Semi-Markov models for performance evaluation of failure-prone IP multimedia subsystem core networks

Abstract

Next generation telecommunication core networks are typically based on the Third Generation Partnership Project Internet protocol (IP) multimedia subsystem (IMS). Their planning and deployment must take into account the occurrence of random failures causing performance degradations, in order to assess and maintain a high level of quality of service. In particular, IMS signalling servers can be modelled as repairable multi-state elements where states correspond to different performance levels. This article provides an evaluation of IMS signalling network performance in long runs in terms of two metrics adopted in the practice, such as the number of call set-up sessions that the network can manage at the same time and the call set-up delay. A semi-Markov model has been adopted for the IMS servers, which allows as well for non-exponential probability distributions of sojourn times, as often observed in real networks. Furthermore, a redundancy optimization problem is solved in an IMS-based realistic scenario, to the aim of minimizing the deployment cost of a telecommunication network with a given availability requirement

Availability Evaluation of Multi-Tenant Service Function Chaining Infrastructures by Multidimensional Universal Generating Function

“…In order to characterize the u-function of a system described by a series/parallel availability model, two operators can be adopted: an operator π to build the u-function of subsystems connected in parallel, and an operator σ to calculate the u-function of elements interconnected in series. Recall that, under the assumption of adopting load balancing techniques for parallel nodes [22], a subsystem constituted by multi-state components with flow dispersion is obtained. According to [21], the u-function of subsystem m with L m parallel nodes is expressed by the π operator that represents the product of the UGFs for each node, namely…”

“…By solving this CTMC with N = 14 states for t → ∞ as described in Sect. 5.1, the steady-state performance distribution (in terms of the number of call set-up sessions) for a single virtualized node is computed, and is given by the collection of pairs {p j , g j }, where p j is the probability (4) correlated to the performance level g j in the set (2) and where Q is given by (22), along with the vector d defined as: Table 1 summarizes the obtained performance levels and pertinent state probabilities. It is useful to note that p 1 +p 2 + p 3 refers to (0, 0) performance vector since HLF, VLF and the state accounting for all failed software instances admit the same performance vector (namely, a completely failed system).…”

“…Approaches exploiting the UGF function (to handle multi-state system representations) are adopted in [21] and [22], where Markov and semi-Markov models are used, respectively, to characterize a single tenant IMS architecture. Another UGF-based method is employed to model physical and virtual machine failures in [23], where a single tenant cloud-based environment is implicitly assumed.…”

An analytical approach for optimization of computer network design considering the integration of the communication and power infrastructures

“…In the last few years, a number of research works 1,[10][11][12][13][14][15][16][17][18][19] have been proposed for system infrastructure planning (e.g., telecommunication, data centers and power). Initially, Sauve et al 1 proposes a design approach that is related to the following areas: BDIM, infrastructure design and SLA design.…”

Single minimal path based backup path for multi-state network

“…Such a network is called a multi-state network (MSN). [8][9][10][11][12][13][14][15][16][17][18][19] There are mainly two ways to evaluate MSN reliability: approximation algorithms [20][21][22][23][24] and exact algorithms. 18,[25][26][27][28][29][30][31][32][33][34][35][36] Among the exact algorithms, one of the most commonly used methods is based on either binary-state minimal path (MP) 9,[32][33][34][35][36] or binary-state minimal cuts (MCs), 27,29,30,31 where a MP/MC is a path/cut set such that if any arc is removed from this proposed an MP-based algorithm to calculate the weighted multi-commodity MSN reliability with unreliable nodes and arcs.…”