Genetic evolutionary algorithm for optimal allocation of wavelength converters in WDM optical networks

In this article, a genetic evolutionary algorithm is proposed for efficient allocation of wavelength converters in WDM optical networks. Since wavelength converters are expensive, it is desirable that each node in WDM optical networks uses a minimum number of wavelength converters to achieve a near-ideal performance. The searching capability of genetic evolutionary algorithm has been exploited for this purpose. The distinguished feature of the proposed approach lies in handling the conflicting circumstances during allocation of wavelength converters considering various practical aspects (e.g., spatial problem, connectivity of a node with other nodes) rather than arbitrarily to possibly improve the overall blocking performance of WDM optical networks. The proposed algorithm is compared with a previous approach to establish its effectiveness and the results demonstrate the ability of the proposed algorithm to efficiently solve the problem of Optimal Wavelength Converters Allocation (OWCA) in practical WDM optical networks.

Uplink Multi-service Capacity and Interference Statistics of City Streets WCDMA Cross-shaped Microcells with Imperfect Power Control and Finite Transmitted Power

The multi-service (combination of more than one service) capacity and interference statistics (mean and variance) of the sectors of the street cross-shaped Wide Band Code Division Access (WCDMA) microcell are studied. A model of 21 cross-shaped microcells is used to analyze the uplink. The two slope propagation model is used to analyze the uplink. The capacity and interference statistics of the sectors are studied for different sector ranges, side lobe levels and power control (PC) standard deviation errors. It is noticed that, increasing the sector range will increase the sector capacity, reducing the side lobe level will also increase the sector capacity. Also it is noticed that, increasing the propagation exponent s ₂ will increase the sector capacity, reducing the standard deviations of the shadowing parameters σ₁ and σ₂ will increase the sector capacity and that reducing the power control standard deviation error σ_c will increase the sector capacity.

Partheno genetic algorithm for dynamic multi-services restoration in WDM networks

Dynamic restoration algorithms which support real-time and multi-services recovery are significant for the survivability of WDM (wavelength division multiplexed) networks. In this article, an intelligent dynamic restoration algorithm for multi-services in WDM networks based on the partheno genetic algorithm is proposed. In these networks, partial wavelength conversion is used. The algorithm is implemented within an interconnected multilayer-graph model and two kinds of optical networks matrix models. Compared with the basic restoration scheme, the proposed algorithm can make use of available network state information and can restore the affected multi-services fast and parallel. Simulation showed that the proposed algorithm can improve the restoration efficiency under high loads and reduce the service disruption ratio on the basis of fully utilizing resources of the network.

Modeling Channel Access Delay and Jitter of IEEE 802.11 DCF

With the ever-increasing requirement of WLAN to support real-time services, it is becoming important to study the delay properties of WLAN protocols. This paper constructs a new model to analyze the channel access delay and delay jitter of IEEE 802.11 DCF in saturation traffic condition. Based on this analytical model, average channel access delay and delay jitter are derived for both basic access and RTS/CTS-based access scheme. The accuracy of the analytical model is validated by simulations and furthermore we discuss the impact of initial contention window, maximal backoff stage, and packet size on channel access delay and delay jitter of 802.11 DCF using the proposed model.

A traffic grooming problem of SONET-WDM rings

In this paper, we present a traffic grooming problem of the SONET-WDM ring. The objective is to minimize the total cost of optical add-drop multiplexers (OADMs) and inter-ring hub equipment, while satisfying intra-ring and inter-ring capacities. We develop integer programming (IP) formulations for the problem and devise some reformulations for enhancing the mathematical representation of the proposed IP model. By investigating the polyhedral structure of the problem, we develop some valid inequalities that provide a tight lower bound for the problem. Dealing with the inherent computational complexity of the problem, we also devise an effective tabu search procedure for finding a feasible solution of good quality within reasonable computation time. Computational results are provided to demonstrate the relative strength of the proposed formulations, and to reveal the efficacy of the lower and upper bound procedures for solving the problem.

Accurate and Scalable Simulation of Network of Heterogeneous Sensor Devices

Simulation is an important tool to study and analyze sensor networks. Prior work in sensor network simulation focuses on homogeneous devices. In this paper, we present a system that performs scalable and accurate simulation of a network of heterogeneous sensor devices, including both Stargate intermediate level devices and mote devices. We study accuracy, performance, and scalability of our system. The results show that we can achieve accurate functional behavior for both standalone Stargate simulation and ensemble simulation of a Stargate and motes. For motes, we have less than 4.06% cycle count error for all benchmarks and for Stargate, we have less than 10% error for most benchmarks, and less than 12.5% error for all benchmarks. We also achieve less than 3.6% error for all benchmarks when simulating an ensemble of Stargate and motes. Our system is also more scalable than prior work. We can simulate 160 sensor nodes in real time speed and 2,048 sensor nodes with ten times slowdown on a 16-node cluster.

