1 - 4 of 4
ItemParticle swarm optimization techniques for solving numerical and engineeing optimization problems(University of Hyderabad, 2011-11-04)The success of technology is mostly dependent on how well the real world appli- cations or problems are formulated, controlled and optimized. The complexities associated with real world problems are increasing day by day. The real world problems are often characterized by noisy, incomplete data or multimodality due to their inflexible construction. This demands a robust and efficient optimization or computational paradigm. Since the conventional optimization algorithms do not provide good solutions while optimizing nondifferentiable, nonseparable, dis- continuous, discrete and multimodal problems, nature inspired algorithms are the most sorted out paradigms for handling such real world problems. The nature inspired algorithms have evolved over decades and often contain many simple in- dividuals which when work together, produce complex emergent behavior and can be used to solve complex optimization problems. Among many nature inspired algorithms, the Swarm Intelligence (SI) is widely used over decades. SI is an in- novative, distributed and intelligent paradigm for solving optimization problems, developed from the inspiration of biological phenomena such as swarming, flocking and herding of different entities. One of the very simple SI based computational algorithm is Particle Swarm Optimization (PSO). It is a decade old concept in the optimization domain introduced in 1995 by Kennedy and Eberhart. Being a stochastic algorithm it exhibits many similarities with the other evolutionary algorithms. PSO essentially imitates the food foraging behavior of swarm of birds or school of fish. The main source of swarm’s search capability is the interaction among the individuals and the reaction to others experience in reaching the goal. Almost all the real world optimization problems can be modeled as any of the four types of optimization problems; i) Single Objective Unconstrained, ii) Single Objective Constrained, iii) Multi Objective Unconstrained and iv) Multi Objective Constrained problem. Though many variants of PSO are being developed to solve these kind of problems, it still suffers with major problems like premature
ItemStudies on spectrum sensing algorithms and real-time implementation for cognitive radio(University of Hyderabad, 2018-12-01)In the recent past Cognitive Radio (CR) technology has received increased at- tention to solve the spectrum scarcity problem using the opportunistic spectrum re-usage technique. It allows secondary users to use the unlicensed channels oppor- tunistically without causing interference to the licensed users. It involves mainly two functionalities namely spectrum sensing and spectrum management. Spec- trum sensing identifies vacant spectrum bands, i.e., spectrum holes for the op- portunistic use of secondary users in the network. Robust spectrum sensing is essential to avoid interference to existing ’licensed’ users and maximize the spec- trum utilization. There exist mainly two class of sensing algorithms: data aided and blind. The data aided algorithm requires the characteristics of the signal a pri- ori for successful detection. However, in the real-time radio environment, it is not always possible to know the signal characteristics a priori. Hence blind algorithms are essential for reliable spectrum sensing in the real-time environment. In the sensing domain, although energy detection is popularly being used due to its low computational complexity, its performance suffers due to noise uncer- tainty. Although it is a blind technique, but it requires information about the noise variance. The estimation accuracy of received signal noise variance affects the de- tection performance. This thesis proposes an improved energy detection technique where the threshold is adjusted with respect to the noise variance. The noise vari- ance is estimated using the Linear predictor method, i.e., the Burg method. This thesis performs a detailed performance analysis of the energy detection algorithm with noise variance estimation in the case of a single node and multi-node sensing under white and colored noise characteristics of the channel. Since the accuracy of sensing depends on the accuracy of the noise variance estimator, a detailed study of different types of noise estimator and its impact on sensing accuracy is also carried out.
ItemDevelopment of FPGA based coprocessors for signal processing applications(University of Hyderabad, 2013-07-01)Now-a-days there has been an increase in demand for designing recon- gurable embedded systems in signal processing, multimedia and evolutionary computation applications. Embedded processors alone, cannot achieve the desired computational capability to ful ll the requirements of massive parallelism, higher memory bandwidth, higher execution speed to execute these applications. In order to meet these requirements, Field Programmable Gate Arrays (FPGAs) are used by exploiting the recon gurable resources. Beyond their well-known exibility, FPGAs o er the versatility of running software applications on embedded processors and at the same time taking the advantage of available recon gurable resources, all on same package. FPGA based System on Chip (SoC) design solution replaces traditional System on Board (SoB) design concept and is often referred as Programmable SoC (PSoC). This platform consists of hard/soft embedded processors, external memory and custom Intellectual Properties (IPs). These IPs are used to accelerate the computational task of an algorithm. This involves developing dedicated IP and its integration in SoC platform. There are mainly two types of IP interfacing techniques, i.e., Slave Unit (SU) and Auxiliary Processing Unit (APU). The SU interface has Register/First-In-First-Out (FIFO) connected to the processor through shared system bus (Processor Local Bus (PLB)). Although this interface is simpler in design, the main bottleneck is bus arbitration, which lowers the total execution speed. The other bus interface is APU (only for PowerPC440), which can be directly connected to the custom IP through a dedicated Fabric coprocessor Bus (FCB). This interface has no communication overhead and allows quick synchronization between the processor and IP. Custom IPs have been developed and integrated using APU interface for maximizing the portability and modularity.
ItemStudies on development of differential evolution based spectrum allocation algorithms and field programmable gate array omplementation for cognitive radio networks(University of Hyderabad, 2015-11-10)Recent trends in wireless communication technologies claim a rapid in- crease in demand of radio spectrum. In the current spectrum allocation scheme, it is difficult to accommodate the demand of radio spectrum. Moreover the desig- nated spectrum are not efficiently exploited, resulting its poor utilization. Stud- ies have demonstrated that reuse of the un-utilized spectrum provides a signifi- cant improvement in network capacity. Recently, a new dynamic spectrum access paradigm called Cognitive Radio (CR) has gained popularity to solve the short- comings of spectrum under-utilization and spectrum scarcity. In CR technology, unlicensed users (secondary users) make use of the unused spectrum of licensed users (primary users), thereby discovering a new capacity and commercial value from the existing unused spectrum. The main functions of the CR are spec- trum sensing, spectrum management, spectrum mobility and spectrum sharing. Spectrum sensing deals with the detection of vacant spectrum bands known as spectrum holes and these detected holes are assigned to the secondary users (SUs) during spectrum management phase. It uses different spectrum allocation (SA) algorithms for allocating spectrum to SUs. The present thesis mainly concen- trates on spectrum allocation phase. The objectives of SA phase are a) maximize the spectrum utilization, b) minimize interference to primary users (PUs) and neighbor secondary users and c) maintain fairness across the users. To achieve these goals, an efficient SA technique is required for making deci- sions within a stipulated time. For this purpose, various techniques like graph col- oring, game theory, evolutionary algorithms, local bargaining, auction and pricing mechanisms and stochastic search methods have been reported in the literature. The problem of allocating channels amongst the secondary users in the network is considered as a NP-hard problem. In this work, evolutionary algorithms, namely Differential Evolution (DE), firefly and particle swarm intelligence are applied to find an efficient channel assignment solution. Further, the performance of three algorithms in terms of quality of solution and time complexity are compared to find the best solution