Eduard Sopin

Analysis of an M|G|1|R queue with batch arrivals and two hysteretic overload control policies

Abstract Hysteretic control of arrivals is one of the most easy-to-implement and effective solutions of overload problems occurring in SIP-servers. A mathematical model of an SIP server based on the queueing system M[X]|G|1(L,H)|(H,R) with batch arrivals and two hysteretic loops is being analyzed. This paper proposes two analytical methods for studying performance characteristics related to the number of customers in the system. Two control policies defined by instants when it is decided to change the system’s mode are considered. The expression for an important performance characteristic of each policy (the mean time between changes in the system mode) is presented. Numerical examples that allow comparison of the efficiency of both policies are given

Sojourn Time Analysis for Processor Sharing Loss System with Unreliable Server

Processor sharing (PS) queuing systems and particularly their well-known class of egalitarian processor (EPS) sharing are widely investigated by research community and applied for the analysis of wire and wireless communication systems and networks. The same can be said for queuing systems in random environment, with unreliable servers, interruptions, pre-emption mechanisms. Nevertheless, only few works focus on queues with both PS discipline and unreliable servers. In the paper, compared with the previous results we analyse a finite capacity PS queuing system with unreliable server and an upper limit of the number of customers it serves simultaneously. For calculating the mean sojourn time, unlike a popular but computational complex technique of inverse Laplace transform we use an effective method based on embedded Markov chains. The paper also includes a practical numerical example of web browsing in a wireless network when the corresponding low priority traffic can be interrupted by more priority applications.

LTE performance analysis using queuing systems with finite resources and random requirements

Heavy traffic load in current LTE networks calls for effective radio resource allocation methods and tools for performance evaluation. In this work, we provide an analytical framework for LTE resource allocation in terms of queuing theory. We consider a multiservice queuing system with a finite amount of resources of several types, and allow the customers occupy a random amount of resources upon their arrival. Random resource requirements lead to more accurate performance evaluation compared to conventional multiservice models. For the considered model, we prove that the stationary probability distribution has a multiplicative form. Our findings are illustrated with a numerical example.

Evaluating a Case of Downlink Uplink Decoupling Using Queuing System with Random Requirements

The need for efficient resource utilization at the air interfaces in heterogeneous wireless systems has recently led to the concept of downlink and uplink decoupling (DUDe). Several studies have already reported the gains of using DUDe in static traffic conditions. In this paper we investigate performance of DUDe with stochastic session arrivals patterns in LTE environment with macro and micro base stations. Particularly, we use a queuing systems with random resource requirements and to calculate the session blocking probability and throughput of the system. Our results demonstrate that DUDe association approach allows to significantly improve the metrics of interest compared to conventional downlink-based association mechanism.

Sojourn Time Analysis for Processor Sharing Loss Queuing System with Service Interruptions and MAP Arrivals

Processor sharing (PS) queuing systems are widely investigated by research community and applied for the analysis of wire and wireless communication systems and networks. Nevertheless, only few works focus on finite queues with both PS discipline and service interruptions. In the paper, compared with the previous results we analyze a finite capacity PS queuing system with Markovian arrival process, unreliable server, service interruptions, and an upper limit of the number of customers it serves simultaneously. For calculating the mean sojourn time, unlike a popular but computational complex technique of inverse Laplace transform we use an effective method based on embedded Markov chain. A practical example concludes the paper.

