The evolution of mobile edge networks demands efficient content delivery mechanisms, as traditional caching algorithms struggle with dynamic user demand. We propose a predictive caching framework combining digital twins and machine learning to optimize content delivery. By predicting content popularity and dynamically adjusting Time-To- Live parameters, our system efficiently manages cache size, retaining content based on demand predictions to optimize space utilization and access times. The framework also enhances performance through load balancing and content sharing among edge nodes. Integrating digital twins enables real-time predictive caching, improving cache hit rates, reducing latency by 75%, and achieving a cache hit ratio of 0.83, thereby enhancing user Quality of Service.

A multitude of 6G smart devices will integrate applications that require contextual data processing, such as healthcare remote sensing and interactive applications. These are computationally intensive tasks that require quick execution within stringent time constraints. Given the heterogeneity of these applications as well as the communication and computation uncertainty due to the shared nature of the edge environment, computation competition and access interference pose serious challenges in the offloading of computational tasks. In this respect, the offloading decision-making of edge devices is influenced and shaped by prospect-theoretic characteristics. To address this challenge, a graph neural network (GNN)-based discrete choice experiment (DCE) is proposed to model the choice of the appropriate offloading server based on a feature — attribute matching. An encoded vector that is validated using interference, channel gain, and transmission rates, is encoded into the embedding space, and offloading solution for each edge device is obtained. The performance evaluation results show that the proposed GNN-based DCE lowers the expected overheads, which maximizes the amount of offloaded computational tasks.

Numerous code constructions which limit the maximum homopolymer run length and constrain GC-content have been proposed in literature for DNA data storage. These codes help to reduce insertion, deletion and substitution errors in DNA strands, since strands with large homopolymer runs and a large unbalance in GC-content, i.e., GC-content is too high or too low, are more prone to the previously mentioned errors during synthesis and sequencing. While the capacity of non-binary runlength-limited codes is known, the capacity of non-binary balanced runlength-limited codes is unknown. Here, we generalize the capacity result of binary balanced runlength-limited codes to the non-binary alphabet and demonstrate that the capacity of non-binary balanced (equal number of occurrences of each non-binary symbol) runlength-limited codes is equal to the capacity of the corresponding runlength-limited codes without the balancing constraint. An immediate consequence of the above result when applied to the quaternary alphabet of DNA data storage is that the capacity of runlength-limited and GC-content constrained codes for any arbitrary range of permitted GC-content unbalance which includes the balanced case is equal to the capacity of the corresponding runlength-limited codes without the GC-content constraint, i.e., the imposition of any GC-content constraint to runlength-limited codes does not induce a capacity penalty.

Due to the increase in resource-constrained internet of things (IoT) devices, the multi-access edge computing (MEC) have become very competitive environments in terms of successful data offloading and allocation of computational resources. This competition together with the varying workload and service requests makes the real deployment of computational resources in edge servers a major challenge. In order to address this challenge, an intelligent offloading and container provisioning scheme is developed using a prospect-theoretic deep reinforcement learning (DRL) strategy. An offloading utility function is formulated by exploiting offloading overhead options, and a neural network (NN) is used to monitor queue states and workload matching to construct a behavioral function that will influence accurate container provisioning. A scaled cost function is formulated to balance the energy consumption and the quality of service (QoS) cost subject to hard per-task latency constraints, which lead to better container utilization and energy consumption.

Interference modeling in cognitive radio network is important to ensure adequate coverage in the network. A reliable interference model, however, depends on accurately characterizing the distribution of users. In this paper, the dependence between primary and secondary networks is examined in order to capture more system parameters related to system characterization. Hence, two cases are considered - primary user (PU) interference control and PU with secondary user (SU) interference control mechanisms. Under PU interference control, distributions of PUs follow the Matern hard core process while the distribution of SUs follow the Poisson hole process (PHP). However, under PU with SU interference control, the distribution of active SUs follow a modified PHP. Bound and approximate expressions were derived for coverage probability at both primary and secondary networks, while simple yet accurate expressions were obtained to depict the number of simultaneous active users supported for the two cases. The tight closeness between the bound and the approximate expressions shows the reliability of the presented theoretical analysis. Furthermore, the bipolar network model assumption was relaxed while the case of independence assumption among users was also considered. Numerical result showed close tightness when the bipolar network model assumption was relaxed while the independence assumption was shown to overestimate users' coverage probability.

The increased densification of wireless networks makes it difficult to provide traditional fiber backhaul access for every small cell base station. The increasing maturity of millimeter-wave communication and its suitability for wireless backhaul links has led to the development of integrated access and backhaul (IAB) for 5G networks. This paper proposes a fully autonomous and distributed reinforcement learning scheme for user association in a congested IAB network that uses instantaneous load-based bandwidth partitioning, where a cognitive engine judiciously determines network congestion to initiate a bandwidth split. Firstly, the effect of learning the congestion rate using Q-tables is determined in terms of the learning rate and discount factor. Secondly, reactive load balancing, which considers the effect of neighboring slave base stations' (SBSs') loads using exponentially weighted moving average, was applied. The results indicate that even at higher neighboring load intervals, the current SBS load can stay below the predetermined threshold, which gives insight on how much extra bandwidth needs to be shared among different SBSs according to the instantaneous load-based partitioning scheme.

