Mohd Alaudin Mohd Ali

Hybrid SDF-HDF Cluster-Based Fusion Scheme for Cooperative Spectrum Sensing in Cognitive Radio Networks

In cognitive radio networks, cooperative spectrum sensing schemes are proposed to improve the performance of detecting licensees by secondary users. Commonly, the cooperative sensing can be realized by means of hard decision fusion (HDF) or soft decision fusion (SDF) schemes. The SDF schemes are superior to the HDF ones in terms of the detection performance whereas the HDF schemes are outperforming the SDF ones when the traffic overhead is taken into account. In this paper, a hybrid SFD-HDF cluster-based approach is developed to jointly exploit the advantages of SFD and HDF schemes. Different SDF schemes have been proposed and compared within a given cluster whereas the OR-rule base HDF scheme is applied to combine the decisions reported by cluster headers to a common receiver or base station. The computer simulations show promising results as the performance of the proposed scenario of hybridizing soft and hard fusion schemes is significantly outperforming other different combinations of conventional SDF and HDF schemes while it noticeably reduces the network traffic overhead.

Chemical field-effect transistor with constant-voltage constant-current drain-source readout circuit

Response of Chemical Field-Effect Transistor (CHEMFET) electrochemical sensors are taken from the output of a readout interface circuit that maintains constant drain-source voltage and current levels. We employ the readout circuit for the purpose of supervised learning training data collection. Sample solutions are prepared by keeping the main ion concentration constant while the activity of an interfering ion varied based on the fixed interference method. Results show that the voltage response demonstrates linear relationship to the ion concentration within the detection limit. However, noise in the form of abrupt and random changes in DC levels reduces correlation and increases mean square error between similarly repeated measurements. We find that referencing the voltage response to the sensor response in DIW prior to measurement greatly improves the repeatability. The process of approximating ionic concentration level is achieved up to 80% recognition by feeding the readout circuit output to a neural network post-processing stage.

Genetic algorithm-assisted soft fusion-based linear cooperative spectrum sensing

In this letter, a cooperative spectrum sensing scheme for cognitive radio networks based on genetic algorithm (GA) is proposed. The proposed scheme lies in optimizing the weighting coefficients vector used in a linear soft-decision fusion- (SDF-) based framework. The detection performance of the proposed GA-assisted SDF-based scheme is compared with the conventional SDF schemes as well as the OR-logic hard-decision fusion (HDF) scheme. The simulation results show that the proposed scheme outperforms all other schemes and can achieve higher probability of detection given the same probability of false alarm.

Pragmatic trellis coded modulation for adaptive multi-objective genetic algorithm-based cognitive radio systems

The cognitive radio (CR) has recently been proposed as one of the main opportunistic spectrum access solutions for dynamic spectrum allocation. In the CR context, an adaptive wireless node changes its transmission parameters to communicate efficiently and avoid interference with licensed users. In this paper, an adaptive pragmatic trellis coded modulation (TCM) scheme is proposed for CR systems. The proposed scheme employs the family of pragmatic TCM to combine different coded modulation modes to serve the CR node adaption while reducing the hardware complexity. The tradeoff of the error-throughput-power consumption performance is first analyzed using the so-called Pareto front concept. Then, the pragmatic TCM modes are implemented into a multi-objective genetic algorithm (MOGA) based CR decision engine. It is demonstrated that the proposed system is able to evolve to an appropriate transmission mode based on the wireless environmental conditions.

Chemical Field Effect Transistor Response with Post Processing Supervised Neural Network

This work presents the classification of potassium ion concentration in the presence of interfering ammonium ions from Chemical Field-Effect Transistor (CHEMFET) sensors involving neural network post-processing stage. Data collection for the purpose of supervised learning training data is obtained from sample solutions prepared by keeping the main ion concentration constant while the activity of the interfering ions based on the fixed interference method. The measurement setup includes a readout interface circuit that ensures constant-current constant-voltage across the drain-source for isothermal point operation. The training algorithm is back-propagation with generalized delta rule on a multilayer feed-forward network. Activation function based on the MOSFET drain current equation in the linear region is attempted in the hidden layer. Using function fitting approach, the network aims to find the potassium ion concentration despite the presence of interfering ion, without having to estimate device and chemically related parameters that would otherwise require further experiments.

