder from the style embedding module of Global Style Tokens (GST) Tacotron. The semantic decoder integrates a phoneme recognition module and a GST-Tacotron-based text-to-speech (TTS) module to facilitate accurate and expressive reconstruction of the original speech signal, with the incorporation of prosodic features enhancing the naturalness and intelligibility of the synthesized speech. The proposed system has been tested in noisy channels. It demonstrates that the system maintains superior and robust performance even at Bit Error Rate (BER) of 10−1, as reflected by a stable Character Error Rate (CER) approximately 0.0940 and 0.0649 for the base and large versions of wav2vec 2.0 respectively in speech recognition, and consistent cFDSD scores approximately 0.9 in speech quality assessment. This performance surpasses that of the existing semantic communication systems, while also providing reliable support for a wider range of downstream speech applications.

Millimeter-Wave (mmWave) radar stands as a promising candidate for high-precision indoor passive localization, exhibiting robust performance across diverse environmental conditions, including those with smoke and limited light. We introduce a mmWave radar-based full scene adaptive localization (FSAL) system that adeptly tackles both line-of-sight (LOS) and nonline-of-sight (NLOS) scenarios. The system boasts several key features. Notably, it includes a point-cloud reflection scenario network, which effectively differentiates the geometric characteristics of sparse mmWave point cloud to achieve an impressive 97% accuracy in distinguishing between various types of scenarios. In addition, it incorporates an adaptive localization algorithm that leverages an adaptive Kalman filtering for real-time tracking and smoothing of motion trajectories. Furthermore, the system is underpinned by an integrated framework that ensures comprehensive localization. This FSAL system demonstrates remarkable capabilities in real-time tracking and accurate localization, i.e., a mean accuracy of 5 cm in a mixed LOS and NLOS conditions, making it particularly well-suited for indoor environments.

Modulation recognition has emerged as an intensive research topic to improve the communication efficiency in the future 6G network and plays an important role in the security of electromagnetic spectrum. Various pattern recognition methods have been proposed to enhance the performance of modulation recognition, especially deep learning models showing their emerging performance. In this work, we design a modulation recognition model based on an enhanced Transformer, namely MR-Transformer, which is accelerated on a Field Programmable Gate Array (FPGA). The design of MR-Transformer targets on high recognition accuracy, low power consumption, and high computation efficiency, which is suitable for modulation recognition at edge devices. MR-Transformer leverages attention mechanism to extract global features and correspondingly enhance the recognition accuracy. An improved matrix multiplication operation and enhanced Design Space Exploration (DSE) method are proposed in MR-Transformer to improve the computation efficiency and reduce resource consumption. We conduct comprehensive experiments to evaluate the performance of MR-Transformer on three platforms, i.e. Central Processing Unit (CPU), Graphics Processing Unit (GPU), and FPGA, based on two open-source datasets. According to the evaluation results, the MR-Transformer based on FPGA shows the best performance compared with the baseline models considering accuracy, power consumption, and computation efficiency.

The fingerprint positioning scheme has important application value and development prospects in indoor positioning. However, traditional fingerprint positioning schemes often require time-consuming and laborious fingerprint collection to improve positioning accuracy. In this study, we introduce the BERT model in Natural Language Processing (NLP) into the field of indoor positioning and propose a semi-supervised indoor positioning with masked crowdsourcing fingerprints. First, we convert the fingerprint data into text-like data to fit the model input. Then we design the enhanced BERT as a feature extractor to pretrain the masked unlabeled crowdsourcing fingerprints, namely CrowdBERT. Finally, CrowdBERT is fine-tuned with sparsely labeled fingerprints to achieve high-precision indoor positioning. Experimental results demonstrate that CrowdBERT significantly outperforms the other traditional positioning algorithms such as KNN, DNN, ResNet, SAE and VAE.