Secondary user (SU)-experience-oriented resource allocation (RA) will become increasingly important in cognitive radio networks (CRNs) in future wireless networks. For efficient real-time processes, cognitive radios (CRs) are usually combined with artificial intelligence (AI) to improve better adaptation and intelligent RA. However, deep learning (DL), which is a key AI strategy with remarkable capabilities towards advancing this vision, has several built-in limitations. Firstly, the most successful DL applications require training with large amounts of data; secondly, they assume that the data samples to be independent, while in CRNs one typically encounters sequences of highly correlated states. To circumvent this issue, this paper introduces a deep neuroevolution (DNE) technique for dynamic RA. Using this technique, a stable learning framework was achieved by introducing the phenotypic plasticity of transmission rates and delay constraints inside a multi-layer perceptron (MLP). The stability of SU satisfaction as they increased in number was achieved at 36 SUs, which is a 13.3% decrease from when they were only 6 SUs in the CRN for all learning mechanisms.

Integration of mobility into system characterization can complicate system analysis, as characterizing interference in mobile wireless networks is more difficult, especially when the assumption of independence of users' distributions is relaxed. However, in some practical systems, quite a number of users may be mobile, making the integration of mobility important. In this paper, mobile users' coverage in cognitive radio networks is investigated under the assumption that primary users are distributed following the Matern hard core process, while secondary users are distributed following the extended Poisson hole process. A relaying based mobility technique was proposed to sustain coverage at any tagged receiver when such receiver is located outside the coverage of its corresponding tagged transmitter. Analyses of outage probability and average throughput were derived in both primary and secondary networks and verified through simulations. The outcomes of numerical results show that the proposed mobility scheme is capable of improving coverage of mobile users in cognitive networks.

Malicious users (MUs) have the tendency to disrupt the activities of honest users in the network if not properly controlled. In a massive cognitive radio network (CRN) with priority queues, malicious secondary users (SUs) can manipulate their priority queue requirements and mislead legitimate SUs to vacate the channels. In this paper, a game theoretic based signal detection approach is proposed to control the presence of MUs in CRN. If the received signal strength is less than the predefined threshold for primary transmissions in the presence of interference and noise, such a user is marked to be malicious and its payoff table is updated. Through the mixed strategy Nash equilibrium method, the payoff table of each user can be updated to aid removal of MUs from the network. The outcome of the simulation results shows that such an approach can reduce the impact of malicious activities in the massive CRN where SUs are expected to be low-power energy-efficient devices.

Information reception in artificially synthesized molecular communication (MC) systems ideally follows the mechanisms employed by natural nanosystems to communicate. One of such reception mechanism is the so called ligand-receptor binding. Contemporary research in MC has considerably discussed this mechanism; however, the impact of a crucial parameter associated with the ligand-receptor binding action has not been given appropriate attention in the MC literature. This parameter is termed the residence time, and has played very crucial role in defining for instance, the efficacy of drugs in therapeutic processes; hence, it is critical in the performance of MC. In this paper, we employ biophysical approach to model and discuss the influence of the ligand residence time on the performance of MC systems. The performance metrics considered here are the receiver sensitivity and the intersymbol interference. Numerical results that expose the impact of the residence time on these metrics, and the interrelationships between these metrics in MC system, are discussed.

The recent surge in real-time traffic has led to serious energy efficiency concerns in cognitive radio networks (CRNs). Network infrastructure such as base stations (BSs) host different service classes of traffic with stringent quality-of-service (QoS) requirements that need to be satisfied. Thus, maintaining the desired QoS in an energy efficient manner requires a good trade-off between QoS and energy saving. To deal with this problem, this paper proposes a deep learning-based computational-resource-aware energy consumption technique. The proposed scheme uses an exploration technique of the systems' state-space and traffic load prediction to come up with a better trade-off between QoS and energy saving. The simulation results show that the proposed exploration technique performs 9% better than the traditional random tree technique even when the provisioning priority shifts away from energy saving towards QoS, i.e., α ≥ 0.5.

A novel empirical analysis of a decentralised network under a probabilistic Byzantine attack was performed. A decentralised cognitive radio network was simulated using the Neyman-Pearson Belief Propagation (NP-BP) algorithm for cooperative spectrum sensing (CSS). The network was exposed to malicious users (MUs) that attempted to increase the false alarm or missed detection rates in the sensing of the primary user (PU). The NP-BP algorithm was seen to reduce the occurrence of spectrum sensing (SS) errors (in comparison to likelihood detection) when no MUs were present. The performance of the data fusion algorithm was very sensitive to MUs that increased the false alarm rate of PU detection. Conversely, the network performed much better when missed detection attacks were conducted by the MUs. The NP-BP algorithm tends to prefer moderate network connectivity for optimal SS results.

