This paper presents a Path Bandwidth Management (PBM) model for large-scale networks that leads to an almost optimal PB allocation, under constraints posed by the installed bandwidth in the transmission links of the network. The PB allocation procedure is driven from a traffic demand matrix and consists of three phases. In the first phase, a suitable decomposition of the whole network takes place, where the large-scale network is split to a set of one-level sub-networks. In the second phase, the optimization algorithm developed for one-level telecommunication networks is applied to each sub-network in order to define the optimal PB allocation. The criterion for optimization is to minimize the worst Call Blocking Probability (CBP) of all switching pairs of the sub-network. In the third phase, composition of the sub-networks takes place in a successive way, which leads to the final PB allocation of the large-scale network. As the large-scale network is built up from optimized sub-networks, an almost optimal PB allocation is anticipated. For evaluation, the worst resultant CBP of the proposed scheme is compared with that obtained by the optimal PB allocation procedure in order to prove its optimality and efficiency. We choose a set of large-scale networks whose size is not very large so that we can apply the optimization algorithm developed for one-level telecom networks for defining its optimal bandwidth allocation. Extensive evaluation of the PBM model has showed that the worst resultant CBP is about 2% above the optimal value, which is a satisfactory result. The proposed PBM scheme is explained by means of an application example.

Bursty traffic is dominant in modern communication networks and keeps the call-level QoS assessment an open issue. ON-OFF traffic models are commonly used to describe bursty traffic. We propose an ON-OFF traffic model of a single link which accommodates service-classes of finite population (f-ON-OFF). Calls compete for the available link bandwidth under the complete sharing policy. Accepted calls enter the system via state ON and then may alternate between ON-OFF states. When a call is transferred to state OFF it releases the bandwidth held in state ON, while when a call tries to return to state ON, it re-requests its bandwidth. If it is available a new ON-period (burst) begins; otherwise the call remains in state OFF (burst blocking). We prove that the proposed f-ON-OFF model has a product form solution, and we provide an accurate recursive formula for the call blocking probabilities calculation. For the burst blocking probabilities calculation we propose an approximate but robust formula. In addition, we show the relation between the f-ON-OFF model and other call-level loss models. Furthermore, we generalize the f-ON-OFF model to include service-classes of both finite and infinite population. Simulation results validate our analytical methodology.

Fast proliferation of mobile Internet and high-demand mobile applications necessitates the introduction of different priority classes in next-generation cellular networks. This is especially crucial for efficient use of radio resources in the heterogeneous and virtualized network environments. Despite the fact that many analytical tools have been proposed for capacity and radio resource modelling in cellular networks, only a few of them explicitly incorporate priorities among services. We propose a novel analytical model to analyse the performance of a priority-based cellular CDMA system with finite source population. When the cell load is above a certain level, low-priority calls may be blocked to preserve the quality of service of high-priority calls. The proposed model leads to an efficient closed-form solution that enables fast and very accurate calculation of resource occupancy of the CDMA system and call blocking probabilities, for different services and many priority classes. To achieve them, the system is modelled as a continuous-time Markov chain. We evaluate the accuracy of the proposed analytical model by means of computer simulations and find that the introduced approximation errors are negligible.

The Generalized Max-Min Fairness policy (GMM) allocates in a fair way the available bandwidth among elastic calls by taking into account their minimum and maximum rate requirements. The GMM has been described in a five-step procedure, which has the advantage of an easy presentation, but does not come into details, as far as its computer implementation is concerned, and fails to describe the policy in a clear mathematical way. We propose a new algorithm for the GMM policy, in a clear mathematical way, based on Linear Programming (LP). The new algorithm is directly convertible into software. Numerical examples clarify our algorithm.

The call-level performance modelling is a challenge in the highly heterogeneous environment of modern telecom networks, due to the presence of elastic traffic. In this paper, we review existing teletraffic loss models and propose a model for elastic traffic of service-classes with finite population (quasi-random call arrival process). Upon arrival, calls have contingency alternative bandwidth requirements that depend on thresholds which indicate the available/occupied link bandwidth (state dependent model). Calls are admitted under the complete sharing policy, and can tolerate bandwidth compression, while in-service. We prove a recurrent formula for the efficient calculation of the link occupancy distribution and consequently the call blocking probabilities and link utilization. The accuracy of the proposed model is verified by simulation and is found to be quite satisfactory. Comparative results with other existing models show the necessity and the effectiveness of the proposed model. Its potential applications are mainly in the environment of wireless networks.