W PerfectSoft łączymy inżynierię oprogramowania z rygorem analitycznym. Założyciel i główny inżynier firmy jest jednocześnie aktywnym naukowcem, prowadzącym badania w obszarze teleinformatyki oraz sztucznej inteligencji. Jego dorobek, publikowany w międzynarodowych wydawnictwach (m.in. IEEE, MDPI, Springer, Elsevier), stanowi otwarty, niezależny wkład w rozwój nauki i technologii.
To akademickie doświadczenie jest dla nas unikalną wartością dodaną. Wiedza ekspercka pozwala nam z szerszej perspektywy projektować dedykowane systemy IT dla biznesu. W realizowanych w PerfectSoft niezależnych projektach komercyjnych – od zaawansowanych systemów ERP po autorskie symulatory (NDE) i modele AI – do każdego wyzwania aplikujemy zaawansowaną metodykę badawczą.
Dwukierunkowa synergia kompetencji. Nasza przewaga technologiczna opiera się na ciągłym przenikaniu się dwóch światów. Z jednej strony, akademicki rygor badawczy i zaawansowane metody analityczne bezpośrednio podnoszą jakość codziennych praktyk inżynieryjnych w PerfectSoft. Z drugiej – twarde realia biznesowe, bezkompromisowe dążenie do skutecznego wdrożenia i weryfikacja przez wymagających klientów nadają badaniom naukowym założyciela unikalny, pragmatyczny wymiar. To wyjątkowe sprzężenie zwrotne pozwala nam projektować innowacyjne oprogramowanie oraz systemy B+R (w tym fazy Proof of Concept) całkowicie od podstaw, dając pełną gwarancję autorskiego kodu i wyłączności praw własności intelektualnej (IP).
Naukowa wizytówka kompetencji
Poniższe zestawienie prezentuje dorobek publikacyjny założyciela, realizowany we współpracy z niezależnymi badaczami i ośrodkami akademickimi. Udostępniamy je jako potwierdzenie najwyższych kompetencji analitycznych, na których opiera się jakość technologiczna PerfectSoft.
Aby pobrać formatkę cytowania (BibTeX, RIS, RTF), prosimy rozwinąć opcje pod wybranym artykułem.
Wildfire mitigation depends critically on minimizing Time-to-Detect (TTD). While lightweight convolutional neural networks (CNNs) enable visual detection of early-stage wildfire smoke, deploying them in practical, autonomous off-grid monitoring towers remains a system engineering challenge due to strict power and hardware constraints. This paper presents a proof-of-concept (PoC) implementation of a deterministic Edge-AI architecture designed for reliable wildfire surveillance. The proposed approach introduces an application-layer scheduler guaranteeing bounded system latency, as well as a hybrid approach connecting neural network output with heuristic spatio-temporal logic reducing false positives. Preliminary experimental results demonstrate that the proposed system achieves an F1-Score of 92.5% while maintaining stable latency for up to nine concurrent streams, proving that a multi-camera wildfire detection is achievable in practice, also on low-power edge hardware.
This article proposes an analytical model of a multiservice hierarchical system with multi-service overflow traffic. To model the primary and secondary resources of this system, the state-dependent queue service discipline was used. In order to model the secondary resources with the dedicated queue for overflow traffic, the Hayward’s approach was generalized and applied. To evaluate its accuracy, the results of analytical modeling were compared with the data obtained during the simulation experiments carried out in the study. Both the data presented in the article and the results obtained by the present authors in numerous comparative studies clearly indicate that the proposed model makes it possible to evaluate the values of the blocking probability with the accuracy that provides its reliable practical application at the stage of network dimensioning. The overflow mechanism has particular significance in networks with limited resources, such as mobile networks.