It is known that, for large user word lengths, the auxiliary data can be used to recover most of the redundancy losses of Knuth's simple balancing method compared with the optimal redundancy of balanced codes for the binary case. Here, this important result is extended in a number of ways. First, an upper bound for the amount of auxiliary data is derived that is valid for all codeword lengths. This result is primarily of theoretical interest, as it defines the probability distribution of the number of balancing indices that results in optimal redundancy. This result is equally valid for particular non-binary generalizations of Knuth's balancing method. Second, an asymptotically exact expression for the amount of auxiliary data for the ternary case of a variable length realization of the modified balanced code construction is derived, that, in all respects, is the analogue of the result obtained for the binary case. The derivation is based on a generalization of the binary random walk to the ternary case and a simple modification of an existing generalization of Knuth's method for the non-binary balanced codes. Finally, a conjecture is proposed regarding the probability distribution of the number of balancing indices for any alphabet size.

A novel Knuth-like balancing method for runlength-limited words is presented, which forms the basis of new variable- and fixed-length balanced runlength-limited codes that improve on the code rate as compared to balanced runlength-limited codes based on Knuth's original balancing procedure developed by Immink et al. While Knuth's original balancing procedure, as incorporated by Immink et al., requires the inversion of each bit one at a time, our balancing procedure only inverts the runs as a whole one at a time. The advantage of this approach is that the number of possible inversion points, which needs to be encoded by a redundancy-contributing prefix/suffix, is reduced, thereby allowing a better code rate to be achieved. Furthermore, this balancing method also allows for runlength violating markers which improve, in a number of respects, on the optimal such markers based on Knuth's original balancing method.

In an overlay/underlay cognitive radio network the channel can primarily be in two modes; occupied or unoccupied by a primary user, thus being in an overlay or underlay mode for the secondary user. However, during the underlay mode, the channel can present different capacities for the secondary user due to the varying primary user power levels. In this paper we introduce a method for modeling the multi-mode overlay/underlay levels for the secondary users and beyond the traditional 3-mode levels applied in the previous work. We essentially consider the case where the service levels of the channel can switch between the different modes, hence the channel availability and effective capacity to a secondary user are dynamic. We further present a discrete time queueing model that captures this scenario through the use of a discrete time Markov chain for analyzing the behavior of the system.

Local caching of popular content files at the wireless edge is a promising way to address several challenges for the next mobile radio generation. Apart from reducing latency, it also helps save energy and better control the available end-to-end power budget. In this paper, we take advantage of dense small cell networks' architecture to further reduce the power consumption of the system, using a graph-theoretic approach. The basic idea is to download the requested content file from a neighboring small cell rather than macro cell or backhaul, if it is not cached at the serving base station. More precisely, we build a graph to fetch the requested content file from the nearest less power consuming small cells. Special attention is paid to the adequate number of small base stations involved in the graph to ensure an energy efficient solution. In addition, we apply our graph theory based approach to multicast transmissions and show that our built graph outperforms the conventional unicast/multicast caching transmission in terms of energy efficiency. Simulation results also showed the effectiveness of our approach for different cache size and steepness parameter.

Massive multiple-input multiple-output (MIMO) is an emerging technology, which enables future broadband wireless connections. In this paper we investigate the optimization problem of signal-to- interference-plus-noise ratio (SINR) balancing for users in a typical massive MIMO network. We present a fast converging robust beamforming solution for the case of imperfect channel state information at the transmitter. The proposed method applies the Smith form reformulation to transform the original non-convex optimization problem into a standard second-order cone program. The matrix stuffing technique and the alternative direction method of multipliers are then used to provide an efficient solution to the problem. Simulation results show that the proposed solution gives SINR performance with lower accuracy averaging 97%, while it converges to a solution more than twice faster than the optimal solution, which uses CVX and the SeDuMi solver.

A simple and general method is presented for enumerating the number of balanced (zero-disparity) runlength-limited words based on bivariate generating functions and the symbolic methods of analytic combinatorics. This procedure scales well with increasing word length, allowing for exact enumerations for very large word lengths. As a result, it serves as a precise benchmark against which to evaluate the performance, in terms of code rate, of balanced runlength-limited coding schemes.

