Phuong Nga Tran

Optimal mapping of virtual networks considering reactive reconfiguration

Network Virtualization (NV) has recently been considered as a key feature of the Future Internet because it allows easy transition from existing IPv4 based Internet to new mechanisms and protocols, and a more efficient usage of infrastructure by sharing it with different virtual network operators while deploying new services. However, Network Virtualization also raises new challenges for network operators [1]. One of the main challenges is the efficient resource allocation, known as virtual network mapping problem, in which virtual nodes and links are embedded to specific substrate nodes and paths. In this paper, we study this problem for the online scenario, where virtual network requests come and leave dynamically while considering the ability to reconfigure (re-arrange) the currently mapped virtual networks. All the recent research on this topic just focuses on the acceptance ratio of the virtual networks but do not address the possible service disruption during the reconfiguration. This paper proposes a so-called reactive reconfiguration mechanism, which reacts to any rejection of new coming virtual network requests. If a virtual network is rejected due to the lack of physical resources, the mechanism is triggered to reconfigure the currently-mapped networks so that the new request can be embedded. At the same time, it minimizes the number of necessary changes to reduce the service disruption. The mechanism is mathematically formulated as Integer Linear Programming (ILP) problem and solved by CPLEX [2]. Its performance is then evaluated through extensive simulations and compared with the case of no reconfiguration. A heuristic is then proposed to speed up the solving time of the ILP problem while maintaining its performance.

Energy consumption optimization for software defined networks considering dynamic traffic

Todays networking hardware (e.g. switches, routers) is typically running 24/7, regardless of the traffic volume. This is because in current networks, the controlling and data forwarding functions are embedded in the same devices, and all L2/L3 network protocols are designed to work in a distributed manner. Therefore, network devices must be switched on all the time to handle the traffic. This consequently results in very high global energy consumption of communication networks. Software Defined Networking was recently introduced as a new networking paradigm, in which the control plane is physically separated from the forwarding plane and moved to a globally-aware software controller. As a consequence, traffic can be monitored in real time and rerouted very fast regarding certain objectives such as load balancing or QoS enhancement. Accordingly, it opens new opportunities to improve the overall network performance in general and the energy efficiency in particular. This paper proposes an approach that reconfigures the network in order to reduce the energy consumption, based on the current traffic load. Our main idea is to switch on a minimum amount of necessary switches/routers and links to carry the traffic. We first formulate the problem as a mixed integer linear programming (MILP) problem and further present a heuristic method, so called Strategic Greedy Heuristic, with four different strategies, to solve the problem for large networks. We have carried out extensive simulations for a typical campus network and arbitrary mesh networks with realistic traffic information and energy consumption, to prove the potential energy saving of the proposed approach. The results showed that we can save up to 45% of the energy consumption at nighttime.

An Exact ILP Formulation for Optimal Wavelength Converter Usage and Placement in WDM Networks

In WDM all-optical networks, the use of wavelength converters can increase the wavelength resource efficiency and reduce the blocking probability. However, all-optical wavelength converters are likely to remain costly devices. Moreover, the blocking probability does not decrease linearly with the number of converters. Hence, it is desirable that just a limited amount of wavelength converters are used in the whole network. In this case, a vital question arises: how many converters are enough and where to place these converters? In this paper, we introduce for the first time an exact ILP formulation to minimize the number of necessary wavelength converters as well as to place these converters in the network for a static traffic pattern, when its Routing Wavelength Assignment (RWA) is infeasible if no wavelength converter is used but feasible if the wavelength continuity constraint is relaxed. By running the optimization for numerous different random traffic patterns, we can at the end find the nodes with highest probability to place wavelength converters.

Distributed algorithm for dynamic logical topology reconfiguration in IP over WDM networks

In this paper, we present a novel distributed algorithm to solve the logical topology reconfiguration problem for IP over WDM networks under dynamic traffic. From the optimization model proposed in [1], we apply the Lagrangian relaxation to eliminate the capacity constraint and form a dual problem whose optimal solution is the lower bound of the original one. The dual problem is then decomposed into subproblems, which can be solved in a distributed manner. Based on the solution of the dual problem, a heuristic algorithm is proposed to obtain a good feasible solution for the orginal reconfiguration problem. Our distributed algorithm fulfils all requirements of the reconfiguration process, such as no service disruption, load balancing in the new logical topology while allowing to add and to delete multiple lightpaths. Moreover, the distributed approach offers a high scalability. Not limited to small or medium-sized networks as the centralized model in [1], this distributed algorithm can be applied to large networks with the computation time increasing linearly with the network size. The performance of the proposed distributed algorithm, which lends itself to a protocol implementation, is investigated and compared with the corresponding centralized approach.

