The concepts of a digital twin and extended reality have recently emerged, which require a massive amount of sensor data to be transmitted with low latency and high reliability. For low-latency communications, joint source-channel coding (JSCC) is an attractive method for error correction coding and compared to highly complex digital systems that are currently in use. I propose the use of complex-valued and quaternionic neural networks (QNN) to decode JSCC codes, where the complex-valued neural networks show a significant improvement over real-valued networks and the QNNs have an exceptionally high performance. Furthermore, I propose mapping encoded JSCC code words to the baseband of the frequency domain in order to enable time/frequency synchronization as well as to mitigate fading using robust estimation theory. Additionally, I perform robust statistical signal processing on the high-dimensional JSCC code showing significant noise immunity with drastic performance improvements at low signal-to-noise ratio (SNR) levels. The performance of the proposed JSCC codes is within 5 dB of the optimal performance theoretically achievable and outperforms the maximum likelihood decoder at low SNR while exhibiting the smallest possible latency. I designed a Bayesian minimum mean square error estimator for decoding high-dimensional JSCC codes achieving 99.96% accuracy. With the recent introduction of electromagnetic reconfigurable intelligent surfaces (RIS), a paradigm shift is currently taking place in the world of wireless communications. These new technologies have enabled the inclusion of the wireless channel as part of the optimization process. In order to decode polarization-space modulated RIS reflections, robust polarization state decoders are proposed using the Weiszfeld algorithm and an generalized Huber M-estimator. Additionally, QNNs are trained and evaluated for the recovery of the polarization state. Furthermore, I propose a novel 64-ary signal constellation based on scaled and shifted Eisenstein integers and generated using media-based modulation with a RIS. The waveform is received using an antenna array and decoded with complex-valued convolutional neural networks. I employ the circular cross-correlation function and a-priori knowledge of the phase angle distribution of the constellation to blindly resolve phase offsets between the transmitter and the receiver without the need for pilots or reference signals. Furthermore, the channel attenuation is determined using statistical methods exploiting that the constellation has a particular distribution of magnitudes. After resolving the phase and magnitude ambiguities, the noise power of the channel can also be estimated. Finally, I tune an Sq-estimator to robustly decode the Eisenstein waveform. / Doctor of Philosophy / This dissertation covers three novel wireless communications methods; analog coding, communications using the electromagnetic polarization and communications with a novel signal constellation. The concepts of a digital twin and extended reality have recently emerged, which require a massive amount of sensor data to be transmitted with low latency and high reliability. Contemporary digital communication systems are highly complex with high reliability at the expense of high latency. In order to reduce the complexity and hence latency, I propose to use an analog coding scheme that directly maps the sensor data to the wireless channel. Furthermore, I propose the use of neural networks for decoding at the receiver, hence using the name neural receiver. I employ various data types in the neural receivers hence leveraging the mathematical structure of the data in order to achieve exceptionally high performance. Another key contribution here is the mapping of the analog codes to the frequency domain enabling time and frequency synchronization. I also utilize robust estimation theory to significantly improve the performance and reliability of the coding scheme. With the recent introduction of electromagnetic reconfigurable intelligent surfaces (RIS), a paradigm shift is currently taking place in the world of wireless communications. These new technologies have enabled the inclusion of the wireless channel as part of the optimization process. Therefore, I propose to use the polarization state of the electromagnetic wave to convey information over the channel, where the polarization is determined using a RIS. As with the analog codes, I also extensively employ various methods of robust estimation to improve the performance of the recovery of the polarization at the receiver. Finally, I propose a novel communications signal constellation generated by a RIS that allows for equal probability of error at the receiver. Traditional communication systems utilize reference symbols for synchronization. In this work, I utilize statistical methods and the known distributions of the properties of the transmitted signal to synchronize without reference symbols. This is referred to as blind channel estimation. The reliability of the third communications method is enhanced using a state-of-the-art robust estimation method.

