Antnet Routing Algorithm with Link Evaporation and Multiple Ant Colonies to Overcome Stagnation Problem

Antnet is a software agent-based routing algorithm that is influenced by the unsophisticated and individual ant’s emergent behaviour. The aim of this chapter is twofold, firstly to introduce improvements to the antnet routing algorithm and then to critically review the work that is done around antnet and reinforcement learning in routing applications. In this chapter a modified antnet algorithm for packet-based networks has been proposed, which offers improvement in the throughput and the average delay by detecting and dropping packets routed through the non-optimal routes. The effect of traffic fluctuations has been limited by applying boundaries to the reinforcement parameter. The round trip feedback information supplied by the software agents is reinforced by updated probability entries in the distance vector table. In addition, link usage information is also used to prevent stagnation problems. Also discussed is antnet with multiple ant colonies applied to packet switched networks. Simulation results show that the average delay experienced by data packets is reduced for evaporation for all cases when non-uniform traffic model traffic is used. However, there is no performance gain on the uniform traffic models. In addition, multiple ant colonies are applied to the packet switched networks, and results are compared with the other approaches. Results show that the throughput could be increased when compared to other schemes, but with no gain in the average packet delay time.

Topological Design Using Genetic Algorithms

The proliferation of Internet access and the appearance of new telecommunications services are originating a demand for resilient networks with extremely high capacity. Thus, topologies able to recover connections in case of failure are essential. Given the node location and the traffic matrix, the survivable topological design is the problem of determining the network topology at minimum capital expenditure such that survivability is ensured. This problem is strongly NP-hard and heuristics are traditionally used to search near-optimal solutions. The authors present a genetic algorithm for this problem. As the convergence of the genetic algorithm depends on the used operators, an analysis of their impact on the quality of the obtained solutions is presented as well. Two initial population generators, two selection methods, two crossover operators, and two population sizes are compared, and the quality of the obtained solutions is assessed using an integer linear programming model.

Monitoring Devices for Providing Network Intelligence in Optical Packet Switched Networks

The development of all-Optical Packet Switching (OPS) networks brings about new challenges in the topic of Optical Performance Monitoring (OPM). The objectives of this chapter are addressed to the proposal of new monitoring techniques capable of packet-by-packet monitoring in the optical domain to preserve packet transparency. Moreover, new optical layer functionalities such as dynamic reconfiguration and link level restoration also introduce a level of complexity that may require advanced OPM capabilities. In this chapter, an OSNR monitoring technique and its application for providing network intelligence are explained in detail. In particular, the integration of the monitoring system with the control and management planes is investigated to perform other functions such as quality of service implementation, OSNR-assisted routing, and backup route selection.

Optical Label Processing Techniques for Intelligent Forwarding of Packets in All-Optical Packet Switched Networks

In this chapter, the authors review several optical label processing techniques providing a comparison based on the potential for each technique to allow for implementation of a scalable and low latency optical packet switching cross-connect node. They present and demonstrate an optical packet switch sub-system employing in-band labeling to allow for transparent forwarding of multi-wavelength packets with multiple data formats at multiple data bit-rates. The optical packet switching sub-system employs a scalable, asynchronous, and low latency label processor. Experimental results are provided that confirm the operation of the label processor in optical packet switching system testbeds. Moreover, the authors discuss applications of the optical packet switching node based on optical label processor and the potential to allow the implementation of intelligent systems for optimal routing of the packets in the optical domain.

Optical Network Optimization

Optical networks form the foundation of the global network infrastructure; hence, the planning and design of optical networks is crucial to the operation and economics of the Internet and its ability to support critical and reliable communication services. This book chapter covers various aspects of optimal optical network design, such as wavelength-routed Wavelength Division Multiplexing (WDM) optical networks, Spectrum-Sliced Elastic (SLICE) optical networks. As background, the chapter first briefly describes optical ring networks, WDM optical networks, and SLICE optical networks, as well as basic concepts of routing and wavelength assignment and virtual topology design, survivability, and traffic grooming in optical networks. The reader is referred to additional references for details. Many optical network design problems can be formulated as sophisticated optimization problems, including (1) Routing and Wavelength Assignment (RWA) and virtual topology design problem, (2) a suite of network design problems (such as variants of traffic grooming, survivability, and impairment-aware routing), (3) various design problems aimed at reducing the overall energy consumption of optical networks for green communication, (4) various design optimization problems in SLICE networks that employ OFDM technologies. This chapter covers numerous optical network design optimization problems and solution approaches in detail and presents some recent developments and future research directions.

