Bodhaswar Tikanath Jugpershad Maharaj

Multiview Deep Learning for Land-Use Classification

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.

A cost-effective wideband MIMO channel sounder and initial co-located 2.4 GHz and 5.2 GHz measurements

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.

A novel and secure IoT based cloud centric architecture to perform predictive analysis of users activities in sustainable health centres

Diabetes, blood pressure, heart, and kidney, some of the diseases common across the world, are termed ’silent killers’. More than 50 % of the world’s population are affected by these diseases. If suitable steps are not taken during the early stages then severe complications occur from these diseases. In the work proposed, we have discussed the manner in which the Internet-of-Things based Cloud centric architecture is used for predictive analysis of physical activities of the users in sustainable health centers. The architecture proposed is based on the embedded sensors of the equipment rather than using wearable sensors or Smartphone sensors to store the value of the basic health-related parameters. Cloud centric architecture is composed of a Cloud data center, Public cloud, Private cloud, and uses the XML Web services for secure and fast communication of information. The architecture proposed here is evaluated for its adoption, prediction analysis of physical activities, efficiency, and security. From the results obtained it can be seen that the overall response between the local database server and Cloud data center remains almost constant with the rise in the number of users. For prediction analysis, If the results collected in real time for the analysis of physical activities exceed any of the parameter limits of the defined threshold value then an alert is sent to the health care personnel. Security analysis also shows the effective encryption and decryption of information. The architecture presented is effective and reduces the proliferation of information. It is also suggested, that a person suffering from any of the diseases mentioned above can defer the onset of complications by doing regular physical activities.

Performance of a Hidden Markov channel occupancy model for cognitive radio

This paper investigates the effect that various training algorithms have on the performance of a primary user (PU) channel occupancy model for cognitive radio. The model assumes that PU channel occupancy can be described as a binary process. A two state Hidden Markov Model (HMM) has thus been chosen and it is shown that the performance of the model is influenced by the algorithm employed for training the model. Traditional training algorithms are compared to certain evolutionary based training algorithms in terms of the resulting prediction accuracy and convergence time achieved. The performance of the model is important since it provides secondary users (SU) with a basis upon which channel switching and future channel allocation may be performed. Further simulation results illustrate the positive effect that our model has on channel switching under both heavy and light traffic density conditions.

PAPR reduction in FBMC using an ACE-based linear programming optimization

This paper presents four novel techniques for peak-to-average power ratio (PAPR) reduction in filter bank multicarrier (FBMC) modulation systems. The approach extends on current PAPR reduction active constellation extension (ACE) methods, as used in orthogonal frequency division multiplexing (OFDM), to an FBMC implementation as the main contribution.The four techniques introduced can be split up into two: linear programming optimization ACE-based techniques and smart gradient-project (SGP) ACE techniques. The linear programming (LP)-based techniques compensate for the symbol overlaps by utilizing a frame-based approach and provide a theoretical upper bound on achievable performance for the overlapping ACE techniques. The overlapping ACE techniques on the other hand can handle symbol by symbol processing. Furthermore, as a result of FBMC properties, the proposed techniques do not require side information transmission. The PAPR performance of the techniques is shown to match, or in some cases improve, on current PAPR techniques for FBMC. Initial analysis of the computational complexity of the SGP techniques indicates that the complexity issues with PAPR reduction in FBMC implementations can be addressed.The out-of-band interference introduced by the techniques is investigated. As a result, it is shown that the interference can be compensated for, whilst still maintaining decent PAPR performance. Additional results are also provided by means of a study of the PAPR reduction of the proposed techniques at a fixed clipping probability. The bit error rate (BER) degradation is investigated to ensure that the trade-off in terms of BER degradation is not too severe.As illustrated by exhaustive simulations, the SGP ACE-based technique proposed are ideal candidates for practical implementation in systems employing the low-complexity polyphase implementation of FBMC modulators. The methods are shown to offer significant PAPR reduction and increase the feasibility of FBMC as a replacement modulation system for OFDM.