Radio resource allocation in LTE using utility functions based on moving average rates

In this paper, we propose a new algorithm to solve the downlink resource allocation problem in LTE networks taking the channel conditions into account. We first formulate the problem as a convex optimization problem, which maximizes the aggregated utility of all users. Different from other papers of the same topic, we construct the utility function based on the Exponential Moving Average (EMA) rate instead of the instantaneous data rate. The advantage of this approach is that it guarantees the users with very bad channel conditions still to be scheduled, even if the number of Physical Resource Blocks (PRBs) is smaller than the number of users. The problem is solved optimally with the Lagrangian decomposition method. Extensive simulations have been carried out to compare our approach to the instantaneous data rate approach and to analyze the influence of the system parameters on the behavior of the algorithm.

Network Planning for Stochastic Traffic Demands

Traffic in communication networks is not constant but fluctuates heavily, which makes the network planning task very challenging. Overestimating the traffic volume results in an expensive solution, while underestimating it leads to a poor Quality of Service (QoS) in the network.

Resource efficient logical topology design for IP-over-WDM backbone networks

With the advancements in WDM technology to provide huge capacity and the Internet Protocol playing a dominant role in networking technology, IP-over-WDM network becomes a potential candidate for next generation networks. This article studies the problem of designing logical topologies for IP-over-WDM backbone networks. One important feature of IP-over-WDM networks is the reconfigurability when traffic demands change over time. If the network resource is not used efficiently, the reconfiguration can cause network disruption because no new wavelength channels (lightpaths) can be added before deleting some working channels. Moreover, more than one logical topology for different services or from different ISPs can be simultaneously implemented in an existing WDM network. Hence using resources efficiently is an important issue while designing logical topologies. In this article, we present a new approach to design logical topologies so that the network resource is used most efficiently while the network performance is guaranteed. The problem is formulated as an MILP problem. Since this MILP problem is NP-hard and can be applied for small networks only, a heuristic algorithm based on iterative optimization is proposed to solve the problem for real-size networks.

QoE-based access network dimensioning

Optimum utilisation of the transport links in the access networks as well as satisfaction of the end-user service quality (Quality of Experience - QoE) is important for mobile operators. The necessary prerequisites for proper dimensioning (capacity planning) of the access network are careful modeling of the arrival traffic characteristics and proper QoE to QoS (Quality of Service, e.g., delay) mapping. The Internet traffic is known to be bursty and self-similar and therefore in this paper an analytical modeling of the arrival traffic is developed based on a two-state Markov Modulated Poisson Process, MMPP(2). The performance in terms of inter-arrival time of the packets and the resulting queueing delay distribution of the MMPP(2)/D/1 system is found to be a good approximation for the packet delays of the VoIP, web and video traffic.

QoE-based radio resource allocation in LTE femtocell considering transport limitations

In this paper, we study the QoE-based resource allocation problem for downlink direction in LTE femtocells, considering the transport network limitation. We first formulate the problem as a convex optimization problem, which maximizes the aggregated QoE of all users and solve it using Lagrangian relaxation method. This approach provides an ideal resource allocation solution but it is not practical, because the allocated bandwidth in the model is assumed to be a real value, while in practice, only an integer amount of Physical Resource Blocks (PRBs) can be assigned to a user. However, it is considered as a benchmark for our new developed algorithm. Then we propose an efficient heuristic, which can be applied in real LTE systems. The heuristic is mathematically proven to provide an optimal solution in case there is no limitation in the transport network. In case of transport network limitation, the simulation results show that it gives almost identical performance (ca. 1%) as the optimal solution in the investigated scenario.

Coordinated radio resource allocation in LTE femtocell cluster considering transport limitations

In this paper, we study the coordinated QoE-based resource allocation problem for downlink direction in a LTE femtocell cluster, considering the transport network limitation. We first formulate the problem as a convex optimization problem, which maximizes the aggregated QoE of all users and solve it using Lagrangian relaxation method. This approach provides an ideal resource allocation solution but it is not practical, because the allocated bandwidth in the model is assumed to be a real value, while in practice, only an integer amount of Physical Resource Blocks (PRBs) can be assigned to a user. However, it is considered as a benchmark for our new developed algorithms. Afterwards, we propose efficient heuristics, which can be applied in real LTE systems due to their low complexity. The performance of our algorithms is investigated and evaluated by simulations. The simulations show that the coordinated resource allocation scheme considering transport limitation results in a better average QoE compared the scheme that does not take into account the transport network capacity. Furthermore, our proposed heuristics are very close to the theoretical optimal bound.