A channel share scheme is proposed as a solution to the resource allocation problem in cognitive radio networks whereby two secondary users can occupy the same channel simultaneously with the primary user. The solution will increase utilisation thus freeing up much needed spectrum to accommodate many more users. Processor sharing techniques have been developed but have not been extensively applied in cognitive radio networks, specifically, the weighted head of line processor sharing scheme has the potential of improving spectrum usage in networks. In this paper, we propose using transmission power as the basis for the weighting parameter. We then develop the state transition matrices for the model using continuous time Markov chains. Queues are imposed on both secondary users. The model is evaluated through simulations done on various network conditions. The results show that the proposed model offers a significant improvement over an ordinary M/M/1 two priority system. The applications can lead to improved network efficiency by allowing more users to transmit within a network.

This paper presents a practical feasibility study of a novel two-phase three-part-message strategy for half-duplex relaying, which features superposition coding and interference-aware cancellation decoding. Aiming to analyze the performance of the proposed scheme in the non-asymptotic regime, this paper evaluates the spectral efficiency with finite block-length and discrete constellation signaling and compares it with the theoretical performance of Gaussian codes with asymptotically large block-lengths. The performance evaluation is carried out on an LTE simulation test bench. During each transmission phase, the modulation and coding scheme is adapted to the channel link qualities to enhance the overall spectral efficiency. A single-antenna source and relay, and a multi-antenna destination are assumed. The static Gaussian and two frequency selective channel models are considered for the proposed scheme. A spectral efficiency comparison with a baseline scheme (non-cooperative two-hop transmission, i.e., the source-destination link is absent) and with the point-to-point transmission strategy (no relay) is presented. The results confirm that physical-layer cooperation and multi-antennas are critical for performance enhancement in heterogeneous networks. Moreover, they show that physical layer cooperation advantages are within practical reach with existing LTE coded-modulation and interference-mitigation techniques, which are prevalent in modern user-equipment.

Resources available for operation in cognitive radio networks (CRN) are generally limited, making it imperative for efficient resource allocation (RA) models to be designed for them. However, in most RA designs, a significant limiting factor to the RA’s productivity has hitherto been mostly ignored, the fact that different users or user categories do have different delay tolerance profiles. To address this, in this paper, an appropriate RA model for heterogeneous CRN with delay considerations is developed and analysed. In the model, the demands of users are first categorised and then, based on the distances of users from the controlling secondary user base station and with the assumption that the users are mobile, the user demands are placed in different queues having different service capacities and the resulting network is analysed using queueing theory. Furthermore, to achieve optimality in the RA process, an important concept is introduced whereby some demands from one queue are moved to another queue where they have a better chance of enhanced service, thereby giving rise to the possibility of an improvement in the overall performance of the network. The performance results obtained from the analysis, particularly the blocking probability and network throughput, show that the queueing model incorporated into the RA process can help in achieving optimality for the heterogeneous CRN with buffered data.

Most research on the mathematical modelling of wireless sensor networks (WSNs) have focussed mainly on the optimization aspects, such as those relating to sensor placements, data routing, reliability, etc. Surprisingly the issue relating to performance analysis of data processing and transmission at the nodes, have not received as much attention. A considerable amount of delay to data actually happens at the nodes as a result of queue build up. Hence, understanding the role of queueing in WSN modelling is very important. In this paper we study the literature of queueing as applied to WSNs and provide insight to the current state of the art and directions for the future. The utilization of queueing theory in WSNs is broadly classified into four categories, namely, congestion control methods, power allocation schemes, network performance evaluation techniques and scheduling schemes.

In a cognitive radio network (CRN) of multiple secondary users (SUs) and multiple primary user channels, the protocol of allocating channel resources to SUs and the evaluation of the effectiveness of the protocol are still open issues.In this paper, we developed a novel protocol to allocate channels to SUs and an analytical criterion to evaluate the performance of this protocol.The system settings and environmental parameters, including spectrum sensing errors of SUs, adaptive modulation and coding schemes, and finite SU buffer size, are taken into consideration in our framework.By modeling the behavior of each SU as a discrete-time queue, we obtained a performance evaluation function (PEF) considering the rejected packets, collision packets and error packets.Simulation and numerical results showed that the protocol is better than random allocation in terms of PEF. The impact of different system settings on the optimal protocol settings and the optimal PEF was also investigated.

Massive multiple-input multiple-output (MIMO) is an emerging technology, which is an enabler for future broadband wireless networks that support connection of the internet of people and internet of things. The use of massive MIMO base stations and heterogeneous networks offers an increase in throughput without increasing the bandwidth, but with reduced power consumption. This makes the two technologies very attractive for future wireless mobile communication systems. We investigate the optimization problem of maximizing the data rate whilst mitigating inter-tier interference for macrocell users in a typical heterogeneous network. We present a first-order solution for maximizing the weighted sum rate of the macrocell users. The proposed method uses the matrix stuffing technique and the alternative direction method of multipliers to efficiently transform and solve the optimization problem, respectively. Simulation results of a single-cell heterogeneous network show that the solution gives near-optimal performance, which is achieved by state-of-the-art modelling languages and second-order solvers for an increased number of antennas at the macrocell base station. The proposed method converges much faster than the interior-point methods, which makes it attractive for practical implementation.