As a mature indoor positioning solution, fingerprint-based positioning has been widely applied. However, traditional fingerprint positioning schemes still face the problems of limited hidden spatial feature extraction ability and insufficient fingerprint calibration with unlabeled crowdsourcing data. In order to address the above problems, we refer to the transformer-based deep learning model in natural language processing (NLP) and propose a crowdsourcing indoor positioning model via semi-supervised bidirectional encoder representation for transformer with masking, namely CrowdBERT. First, we tokenize the fingerprint data to adapt to the input form of the model. Then, we design a spatial fingerprint attention encoder as a feature extractor, which internal multihead attention mechanism combined with three-layer spatial feature embedding can fully capture the spatial features of fingerprint sequences. Meanwhile, we propose received signal strength-token masking to help the model perform bidirectional feature extraction so that the pretraining can more efficient use of hidden features from unlabeled crowdsourcing fingerprints. Finally, the limited labeled fingerprints is used to fine tune the downstream network structure to further improve the positioning accuracy. To evaluate our proposed positioning system, we conduct a set of comprehensive experiments on the three different data sets and evaluation results demonstrate that the CrowdBERT model significantly outperforms the other traditional positioning algorithms, such as K nearest neighbor, DNN, residual network, stacked autoencoder, and variational autoencoder.

The millimeter-wave (mmWave) spectrum has become a core of wireless communication, which has the advantages of richer spectrum resources, larger communication bandwidth, and smaller spectrum interference. Human activity recognition (HAR) by mmWave radar based on point cloud attracts significant attention due to its nature of privacy-preserving, which is an important task of realizing integrated sensing and communication (ISAC). This article proposes a framework of spatial–temporal point cloud transformer (ST-PCT) to realize high precision of HAR, based on sequential point cloud after preprocessing from mmWave radar without voxelization. In ST-PCT, it consists of four enhanced components: 1) a framewise spatial neighbor embedding module to extract the local feature; 2) a temporal and spatial attention mechanism module to find connections within and across frames; 3) an optimized attention mechanism to improve the efficiency of feature extraction; and 4) a sensor fusion module with more motion information to improve the difference between activities. We experimentally evaluate the efficiency of our framework compared with several approaches based on the voxelization or point cloud directly. The experimental results have demonstrated that the proposed ST-PCT network greatly outperforms the other approaches in terms of overall accuracy (oAcc), achieving 99.06% and 99.44%, respectively, on two data sets.

Vehicular edge computing (VEC) is considered to be a key technology to improve the processing efficiency of computing tasks for the Internet of Vehicles (IoV). Using roadside units (RSUs) distributed on both sides of a road as edge servers, computation-intensive and latency-sensitive in-vehicle tasks can be responded to quickly. However, some quality of service (QoS) is often difficult to ensure due to clogged dense urban buildings or lack of infrastructure in remote areas. In this paper, we propose a software-defined network (SDN)-driven partial offloading model for unmanned aerial vehicle (UAV)-assisted VEC networks, where the RSUs and UAVs jointly provide computing services to the vehicles and collect global information through centralized control using a SDN controller. To guarantee these vehicles obtain computing results in time and rationally utilize computing resources, we develop an optimal offloading mechanism using age of information (AoI), together with energy consumption and rental price as a comprehensive weighted cost of our above optimization objective. The total system cost of the performing tasks is minimized by jointly optimizing the UAV trajectory, user association, and offloading decision. Considering the mobility of the vehicles and UAVs and the dynamic network environment, we design a deep reinforcement learning (DRL)-based joint trajectory control and offloading allocation algorithm (DRL-TCOA) to solve the proposed computation offloading problem. Experimental results show that the proposed DRLTCOA algorithm maintains better information freshness and lower system cost than the other baseline offloading strategies.

With the fast development of location-based services, an ubiquitous indoor positioning approach with high accuracy and low calibration has become increasingly important. In this work, we target on a crowdsourcing approach with zero calibration effort based on visible light, magnetic field, and WiFi to achieve submeter accuracy. We propose a CrowdFusion simultaneous localization and mapping (SLAM) composed of coarse-grained and fine-grained trace merging, respectively, based on the iterative closest point (ICP) SLAM and GraphSLAM. ICP SLAM is proposed to correct the relative locations and directions of crowdsourcing traces and GraphSLAM is further adopted for fine-grained pose optimization. In CrowdFusion SLAM, visible light is used to accurately detect loop closures and magnetic field to extend the coverage. According to the merged traces, we construct a radio map with visible light and WiFi fingerprints. An enhanced particle filter fusing inertial sensors, visible light, WiFi, and floor plan is designed, in which visible light fingerprinting is used to improve the accuracy and increase the resampling/rebooting efficiency. We evaluate CrowdFusion based on comprehensive experiments. The evaluation results show a mean accuracy of 0.67 m for the merged traces and 0.77 m for positioning, merely replying on crowdsourcing traces without professional calibration.