Στις μέρες μας, οι αυξανόμενες απαιτήσεις για υπηρεσίες υψηλού φασματικού εύρους ζώνης επιβάλλουν την ανάπτυξη νέων τεχνολογιών στο σχεδιασμό των δικτύων νέας γενιάς, ικανές να προσφέρουν α) χαμηλό κόστος κατά το σχεδιασμό του συστήματος, β) μεγάλη απόσταση μετάδοσης, γ) πολλοί χρήστες και δ) υψηλό εύρος ζώνης στην πλευρά του χρήστη για την παροχή των νέων υπηρεσιών. Ωστόσο, τα οπτικά δίκτυα λόγω των αναλογικών χαρακτηριστικών των οπτικών σημάτων τους, υποφέρουν από γραμμικές και μη γραμμικές παραμορφώσεις. Αυτές οι παραμορφώσεις επηρεάζουν άμεσα την απόδοση των συστημάτων και η επίδραση τους αυξάνει με την αύξηση του ρυθμού μετάδοσης. Παραδοσιακά χρησιμοποιούνται οπτικοί τρόποι για την εξομάλυνση των παραμορφώσεων. Ωστόσο, η ραγδαία ανάπτυξη στον τομέα των ηλεκτρονικών αναδεικνύει την ηλεκτρονική εξομάλυνση των παραμορφώσεων ως μία ευέλικτη, χαμηλού κόστους ολοκληρωμένη και βιώσιμη λύση που αποφεύγει τις επιπρόσθετες οπτικές απώλειες. Σκοπός της διδακτορικής διατριβής είναι η εξομάλυνση με αποδοτικό τρόπο των πιο σημαντικών παραμορφώσεων (χρωματική διασπορά, αυτοδιαμόρφωση φάσης και φαινόμενο αλληλουχίας φίλτρων) που δημιουργούνται στα οπτικά δίκτυα και ειδικότερα στα μητροπολιτικά δίκτυα, στα δίκτυα πρόσβασης, και στα παθητικά δίκτυα. Από σχεδιαστικής πλευράς του συστήματος, αυτή η διατριβή προτείνει τη βέλτιστη χρησιμοποίηση λύσεων χαμηλού κόστους, ικανές να επεκτείνουν (σε ρυθμό μετάδοσης και απόσταση) την χρησιμοποίησή τους σε οπτικά δίκτυα νέας γενιάς. Πιο συγκεκριμένα, η απόδοση της ηλεκτρονικής αντιστάθμισης μελετάται για συστήματα που χρησιμοποιούν χαμηλού κόστους, συμβατικούς πομπούς laser άμεσης διαμόρφωσης (DML), που οδηγούνται στα 10 Gb/s. Σκοπός σε αυτήν την περίπτωση είναι η αύξηση της απόστασης και του ρυθμού μετάδοσης που μπορεί να επιτευχθεί, εξομαλύνοντας τις παραμορφώσεις που δημιουργούνται εξαιτίας των χαρακτηριστικών των πομπών και αυτών που δημιουργούνται κατά τη μετάδοση του σήματος (χρωματική διασπορά, αυτοδιαμόρφωση φάσης και φαινόμενο αλληλουχίας φίλτρων) με την βέλτιστη χρησιμοποίηση ηλεκτρονικού εξισωτή. Επιπλέον, όσον αφορά τα παθητικά δίκτυα πρόσβασης νέας γενιάς, μελετάται μία αποδοτική και χρήσιμη τεχνική, χρησιμοποιώντας τα πλεονεκτήματα της χρήσης του ηλεκτρονικού εξισωτή στην πλευρά του δέκτη (OLT). Η πειραματική μελέτη εστιάζει στα παθητικά οπτικά δίκτυα (PON) στα 10 Gb/s χρησιμοποιώντας χαμηλού κόστους, χαμηλού εύρους ζώνης RSOA στην πλευρά του χρήστη (ONU), και ηλεκτρονικό εξισωτή στην πλευρά του δέκτη (OLT). Αυτή η τεχνική προσφέρει την απαιτούμενη ευελιξία για την προσαρμογή στις καινούργιες συνθήκες του συστήματος και την υλοποίηση των απαιτήσεων (πολύ μεγάλες αποστάσεις μετάδοσης, αριθμό χρηστών και ρυθμό μετάδοσης), ενώ ταυτόχρονα μπορεί και εκπληρώνει τις απαιτήσεις χαμηλού κόστους στην ανάπτυξη των μελλοντικών δικτύων πρόσβασης νέας γενιάς. / Nowadays, the rapid increase in bandwidth demanding services imposes new technological directions in the design of next generation optical networks with the purpose to achieve: a) reduced cost, b) larger transmission distances, c) larger number of users and d) higher bandwidth connectivity to the end user. However, due to the analogue nature of the optical signals, the optical networks suffer from a variety of linear and non-linear impairments. These impairments have a direct impact in the signal’s bit error rate performance, while their effect increases as bit rate increases. The compensation of impairments has been traditionally performed by optical means. However, the rapid increase in available electronic processing power has made electronic mitigation of impairments a viable option, leading to an adaptive, low cost and integrated solution which avoids additional optical losses. The goal of this thesis is to study the effective mitigation by electronic means of the most important impairments (i.e. chromatic dispersion, self phase modulation and filter concatenation) that are related with optical networks and