With the advent of nanotechnology, concepts related to the Internet of Things, such as the Internet of NanoThings and Internet of Bio-NanoThings (IoBNT) have also emerged in the classical literature. The main concern of this paper is the IoBNT, which projects the prospective application domain where the activities of very tiny, biocompatible, and non-intrusive devices operating in an in-body nanonetwork can be monitored and controlled through the Internet. In this paper, we present an illustrative scenario and system model of an IoBNT for application in an advanced healthcare delivery system. To address one of the major challenges of the IoBNT, we present an exemplary architecture and model of a bio-cyber interface for connecting the conventional electromagnetic-based Internet to the biochemical signaling-based bionanonetwork. The bio-cyber interface is designed and modeled by employing biological concepts, such as the responsiveness of certain biomolecules to thermal and light stimuli, and the bioluminescence phenomenon of some biochemical reactions. The analysis in this paper focuses on the system that comprises the bio-cyber interface and the information propagation network of the blood vessel that leads to the in-body nanonetwork location. The effects of the system and design parameters associated with the IoBNT models presented are numerically evaluated.

This paper presents a practical implementation of a novel three-message transmission strategy for the Gaussian half-duplex relay channel based on Turbo code superposition encoding and interference-aware successive interference cancellation. The impact of finite block-length and discrete input constellations on the Block Error Rate (BLER) performance is evaluated through extensive simulations on an LTE simulation test bench and compared to the theoretical performance of asymptotically large block-length Gaussian codes. For the practically relevant BLER value of 10-2and by varying the direct source-destination link strength, the maximum spectral efficiency gap between theory and the presented implementation is found to be of 0.458 bits/dim when the strength of the source-destination and relay-destination links is the same and of 0.681 bits/dim when the relay-destination link is 5 dB stronger than the source-destination link. These values indicate that practical implementations of high-performing HD relay techniques for future Heterogeneous Network deployments are within reach. A comparison with a baseline strategy without direct source-destination transmission, as currently proposed in the LTE standard for relay scenarios, shows superior performances of the proposed scheme. In particular, the rate gain is of a factor of 2 when the strength of the source-destination and relay-destination links is the same and of a factor of 1.2 when the relay-destination link is 5 dB stronger than the source-destination link, thereby highlighting the critical importance of physical-layer cooperation in broadband wireless systems.

A novel physical-layer based security method that utilizes jammers to generate additional interference for devices that are eavesdropping on wireless network communication is proposed. The scheme involves the intelligent placement of continuous jammers in order to achieve maximum protection and data confidentiality for wireless mesh networks with multiple eavesdroppers, sources and destinations. Furthermore, the scheme is optimized in terms of the transmitting power associated with each jammer so that the energy expended by the jammers is kept at a minimum. The protection scheme precludes the use of any cryptographic techniques and is only physical layer based. The security method is modeled as a minimization mixed integer non-linear problem and is approximated as the sequential solution of two linear optimization sub-problems relating to the placement and power allocation of the wireless jammers. The placement of the jammers takes the form of a multiple demand multidimensional knapsack problem with a minimization objective. The power allocation problem is modeled as a linear real-valued minimization optimization problem. The performance of the proposed security method is evaluated through appropriate simulations conducted on random network instances.

Cognitive radio (CR) network is currently an emerging technology for future wireless communications. In this paper, the problem of resource allocation for heterogeneous users in an underlay CR network is addressed. The objective is to maximize the weighted sum of the data rate for all the categories of secondary users within the network. The problem is first formulated as a non-linear optimization problem which is non-deterministic polynomial-time (NP) hard. The problem is then cleverly re-formulated as an integer linear programming problem and solved through branch-and-bound method. The impact of the weights on the different categories of users is also investigated. The results obtained show that through the approach developed by the authors, optimal resource allocation is achievable for heterogeneous CR networks.

A multiscale input strategy for multiview deep learning is proposed for supervised multispectral land-use classification, and it is validated on a well-known data set. The hypothesis that simultaneous multiscale views can improve composition-based inference of classes containing size-varying objects compared to single-scale multiview is investigated. The end-to-end learning system learns a hierarchical feature representation with the aid of convolutional layers to shift the burden of feature determination from hand-engineering to a deep convolutional neural network (DCNN). This allows the classifier to obtain problem-specific features that are optimal for minimizing the multinomial logistic regression objective, as opposed to user-defined features which trade optimality for generality. A heuristic approach to the optimization of the DCNN hyperparameters is used, based on empirical performance evidence. It is shown that a single DCNN can be trained simultaneously with multiscale views to improve prediction accuracy over multiple single-scale views. Competitive performance is achieved for the UC Merced data set, where the 93.48% accuracy of multiview deep learning outperforms the 85.37% accuracy of SIFT-based methods and the 90.26% accuracy of unsupervised feature learning.

