Wireless networks like many multi-user services have to balance limited resources in real-time. In 6G, increased network automation makes consumer trust crucial. Trust is reflect in both a personal emotional sentiment as well as a physical understanding of the transparency of AI decision making. Whilst there has been isolated studies of consumer sentiment to wireless services, this is not well linked to the decision making engineering. Likewise, limited recent research in explainable AI (XAI) has not established a link to consumer perception.Here, we develop a Quality-of-Trust (QoT) KPI that balances personal perception with the quality of decision explanation. That is to say, the QoT varies with both the time-varying sentiment of the consumer as well as the accuracy of XAI outcomes. We demonstrate this idea with an example in Neural Water-Filling (N-WF) power allocation, where the channel capacity is perceived by artificial consumers that communicate through Large Language Model (LLM) generated text feedback. Natural Language Processing (NLP) analysis of emotional feedback is combined with a physical understanding of N-WF decisions via meta-symbolic XAI. Combined they form the basis for QoT. Our results show that whilst the XAI interface can explain up to 98.9% of the neural network decisions, a small proportion of explanations can have large errors causing drops in QoT. These drops have immediate transient effects in the physical mistrust, but emotional perception of consumers are more persistent. As such, QoT tends to combine both instant physical mistrust and long-term emotional trends.
I. Abstract: Despite recent advances in sensor and mobile technology, there is still a lack of an accurate, scalable, and non-intrusive way to knowing how much sunlight we are exposed to. For the first time, we devise a mobile phone software application (SUN BATH), that utilizes a variety of on-board sensors and data sets to accurately predict the sunlight exposure each person is exposed to. The algorithm is able to take into account the sunlight exposure based on the person location, the local weather, sun location, and shadow from buildings. The algorithm achieves this by using the mobile user's location and other sensors to determine whether it is indoors or outdoors, and uses building data to calculate shadow effects and weather data to calculate diffused light contributions. This will ultimately allow the user to be more informed about sunlight exposure and compare it with daily recommended levels to encourage positive behaviour change. In order to show the value added by the application, SUN BATH is distributed to a sample of students population for benchmarking and user experience trials. The latest stable version of the application, suggests a scalable and affordable way compared to survey or physical sensing methods. II. Introduction: In this particular proposal, we examine how to live healthily in cities using a data-driven mobile-sensing approach. Cities are partly defined by a high building concentration and a human lifestyle that is predominantly spent indoors or in the shadow of buildings. Some cities in particular, also suffer from heavy pollution effects that significantly reduces the level of direct solar radiation. As a result, one area of concern is the urban dwellers lack of exposure to the ultra-violet (UV) band of sunlight and the wide range of associated health problems. The large scale and chronic nature of the health problems can lead to a time bomb in the National Health Service and cause irreversible future damage to the economy. This article proposes using the ray tracing SORAM model by Erdelyi et al. as an innovative and flexible technique for modelling and estimating the amount of solar irradiation can be collected at a time and certain location. SORAM module is already benchmarked against real measurement data, hence, our work here will benefit from this by taking the calculated ray-tracing information as a primary filter. The aim is to devise an affordable and accurate way of continuously estimating each person's UV exposure. Primarily, this is achieved by developing an Android smartphone application that uses the SORAM advanced modelling techniques to estimate the level of UV exposure each person is subjected to at any given time and location. The research novelty is that the proposed solution does not require additional purpose-built hardware such as a photovoltaic sensor, but instead utilizes a combination of accurate wireless localization, and weather-/terrain-informed sunlight propagation mapping. The challenges addressed include how to accurately locate a human and how to model the propagation of sunlight in complex urban environments. The latest stable version of the application, suggests a novel and affordable way compared to traditional or physical methods when calculating the amount of sunshine we are exposed to. III. System Overview: We implemented and evaluated SUN BATH application with the Android platform using different mobile phone models such as Samsung S5, Asus Zen5 and Archos tablet. The application is developed using Android Studio as IDE for Android application development. The application allows the user to create a profile using a user name and some information such as date of birth, height, weight, skin colour, country of origin, and level of income. To be used later for future detailed reporting with relation to the amount of sun bathing for different groups and ethnicities. SUN BATH only relies on lightweight “sensors to server” modelling which allows continuous low-energy and low-cost tracking of the user location and state transitions. In particular, we will present the process to show that we were able to use SORAM within the smart-phone environment to accurately infer the amount of sunshine in a user is exposed to based on the accuracy of the GPS and other location modules for