Smart Vehicle Navigation System Using Hidden Markov Model and RFID Technology

The road transport of dangerous goods has been the subject of research with increasing frequency in recent years. Global positioning system (GPS) based vehicle location devices are used to track vehicles in transit. However, this tracking technology suffers from inaccuracy and other limitations. In addition, real-time tracking of vehicles through areas shielded from GPS satellites is difficult. In this paper, the authors have addressed the implementation of a smart vehicle navigation system capable of using radio frequency identification based on information about navigation paths. For prediction of paths and accurate determination of navigation paths in advance, predictive algorithms have been used based on the hidden Markov model. At the core of the system there is an existing field programmable gate array board and hardware for collection of navigation data. A communication protocol and a database to store the driver’s habit data have been designed. From the experimental results obtained, an accurate navigation path prediction is consistently achieved by the system. In addition, once-off disturbances to the driver habits have been filtered out successfully.

PAPR reduction in FBMC systems using a smart gradient-project active constellation extension method

The filter bank multicarrier (FBMC) modulation scheme has recently seen renewed interest and is being considered as a viable alternative to orthogonal frequency division multiplexing (OFDM). FBMC, however suffers from the same high peak-to-average power ratio (PAPR) drawback as OFDM systems. Conventional OFDM PAPR reduction techniques cannot be directly applied to FBMC due to the overlapping nature of FBMC symbols. A novel technique is proposed based on an evolution of the smart-gradient project active constellation extension (SGP-ACE) PAPR reduction method used for OFDM systems, namely the FBMC SGP-ACE method. The proposed method is applied to a set of contiguous FBMC symbols, thereby compensating for the overlapping nature of FBMC modulation. The proposed method requires less iterations than a projection-onto-convex-sets (POCS) ACE approach to converge to a lower PAPR and significant reduction in complexity can be achieved as opposed to current FBMC PAPR reduction techniques which tend to require the addition of advanced signal processing. The proposed FBMC SGP-ACE method outperforms a conventional FBMC POCS-ACE method by 2.6dB in PAPR reduction at a clip probability of 10-4 on the 1st iteration.

Accurate MIMO Channel Modeling: Correlation Tensor vs. Directional Approaches

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.

Diffusion-controlled interface kinetics-inclusive system-theoretic propagation models for molecular communication systems

Inspired by biological systems, molecular communication has been proposed as a new communication paradigm that uses biochemical signals to transfer information from one nano device to another over a short distance. The biochemical nature of the information transfer process implies that for molecular communication purposes, the development of molecular channel models should take into consideration diffusion phenomenon as well as the physical/biochemical kinetic possibilities of the process. The physical and biochemical kinetics arise at the interfaces between the diffusion channel and the transmitter/receiver units. These interfaces are herein termed molecular antennas. In this paper, we present the deterministic propagation model of the molecular communication between an immobilized nanotransmitter and nanoreceiver, where the emission and reception kinetics are taken into consideration. Specifically, we derived closed-form system-theoretic models and expressions for configurations that represent different communication systems based on the type of molecular antennas used. The antennas considered are the nanopores at the transmitter and the surface receptor proteins/enzymes at the receiver. The developed models are simulated to show the influence of parameters such as the receiver radius, surface receptor protein/enzyme concentration, and various reaction rate constants. Results show that the effective receiver surface area and the rate constants are important to the system’s output performance. Assuming high rate of catalysis, the analysis of the frequency behavior of the developed propagation channels in the form of transfer functions shows significant difference introduce by the inclusion of the molecular antennas into the diffusion-only model. It is also shown that for t > > 0 and with the information molecules’ concentration greater than the Michaelis-Menten kinetic constant of the systems, the inclusion of surface receptors proteins and enzymes in the models makes the system act like a band-stop filter over an infinite frequency range.

A Technology Selection Framework for the Telecommunications Industry in Developing Countries

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.