Hopfield Neural Networks for Routing in Communication Networks

Although some interesting routing algorithms based on HNN were already proposed, they are slower when compared to other routing algorithms. Since HNN are inherently parallel, they are suitable for parallel implementations on parallel platforms, such as Field Programmable Gate Arrays (FPGA) and Graphic Processing Units (GPU). In this chapter, the authors show parallel implementations of a routing algorithm based on Hopfield Neural Networks (HNN) for GPU and for FPGAs, considering some implementation issues. They analyze the hardware limitation on the devices, the memory bottlenecks, the complexity of the HNN, and, in the case of GPU implementation, how the kernel functions should be implemented, as well as, in the case of the FPGA implementation, the accuracy of the number representation and memory storage on the device. The authors perform simulations for one variation of the routing algorithm for three communication network topologies with increasing number of nodes. They achieved speed-ups up to 78 when compared the FPGA model simulated to the CPU sequential version and the GPU version is 55 times faster than the sequential one. These new results suggest that it is possible to use the HNN to implement routers for real networks, including optical networks.

Artificial Bee Colony Approach for Routing and Wavelength Assignment in Optical WDM Networks

Routing and Wavelength Assignment (RWA) of lightpaths in optical WDM networks is a challenging task that belongs to a class of complex combinatorial problems. To solve the RWA problem of realistic size, heuristic or meta-heuristic approaches have to be used. In this chapter, an artificial bee colony metaheuristic approach, known as the Bee Colony Optimization (BCO), is used to solve the RWA problem for static lightpath establishment in wavelength routed optical WDM networks. The BCO metaheuristic is tailored here to solve the Max-RWA problem in which the objective is to maximize the number of established lightpaths for a given number of wavelengths. Behind a comprehensive description of the proposed BCO-RWA algorithm, the numerical results obtained by numerous simulations performed over widely used real world European Optical Network (EON) topology are given and compared with some other approaches used to solve the same problem.

Energy-Efficient Optical Transport Networks with Mixed Regenerator Placement

In this chapter, the authors first discuss the background and basic concepts of all-optical regeneration and translucent optical networks. Since all-optical regenerator is much more energy-efficient than traditional O/E/O ones, they investigate translucent lightpath arrangement that involves mixed placement (MRP) of optical inline amplifiers (1R), all-optical 2R regenerators, and O/E/O 3R regenerators for energy-saving. In order to make sure that the end-to-end transmission performance requirement can still be satisfied with this arrangement, the authors analyze the signal BER evolution through fiber links and different types of regenerators, and propose a theoretical model. They then develop search strategies based on exhaustive search and genetic algorithms, and discuss how to use them to optimize the energy-efficiency of lightpaths using MRP. Finally, the authors move to the network design using MRP, in which they considered both offline network design and online network provisioning.

Wavelength and Routing Assignment in All Optical Networks Using Ant Colony Optimization

Routing and Wavelength Assignment (RWA) in an arbitrary mesh network is an NP-complete problem. So far, this problem has been solved by linear programming for network topologies with a few nodes, and sub-optimally solved for larger networks by heuristic strategies and the application of optimization algorithms such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Differential Evolution (DE), etc. In this chapter, the authors present the use of Ant Colony Optimization (ACO) to find near optimal solutions to the routing and wavelength assignment problem in real sized networks with up to 40 nodes and 65 connecting links. They compare their results to the lower bounds obtained by the Nagatsu’s method, finding them to be equal or very close (one wavelength over) to them.

Applications of Computational Intelligence to Impairment-Aware Routing and Wavelength Assignment in Optical Networks

Computational intelligence techniques have been used to solve hard problems in optical networks, such as the routing and wavelength assignment problem, the design of the physical and the logical topology of these networks, and the placement of some high cost devices along the network when it is necessary, such as regenerators and wavelength converters. In this chapter, the authors concentrate on the application of computational intelligence to solve the impairment-aware routing and wavelength assignment problem. They present a brief survey on this topic and a detailed description and results for two applications of computational intelligence, one to solve the wavelength assignment problem with an evolutionary strategy approach and the other to tackle the routing problem using ant colony optimization.