In this article, we present an analysis of the accuracy level of methods for modeling the multi-service overflow systems that service Erlang, Engset, and Pascal traffic. In systems with traffic overflow, new calls that cannot be serviced by the primary resources are overflown (directed) to other available resources that can service a given call, that is, to the secondary resources (alternative resources). In the article, we focus on studying the influence of methods for determining the parameters of traffic that overflows to the secondary resources on the accuracy of determining the traffic characteristics of overflow systems. Our analysis revealed that the main source of the inaccuracy of the existing methods is their approach to determining both the average value and the variance of multi-service Pascal traffic streams offered to the secondary resources. Therefore, we proposed a new method for determining the parameters of Pascal overflow traffic. The method is based on the decomposition of multi-service primary resources into single-service resources and the subsequent conversion of Engset and Pascal streams into equivalents of Erlang traffic. The results of the analytical calculations obtained on the basis of the new method are then compared with the results of simulation experiments for a number of selected structures of overflow systems that service Erlang, Engset, and Pascal traffic. The results of the study indicate that the proposed theoretical model has a significantly higher accuracy than the models proposed in the literature. The method can be used in the analysis, dimensioning, and optimization of multi-service telecommunication systems composed of separated resources, for example, mobile cellular systems.
Abstract—In this paper, impact of changes in parameters of offered traffic on the accuracy of determining the parameters of overflow traffic in hierarchical systems with multi-service traffic was presented. Pascal type traffic streams were offered to the considered systems. The study investigated the impact of changes in the number of sources, intensity of traffic offered by individual classes, as well as changes in the traffic offered by a single free source. The presented results are based on determined relative errors of the values of overflow traffic obtained in simulations and on the basis of calculations.
This article proposes an analytical model of a multiservice hierarchical system with multi-service overflow traffic. To model the primary and secondary resources of this system, the state-dependent queue service discipline was used. In order to model the secondary resources with the dedicated queue for overflow traffic, the Hayward’s approach was generalized and applied. To evaluate its accuracy, the results of analytical modeling were compared with the data obtained during the simulation experiments carried out in the study. Both the data presented in the article and the results obtained by the present authors in numerous comparative studies clearly indicate that the proposed model makes it possible to evaluate the values of the blocking probability with the accuracy that provides its reliable practical application at the stage of network dimensioning. The overflow mechanism has particular significance in networks with limited resources, such as mobile networks.
Głąbowski, Mariusz, Kmiecik, Damian, Stasiak, Maciej
This article proposes a new method to determine the characteristics of multiservice overflow systems with queueing systems. A number of methods have been developed that have the advantage of determining the parameters of traffic directed to secondary resources as well as providing a way to model these resources. The accompanying assumption is that queues with limited capacities are used in primary resources. The results of analytical calculations are compared with the results of simulation experiments for a number of selected structures of overflow systems with queueing in primary resources. The results of the study confirm high accuracy of the proposed method and, in consequence, the accuracy of the theoretical assumptions adopted for the proposed method.
W artykule zaprezentowano rezultaty badań dotyczące wpływu zmian własności ruchu oferowanego na dokładność wyznaczania parametrów ruchu przelewanego w hierarchicznych systemach z ruchem wielousługowym. Rozważanym systemom oferowane były strumienie ruchu typu Pascala. Podczas badań określono wpływ zmian ilości źródeł, natężenia ruchu oferowanego przez poszczególne klasy, a także zmian natężenia ruchu oferowanego przez pojedyncze wolne źródło. Wnioski prezentowane w artykule zostały oparte na wyznaczonych błędach względnych wartości ruchu spływającego otrzymanego symulacyjnie oraz na podstawie obliczeń analitycznych.
In this paper, impact of changes in parameters of offered traffic on the accuracy of determining the parameters of overflow traffic in hierarchy systems with multiservice traffic was presented. Pascal type traffic streams were offered to the considered systems. The study investigated the impact of changes in the number of sources, intensity of traffic offered by individual classes, as well as changes in the traffic offered by single free source. The presented results are based on determined relative errors of the values of overflow traffic obtained in simulations and on the basis of calculations.
In most of the hitherto works devoted to modeling overflow systems, both single-service and multi-service, it has been assumed that streams of calls (flows, requests) offered to primary resources of networks with traffic overflow are consistent with Poisson distribution or binomial distribution [1]–[6]. However, the diversity of network services observed today as well as an increasing share of overflow traffic, in particular in multi-tier cellular networks, makes calls streams more and more often corresponding to the negative binomial distribution (Pascal). In order to enable analytical modeling of the networks with different multiservice traffic streams, the paper presents a simple method of determining the first two moments of the overflow traffic, which – offered to the primary resources of the network – is consistent with Pascal distribution.
