No description available.

No description available.

No description available.

No description available.

No description available.

This thesis proposes a new overlay approach to converge fixed-aggregation and mobile-access networks in terms of resource management and Quality of Service (QoS). Convergence, in the context of this work, refers to end-to-end status information exchange between fixed and mobile networks resulting in enhanced resource management and better QoS. The main motive for such an approach was driven from the fast evolution of both wireless and fixed networks, showing the need of a technology independent, scalable and extensible system. Recent literature in this area, on one hand reveals the benefits of converging aggregation and access networks but on the other hand highlights the lack of such an overlay mechanism. At first, a protocol was designed in order to perform the signaling between networks carrying bandwidth request/allocation, queue and service status information. This information is used from the Dynamic Bandwidth Allocation (DBA) and Call Admission Control (CAC) functions in the aggregation and access network respectively. A fairness-based DBA algorithm that is solely based on the signaling information, was designed for use in the aggregation network. Additionally, two CAC schemes were designed to operate in the Base Stations (ESs) of the mobile networks, each one taking into consideration different portions of the available information. For the validation of the proposed system, Ethernet Passive Optical Network (EPON) and Worldwide Interoperability for Microwave Access (WiMAX) networks were employed in OMNet++ simulation environment. Simulations performed using different mobility and traffic parameters, showed that the proposed overlay approach is viable and always beneficiary in terms of network performance. Furthermore, it was shown that the CAC function performs better when more information about the aggregation network is made available to it. The fully converged method achieved to limit the packet drop rate to 10% under constant and heavy network load and even less for periodically congesting networks. In both cases, the overall network throughput was not affected. However, the previous approach was not able to guarantee end-ta-end packet delay. For this reason a scaling parameter was introduced that efficiently controls the major trade-off in converged environments between QoS and network throughput. This proved to be able to bound the packet delay at the expense of less throughput

This document introduces a Self-Reconfigurable system for Software defined radio (SR.S). The SRS is a modular processing platform which extends an ordinary PC system with Software Defined Radio capabilities. It consists of several daughter cards and one main board. The daughter cards connect the main board to external entities such as an analogue front end and the PC system. The main board hosts a single processor which executes signal processing algorithms and interfaces to external entities. The PC system controls which algorithms the processor executes and if necessary it initiates an algorithm change during runtime. The SRS system uses a reconfigurable architecture for the processor on the main board. This architecture provides a mechanism known as runtime reconfiguration to change the functionality or parts of the functionality of the processor. The PC system communicates with one part of the reconfigurable processor. This part reconfigures the signal processing part of the processor if the PC1 system initiates an algorithm change. I call this mechanism partial self-reconfiguration. Several components within the reconfigurable processor abstract the interfacing and signal processing tasks. The processor executes these components in parallel. I used a design methodology based on formal methods to design this component network. To be specific. I employed a model checker to validate a Communicating Sequential Processes (CSP) model of the component network. In a second step this CSP model was translated into hardware description code. The reconfigurable processor accelerates the baseband algorithms of digital communication standards. I studied the baseband algorithms which define the Orthogonal Frequency Division Multiplex (OFDM) modulation scheme. These algorithms have specific processing requirements. The specific-processor for the main board was selected based on the processing requirements of these algorithms. The SRS processing platform copes with the increasing algorithm complexity, higher demands for energy efficiency and flexibility. The modular system design provides a great deal of connection flexibility. This actually extends the idea of Software Defined Radio to wire or bus standards, i.e. software defines the bus standard. At the same time, formal methods ensure that the design, which defines the functionality of the reconfigurable processor, is reliable. Moreover, formal methods describe architecture and functionality of the system in a very compact manner. This provides the opportunity to document a hardware system completely. The OFDM algorithms provide a. deep insight into modern digital algorithms and how to accelerate these algorithms in a parallel architecture. This insight is necessary to draw up the specification for the system design.  The functionality of the reconfigurable processor was created with a design methodology based on formal methods. Compared with standard design methodologies the incorporation of formal methods speeds up the design process and the results are more reliable. The SRS system executes digital communication algorithms. Time and frequency synchronisation algorithms are amongst the most demanding, because they require realtime processing and they are very complex. The OFDM time and frequency synchronisation algorithms were improved with a technique called multi-rate signal processing. In a nutshell, multi-rate signal processing allows increased the time resolution and therefore it is possible to track a signal with finer steps.

