1,721,010 research outputs found
Wireless Localization Systems: Statistical Modeling and Algorithm Design
No full textWireless localization systems are essential for emerging applications that rely on
context-awareness, especially in civil, logistic, and security sectors. Accurate localization in indoor environments is still a challenge and triggers a fervent research
activity worldwide. The performance of such systems relies on the quality of range
measurements gathered by processing wireless signals within the sensors composing
the localization system. Such range estimates serve as observations for the target
position inference. The quality of range estimates depends on the network intrinsic
properties and signal processing techniques. Therefore, the system design and analysis call for the statistical modeling of range information and the algorithm design
for ranging, localization and tracking. The main objectives of this thesis are: (i) the
derivation of statistical models and (ii) the design of algorithms for different wire-
less localization systems, with particular regard to passive and semi-passive systems
(i.e., active radar systems, passive radar systems, and radio frequency identification
systems). Statistical models for the range information are derived, low-complexity
algorithms with soft-decision and hard-decision are proposed, and several wideband
localization systems have been analyzed. The research activity has been conducted
also within the framework of different projects in collaboration with companies and
other universities, and within a one-year-long research period at Massachusetts Institute of Technology, Cambridge, MA, USA. The analysis of system performance,
the derived models, and the proposed algorithms are validated considering different case studies in realistic scenarios and also using the results obtained under the
aforementioned projects
Innesto di tessuto cerebrale bioingegnerizzato e substrati flessibili: primi passi verso una medicina rigenerativa potenziata nell'epilessia del lobo temporale
No full textL'epilessia è una malattia cronica progressiva e potenzialmente fatale caratterizzata da crisi spontanee e ricorrenti causate da un’attività cerebrale ipersincrona. L'epilessia del lobo temporale (TLE) è il tipo più comune di epilessia focale, tipica dell’adulto, spesso farmaco-resistente.
L’epilessia del lobo temporale mesiale (MTLE) è la forma più grave di TLE ed è caratterizzata da sclerosi ippocampale, in particolare nella zona del CA3 ventrale (CA3V). Non tutti i pazienti farmaco-resistenti sono candidabili alla resezione chirurgica del fuoco epilettico a causa di foci multipli o all’insorgenza del fuoco in una regione eloquente del cervello.
Recentemente, sono state sviluppate strategie terapeutiche alternative quali la terapia con cellule staminali e la stimolazione cerebrale profonda, ma entrambi questi interventi presentano limitazioni e non portano a una risoluzione delle crisi. Il progetto HERMES "Hybrid enhanced regenerative medicine systems" si propone di guarire la TLE attraverso il paradigma della medicina rigenerativa potenziata, basata sull'integrazione simbiotica di 3 componenti: tessuto cerebrale bioingegnerizzato costituito da cellule staminali neurali (NSCs) e matrice extracellulare (ECM) per ricostruire la materia cerebrale, una componente neuromorfica (NCS) per emulare e integrare la funzione cerebrale, e l'intelligenza artificiale (AI) come coordinatore super partes.
La mia ricerca di dottorato fa parte del progetto HERMES finanziato dall'UE per 6 anni. In particolare, gli obiettivi del progetto di dottorato sono: I) il trapianto di tessuto bioingegnerizzato in CA3V; II) l’inserimento di un dispositivo neuromorfico flessibile custom-made.
Gli esperimenti in vivo sono stati eseguiti sul modello pilocarpina di TLE indotta in ratti maschi adulti Sprague-Dawley con un’iniezione sistemica di pilocarpina, un agonista muscarinico.
Per limitare il numero di chirurgie e ridurre il dolore e la sofferenza nei ratti, sono state eseguite iniezioni multiple mediante cannule guida. Nella prima parte del progetto abbiamo messo a punto una chirurgia mininvasiva al fine di impiantare cannule guida che consentissero iniezioni in strutture profonde. A questo scopo, abbiamo testato due strategie e identificato quella che consente una precisa iniezione di NSCs nella posizione desiderata con danni cerebrali minimi.
Per evitare l'aumento della pressione intracranica causata dal trapianto, ai ratti epilettici è stato iniettato un agente citotossico, acido ibotenico, la cui azione impedirebbe inoltre alle cellule epilettiche native di indurre le NSCs verso un fenotipo epilettico.
