Dobri Atanassov Batovski

• Ph.D.
• Lecturer at Assumption University

The advances in communications technology depend on the development of tiny devices being capable of transmitting and receiving data at the highest possible data rates. The digital signal processing at gigabit-per-second rates is soon to be upgraded to the terabit-per-second range. Nanotechnology provides a variety of alternatives for the design of...

The invariance of the hypercube to discrete rotation (cyclic shift) is used to prove the existence of cyclic modular destination graphs (cyclic MDG, or CMDG) for optimal oblivious many-one packet routing for arbitrary dimensions. The proof is constructive, thus allowing a straightforward generation of said CMDGs. The optimality of the graphs is mea...

The performance evaluation of optical sparse codes in code-division multiple-access (CDMA) networks is considered for the case of multi-access interference (MAI). A comparison with the standard Gaussian approximation (SGA) shows that the said approximation overestimates the performance of the codes and should not be used for small bit error rates (...

The resultant traffic flows in a preferred direction within a set of cooperative gateway nodes is optimized taking into consideration the internal rate-delay products. The internal rates on directed links among adjacent gateway nodes are included in the set of control variables. It is also shown that the optimal internal rate-delay product of a set...

The controlled addition to the information-bearing data sequence of a pattern of a priori known symbols, referred to as a priori information, makes it possible to improve the error-correcting capabilities of the main coding scheme in use. The introduced redundancy makes it easier to eliminate burst errors and establish error-free data transmission....

This study introduces an efficient algorithm for local packet routing using two-hop splitting of packet flows within a narrow (one hop away) routing corridor along a global end-to-end routing path. The main concept of the proposed algorithm is based on an attempt for simultaneous transmission of same packet from a local relaying node to two (or mor...

The optimization of packet flows in a set of cooperative nodes with single output ports is considered. A single output port relays a packet to a single connected node at a time. The different service time distributions to distinct connected nodes are considered in terms of multiclass queuing with a single first-come first-serve queue and a single s...

The low-density parity-check (LDPC) codes have a relatively simple algorithm for error correction. The best performance of such codes is demonstrated for large frame sizes and a substantial number of iterations. The decoding delay depends on the number of iterations being performed with the received noisy data in order to obtain the best estimate t...

This contribution is concerned with the performance evaluation of topology control algorithms for indoor ad hoc networks in multi-storey buildings. Several proposed topology control algorithms based on discrete sets are compared. The topology control algorithms operate with wireless links having variable transmission rates according to existing wir...

The duals of finite impulse response filters constructed of binary matrix sequences are used to design an eight-channel transmultiplexer. The binary matrix sequences of analysis and synthesis filters are obtained from corresponding complementary sequences which satisfy the conditions for perfect reconstruction. The distortions of eight transmultipl...

The series of soft values of consecutive iterations of turbo decoding are used for convergence acceleration with the Epsilon algorithm. It is shown that a certain improvement of the bit error rate (BER) can be obtained compared to the standard turbo decoding. Turbo codes for third generation communication systems are used for the computational expe...

The overlapping of multihop paths in decentralized networks may cause congestion at certain nodes and underutilization of other neighbor nodes. The choice of a given path for the delivery of packets from source to destination makes other connected nodes, located one hop away from the said path, prospective candidates for local rerouting in case of...

The performance evaluation of large multihop networks in terms of the standard quality of service parameters such as latency, throughput, network utilization, etc., is often performed with the use of discrete-event simulation tools which would require substantial computational resources with the increase of the number of nodes in the network. An al...

It is shown that a significant part of the text of the paper entitled “Performance Evaluation of VDSL Network with Fuzzy Control Policing Mechanisms” by Lekcharoen and Jittawiriyanukoon (2005), published in Assumption University Journal of Technology 8(4):172-7, is constructed by directly copying texts from a set of selected papers.

The three-color ability of the binary de Bruijn network is used to establish conflict-free oblivious packet routing within the all-port model. Tight multi-plane routing is implemented to obtain the optimal number of planes for permutation routing for a number of dimensions from two to six.

The packet length in wireless ad hoc networks is a subject of conflicting requirements imposed by the data link layer and the network layer. The dynamically changing signal-to-noise (SNR) ratio in wireless channels requires adaptive selection of the short frame length and the transmission rate at the data link layer. Most frequently, maximum throug...

The oblivious edge-disjoint routing in the directed binary de Bruijn network is considered on the basis of the all-port model. Packet burst switching is used to establish tight permutation routing. Permutation circuit switching within the all-port model is demonstrated with the use of multi-wavelength routing. Tight multi-plane routing is establish...

The concept of multi-plane ad hoc networks is introduced for two emerging applications, namely, multi-plane three-dimensional indoor topologies and virtual multi-plane routing in dense networks. The connectivity between different planes is established with the use of distributed gateways consisting of a prescribed number of interconnected nodes on...

