Currently, the Open Radio Access Network (O-RAN) architecture is the best solution for deploying Radio Access Network (RAN). This architecture presents several challenges, i.e., function splitting, function placement and energy consumption. The aim of this study is to satisfy user demand while minimizing costs. In this work we combine the O-RAN architecture and the multiband system. We consider that the RAN network load varies during the day and we study the optimal placement of the Distributed Unit (DU) function and switch off unnecessary frequency bands. First, we propose an algorithm to find the number of frequency bands needed to satisfy user demand for different smoothing periods. Based on these results, we formulate the problem of DU placement and frequency band switch-off as an Integer Linear Programming (ILP) whose objective is to minimize computing and routing cost while respecting delay and capacity constraints. Evaluation of our model on real topology has shown that our model with frequency band extinction has a much lower system cost than our model without band extinction.

We consider the placement of jobs inside a data center. Traditionally, this is done by a task orchestrator without taking into account network constraints. According to recent studies, network transfers may account for up to 50% of the completion time of classical jobs. Thus, network resources must be considered when placing jobs in a data center. In this paper, we propose a new scheduling framework, introducing network tasks that need to be executed on network machines alongside traditional (CPU) tasks. The model takes into account the competition between communications for the network resources, which is not considered in the formerly proposed scheduling models with communication. Network transfers inside a data center can be easily modeled in our framework. As we show, classical algorithms do not efficiently handle a limited amount of network bandwidth. We thus propose new provably efficient algorithms with the goal of minimizing the makespan in this framework. We show their efficiency and the importance of taking into consideration network capacity through extensive simulations on workflows built from Google data center traces.

Multi-topology routing (MTR) provides an attractive alternative to segment routing for traffic engineering when network devices cannot be upgraded. However, due to a high overhead in terms of link state messages exchanged by topologies and the need to frequently update link weights to follow evolving network conditions, MTR is often limited to a small number of topologies and the satisfaction of loose QoS constraints. To overcome these limitations we propose vMTR, an MTR extension where demands are routed over virtual topologies that are silent, i.e., they do not exchange LSA messages, and that are continuously derived from a very limited set of real topologies, optimizing each a QoS parameter. In this context, we present a polynomial and exact algorithm for vMTR and, as a benchmark, a local search algorithm for MTR. We show that vMTR helps reducing drastically the number of real topologies and that it is more robust to QoS changes.

The unavoidable transition from rigid dedicated hardware devices towards flexible containerized network services, introduced by Network Function Virtualization (NFV), brings novel opportunities while presenting several new challenges. Indeed, meeting the expectations of NFV in post-5G networks depends on the efficient placement of the services. The online placement of network services, demanding strict end-to-end latency requirements, with restricted computing resources presents a challenging problem which is worth investigating. We propose a Branch-and-Bound search approach for finding optimal placements of the network services by applying several cost functions to maximize the service acceptance. Extensive evaluations have been carried out, and the results confirm significant improvements when we consider a fair distribution of the resources on the edge.

Open Radio Access Network (O-RAN) is very promising for flexible and efficient 5G and 6G wireless networks. The O-RAN architecture consists of three main units: Radio Unit (RU), Distributed Unit (DU), and Centralized Unit (CU). In this paper, we study the placement of virtualized DUs. This placement has strong consequences on cost and delay, among others, and is thus an important challenge. First, we analyze the throughput between the O-RAN interfaces. Based on our analysis, we propose an efficient Integer Linear Programming (ILP) model. The objective is to minimize the O-RAN cost depending on the DU placement while respecting the delay and capacity constraints. We evaluate our model on a real topology. Our results provide interesting insights into the cost savings with regard to a legacy architecture. Moreover, the proposed model provides solutions in a configuration where a fully centralized Cloud RAN architecture would not. We also estimate the limits of capacity of a given configuration.

The inevitable transition from physical dedicated hardware devices towards lightweight containerized reusable software modules with Network Function Virtualization (NFV) introduces countless opportunities while presenting several unprecedented challenges. Satisfying NFV expectations in post-5G networks heavily depends on the efficient placement of network services. In this paper, after modeling the placement problem and proposing the exact resolutions using Integer Linear Programming (ILP) and Column Generation (CG), we propose our deterministic placement solution, capable of obtaining optimal results with the scalability of a heuristic-grade approach. Our method is organized as a Branch and Bound (BnB) structure, applying Artificial Intelligence (AI) search strategies (especially A*) to address the problem of network service placement. We believe that it is suitable for a range of applications in online placement scenarios, whether we concentrate on the quality of the results or on the strict time constraints. We are interested in the popular objective of Service Acceptance (SA) maximization and have carried out several extensive evaluations. The obtained results confirm the effectiveness of our solution.