particularly metropolitan, access and passive optical networks. From the network (and system) design point of view, this study proposes the optimum use of certain low cost solutions able to extend (in bit rate and coverage) the applicability of next generation optical networks. More specifically, the effectiveness of electronic equalization is examined for systems utilizing low cost, conventional directly modulated laser (DML) sources that are operated at 10 Gb/s. The purpose in this case is to extend the reach and operating data rate of these systems by mitigating the transmission limiting effects due to the source characteristics and the link impairments (dispersion, self-phase modulation, and filter concatenation) with the optimum use of electronic equalization. Moreover, with respect to next generation optical access networks an effective and useful design approach on PON systems is fully investigated, by using the benefits of electronic equalization at the receiver side (ΟLT). This experimental system studies are focusing on PON systems operated at 10 Gb/s by using low cost and low bandwidth RSOAs at the ONU side assisted by electronic equalization at the receiver (ΟLT). This technique offers the required flexibility for the optimum adaptation on the specific network characteristics (in terms of covered distance, number of users and bit rate) and additionally meets the requirements for the development and further extension of future low cost optical access networks.

The continuous growth of traffic demand driven by the brisk increase in number of Internet users and emerging online services creates new challenges for communication networks. The latest advances in Wavelength Division Multiplexing (WDM) technology make it possible to build Transparent Optical Networks (TONs) which are expected to be able to satisfy this rapidly growing capacity demand. Moreover, with the ability of TONs to transparently carry the optical signal from source to destination, electronic processing of the tremendous amount of data can be avoided and optical-to-electrical-to-optical (O/E/O) conversion at intermediate nodes can be eliminated. Consequently, transparent WDM networks consume relatively low power, compared to their electronic-based IP network counterpart. Furthermore, TONs bring also additional benefits in terms of bit rate, signal format, and protocol transparency. However, the absence of O/E/O processing at intermediate nodes in TONs has also some drawbacks. Without regeneration, the quality of the optical signal transmitted from a source to a destination might be degraded due to the effect of physical-layer impairments induced by the transmission through optical fibers and network components. For this reason, routing approaches specifically tailored to account for the effect of physical-layer impairments are needed to avoid setting up connections that don’t satisfy required signal quality at the receiver. Transparency also makes TONs highly vulnerable to deliberate physical-layer attacks. Malicious attacking signals can cause a severe impact on the traffic and for this reason proactive mechanisms, e.g., network design strategies, able to limit their effect are required. Finally, even though energy consumption of transparent WDM networks is lower than in the case of networks processing the traffic at the nodes in the electronic domain, they have the potential to consume even less power. This can be accomplished by targeting the inefficiencies of the current provisioning strategies applied in WDM networks. The work in this thesis addresses the three important aspects mentioned above. In particular, this thesis focuses on routing and wavelength assignment (RWA) strategies specifically devised to target: (i) the lightpath transmission quality, (ii) the network security (i.e., in terms of vulnerability to physical-layer attacks), and (iii) the reduction of the network energy consumption. Our contributions are summarized