The exchange of information in networks deals with various types of routing such as one-one (permutation), many-one (or one-many), and many-many packet routing as well as circuit-switching. When using oblivious techniques, the problem statement is to find a conflict-free graph for a given network topology under various optimization constraints. Wit...

This paper considers the necessary conditions for the existence of oblivious circuit switching in multi-link binary hypercubes. It is also shown that 1(2,3,4,7)-link edge-disjoint, many-one routing exists in the 1(2,3,4,5)-cube. Modular destination graphs are decomposed for the said number of links, where the physical paths are obtained using quasi...

A study of the permutation capability of a binary hypercube under the commonly used dimension-order routing is presented. Two modes of the basic routing algorithm have been explored based on non-skipping or skipping identical bits in source and destination addresses when adjusting a route. All the study was done computationally. It has been found t...

The paper considers the existence of oblivious many-many packet routing in multi-link binary hypercubes. Modular destination graphs are decomposed according to a certain number of links for a particular dimension n of the hypercube in order to establish conflict-free tight minimum routing within a network cycle consisting of n hops. Quasi-dimension...

An idea for superposition of error patterns onto neural network configurations is presented. The perturbed neural scheme will produce different final responses simulating the work of real systems for which management control is needed. The management analysis in this case consists of collecting information about variety of situations, which can ari...

This paper considers data processing of time-resolved photoluminescence (PL) spectra obtained with two-dimensional streak camera. Effective numerical algorithms are implemented in an applied software product (program LumFit).

We report on time-resolved optical experiments performed resonantly onto the fundamental excitonic transition on several (Zn,Cd)Se/ZnSe quantum wells, for temperatures from 10 K up to room temperature. We show that the thermal quenching of the photoluminescence intensity is due to extrinsic non-radiative channels, whose activation energy is found t...

A general method for numerical solution of eigenfunction problems is presented. The method is based on differential T-transform technique and the acceleration properties of the Padé approximation. The advantages of the method consist of its quick-action and the possibility to find precisely the searching solution, especially in cases when the conve...

We report on time-integrated and time-resolved optical experiments performed on a 26 Å thick Zn0.85Cd0.15Se/ZnSe quantum well, for temperatures in the 10-200 K range. Excitation spectroscopy allows an estimation of the relative valence band offset, which is found to be 10%. From the temperature variation of the decay time of the photoluminescence p...

The thesis has been discussed and has been appointed for defense by the Departmental Council of the Department of Solid State Physics and Microelectronics, Faculty of Physics, Saint Kliment Ohridski University. The investigations presented in the thesis have been made in the aforementioned department, where Mr. Dobri Batovski has been studying as a...

A general method for processing deep level transient spectroscopy (DLTS) data is presented. It is shown that the theoretical model is a two-dimensional Fredholm equation of the first kind with a four-dimensional integral kernel. The method is based on the numerical solving of the inverse ill-posed problem of the theoretical model via regularization...

The photovoltage shows a very complicated behavior and could be considered as a new approach to the investigation of the excess carrier transfer in semi‐insulating (SI) semiconductors. Both time‐resolved short‐circuit (SC) photovoltage and photocurrent at the same energies in the range 1.13–1.49 eV for SI GaAs were measured. The peculiarities of th...

This paper studies theoretically the problem of a semiconductor in a two-dimensional (2D) constraining potential (space charge region for instance) and translational invariance in the third direction (an example of this would be a quantum wire or a gate). We propose an analytical solution to the Schrödinger equation of the electrons using the separ...

The photovltage developed along the plane of the translational invariance (longitudinal one) of the semiconductor layers due to the illumination was widely investigated in variety of GaAs/AlGaAs multi-quantum well structures. In the structures with perfect quality there should be no photovoltage, so an criterium of material quality can be derived f...

An experimental test is proposed to check the validity of the conventional theoretical model used to interpret deep level transient spectroscopy (DLTS) spectra. The test is based on the conventional DLTS technique with time points determined from a corresponding Diophantine equation. The test is simple, fast, and can be realized during each experim...

An algorithm for fast calculation of theoretical deep level photoluminescence (PL) lineshapes is presented. The algorithm can be used for effective comparison with experimental spectra in cases of bulk photoluminescence. The method was expanded and for the cases of deep level photoluminescence in quasi-two and quasi-one dimensional electron systems...

It is shown that a recently proposed comment on our paper [J. Appl. Phys. 74, 291 ( 1993)] cannot be considered as a significant simplification by obtaining the deep level energy spectrum D(E).

A new method for nonexponential deep level transient spectroscopy (DLTS) theoretical analysis is proposed. This method is based on the numerical solving of the inverse problem for the theoretical model via regularization algorithms. It is shown that it is possible to obtain the shape of the deep level energy spectrum from the noisy DLTS spectra in...