Android mobile phones. To meet stringent design requirements, SUN BATH utilizes a series of lightweight ‘sensors – server’ for a fault-tolerant location detection. SUN BATH primarily makes use of three types of location-aware detectors: WiFi, cellular-network, and GPS. The aforementioned three wireless location detectors are used in conjunction to improve resolution and resilience. WiFi hub SSID identifiers are used to locate the hub in known open and commercial databases up to an accuracy of a few metres. In the absence of WiFi, a combination of cell tower location area and assisted GPS is used to get an accuracy of 10–15 m in urban areas with shadow effects. WiFi detector adopts the distributed IP address to capture the source location to determine the region the user is in. Cellular-network detector detects the source and attenuation of signals caused by objects on its path (e.g., trees, buildings). It normally help to indicate the movement of the user as the mobile signal gets handed from one network to another. The Application utilizes the GPS sensor to exactly pinpoint the coordinates of the user location i.e. Latitude and Longitude. The system time clock is also used to assist the detection of the local time. The App cache-in those parameters and sends it to a remote server whenever there is an Internet connection. The server hosts the SORAM calculation algorithm which generates a live estimate of the amount of sun exposure the user is experiencing. The results are then passed back to the applications through the Open Database Connectivity (ODBC) middleware service to permanently store the results in a secure database management system (DBMS). IV. Soram Ray-tracing Methodology: A person positioned in an out-door environment is surrounded by solar radiation, which consists of direct and diffuse rays. Direct and diffuse radiation data on a horizontal surface are usually collected at various locations and weather stations across the universe. The raw datasets collected can be used to estimate the amount of global radiation at any point of earth of a given slope and azimuth. Due to cost and scarcity of live data, the SORAM algorithm embed and made use of the Reindel Model, to estimate the direct and diffuse irradiation from hourly horizontal global radiation data. In addition and to go light on computation and avoid calculations for the nighttime hours, the SORAM determines the sunrise time for each day of the year and the amount of solar radiation data from that point onwards which is then calculated until sunset. The algorithm also estimates and with high accuracy direct and diffuse radiation on a surface of given slope and azimuth from their counterparts on a horizontal access considering surrounding shading conditions. We tested SUN BATH in simulated and real locations for five continuous days from sunrise to sunset in around the School of Engineering building complex at the University of Warwick campus. Simulated tests were carried manually, using two fixed location parameters i.e Longitude and Latitude. A quick memory and CPU monitor view revealed that the SUN BATH energy consumptions and resource-constraint on the used smartphone devices were moderate. A full memory and CPU monitor view can be easily produced, but it is beyond the scope of this article. V. Conclusions and Future Work: This research presented the architecture of SUN BATH mobile sensing application that gathers a variety of lightweight sensors information and utilised ray-tracing algorithm to derive the level of human sun exposure in urban areas. The application has demonstrated that it can be an affordable and pervasive way of accurately measuring the level of sunlight exposure each person is exposed to. Further work is required to scale the project to the global level, which requires big data sets on urban building maps and meterological data from all the cities.
Drones are set to penetrate society across transport and smart living sectors. While many are amateur drones that pose no malicious intentions, some may carry deadly capability. It is crucial to infer the drone's objective to prevent risk and guarantee safety. In this article, a policy error inverse reinforcement learning (PEIRL) algorithm is proposed to uncover the hidden objective of drones from online data trajectories obtained from cooperative sensors. A set of error-based polynomial features are used to approximate both the value and policy functions. This set of features is consistent with current onboard storage memories in flight controllers. The real objective function is inferred using an objective constraint and an integral inverse reinforcement learning (IRL) batch least-squares (LS) rule. The convergence of the proposed method is assessed using Lyapunov recursions. Simulation studies using a quadcopter model are provided to demonstrate the benefits of the proposed approach.
In this paper, we consider a molecular diffusion based communications link that can reliably transport data over-the-air. We show that the system can also reliably transport data across confined structural environments, especially in cases where conventional electromagnetic (EM) wave based systems may fail. In particular, this paper compares the performance of our proprietary molecular communication test-bed with Zigbee wireless sensors in a metal pipe network that does not act as a radio wave-guide. The paper first shows that a molecular-based communication link's performance is determined primarily by the delay time spread of the pulse response. The paper go on to show that molecular-based systems can transmit more reliably in complex and confined structural environments than conventional EM-based systems. The paper then utilizes empirical data to find relationships between the received radio signal strength, the molecular pulse spread, data rate (0.1 bits/s) and the structural propagation environment.