Gesture recognition has been a hot research topic in human–computer interaction, since contactless gesture recognition will provide increasing applications in many fields. Millimeter-wave (mmWave) radar well serves this technology because of its high accuracy, easy integration, and strong anti-jamming ability in moving object detection. However, it is still challenging to meet the requirement of high precision in subtle gesture recognition based on traditional methods via point cloud or Range-Doppler heat map of mmWave radar. Considering the raw data from mmWave radar with more information, such as phase, we propose a system that uses the constructed mmWave radar data cube sequence and timedistributed-CNN-transformer network (CTN), called DCS-CTN system, to get higher hand gesture recognition accuracy. In this system, we introduce a time-distributed wrapper (TD) and convolutional neural network (CNN) to extract local features of the data cube sequence, a position encoder to retain time information of the sequence, and a transformer network to get global features of the sequence. The experiments results show that this system can achieve hand gesture recognition accuracy of 99.75%, which is significantly higher than the other traditional approaches.

In recent years, the demand for indoor positioning systems has grown rapidly with regard to location-based services. As a cost-effective choice, WiFi-based indoor positioning has attracted great increasing research attentions because it does not require external devices installed in the target environment. Although extensive research has been conducted on WiFi fingerprint matching, the problem of automatically adapting radio maps to fresh signal space environment still exists. The traditional methods often conduct site surveys regularly to update the outdated radio maps, which is time consuming and laborious. In this work, we propose an indoor positioning system AAIMSS to automatically update the radio maps based on an enhanced transfer learning (TL) approach with altered access points (APs) identification and mapping space searching. In the system, the proposed TL approach removes the outlier features and then searches a more accurate mapping space between the original radio map and crowdsourcing data. Our lightweight solution does not rely on additional devices and inertial sensors with high-power consumption. A set of experiments has been conducted in a teaching building to evaluate AAIMSS. The results show that this AAIMSS is robust to locate users in a dynamic environment. The average positioning accuracy achieves 2.5m, which significantly outperforms the positioning strategies with the original radio map by 65.3%, the radio map by directly removing the altered APs by 19.4%, and the radio map by the traditional TL by 85.2%.

Recently, indoor positioning has played a critical role in many emerging indoor applications. However, due to complicated indoor environments, it is still challenging to develop an indoor positioning system with high positioning accuracy and low deployment efforts. In this work, an indoor positioning system based on visible light fingerprinting is proposed by leveraging a novel visible light GraphSLAM (VL-GraphSLAM) technique. The proposed VL-GraphSLAM provides enhanced solutions at both frontend and backend to improve the accuracy of estimated trajectories. Then, the estimated trajectory is anchored in floor map based on a novel door detection method to recover indoor walking paths. Based on VL-GraphSLAM, we construct a database with the visible light received signal strength labeled by the locations of walking paths, which is called visible light map. Moreover, a Kalman filter is adopted to fuse the visible light fingerprinting and inertial sensors to locate users. Comprehensive experiments illustrate that our proposed system can accurately recover walking paths (0.4 m) and locate users (0.9 m) in an accuracy of submeter, which significantly outperforms a traditional WiFi-based fingerprinting system and is more convenient to deploy than a traditional visible light positioning based on ranging.

Traditional WiFi positioning relies on a predefined radio map, which is labor-intensive and time-consuming for professionals. Recently, crowdsourcing has emerged as a promising solution for facilitating WiFi positioning. To crowdsense a radio map, traces collected from normal users are merged to recover the original walking paths. In this work, we design a robust iterative trace merging algorithm called WiFi-RITA based on WiFi access points as signal-marks. The algorithm formulates the trace merging problem as an optimization problem in which each trace is translated and rotated to minimize the limitation of distances among traces defined by WiFi access points. WiFi-RITA is further enhanced by removing outliers. WiFi-RITA is robust to the rotation errors of traces and efficient for a large number of short traces. According to the crowdsensed radio map, a sensor fusion approach based on particle filter by fusing inertial sensors and a multivariate Gaussian fingerprinting is proposed to enhance the accuracy of crowdsourcing indoor positioning. The experiment results in two large-scale environments demonstrate that WiFi-RITA positioning with zero-effort calibration achieves high positioning accuracy, which outperforms Pedestrian Dead Reckoning (PDR) and fingerprinting with K Nearest Neighbor.