The authors have proposed a fair solution to optimal rate adaptation. The problem has been described in terms of the objective function, decision variables and constraints. Furthermore, using this problem definition, a rate adaptation heuristic was developed and divided into two sub- problems; part one requires finding the optimal rate allocation within the network, part two continues by finding a fair solution whilst still keeping an optimal rate allocation. The heuristic relies on cooperation in the network, and information regarding selected rates and loss due to interference, is distributed between neighbouring nodes. Furthermore, the heuristic is modelled as a repeated game with infinite horizon and it is shown how cooperation can be enforced. A Stack topology has been used for analysing and comparing performance and OMNeT++ 4.2.2 has been selected as simulation platform. The authors have shown that the heuristic effectively reduces the packet loss ratio (PLR). Thereafter, it was shown that the solution is both fair and optimal in terms of data rate.

Multitemporal land-use analysis is becoming increasingly important for the effective management of Earth resources. Despite that, consistent differences in the viewing and illumination geometry in satellite-borne imagery introduce some issues in the creation of land-use classification maps. The focus of this letter is settlement classification with high-resolution panchromatic acquisitions, using texture features to distinguish between settlement classes. The important multitemporal variance component of shadow is effectively removed before feature determination, which allows for minimum-supervision across-date classification. Shadow detection based on local adaptive thresholding is employed and experimentally shown to outperform existing fixed threshold shadow detectors in increasing settlement classification accuracy. Both same- and across-date settlement accuracies are significantly improved with shadow masking during feature calculation. A statistical study was performed and found to support the hypothesis that the increased accuracy is due to shadow masking specifically.

Classifier-generic domain adaptation based on feature space matching is applied in this study, with the aim of correcting dataset shifts consisting of both covariate and concept shifts. The feature space transformation between training and test samples is estimated as a set of partition translations, where each transformed partition mean coincides with the mean of a paired target partition. Various feasible instantiations of the generalized transformation estimate are used to characterize the spatial and temporal feature variance present in a settlement classification problem using panchromatic across-area and across-date high resolution QuickBird imagery. A numerical analysis indicates that a significant settlement classification accuracy improvement is possible with the application of feature space matching, where texture features are used to describe settlement characteristics.

QuickBird imagery acquired on separate dates may have significant differences in viewing- and illumination geometries, which can negatively impact across-date settlement type classification accuracy. The effect of cast shadows on classification accuracy is specifically investigated in this study by performing shadow removal, before the calculation of texture features which are then used as classifier inputs. It is shown that popular texture features are sensitive to differences in shadow profiles, and that across-date classification accuracy can be improved with shadow removal. The strong relationship between the best shadow detection thresh-old and the best settlement classification accuracy is indicated for the study data set. A refined shadow correction algorithm is proposed, and demonstrated to be more effective than basic shadow correction methods.

Spectrum, environment and location awareness capabilities play a vital role in the autonomous and adaptive characteristics of cognitive radio. Knowledge of a cognitive radio (CR) user's location and type of environment may enhance optimisation of dynamic spectrum management and channel allocation. In this paper, a method for determining the near future location of a secondary user (SU) in a CR network, is proposed. This method incorporates a Time-of Arrival (TOA) location estimation algorithm with a Hidden Markhov Model (HMM) based channel occupancy prediction model. Using this method, location estimation under various channel conditions and for a range of available channel bandwidths, is investigated. Results indicate that non-line-of-sight conditions have a profoundly negative effect on location estimation accuracy, but that employing multiple future bandwidths as opposed to a single set of instantaneous bandwidths, does improve the situation, i.e. an average relative improvement in positional accuracy of 27%.

The application of joint source-channel coding, based on fountain codes, in the broadcast timeslot of wireless two-way network coding is investigated in this paper. A computationally efficient subroutine is contributed to the implementation of the fountain compressor, and an error analysis is done. Motivated to develop a true joint source-channel-network code that compresses, adds robustness against channel noise and network codes two packets on a single bipartite graph and iteratively decodes the intended packet on the same Tanner graph, an adaptation of the fountain compressor is presented. The proposed code is shown to outperform a separated joint source-channel and network code in high source entropy and high channel noise regions.

Accurate information regarding a cognitive radio user's location and environment can enhance the adaptive and spectral awareness capabilities of cognitive radio systems. In this paper, a single-path multiband Time-of-Arrival (TOA) positioning technique for cognitive radio is proposed and the performance evaluated using maximum-likelihood (ML) location estimation for a typical rural scenario where signal line-of-sight (LOS) between a transmitter and receiver is prevalent. The multiband Cramer Rao Lower Bound (CRLB) time-delay and channel fading coefficient estimates are derived followed by an estimation combining technique which involves selecting an overall optimum estimate using all the utilized bands. It is observed that the improvement in positioning accuracy for the multiband system depends primarily on the number of utilized bands as well as the (signal-to-noise ratio) SNR of these bands.