The development of models to capture large-scale dynamics in human history is one of the core contributions of cliodynamics. Most often, these models are assessed by their predictive capability on some macro-scale and aggregated measure and compared to manually curated historical data. In this report, we consider the model from Turchin et al. (2013), where the evaluation is done on the prediction of “imperial density”: the relative frequency with which a geographical area belonged to large-scale polities over a certain time window. We implement the model and release both code and data for reproducibility. We then assess its behavior against three historical datasets: the relative size of simulated polities versus historical ones; the spatial correlation of simulated imperial density with historical population density; and the spatial correlation of simulated conflict versus historical conflict. At the global level, we show good agreement with population density (R2<0.75), and some agreement with historical conflict in Europe (R2<0.42). The model instead fails to reproduce the historical shape of individual polities. Finally, we tweak the model to behave greedily by having polities preferentially attacking weaker neighbors. Results significantly degrade, suggesting that random attacks are a key trait of the original model. We conclude by proposing a way forward by matching the probabilistic imperial strength from simulations to inferred networked communities from real settlement data. Page numbers for this article were updated on 01/05/2021.
Trajectory inference is a hard problem when states measurements are noisy and if there is no high-fidelity model available for estimation; this may arise into high-variance and biased estimates results. This article proposes a physics informed trajectory inference of a class of nonlinear systems. The approach combines the advantages of state and parameter estimation algorithms to infer the trajectory that follows the nonlinear system using online noisy state measurements. The algorithm is composed of a parallel estimated model constructed in terms of a low-pass filter parameterization. The estimated model defines a physics informed model that infers the trajectory of the real nonlinear system with noise attenuation capabilities. The parameters of the estimated model are updated by a closed-loop output error identification algorithm which uses the estimated states instead of the noisy measurements to avoid biased estimation. Stability and convergence of the proposed technique is assessed using Lyapunov stability theory. Simulations studies are carried out under different scenarios to verify the effectiveness of the proposed inference algorithm.
Internet-of-Things (IoT) devices are often used to transmit physical sensor data over digital wireless channels. Traditional Physical Layer Security (PLS)-based cryptography approaches rely on accurate channel estimation and information exchange for key generation, which irrevocably ties key quality with digital channel estimation quality. Recently, we proposed a new concept called Graph Layer Security (GLS), where digital keys are derived from physical sensor readings. The sensor readings between legitimate users are correlated through a common background infrastructure environment (e.g., a common water distribution network or electric grid). The challenge for GLS has been how to achieve distributed key generation. This paper presents a Federated multi-agent Deep reinforcement learning-assisted Distributed Key generation scheme (FD2K), which fully exploits the common features of physical dynamics to establish secret key between legitimate users. We present for the first time initial experimental results of GLS with federated learning, achieving considerable security performance in terms of key agreement rate (KAR), and key randomness.
View Video Presentation: https://doi.org/10.2514/6.2023-0193.vid Uncertainty-based sensor management for positioning is an essential component in safe drone operations inside urban environments with large urban valleys. These canyons significantly restrict the Line-Of-Sight signal conditions required for accurate positioning using Global Navigation Satellite Systems (GNSS). Therefore, sensor fusion solutions need to be in place which can take advantage of alternative Positioning, Navigation and Timing (PNT) sensors such as accelerometers or gyroscopes to complement GNSS information. Recent state-of-art research has focused on Machine Learning (ML) techniques such as Support Vector Machines (SVM) that utilize statistical learning to provide an output for a given input. However, understanding the uncertainty of these predictions made by Deep Learning (DL) models can help improve integrity of fusion systems. Therefore, there is a need for a DL model that can also provide uncertainty-related information as the output. This paper proposes a Bayesian-LSTM Neural Network (BLSTMNN) that is used to fuse GNSS and Inertial Measurement Unit (IMU) data. Furthermore, Protection Level (PL) is estimated based on the uncertainty distribution given by the system. To test the algorithm, Hardware-In-the-Loop (HIL) simulation has been performed, utilizing Spirent's GSS7000 simulator and OKTAL-SE Sim3D to simulate GNSS propagation in urban canyons. SimSENSOR is used to simulate the accelerometer and gyroscope. Results show that Bayesian-LSTM provides the best fusion performance compared to GNSS alone, and GNSS/IMU fusion using EKF and SVM. Furthermore, regarding uncertainty estimates, the proposed algorithm can estimate the positioning boundaries correctly, with an error rate of 0.4% and with an accuracy of 99.6%.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