Seamless outdoor-indoor positioning plays a critical role in many emerging applications, e.g., large-coverage user navigation in cities, smart buildings, and analytics of user spatial location big data. It is still challenging to construct a large-scale seamless outdoor-indoor positioning system due to the limited coverage of indoor positioning. In this paper, we propose a seamless outdoor-indoor crowdsensing positioning (SoiCP) system in which a radio map is automatically constructed based on crowdsourcing pedestrian dead reckoning (PDR) traces without professional site surveying. The constructed radio map is robust to inaccurate PDR traces and does not rely on prior knowledge of floor plans. In SoiCP, the crowdsensed radio map is obtained by a proposed three-step trace matching algorithm. This algorithm leverages building gates and WiFi fingerprints as landmarks to merge the noisy crowdsourcing traces and accurately construct the user walking paths. Moreover, following the crowdsensed radio map, SoiCP uses an enhanced particle filter to fuse PDR, GPS, and WiFi fingerprinting for seamless outdoor-indoor positioning with high accuracy. The comprehensive real-world experiments in two large-scale shopping malls demonstrate that SoiCP can effectively crowdsense the walking paths and track moving users with high accuracy.

Wireless body area networks (WBANs) are not only an extension and branch of wireless sensor networks (WSNs) but also a practical application area of Internet of Things (IoT). With the extensive development of IoT technology, WBAN can monitor human physiological parameters in real time. Reliable information transmission is an important factor limiting the development of WBAN owing to special path loss and shadowing effect. Therefore, maximizing throughput is a pivotal part of improving system performance. In this paper, a multipoint WBAN (MP-WBAN) with energy harvesting for normal and abnormal scenarios is studied. We propose two different protocols, including a time switching (TS) strategy and a hybrid TS and power splitting (PS) strategy, respectively. In the abnormal scenarios, the access point (AP) harvests independent command signals from sensor nodes in the uplink (UL) and broadcasts dedicated energy signals to all sensor nodes in the downlink (DL). At the same time, the AP simultaneously broadcasts wireless command and energy signals to all sensor nodes in the normal situation. After all sensors harvest energy from the radio frequency (RF) signals, physiological datas can be transfered to the AP in a specific time sequence. We optimize TS ratios to achieve the abnormal situation sum-throughput maximization by utilizing convex optimization techniques. For sum-throughput maximization in normal situation, a near-optimal solution can be acquired by iteratively updating TS ratios and PS ratios. Numerical simulation results show the system performances of sum-throughput can be significantly improved by the proposed algorithms.

Positioning is envisioned as an essential enabler of future fifth generation (5G) mobile networks due to the massive number of use cases that would benefit from knowing users' positions. In this work, we propose a particle filter-based reinforcement learning (PFRL) approach for the robust wireless indoor positioning system. Our algorithm integrates information of indoor zone prediction, inertial measurement units, wireless radio-based ranging, and floor plan into an particle filter. The zone prediction method is designed with an ensemble learning algorithm by integrating individual discriminative learning methods and Hidden Markov Models. Further, we integrate the particle filter approach with a reinforcement learning-based resampling method to provide robustness against localization failure problems such as the kidnapping robot problem. The PFRL approach is validated on a two-tier architecture, in which distributed machine learning tasks are hosted at client and edge layer. Experiment results show that our system outperforms traditional terminal-based approaches in both stability and accuracy.