Experimental Evaluation of Dynamic Licensed Shared Access Operation in Live 3GPP LTE System

As next-generation mobile networks are rapidly taking shape driven by the target standardization requirements and initial trial implementations, a range of accompanying technologies prepare to support them with more reliable wireless access and improved service provisioning. Among these are more advanced spectrum sharing options enabled by the emerging Licensed Shared Access (LSA) regulatory framework, which aims to efficiently employ the capacity of underutilized frequency bands in a controlled manner. The concept of LSA promises to equip network operators with the much needed additional spectrum on the secondary basis and thus brings changes to the existing cellular network management. Hence, additional research is in prompt demand to determine the required levels of Quality of Service (QoS) and service provisioning reliability, especially in cases of dynamic geographical and temporal LSA sharing. Motivated by this recent urge and having at our disposal a fully-functional 3GPP LTE cellular deployment, we have committed to implement and trial the principles of dynamic LSA-compatible spectrum management. This paper is our first disclosure on the comprehensive experimental evaluation of this promising technology. We expect that these unprecedented practical results together with the key lessons learned will become a valuable reference point for the subsequent integration of flexible LSA-based services, suitable for inter-operator and multi-tenant spectrum sharing.

Threshold-based queuing system for performance analysis of cloud computing system with dynamic scaling

Cloud computing is promising technology to manage and improve utilization of computing center resources to deliver various computing and IT services. For the purpose of energy saving there is no need to unnecessarily operate many servers under light loads, and they are switched off. On the other hand, some servers should be switched on in heavy load cases to prevent very long delays. Thus, waiting times and system operating cost can be maintained on acceptable level by dynamically adding or removing servers. One more fact that should be taken into account is significant server setup costs and activation times. For better energy efficiency, cloud computing system should not react on instantaneous increase or instantaneous decrease of load. That is the main motivation for using queuing systems with hysteresis for cloud computing system modelling. In the paper, we provide a model of cloud computing system in terms of multiple server threshold-based infinite capacity queuing system with hysteresis and noninstantanuous server activation. For proposed model, we develop a method for computing steady-state probabilities that allow to estimate a number of performance measures.

Approximating performance measures of radio admission control model for non real-time services with maximum bit rates in LTE

An important feature of next generation networks is a possibility to provide services with various requirements of quality of service (QoS) to users. In accordance with the 3GPP specifications, two types of services are defined – guaranteed bit rate (GBR) services, i.e. real time gaming, and non-guaranteed bit rate (non-GBR) services, i.e. e-mail, data communication. Different types of non-GBR services differ one from other, first, in a data volume defining data transfer time. The data transfer time also depends on subscriber devices, which not necessarily support LTE. This fact superimposes restrictions on the possible maximum bit rate. We analyze a multi-rate model of a LTE network with two types of non-GBR services. The radio admission control scheme assumes restrictions on the maximum bit rate and adaptive bit rate changes. These changes depend on the cell load. We apply teletraffic and queuing theories to get an approximate method for calculating the blocking probability.

Analyzing Blocking Probability in LTE Wireless Network via Queuing System with Finite Amount of Resources

According to analytics, the global mobile date traffic will grow three times faster than fixed traffic by 2019. The number of user’s mobile devices is supposed to increase from 4.1 billion to 4.9 billion while the number of mobile device connections can reach even 10 billion. Broadband speeds in wireless networks are expected to double from 1.7 Mbps to 4.0 Mbps to the end of 2019. As it is noted the mobile video traffic will be up to 72 percent of the global mobile traffic. As the number of wireless connections tends to increase significantly, it results in dramatic mobile traffic growth. Mobile service providers face the challenge to utilize limited radio resources efficiently. In this paper, we propose a mathematical model of radio resources allocation in broadband wireless networks such as LTE-Advanced in terms of queuing systems and evaluate blocking probability and average amount of occupied resources.

Two approaches to analyzing dynamic cellular networks with limited resources

In current cellular networks, each user session requires a share of radio resources to be served. Performance analysis of such networks can generally be conducted by employing queuing systems with limited resources. However, with the ongoing network densification, the loading on future networks is expected to vary significantly and unpredictably. This calls for new methods to model flow-level dynamics in next-generation dense heterogeneous networks. In this paper, we discuss two new approaches that simplify the analysis of dynamic cellular systems by making it analytically tractable. Our findings are confirmed by simulations.