This paper investigates the performance effects of packet depletion on multihop wireless networks. The resulting capacity usage inefficiencies of link assigned schedules are shown even at the maximum supportable uniform network load. Hybrid node-link assignment is employed to extend transmission rights, allowing nodes to transmit on sets of reserve links when the primary scheduled link experiences packet depletion. Improvements in network performance with the extension of transmission rights are shown for proportionally fair cross-layer optimized link assigned schedules. The achievable capacity of spatial reuse networks affected by packet depletion is defined in a new rate region and a computationally feasible suboptimum is proposed. Cross-layer optimization in this rate region demonstrates further improvements in network throughput and fairness. The hybrid assignment facilitates network coding and the increased benefits thereof are shown for schedules optimized in the proposed rate region.

In this paper, a novel method called offset modulation (OM-OFDM) is proposed, to control the peak-to-average power ratio of an OFDM signal. The theoretical bandwidth occupancy of the proposed offset modulated signal is derived. Using this bandwidth occupancy a closed form theoretical bit error rate (BER) expression for an offset modulated transmission is further derived, validated and compared to OFDM. Thereafter, the authors present a newly applied decision metric, which is used to compare the offset modulated method to the traditional OFDM transmission. After studying these results, OM-OFDM is shown to offer between 1.23 dB - 1.65 dB net performance gain at a BER of 10-4when compared to an OFDM transmission in frequency selective fading and additive white Gaussian noise channel conditions respectively.

In this paper, a reduced complexity soft-input soft-output MIMO detector is presented. The detector is intended to be used in conjunction with an error correction code. The detector combines a Sphere Decoder with a Hopfield network to calculate a max-log-map approximation. It is then combined with the error correction code in an iterative structure (turbo). The code used is a quasi-cyclic non-binary LDPC code. The simulation results demonstrate that with less computational complexity, the proposed system's performance equals that of an optimal sphere decoder based detector.

This paper introduces a STFC-LDPC (space-time- frequency code low density parity check) concatenated coding scheme for high performance MIMO (multiple input, multiple output) systems. The coding design and performance evaluation are illustrated for various frequency-selective block fading conditions and compared with a STC-LDPC (space-time code LDPC) concatenated scheme over quasi-static fading conditions. The codes were analyzed and compared in space-selective conditions using various PDPs (power-delay profiles) to illustrate the performance gain of this newly proposed coding technique. The proposed STFC-LDPC showed a performance increase compared to the STC-LDPC scheme due to the inherent higher diversity of the STFC and is a suitable coding technique for high performance MIMO systems.

In this paper, we propose a downlink multi-antenna multi-carrier direct-sequence code division multiple access (MC-DS-CDMA) transmission scheme, incorporating space-time-frequency block (STFB) codes. The proposed scheme achieves time, frequency and spatial diversity using non-redundant algebraic constellation precoders. We simulate the proposed system's bit-error-rate (BER) performance in a statistically accurate multiple-input multiple-output (MIMO) channel model, capable of simulating spatially-selective, frequency-selective and time-selective channel conditions. We evaluate the system's performance when antenna correlation is introduced. Simulation results reveal that broadband MC-DS-CDMA using STFB codes, constitutes a promising downlink transmission scheme.

Little results exist in the literature regarding modelling the 'real world' industrial MIMO channel. This study models a dual-polarized MIMO channel using the Kronecker and Weichselberger models for two different indoor environments. The author's contribution here is based on a measurement campaign employing dual-polarized 8times8 patch arrays, and is analyzed at 2.4 and 5.2 GHz, where measurements were performed in an office building and an industrial environment. The average capacities attained from the measured data is compared with that of the models. Although the Weichselberger model performs better than the Kronecker model in both environments and at both frequencies, both models are found to underestimate the measured data. The Kronecker model is known to perform poorly for large antenna sizes and the performance of the Weichselberger model can be attributed to certain parts of the channel not fading enough.

It is becoming ever more difficult to clarify the right technology alternatives because the number of technologies is increasing and technologies are becoming very complex. Technology selection describes the process of making a choice between a number of alternatives. This paper proposes a new technology selection model which can be added to the already existing pool of models, with particular reference to South Africa. To test the model, two case studies were conducted. The results obtained from the study are also discussed in this paper.

This paper proposes a new noise-robust lossless compression algorithm, for binary memoryless sources, based on a decremental redundancy approach with Fountain codes. Both the whole binary entropy range compression performance and the noise-robustness of an existing incremental redundancy Fountain code compression technique are exceeded. An input source is coded with a new modified incremental degree LT-code (Luby transform) and a decremental redundancy algorithm is used to compress the Fountain-coded source. The performance variations with a systematic precode, a constant input degree distribution and a low-complexity puncturing distribution are investigated as possible amendments to the LT-code bipartite graph, for the case of decremental redundancy. The improved compression and robustness against transmission errors with our novel incremental degree puncturing decremental redundancy algorithm over that of the existing incremental redundancy closed-loop iterative doping (LT-CLID) algorithm is shown.