Reliability-resources tradeoffs in cluster-based cooperative spectrum sensing

Cognitive radio (CR) networks are attracting massive attention as means to alleviate spectrum scarcity. In order for the CR users to opportunistically utilize the spectrum in absence of primary users (PUs), a reliable cooperative spectrum sensing scheme should be employed to assure service quality for the PUs as well as spectrum accessibility for the CRs. In this paper, a hybrid soft-hard decision fusion scheme of cooperative sensing is proposed. The hybridization of soft-hard fusion is realized by means of two-stage hierarchical clustering to jointly exploit the advantages of soft and hard fusion schemes. Through computer simulations, the proposed cooperation scheme is proven to have promising overall balanced-performance. However, it has been also shown that the overall performance of the CR network would be very much dependent on some existing tradeoffs between detection reliability and radio resources.

CHEMFET Reponse for Supervised Learning of Neural Network

Electrical response from Chemical Field-Effect Transistor (CHEMFET) sensors intended to be selective to a specific ion is influenced by interfering chemical ions present in the solution. To be able to detect the main chemical ion of interest, we include a neural network post-processing stage after a readout interface circuit. This work focuses on the training data collection of potassium sensors in the presence of ammonium ions intended for the supervised learning of the neural network module. Using function fitting approach, the network aims to find the potassium ion concentration. Training data is obtained from sample solutions prepared by keeping the main ion concentration constant while the activity of the interfering ions based on the fixed interference method. The training algorithm is back-propagation with generalized delta rule on a multilayer feed-forward network. Activation function based on the MOSFET drain current equation in the linear region is attempted in the hidden layer. We find that referencing voltage readings to sensor response in deionized water prior to measurement improves repeatability of measured training data.

Ion-selective field-effect transistor sensor response for neural network supervised learning

Ion-selective field transistors (ISFETs) are electrochemical sensors that can detect ion activities but have low selectivity issues for mixed-ion environments. This paper presents the data extraction and investigation of K+ ISFET sensors for neural network as post-processing stage. The environment for the sensor application is potassium and calcium mixed ions that is represented by solutions prepared based on the fixed interference method. The device measurement approach used is MOSFET semiconductor characterization technique, with further extracted data from the transfer characteristics subjected to statistical analysis. Results show that interfering ions influence the sensitivity graph of sensors detecting the main ion by shifting the gradient by 17%. Mean value of voltage response across the interfering ion range results in shifts up to 60 mV. Analysis of variance test gives a small -value indicating noticeable mean value of voltage response relating to change of main ion activity despite a large error variance possibly from the interfering ionic activity purposely added to the solutions. Extracted data from the solutions is then subjected to neural network pattern recognition by supervised learning method giving 73.7% correct recognition.

Ion-sensitive field-effect transistor selectivity with back-propagation neural network

The ion-sensitive field-effect transistor (ISFET) produces voltage signals, in a similar manner to the metal-oxide field-effect transistor, sensitive to ionic concentration change. When immersed in ionic solution with mixed ions of similar chemical characteristics, ISFETs respond with deceptive voltage signals due to the interfering ion contribution over the main ion of interest. In this paper, we applied back-propagation neural network to data acquired from titration of potassium ion (K+) and ammonium ion (NH4+). The role of the post-processing is to extract main ionic concentration level in the presence of an interfering ion. Primary data from measured observations with actual device variation and background ion was fed to a feedforward multilayer perceptron trained with several methods of back-propagation. Results show that neural network trained with backpropagation algorithm is able to improve concentration information by gives 15% improvement with 4 sensor array compared to direct estimation without post-processing. Additionally, averaging from multiple classifiers is shown to give a further 5% improvement on the regression factor between output and targeted values.

Multiple feedforward classifiers by bagging for ion-sensitive field effect transistor sensor response

This paper presents the classification of ionic concentration using ion-sensitive field-effect transistor (ISFET) sensors with post-processing neural network ensemble. ISFETs are electrochemical potentiometric sensors that produce voltage response indicative of ionic concentration change. However, in the presence of ions of similar charge, the voltage levels tend to be influenced by the interfering ions. The training dataset is based on actual measured data and are collected from dc response of sensors from titration. The multiple classifier system consists of feedforward multilayer perceptron weak learners constructed by bagging algorithm. Diversity is achieved by randomness of free parameters and resampling techniques of datasets by bootstrapping. Results demonstrate that multiple classifier system improves the ionic concentration classification of the main ion by additional 5% as compared to the average of the individual classifier performance.