Crowdsensing indoor walking paths based on crowdsourcing traces collected from normal users has recently become an emerging topic for indoor positioning, which can reduce the labor effort of building radio maps and improve the positioning accuracy when a floor plan is unavailable. In this work, we design an indoor walking path crowdsensing system with massive noisy crowdsourcing traces. In this system, we propose a robust iterative trace merging algorithm based on WiFi access points as markers (named 'WiFi-RITA') to merge massive noisy traces. The algorithm formulates the trace merging problem as an optimization problem in which each trace is controlled to translate and rotate to minimize the limitation of distances among traces defined by WiFi access points as markers. WiFi-RITA is robust to the rotation errors and uncertain absolute locations of user traces, and can efficiently work for a large number of user traces. We further adopt a landmark matching algorithm to match the merged traces to the target building and adopt a 2-dimensional histogram approach to remove outlier traces. With such procedures, we generate walking paths of a large-scale building with a mean accuracy of 2.1m.

In this work, we propose an automatic radio map construction system based on crowdsourcing Pedestrian Dead Reckoning (PDR) traces, which does not rely on priori knowledge of floor plans and is robust to inaccurate PDR traces. In this system, we propose to process some opportunistic PDR traces, in which users walk through the building, to generate parts of road paths by translating, rotating and scaling the traces based on the opportunistic GPS locations and gate points as landmarks. Then, we further extend the coverage of road paths by processing the PDR traces entirely obtained indoor by compensating the turning errors and merging the PDR traces based on the similarity of WiFi fingerprints. With such procedures, we can accurately generate indoor road paths of a large-scale building and construct the radio map based on these road paths. Our proposed method achieves a median accuracy of 2.8m and mean accuracy of 2.9m for the constructed road paths. By fusing GPS, PDR, and WiFi fingerprinting with the crowdsourcing radio map, we achieve a median positioning accuracy of 2.9m and mean accuracy of 3.4m without site surveying, which significantly outperforms the positioning algorithm by merely fusing GPS and PDR.

Location aware services in the Internet of Things are essential for smart environments. Location awareness enables operational systems to deliver useful information for supplying context-aware applications. We propose an efficient probabilistic model to provide good and stable localization accuracy in smart building environments for smartphones. Our proposed localization method fuses zone detection, radio-based ranging, inertial measurement units and floor plan information into an enhanced particle filter. Zone detection is designed with an ensemble learning algorithm by combining Hidden Markov Models and discriminative learning methods. We first apply ensemble learning models to achieve zone detection. Further, we integrate zone detection and an enhanced ranging model to achieve high and stable localization performance. Experiment results in an office-like indoor environment show that our system outperforms traditional localization approaches considering stability and accuracy. The localization method can achieve performance with an average localization error of 1.26 meters.

Passive positioning systems produce user location information for third-party providers of positioning services. In this paper, we provide a passive tracking system for WiFi signals with an enhanced range-only particle filter using finegrained power. Our proposed particle filter, WVT-bootstrap particle filter, provides improved observation likelihood and is equipped with a single coordinated turn model to address the challenges in passive positioning. The anchor nodes for WiFi signal sniffing use software defined radio techniques to extract channel state information for multipath mitigation and a nonlinear regression method is used for the path-loss model. Our tracking system produces measured positioning errors that, in the 80th percentile, are equal to or less than 2 m; this represents a 33% improvement over the traditional bootstrap particle filter. Additionally, it requires (0.12 s for 1000 particles) only half of the computation efforts as a multi-model particle filter.

Indoor positioning has become an emerging research area because of huge commercial demands for location-based services in indoor environments. Channel State Information (CSI) as a fine-grained physical layer information has been recently proposed to achieve high positioning accuracy by using range-based methods, e.g., trilateration. In this work, we propose to fuse the CSI-based ranges and velocity estimated from inertial sensors by an enhanced particle filter to achieve highly accurate tracking. The algorithm relies on some enhanced ranging methods and further mitigates the remaining ranging errors by a weighting technique. Additionally, we provide an efficient method to estimate the velocity based on inertial sensors. The algorithms are designed in a network-based system, which uses rather cheap commercial devices as anchor nodes. We evaluate our system in a complex environment along three different moving paths. Our proposed tracking method can achieve 1.3m for mean accuracy and 2.2m for 90% accuracy, which is more accurate and stable than pedestrian dead reckoning and range-based positioning.