I ratti sono stati trapiantati con NSCs-GFP positive, in combinazione con alginato marcato, un polisaccaride vegetale che mima la ECM sostenendo crescita e differenziamento cellulare. L'immunofluorescenza eseguita a diversi tempi su sezioni cerebrali ha mostrato la corretta posizione delle NSCs, la loro sopravvivenza e maturazione. Inoltre, abbiamo valutato in vivo l'efficacia antinfiammatoria dell'alginato co-iniettato con NSCs. Parallelamente, abbiamo lavorato sulla metodica di impianto in vivo della componente neuromorfa flessibile, realizzata con materiali biocompatibili. Abbiamo testato due approcci di inserimento della NCS flessibile in strutture cerebrali profonde e abbiamo trovato la migliore strategia che ci consente di impiantarla con successo nella corretta posizione. I risultati del mio progetto di dottorato serviranno per la seconda parte del progetto HERMES che prevede l'integrazione delle parti biologiche e di quelle neuromorfe con l'IA. Abbiamo dimostrato che sono stati compiuti i primi passi verso la medicina rigenerativa potenziata, sostenendo la fattibilità del raggiungimento dell’obiettivo del progetto HERMES di guarire la TLE.Epilepsy is a progressive and potentially fatal chronic disease characterized by spontaneous and recurrent seizures due to hypersynchronous brain activity. Temporal lobe epilepsy (TLE) is the most common type of adulthood drug-resistant focal epilepsy in which seizures may arise from one or more limbic areas. Mesial TLE (MTLE) is the most severe form of TLE and is characterized by hippocampal sclerosis, particularly in ventral CA3 (vCA3) area. Not all drug-resistant patients can undergo surgical resection of the epileptic focus due to multiple foci or focus in an eloquent brain region. Recently, alternative therapeutic strategies have been developed, and stem cell therapy and deep brain stimulation have been widely investigated, but both these interventions presented limitations and do not bring to a seizure-free condition. To overcome these limitations, “Hybrid enhanced regenerative medicine systems - HERMES” project proposes to heal TLE through the paradigm of enhanced regenerative medicine, based on the symbiotic integration of 3 components: bioengineered brain tissue made of neural stem cells (NSCs) and extracellular matrix (ECM) to rebuild the brain matter, neuromorphic computing system (NCS) to emulate and integrate brain function, and the artificial intelligence (AI) as a super partes coordinator.
My PhD research is part of the 6-year EU funded project HERMES. In particular, my PhD project goals were: I) the transplantation of a bioengineered tissue graft into the vCA3; II) the insertion of a custom-made flexible neuromorphic device.
In vivo experiments were performed on the pilocarpine model of TLE induced in adult male Sprague-Dawley rats with the systemic injection of pilocarpine, a muscarinic agonist. In order to limit the number of massive surgeries and reduce pain and discomfort in the rats, multiple injections were performed through guide cannulas. In the first part of the project, we focused to optimize a minimally invasive surgical strategy to implant guide cannula for injections in deep structures. To this purpose, we tested two different strategies and identified the one that allowed a precise injection of NSC in the desired location with minimal brain damage.
To avoid the increase of intracranial pressure provided by the graft, epileptic rats were injected with the cytotoxic agent ibotenic acid. The lesion induced by ibotenic acid would also prevent to have a graft contaminated with epileptic cells which could train injected stem cell to turn into a pathological phenotype. Rats were then transplanted with green fluorescent protein (GFP)-marked NSCs in combination with marked alginate, a vegetal polysaccharide mimicking native ECM to support cell growth and differentiation. Immunofluorescence on brain slices confirmed the correct location of injected NSCs, their survival and ongoing maturation at several time points after injection. In addition, we evaluated for the first time the in vivo anti-inflammatory efficacy of alginate when co-injected with NSC.
In parallel, we worked on the in vivo implantation of the neuromorphic component. Flexible probes were made of biocompatible materials. We tested two different approaches to allow the insertion into the brain of flexible substrates and found the best strategy that allow successful implant of the neuromorphic probes in the correct location.