Thearticle proposes a newmethod of determining traffic characteristics ofmultiservice overflow systems that carry adaptive traffic. When the total offered load in primary resources exceeds a certain value, this type of traffic is admitted for service with lower bitrate. A particular attention is given in the article to a method for a determination of the parameters of traffic that overflows to secondary resources as well as to the way adaptive traffic is serviced. The method takes into consideration three possible types of traffic: Erlang, Engset, and Pascal traffic. It is based on a generalization of Hayward’s concept and its application to model systems with adaptive traffic with threshold compression. The method can be used for optimal dimensioning of logical networks (slices) in modern mobile systems due to possibility of analytical determination of grade of service parameters (blocking probability, carried traffic, and network load). To verify the accuracy of the proposed model the results of analytical calculations, obtained on the basis of the proposedmodel, are then comparedwith the results of simulation experiments for a number of selected structures of overflow systems that service adaptive traffic. The results of the study demonstrate high accuracy of the proposed theoretical model.
Streszczenie: W artykule zaprezentowano rezultaty badań wpływu stosowania kolejkowania zgłoszeń w hierar-chicznych systemach z przelewami ruchu wielousługowego. Rozważane systemy składały się z kilku zasobów pierwot-nych oraz jednego zasobu wtórnego. Systemom tym ofero-wane były wielousługowe strumienie ruchu. W badaniach określono wpływ struktury ruchu, pojemności kolejek, maksymalnego czasu przebywania w kolejce (ograniczony oraz nieograniczony) oraz zasobu, w którym wykorzystano kolejkowanie (zasoby pierwotne, zasoby alternatywne). Wnioski prezentowane w artykule zostały oparte na wy-znaczonych prawdopodobieństwach blokady w systemach przelewowych. Abstract: In this paper, the performance evaluation of call queuing in hierarchical systems with multi-service overflow traffic was presented. The considered system model consisted of a few primary resources and a single secondary resource servicing multi-service traffic streams. The study investigated the impact of the traffic structure, queue capacity, maximum queueing time (limited or unlim-ited) and the placement of queuing resources (primary resources or secondary resources). The presented results are based on determined blocking probabilities in the sys-tems with overflow traffic.
Streszczenie: W pracy zweryfikowano mo˙zliwości zastosowania algorytmów splotowych do modelowania wielousłogowych systemów sieciowych z ró˙znymi, w tym niepoissonowskimi, strumieniami zgłosze ´ n. Badaniom podlegały strumienie zgłoszeń zarówno zale˙zne jak i niezale˙zne od stanu zaj˛etości systemu. Czas pomi˛edzy kolejnymi zgłoszeniami wyznaczano na podstawie naste˛puja˛cych rozkładów prawdopodobieństwa: wykładniczego, równomiernego, normalnego, gamma i Pareto. Rezultaty eksperymentów symulacyjnych zostały porównane w wynikami modelowania na podstawie algorytmu splotowego. Otrzymane rezultaty pozwalaja ˛ na wskazanie granicznych moz˙liwosći zastosowania algorytmu splotowego. Abstract: This article verifies the applicability of the convolution algorithm for modelling multi-service network systems with different call streams, including non-Poissonian streams. The paper includes investigations of traffic streams that are state-dependent and independent. The times between calls ware determined by the following distributions: exponential, uniform, normal, gamma and Pareto. The results of the simulation experiments are compared with the results of modelling by convolution algorithm. The results obtained in this study allow limits for the applicability of the convolution algorithm to be determined and identified.
Abstract. This article presents the results of a study on hierarchical multiservice traffic overflow systems. The systems under investigation were composed of a number of primary resources and one alternative resource. Traffic in the considered systems was generated with the assumption that there was a finite number of traffic sources in individual classes. Further assumption was that offered traffic is elastic traffic for which – with an increase of the load of the system – a change in the volume of allocated resources is possible followed by concurrent extension of their service time. The article includes the results that present the blocking probability in the sample elastic traffic overflow systems. The study focuses on a determination of the influence of the traffic intensity, volume of resources, degree of compression, resource with compression (primary resources, alternative resources) and the cardinality of traffic sources on values of the blocking probability in individual call classes and on the number of calls that overflow to the alternative resource.