This thesis presents our recent researches on the development of coding devices, the investigation of security and the design of systems in the optical cod-division multiple access (OCDMA) systems. Besides, the techniques of nonlinear signal processing used in the OCDMA systems fire our imagination, thus some researches on all-optical signal processing are carried out and also summarized in this thesis. Two fiber Bragg grating (FBG) based coding devices are proposed. The first coding device is a superstructured FBG (SSFBG) using ±π/2-phase shifts instead of conventional 0/π-phase shifts. The ±π/2-phase-shifted SSFBG en/decoders can not only conceal optical codes well in the encoded signals but also realize the reutilization of available codes by hybrid use with conventional 0/π-phase-shifted SSFBG en/decoders. The second FBG based coding device is synthesized by layer-peeling method, which can be used for simultaneous optical code recognition and chromatic dispersion compensation. Then, two eavesdropping schemes, one-bit delay interference detection and differential detection, are demonstrated to reveal the security vulnerability of differential phase-shift keying (DPSK) and code-shift keying (CSK) OCDMA systems. To address the security issue as well as increase the transmission capacity, an orthogonal modulation format based on DPSK and CSK is introduced into the OCDMA systems. A 2 bit/symbol 10 Gsymbol/s transmission system using the orthogonal modulation format is achieved. The security of the system can be partially guaranteed. Furthermore, a fully-asynchronous gigabit-symmetric OCDMA passive optical network (PON) is proposed, in which a self-clocked time gate is employed for signal regeneration. A remodulation scheme is used in the PON, which let downstream and upstream share the same optical carrier, allowing optical network units source-free. An error-free 4-user 10 Gbit/s/user duplex transmission over 50 km distance is reazlied. A versatile waveform generation scheme is then studied. A theoretical model is established and a waveform prediction algorithm is summarized. In the demonstration, various waveforms are generated including short pulse, trapezoidal, triangular and sawtooth waveforms and doublet pulse. ii In addition, an all-optical simultaneous half-addition and half-subtraction scheme is achieved at an operating rate of 10 GHz by using only two semiconductor optical amplifiers (SOA) without any assist light. Lastly, two modulation format conversion schemes are demonstrated. The first conversion is from NRZ-OOK to PSK-Manchester coding format using a SOA based Mach-Zehnder interferometer. The second conversion is from RZ-DQPSK to RZ-OOK by employing a supercontinuum based optical thresholder.

In cellular networks, interference is identified as the major performance bottleneck. Attempts are made in 4G and 5G systems to address this by pooling base stations together to form a network multiple-input multiple-output (MIMO) system. Global coordination in network MIMO systems is known to be highly complex and costly. In this thesis, a scalable solution is proposed by clustering the network into groups of base stations. Interference within the cluster is mitigated by performing network MIMO based signal processing in each cluster independently. Interference between clusters is then cancelled by applying fractional frequency reuse (FFR) on a cluster scale. In FFR systems, a greater reuse factor is used for users near the cell or cluster edge since they are more prone to interference. An important problem in FFR systems is classifying the user location as either being in the centre or near the edge. The conventional technique of using a one-dimensional signal-to-interference-and-noise ratio (SINR) threshold is highly inaccurate and an improved machine learning approach based on logistic regression is proposed. It is shown to improve the accuracy to at least 80% and the cell sum rate performance is shown to be comparable to that of a system with 100% accurate classification.

This thesis focuses on novel interference management and radio access techniques, with no Channel State Information (CSI) at the transmitters, that can be employed in future homogeneous and heterogeneous communication networks, outperforming traditional schemes, such as TDMA.