The results of my PhD project will serve for the second part of the HERMES project which consists in the integration of the biological and neuromorphic parts with AI. We demonstrated that the first steps toward enhanced regenerative medicine have been accomplished supporting the feasibility to achieve the goal of the HERMES project to heal TLE
Passive network localization via UWB wireless sensor radars: The impact of TOA estimation
No full textPassive network localization via wireless sensor radars (WSRs) is rising interest for applications in civilian, commercial, and military sectors. A clear understanding on how impairments and system parameters affect detection and localization accuracy is fundamental for the WSR network design. We introduce a mathematical framework for the analysis of WSR network which allows to consider different scenarios, network topologies, and system configurations. We focus on the effects of ultra-wide bandwidth (UWB) time-of-arrival estimation on the localization accuracy. We present results for a case study of UWB monostatic radars network with application in logistics
Performance analysis of IEEE 802.11p preamble insertion in C-V2X sidelink signals for co-channel coexistence
Full text linkSpectrum scarcity is one of the main challenges of future wireless technologies. When looking at vehicle-to -everything (V2X), this is amplified as spectrum sharing could impact road safety and traffic efficiency. It is therefore of particular importance to study solutions that allow the coexistence, in the same geographical area and in the same channels, of what are today the main V2X access technologies, namely IEEE 802.11p and long term evolution (LTE)-V2X sidelink Mode 4. In this paper, in addition to studying the impact of mutual interference, which is found to have a strong impact especially on the former and under congested channel conditions, a mitigation solution is extensively studied. The solution is based on the insertion of the IEEE 802.11p preamble at the beginning of each LTE-V2X sidelink transmission. The proposal, which is also under discussion within the standardization bodies, requires no changes to the IEEE 802.11p protocol stack and minor changes to LTE-V2X sidelink. This solution is directly applicable to upcoming IEEE 802.11bd and extendable to new radio (NR)-V2X sidelink. The paper shows, through analysis and simulations in free-flow and dense scenarios, that the proposal enables mitigation of collisions caused by co-channel coexistence under low and high load conditions. The improvement is guaranteed even in cases of congestion when combined with additional countermeasures. Regarding the latter aspect, in particular, different approaches are compared, demonstrating that acting on the congestion control mechanisms is a simple but effective solution
UWB passive navigation in indoor environments
No full textLocalization and navigation of passive target objects play a key role in many important applications. An interesting solution for passive localization and navigation is given by monostatic wireless sensor radar (WSR) networks. In this context, ultrawide band (UWB) radar provide fine delay resolution enabling high accuracy localization also in harsh environments such as indoor. We present a mathematical framework for analysis and design of passive navigation based on UWB monostatic WSRs that relies on environment propagation and time-of-arrival estimation characterized by network experiments. A case study where a UWB monostatic WSR network is deployed to infer the position of moving target objects is considered. In particular, Bayesian navigation based on particle filters implementation is analyzed and the role of mobility model for inferring target position is shown
Passive radar via LTE signals of opportunity
No full textPassive radars relying on signals of opportunity enable new applications based on stealth tracking of targets without the need of radar signals emissions. Long term evolution (LTE) base stations employing orthogonal frequency division multiplexing (OFDM) signals are excellent candidates as illuminators of opportunity thanks to their wide availability. The tracking accuracy of such passive radars depends on prior knowledge (e.g., the wireless environment) and signal processing (e.g., clutter mitigation and tracking algorithm). This paper proposes passive radar systems exploiting LTE base stations as illuminators of opportunity to detect and track moving targets in a monitored environment. We analyze such systems based on a Bayesian framework for detection of moving targets and estimation of their position and velocity. A case study accounting for the LTE extended pedestrian model is presented, with various settings in terms of network configuration, wireless propagation, and signal processing
Wideband localization via range likelihood based on reduced dataset
No full textWideband localization commonly relies on accurate range information extracted from received waveforms, which can be obtained via hard-decision or soft-decision ranging. While hard-decision ranging based on energy samples has received much attention because of its low-complexity, soft-decision ranging based on waveform samples can significantly improve the localization accuracy at the cost of higher complexity. This paper proposes new soft-decision ranging techniques with low complexity for wideband localization. The proposed techniques adopt range likelihood functions obtained from a reduced dataset of the received waveform samples. Results show that the proposed soft-decision techniques enable localization with higher accuracy compared to hard-decision ranging
Energy-based order of arrival estimation via UWB-UHF RFID
No full textThe order-of-arrival (OOA) estimation of moving items is essential for various emerging applications, for example, baggage sorting and dispatching in airports. In this paper, we show how OOA estimation can be performed by an UWB-UHF RFID system that allows identification and tracking of passive and semi-passive tags attached to objects moving on a conveyor belt. Particle filtering on the energy samples of the backscattered signal is introduced for OOA estimation. A case study shows the performance of the proposed particle filter for various settings of practical interest
Localization-assisted indoor acoustical data modeling
No full textMonitoring of acoustical data is often done at fixed measurement stations to verify the compliance with environmental noise or work safety regulations. In this work, a different method is explored for monitoring acoustical conditions in indoor environments with the assistance of location information. Such a method exploits a wideband localization system to determine the listener position with high accuracy, which leads, together with prior knowledge of the sound field, to assess the behaviour of the local acoustical data in the indoor environment. An application of this method is developed, considering as case study a long partitioned virtual room, where an ultra- wideband (UWB) localization system is used and the subject is moving along a trajectory characterized by abrupt acoustical changes. First, the behaviour of various acoustical metrics is assessed along the trajectory. Then, the perspectives of this monitoring method are outlined and discussed
Towards counting via passive radar using OFDM waveforms
No full textThe capability of counting targets (people or things) in a monitored area is important for emerging wireless applications. To this aim, passive systems that rely on signals of opportunity and device-free targets are preferred to active systems that rely on dedicated or personal devices for preserving privacy and reducing implementation costs. This paper develops a framework for design and analysis of device-free counting systems via OFDM signals of opportunity. In particular, counting techniques based on model order selection are proposed. Preliminary results show the effectiveness of the proposed techniques in simple use cases
- …