below. A number of Impairment Constraint Based Routing (ICBR) algorithms have been proposed in the literature to consider physical-layer impairments during the connection provisioning phase. Their objective is to prevent the selection of optical connections (referred to as lightpaths) with poor signal quality. These ICBR approaches always assign each connection request the least impaired lightpath and support only a single threshold of transmission quality, used for all connection requests. However, next generation networks are expected to support a variety of services with disparate requirements for transmission quality. To address this issue, in this thesis we propose an ICBR algorithm supporting differentiation of services at the Bit Error Rate (BER) level, referred to as ICBR-Diff. Our approach takes into account the effect of physical-layer impairments during the connection provisioning phase where various BER thresholds are considered for accepting/blocking connection requests, depending on the signal quality requirements of the connection requests. We tested the proposed ICBR-Diff approach in different network scenarios, including also a fiber heterogeneity. It is shown that it can achieve a significant improvement of network performance in terms of connection blocking, compared to previously published non-differentiated RWA and ICBR algorithms.  Another important challenge to be considered in TONs is their vulnerability to physical-layer attacks. Deliberate attacking signals, e.g., high-power jamming, can cause severe service disruption or even service denial, due to their ability to propagate in the network. Detecting and locating the source of such attacks is difficult, since monitoring must be done in the optical domain, and it is also very expensive. Several attack-aware RWA algorithms have been proposed in the literature to proactively reduce the disruption caused by high-power jamming attacks. However, even with attack-aware network planning mechanisms, the uncontrollable propagation of the attack still remains an issue. To address this problem, we propose the use of power equalizers inside the network nodes in order to limit the propagation of high-power jamming attacks. Because of the high cost of such equipment, we develop a series of heuristics (incl. Greedy Randomized Adaptive Search Procedure (GRASP)) aiming at minimizing the number of power equalizers needed to reduce the network attack vulnerability to a desired level by optimizing the location of the equalizers. Our simulation results show that the equalizer placement obtained by the proposed GRASP approach allows for 50% reduction of the sites with the power equalizers while offering the same level of attack propagation limitation as it is possible to achieve with all nodes having this additional equipment installed. In turn, this potentially yields a significant cost saving.    Energy consumption in TONs has been the target of several studies focusing on the energy-aware and survivable network design problem for both dedicated and shared path protection. However, survivability and energy efficiency in a dynamic provisioning scenario has not been addressed. To fill this gap, in this thesis we focus on the power consumption of survivable WDM network with dynamically provisioned 1:1 dedicated path protected connections. We first investigate the potential energy savings that are achievable by setting all unused protection resources into a lower-power, stand-by state (or sleep mode) during normal network operations. It is shown that in this way the network power consumption can be significantly reduced. Thus, to optimize the energy savings, we propose and evaluate a series of energy-efficient strategies, specifically tailored around the sleep mode functionality. The performance evaluation results reveal the existence of a trade-off between energy saving and connection blocking. Nonetheless, they also show that with the right provisioning strategy it is possible to save a considerable amount of energy with a negligible impact on the connection blocking probability. In order to evaluate the performance of our proposed ICBR-Diff and energy-aware RWA algorithms, we develop two custom-made discrete-event simulators. In addition, the Matlab program of GRASP approach for power equalization placement problem is implemented. / <p>QC 20120508</p>