This paper introduces a realistic MIMO platform which was created according to the IEEE 802.16e-2005 standard (WiMAX). The MIMO-WiMAX platform is computationally efficient and capable of simulating space-selective, time-selective and frequency selective fading conditions. The simulator can simulate any arbitrary power-delay profile without increasing the receivers sampling frequency. The simulator has the capability of simulating quasi-static fading conditions as well as blockfading conditions. The platform was used to simulate recent coding technologies which consists of Space-Time codes, Space-Frequency codes and Space-Time-Frequency codes. The results obtained illustrates the effects of simulating these modern coding techniques on a MIMO-WiMAX platform in realistic mobile environments which consists of a large number of multipath delays and correlation between antennas at the transmitter and receiver. The codes simulated form a set of newly defined codes which could be incorporated into the standard complementing the current single-symbol maximum likelihood codes to be used as a guideline in designing real world MIMO-WiMAX systems.

The performance of super-orthogonal space time trellis coded orthogonal frequency division multiplexing (SOTTC-OFDM) systems relies on accurate channel state information at the receiver for proper decoding. In this paper, we propose a channel estimation algorithm that takes advantage of the super-orthogonal properties of the super-orthogonal space time trellis code. The algorithm is tested via a Monte Carlo simulation using frequency selective quasi-static Rayleigh fading channel. The paper presents the frame error (FER) and bit error rate (BER) performance comparison of the SOSTTC-OFDM scheme for various channel delay spread when this algorithm is used to estimate the channel.

A study on the suitability of multiple-input multiple-output (MIMO) architectures employing either single- or dual-polarization antennas is presented, with the purpose of identifying not only which architecture provides better average capacity performance, but also which is more robust for avoiding low channel rank. It is expected that systems with high robustness will be required for MIMO in demanding industrial applications. A measurement campaign employing dual-polarized 8times8 patch arrays at 2.4 and 5.2 GHz is analyzed, where measurements were performed in two very distinct indoor scenarios: an office building and an industrial environment. For both environments the performance of three 4times4 subsystems (dual-polarized, vertical- polarized and horizontal-polarized) are compared in terms of the average capacities attained by these systems and their eigenvalue distributions. Average capacities are found to be only marginally different, indicating little advantage of dual-polarized elements for average performance. However, an eigenvalue analysis indicates that the dual-polarized system is most robust for full-rank MIMO communications, by providing orthogonal channels with more equal gain.

The ability of correlation tensor and directional modeling strategies to accurately capture the spatial behavior of multiple-input multiple-output (MIMO) channels is investigated. Correlation tensor methods are based on the reduced order approximations of the full channel covariance, and do not require any a priori knowledge about the physical scattering mechanism or antenna array. Directional methods require knowledge about the array configuration and usually assume a double-directional wave-propagation mechanism. Five different tensor methods (Kronecker, separable maximum entropy, Weichselberger, principal hyperplane, and sparse core tensor) and one directional method (unstructured diffuse spectrum estimation) are compared in terms of the number of parameters required and the match to the true full covariance matrix. Simulations with a realistic cluster channel model indicate that tensor methods suffer from poor accuracy when too few values in the core tensor are retained, especially when no joint transmit/receive information is available. The directional method, on the other hand, has good accuracy with the same number of parameters as the simplest tensor model. The results stress the importance of including joint transmit/receive information in MIMO models and suggest that directional modeling is a logical choice for high accuracy modeling with few parameters.

A method for determining the maximum entropy (ME) estimate of the full joint transmit/receive MIMO channel covariance matrix is presented, where only knowledge of the separate transmit and receive covariances is assumed. The resulting ME covariance is similar to, but distinct from, the Kronecker modelling strategy, warranting a direct comparison of the two techniques. Application of the method to wideband indoor multiple-input multiple-output (MIMO) measurements taken at 2.4 GHz indicates a slight improvement in the modelling accuracy of the dominant covariance eigenvalues compared to the Kronecker model. On the other hand, the double-directional joint transmit/receive Bartlett spectra are nearly identical, indicating that the ME technique does not significantly improve modelling of the channel multipath structure. This fact suggests that accurate multipath modelling for MIMO systems may not be possible without detailed joint transmit/receive channel knowledge.

A cost-effective, experimental MIMO channel sounder is presented, capable of performing measurements with up to eight transmit and eight receive elements over 80 MHz of system bandwidth. The system provides the wideband channel transfer matrix for both indoor and outdoor wireless communication scenarios, allowing direct measurement of key MIMO system parameters. The system is described, fundamental hardware building blocks are outlined, and post-processing data algorithms are presented. Initial investigations on the effect of frequency scaling and array configuration at 5.2 GHz and 2.4 GHz are also presented.