In this paper, we explore the use of an actor-critic architecture for Deep Reinforcement Learning (DRL) to improve load balancing beyond traditional algorithms. Some centralized Reinforcement Learning (RL) algorithms have targeted in the reward function expression the Quality of Experience (QoE) for video flows, but this requires access to clients, or the Maximum Link Utilization (MLU) for other types of flows. In our approach, we tune the actor-critic algorithm to only leverage on QoS parameters in order to load balance traffic in the network and maximize the QoE experienced by the users. This avoids having to collect observations and performance measurements from client applications, as it only focuses on network metrics that can be easily measured. We explore both centralized and distributed solutions to assess the feasibility of the proposed smart load balancing solutions. We compare them to ECMP, QoE-based reward methods, and RILNET that uses an underlying DDPG optimization approach. The proposed algorithms are shown to outperform previous approaches.

The deployment of 5G networks is paving the road to custom network services. It is now possible to envision the automatic decomposition of a physical network into several virtual networks to serve a wide range of user needs. This technology is also referred to as network slicing. To guarantee the strict isolation of virtual networks, it is possible to rely on underlay technologies such as Flex Ethernet (FlexE). In this demo, we present a slicing solution based on Multi-Topology-Routing (MTR). We will demonstrate how IGP weights can be designed for the embedding of a slice, described by a traffic matrix and end-to-end latency requirements, to minimize the cost of underlay bandwidth reservations.

In the context of multi-agent systems, Automated Mechanism Design (AMD) is the computer-based design of the rules of a mechanism, which reaches an equilibrium despite the fact that agents can be selfish and lie about their preferences. Although it has been shown that AMD can be modelled as a linear program, it is with an exponential number of variables and consequently, there is no known efficient algorithm. We revisit the latter linear program model proposed for the AMD problem and introduce a new one with a polynomial number of variables. We show that the latter model corresponds to a Dantzig-Wolfe decomposition of the second one and design efficient solution schemes in polynomial time for both two models. Numerical experiments compare the solution efficiency of both models and show that we can solve very significantly larger data instances than before, up to 2,000 agents or 2,000 resources in about 35 seconds.

We consider the placement of jobs inside a data center. Traditionally, this is done by a task orchestrator without taking into account network constraints. According to recent studies, network transfers represent up to 50% of the completion time of classical jobs. Thus, network resources must be considered when placing jobs in a data center. In this paper, we propose a new scheduling framework, introducing network tasks that need to be executed on network machines alongside traditional (CPU) tasks. The model takes into account the competition between communications for the network resources, which is not considered in the formerly proposed scheduling models with communication. Network transfers inside a data center can be easily modeled in our framework. As we show, classical algorithms do not efficiently handle a limited amount of network bandwidth. We thus propose new provably efficient algorithms with the goal of minimizing the makespan in this framework. We show their efficiency and the importance of taking into consideration network capacity through extensive simulations on workflows built from Google data center traces.

In the context of Software-Defined Networks (SDN), Network Function Virtualization (NFV) is a new network paradigm in which network functions are implemented in software as Virtual Network Functions (VNFs). To meet the demand, VNFs are next interconnected to form different complete end-to-end services, also known as a Service Function Chains (SFCs). We study the problem of deploying reliable Service Function Chains over a virtualized network function architecture. While there is a need for reliable service function chaining, there is a high cost to pay for it in terms of bandwidth and VNF processing requirements. We investigate two different protection mechanisms and discuss their resource requirements, as well as the latency of their paths. For each mechanism, we develop a scalable exact mathematical model using column generation.

A Service Function Chain (SFC) is an ordered sequence of network functions, such as load balancing, content filtering, and firewall. With the Network Function Virtualization (NFV) paradigm, network functions can be deployed as pieces of software on generic hardware, leading to a flexibility of network service composition. Along with its benefits, NFV brings several challenges to network operators, such as the placement of virtual network functions. In this paper, we study the problem of how to optimally place the network functions within the network in order to satisfy all the SFC requirements of the flows. Our optimization task is to minimize the total deployment cost. We show that the problem can be seen as an instance of the Set Cover Problem, even in the case of ordered sequences of network functions. It allows us to propose two logarithmic factor approximation algorithms which have the best possible asymptotic factor. Further, we devise an optimal algorithm for tree topologies. Finally, we evaluate the performances of our proposed algorithms through extensive simulations. We demonstrate that near-optimal solutions can be found with our approach.