Enhanced efficiency and frequency assignment by optimizing the base stations location in a mobile radio network

Most mobile radio networks have been planned based on the classical cellular concept. However, alternative planning strategies that lead to more efficient network configurations are necessary due to the fact that the traffic density is generally far from constant throughout the service area, making necessary the relocation of base stations inside the traffic hotspots. If the traffic is characterized in a discrete way, the optimization of base stations location resembles vector quantization, a well-known problem in signal processing. In this paper, we use this analogy to propose a mobile radio network planning algorithm. Simulation results show that higher trunking efficiency as well as improved frequency assignment can be obtained if an existing mobile radio network is redesigned using the presented strategy. Raúl Chávez-Santiago was born in Oaxaca City, Mexico. He obtained the B.Sc. degree in communications and electronics engineering in 1997 from the School of Mechanical and Electrical Engineering (ESIME-IPN), and the M.Sc. degree in electrical engineering in 2001 from the Center of Research and Advanced Studies of the National Polytechnic Institute (CINVESTAV-IPN) in Mexico City. He has professional experience as Computer Networking Engineer, and Research and Teaching Assistant. Currently, he is completing his studies toward the Ph.D. degree at Ben-Gurion University (BGU) of the Negev, Israel. In 2002, he received the URSI Young Scientist Award. His main research interests are the optimal planning of radio communication networks, the electromagnetic compatibility of terrestrial and satellite radio systems, and the efficient use of the radio spectrum. He is an IEEE and IEICE student member, and a URSI radio scientist. Avi Raymond was born in Yavne, Israel. He received the B.Sc. and M.Sc. degrees in electrical engineering from Ben-Gurion University (BGU) of the Negev in 2000 and 2005, respectively. He worked for two years in a communication company in the field of Telephony and Broadband Services. He also worked as a Research Assistant in the communications laboratory at BGU. He currently works at Elta Electronic Industries Ltd. as System Engineer and pursues studies toward a second M.Sc. degree in systems engineering at the Technion Institute of Technology, Haifa, Israel. His research interests are the optimization algorithms for frequency assignment in cellular networks. Vladimir Lyandres was born in 1944, in Vologda, Russia. He received the M.Sc. degree in communications engineering in 1966 and the Ph.D. degree in communications theory in 1972 from the State University of Telecommunications (SUT), Saint Petersburg, Russia. He was with SUT until 1990 holding a position of Senior Scientific Associate and working on research and development of digital transmission systems, modeling of radio communication channels and algorithms of frequency planning for broadcasting and cellular systems. Since 1991 he holds a position of Researcher at the communications laboratory and Professor at the Department of Electrical and Computer Engineering, Ben-Gurion University (BGU) of the Negev, Israel. His research interests include synthesis of Markov models, combinatorial optimization, and adaptive power loading. He is a senior member of IEEE and member of IEICE.     

A fast-lock delay-locked loop architecture with improved precharged PFD

In this paper, a dual charge pump architecture for fast-lock delay-locked loop (DLL) is proposed and analyzed. The proposed fast lock DLL uses only one phase frequency detector (PFD) to perform dual path tuning and a lock control circuit to control the switching between the two path. An improved PFD is proposed to reduce the output jitter by reducing the one-shot pulse produced when the reference signal and feedback signal is in phase. The proposed DLL circuit is designed based on the Silterra 0.18-μm 1P6M CMOS process with a 1.8-V supply voltage. The active area of the proposed DLL circuit is 327.46 × 116.16 μm. An experimental chip was implemented and measured. The measurement results show that the proposed DLL has fast locking properties.     

Polarity Analysis of Self-Seeded Aluminum Nitride Crystals Grown by Sublimation

Self-seeded aluminum nitride (AlN) crystals were successfully grown in a tungsten crucible by the sublimation method. The polarities along the growth direction of these AlN samples were characterized by chemical etching, combined with high-resolution transmission electronic microscopy (HRTEM) and Raman spectroscopy. It has been proven by our experimental results that etching in KOH solution and in molten KOH-NaOH can be a reliable method to determine the surfaces with different polarities. Chemically active N-polar (0001) AlN surfaces were removed appreciably, while only etch pits related to threading dislocations were observed on Al-polar surfaces after chemical etching. HRTEM images revealed that there are noticeable inversion domain boundaries in the samples. Raman spectrum measurements showed the vibration mode change before and after etching in the same geometrical arrangement.     

高性能主轴变频器的设计和实现

本文详细介绍了一种基于DSP的全数字主轴电机变频器的设计方案，并且给出了～系列的优化策略以提高系统性能。为了保证主轴电机响应的快速性和运转的平稳性，系统采用了间接矢量控制策略，考虑到该控制策略对于电机参数有一定的依赖性，因此采用离线辨识的方法来获取控制所需的电机参孰为了便于全数字化系统的实现，以及提高母线电压的利用率，系统采用了改进的空间矢量脉宽调制策略；为了提高主轴电机在弱磁区的带载能力，以及延长主轴电机恒功率运转的区间，系统采用了电压最大化利用的优化弱磁控制策略。最后，给出了依据本方案设计的主轴变频器的试验结果。结果证明了本方案是可行的，达到了预期目标。