Abstract—This article proposes a model of a multiservice overflow system supporting elastic traffic. The system can serve as the analytical backbone for an effective analysis of communications network systems, in particular Internet systems and 4G and 5G mobile networks. A particular attention is given to the method for a determination of the parameters of traffic that overflows to secondary resources. The method takes into consideration both service method for elastic traffic handling as well as the level of traffic degeneration. The results of the analytical calculations are then compared with the results of the analytical experiments for a number of selected system structures. The results of the study confirm the accuracy of the proposed method and, as a consequence, a possibility of its application in engineering practice.
S. Hanczewski, D. Kmiecik, M. Stasiak, J. Weissenberg
This article proposes a new model of a multiservice queueing system with elastic traffic. The model makes it possible to determine individual delay characteristics in network nodes for each serviced call class. The proposed model can be used to analyse present-day network systems.
Any analysis of multiservice networks requires a set of appropriate queueing models to be developed. Numerous works, e.g. [1], propose an approximate approach in which all call classes have identical characteristics resulting from a possibility of traffic compression (the so-called elastic traffic). Such an approach eventually leads to the Erlang model which does not allow queues for selected traffic classes to be evaluated. [2] proposes a SDFIFO model (State-Dependent FIFO). This model results from the queueing interpretation of the model with traffic compression [3], [4]. The present article* proposes a generalisation to model [2] to introduce a possibility of servicing elastic traffic.
The paper presents a new approximate model for multi-service, non-full-availability queueing system. This model can be used to model the access nodes of the network. Due to the fact that this model is a approximate model, the results obtained were compared with the results of simulation experiments. The comparison has allowed us to confirm the correctness of the model’s assumptions.
The design and optimization process of modern telecommunications networks is supported by a range of appropriate analytical models. A number of these models are based on the Erlang's Ideal Grading (EIG) model, which is a particular case of non-full-availability groups. A possibility of the application of the EIG model results from the fact that telecommunications systems show properties and features distinctive to non-full-availability systems. No detailed studies that would decisively help determine appropriate conditions for the application of the EIG model for modeling of other non-full-availability groups, that would be models corresponding to real telecommunications systems, have been performed. Therefore, this article attempts to nd an answer to the following question: what are the prerequisite conditions for the application of the EIG model and when the model can be reliably used?
W artykule przedstawiony został nowy, przybli ˙zony model wielousługowego, niepełnodost˛enego systemu kolejkowego. Mo˙ze być on wykorzystywany do modelowania systemów sieciowych o ograniczonym dost˛epie do zasobów. Ze wzgl˛edu na fakt, ˙ze model ten jest modelem przybli˙zonym, rezultaty uzyskane za jego pomoca˛ zostały porównane z wynikami eksperymentów symulacyjnych. Porównanie to pozwoliło na potwierdzić słuszno´sć przyj˛etych zało˙ze ´n modelu.
The paper presents a new approximate analytical model of a multi-service, non-full-availability queueing system. The model can be use to model the access nodes of the network. Taking into consideration that the proposed model is approximate, the results of the analytical calculations obtained on the basis of the proposed in the paper model are compared with the simulation data. The simulation study has confirmed that the proposed model is characterised by high accuracy.
W artykule przedstawiono przybli˙zony model grupy komórek sieci LTE z zaimplementowanym mechanizmem równowaz ˙enia obcia˛z˙enia. Do modelowania grupy komórek zaproponowano wykorzystanie odpowiednio zmodyfikowanego modelu wia˛zki niepełnodoste˛pnej (Erlang’s Ideal Grading). Opracowane rozwia˛zanie umoz˙liwia wyznaczenie podstawowych charakterystyk ruchowych w rozwa˙zanym systemie, tj. rozkładu zaj˛etości oraz prawdopodobieństwa blokady/strat. Ze wzgl˛edu na przybli˙zony charakter proponowanego modelu, rezultaty uzyskane na jego podstawie zostały porównane z wynikami symulacji cyfrowej.