Software-defined Network (SDN) is a new networking paradigm enabling innovation through network programmability. Over past few years, many applications have been built using SDN such as server load balancing, virtual-machine migration, traffic engineering and access control. In this paper, we focus on using SDN for energy-aware routing (EAR). Since traffic load has a small influence on the power consumption of routers, EAR allows putting unused links into sleep mode to save energy. SDN can collect traffic matrix and then computes routing solutions satisfying QoS while being minimal in energy consumption. However, prior works on EAR have assumed that the SDN forwarding table switch can hold an infinite number of rules. In practice, this assumption does not hold since such flow tables are implemented in Ternary Content Addressable Memory (TCAM) which is expensive and power hungry. We consider the use of wildcard rules to compress the forwarding tables. In this paper, we propose optimization methods to minimize energy consumption for a backbone network while respecting capacity constraints on links and rule space constraints on routers. In details, we present two exact formulations using Integer Linear Program (ILP) and introduce efficient heuristic algorithms. Based on simulations on realistic network topologies, we show that using this smart rule space allocation, it is possible to save almost as much power consumption as the classical EAR approach.

Energy aware routing aims at reducing the energy consumption of ISP networks. The idea is to adapt routing to the traffic load in order to turn off some hardware. However, it implies to make dynamic changes to routing configurations which is almost impossible with legacy protocols. The Software Defined Network (SDN) paradigm bears the promise of allowing a dynamic optimization with its centralized controller. In this work, we propose SENAtoR, an algorithm to enable energy aware routing in a scenario of progressive migration from legacy to SDN hardware. Since in real life, turning off network equipments is a delicate task as it can lead to packet losses, SENAtoR provides also several features to safely enable energy saving services: tunneling for fast rerouting, smooth node disabling and detection of both traffic spikes and link failures. We validate our solution by extensive simulations and by experimentation. We show that SENAtoR can be progressively deployed in a network using the SDN paradigm. It allows to reduce the energy consumption of ISP networks by 5 to 35% depending on the penetration of SDN hardware, while diminishing the packet loss rate compared to legacy protocols.

Software-Defined Networking is a new approach to the design and management of networks. It decouples the software-based control plane from the hardware-based data plane while abstracting the underlying network infrastructure and moving the network intelligence to a centralized software-based controller where network services are deployed. The challenge is then to efficiently provision the service chain requests, while finding the best compromise between the bandwidth requirements, the number of locations for hosting Virtual Network Functions (VNFs), and the number of chain occurrences. We propose two ILP (Integer Linear Programming) models for routing service chain requests, one of them with a decomposition modeling. We conduct extensive numerical experiments, and show we can solve exactly the routing of service chain requests in a few minutes for networks with up to 50 nodes, and traffic requests between all pairs of nodes. We investigate the best compromise between the bandwidth requirements and the number of VNF nodes.

We study distributed systems for live video streaming. These systems can be of two types: structured and unstructured. In an unstructured system, the diffusion is done opportunistically. The advantage is that it handles churn, that is the arrival and departure of users, which is very high in live streaming systems, in a smooth way. On the opposite, in a structured system, the diffusion of the video is done using explicit diffusion trees. The advantage is that the diffusion is very efficient, but the structure is broken by the churn. In this paper, we propose simple distributed repair protocols to maintain, under churn, the diffusion tree of a structured streaming system. We study these protocols using formal analysis and simulation. In particular, we provide an estimation of the system metrics, bandwidth usage, delay, or number of interruptions of the streaming. Our work shows that structured streaming systems can be efficient and resistant to churn.

Software Defined Networking (SDN) is gaining momentum with the support of major manufacturers. While it brings flexibility in the management of flows within the data center fabric, this flexibility comes at the cost of smaller routing table capacities. In this paper, we investigate compression techniques to reduce the forwarding information base (FIB) of SDN switches. We validate our algorithm, called MINNIE, on a real testbed able to emulate a 20 switches fat tree architecture. We demonstrate that even with a small number of clients, the limit in terms of number of rules is reached if no compression is performed, increasing the delay of all new incoming flows. MINNIE, on the other hand, reduces drastically the number of rules that need to be stored with a limited impact on the packet loss rate. We also evaluate the actual switching and reconfiguration times and the delay introduced by the communications with the controller.