Στην παρούσα διπλωματική εργασία μελετάμε τα πλεονεκτήματα που επιφέρει η χρήση πολλαπλών κεραιών στον πομπό και στον δέκτη, κατά την μετάδοση, με στόχο την βελτίωση των παρεχομένων υπηρεσιών στο χρήστη. Στο Κεφάλαιο 1, γίνεται μια ιστορική αναδρομή των ασύρματων επικοινωνιών καθώς των σύγχρονων ασύρματων τεχνολογιών και κεραιών που χρησιμοποιούνται. Στη συνέχεια γίνεται μια αναφορά στις έννοιες του διαφορισμού, του κέρδους διάταξης και της χωρικής πολυπλεξίας οι οποίες συνδέονται άρρηκτα με τα συστήματα MIMO. Στο Κεφάλαιο 2, αναφερόμαστε σε όλα εκείνα τα χαρακτηριστικά που περιγράφουν το ασύρματο κανάλι και εξάγουμε την γραμμική σχέση εισόδου-εξόδου του ασύρματου καναλιού. Στην συνέχεια γίνεται μια ανάλυση των στοχαστικών μοντέλων περιγραφής του ασύρματου διαύλου διαλείψεων και πιο συγκεκριμένα των μοντέλων Rayleigh και Rice. Στο Κεφάλαιο 3 εξετάζουμε την αξιοπιστία διαφόρων τύπων κεραιοσυστημάτων, ως προς τον ρυθμό των ρυθμό των λανθασμένων συμβόλων στον δέκτη. Πιο συγκεκριμένα εξετάζεται η τεχνική Maximal Ratio Combining για τα συστήματα SIMO καθώς και του σχήματος Alamouti για τα συστήματα ΜISO. Συνεχίζοντας στα MIMO συστήματα αναλύουμε τις μεθόδους ισοστάθμισης για την ανάκτηση των δεδομένων, και πιο συγκεκριμένα τις τεχνικές Zero Forcing, Minimum Mean Square Error,V-Blast και καθώς και την βέλτιστη τεχνική Maximum Likelihood. Στο τελευταίο μέρος της εργασίας αναλύουμε τα πλεονεκτήματα των MIMO συστημάτων, ως προς την χωρητικότητα που προσφέρουν, σε στοχαστικά κανάλια διαλείψεων.Στη συνέχεια, γίνεται αναφορά στην μέθοδο SVD και στην αναπαράσταση του MIMO καναλιού από έναν αριθμό ανεξάρτητων SISO διαύλων. Κλείνοντας αναφέρουμε την μέθοδο βέλτιστης κατανομής ισχύος στις κεραίες του πομπού Water-filling, και στην περαιτέρω αύξηση της χωρητικότητας του διαύλου που προσφέρει. / In this diploma thesis we are investigating the benefits of using Multiple Input and Multiple Output antennas in information transmission, with final goal to improve Quality of Service. The first Chapter, includes a historical background of the wireless communications but also is a reference to the modern wireless and antenna technologies. Moreover, we introduce the definition of new concepts, such as diversity and array gain and also spatial multiplexing, which are closely connected with MIMO technology. In the second chapter, we introduce the characteristics which they are describe the wireless channel, while simultaneously we mention the linear input-output relationship of the wireless channel. Additionally, we analyze the stochastic wireless channel models, namely the Rayleigh and the Rician fading models. In the third chapter, we investigate the reliability of different types of antenna topologies, regarding the pace of the invalid symbols in the transmitter. More specifically, we examine the Maximal Ratio Combining and Alamouti technique, for SIMO and MISO systems respectively. The next step is to analyze the equalization methods, which are used in MIMO antennas, and more specifically are, Zero Forcing, Minimum Mean Square Error and V-Blast receivers, but also the optimal Maximum Likelihood equalizer. In the last part of this Thesis, we investigate the benefits of MIMO systems regarding the Capacity, in random channels. Also, a reference to the SVD method has been made,which we use to analyze the MIMO channel, in a number of parallel SISO channels. Lastly, we use the water-filling method to allocate, with the optimal way, the given power in the transmit antennas, a fact that leads to even greater Capacity gain.