The rapid growth of ubiquitous applications and location-based services has made indoor navigation an interesting topic. Some indoor localization solutions exploit radio information and Inertial Measurement Units (IMUs), which are embedded in most of the modern smartphones. In this paper, we present a real-time indoor localization approach that fuses WiFi Receiving Signal Strength Indicator (RSSI) readings, IMUs, and floor plan information in an enhanced particle filter. The algorithms are designed and implemented into a terminal-based system, which uses commercial smartphones and WiFi access points. Extensive real-world experiment results show that our tracking method can achieve the average tracking error of 1.7 meters and 90% accuracy of 3.2 meters.

The rapid growth area of ubiquitous applications and location-based services has made indoor localization an interesting topic for research. Some indoor localization solutions for smartphones exploit radio information and Inertial Measurement Units (IMUs), which are embedded in most of the modern smartphones. In this work, we propose to fuse WiFi Receiving Signal Strength Indicator (RSSI) readings, IMUs, and floor plan information in an enhanced particle filter to achieve high accuracy and stable performance in the tracking process. We provide an efficient double resampling method to mitigate errors caused by off-the-shelf IMUs and WiFi sensors embedded in commodity smartphones. The algorithms are designed in a terminal-based system, which consists of commercial smartphones and WiFi access points. We evaluate our system in two complex environments along moving paths. Experiment results show that our tracking method can achieve the average tracking error of $1.01$ meters and $90\%$ accuracy of $1.7$ meters.

Time-based indoor localization has been investigated for several years but the accuracy of existing solutions is limited by several factors, e.g., imperfect synchronization, signal bandwidth and indoor environment. In this paper, we compare two time-based localization algorithms for narrow-band signals, i.e., multilateration and fingerprinting. First, we develop a new Linear Least Square (LLS) algorithm for Differential Time Difference Of Arrival (DTDOA). Second, fingerprinting is among the most successful approaches used for indoor localization and typically relies on the collection of measurements on signal strength over the area of interest. We propose an alternative by constructing fingerprints of fine-grained time information of the radio signal. We offer comprehensive analytical discussions on the feasibility of the approaches, which are backed up by evaluations in a software defined radio based IEEE 802.15.4 testbed. Our work contributes to research on localization with narrow-band signals. The results show that our proposed DTDOA-based LLS algorithm obviously improves the localization accuracy compared to traditional TDOA-based LLS algorithm but the accuracy is still limited because of the complex indoor environment. Furthermore, we show that time-based fingerprinting is a promising alternative to power-based fingerprinting.

Clock synchronization in the order of nanoseconds is one of the critical factors for time-based localization. Currently used time synchronization methods are developed for the more relaxed needs of network operation. Their usability for positioning should be carefully evaluated. In this paper, we are particularly interested in GPS-based time synchronization. To judge its usability for localization we need a method that can evaluate the achieved time synchronization with nanosecond accuracy. Our method to evaluate the synchronization accuracy is inspired by signal processing algorithms and relies on fine-grain time information. The method is able to calculate the clock offset and skew between devices with nanosecond accuracy in real time. It was implemented using software defined radio technology. We demonstrate that GPS-based synchronization suffers from remaining clock offset in the range of a few hundred of nanoseconds but the clock skew is negligible. Finally, we determine a corresponding lower bound on the expected positioning error.

Indoor localization systems become more interesting for researchers because of the attractiveness of business cases in various application fields. A WiFi-based passive localization system can provide user location information to third-party providers of positioning services. However, indoor localization techniques are prone to multipath and Non-Line Of Sight (NLOS) propagation, which lead to significant performance degradation. To overcome these problems, we provide a passive localization system for WiFi targets with several improved algorithms for localization. Through Software Defined Radio (SDR) techniques, we extract Channel Impulse Response (CIR) information at the physical layer. CIR is later adopted to mitigate the multipath fading problem. We propose to use a Nonlinear Regression (NLR) method to relate the filtered power information to propagation distances, which significantly improves the ranging accuracy compared to the commonly used log-distance path loss model. To mitigate the influence of ranging errors, a new trilateration algorithm is designed as well by combining Weighted Centroid and Constrained Weighted Least Square (WC-CWLS) algorithms. Experiment results show that our algorithm is robust against ranging errors and outperforms the linear least square algorithm and weighted centroid algorithm.