Multi-sink wireless sensor networks (WSNs) have been deployed for an ever-widening range of applications, especially due to their low power consumption and reliability. In most applications, the sensor data must be sent to the Internet or core network, so the issue arises of providing a backhaul for WSN traffic. Passive optical networks (PONs), a next-generation access network technology, are well suited to form such a backhaul. In this paper, a new structure is proposed which converges multi-sink WSNs and PONs. It is analyzed and modelled through queueing theory. Based on this analytical model, a DBA algorithm is proposed to optimize the allocation of grant sizes. Numerical results show that the proposed algorithm outperforms other proposals for such a converged network.

One of the major difficulties facing researchers using neural networks is the selection of the proper size and topology of the networks. The problem is even more complex because often when the neural network is trained to very small errors, it may not respond properly for patterns not used in the training process. A partial solution proposed to this problem is to use the least possible number of neurons along with a large number of training patterns. The discussion consists of three main parts: first, different learning algorithms, including the Error Back Propagation (EBP) algorithm, the Levenberg Marquardt (LM) algorithm, and the recently developed Neuron-by-Neuron (NBN) algorithm, are discussed and compared based on several benchmark problems; second, the efficiency of different network topologies, including traditional Multilayer Perceptron (MLP) networks, Bridged Multilayer Perceptron (BMLP) networks, and Fully Connected Cascade (FCC) networks, are evaluated by both theoretical analysis and experimental results; third, the generalization issue is discussed to illustrate the importance of choosing the proper size of neural networks.

An analytical approximation is developed for the probability of sensing coverage in a wireless sensor network with randomly deployed sensor nodes each having an isotropic sensing area. This approximate probability is obtained by considering the properties of the geometric graph, in which an edge exists between any two vertices representing sensor nodes with overlapping sensing areas. The principal result is an approximation to the proportion of the sensing area that is covered by at least one sensing node, given the expected number of nodes per unit area in a two-dimensional Poisson process. The probability of a specified region being completely covered is also approximated. Simulation results corroborate the probabilistic analysis with low error, for any node density. The relationship between this approximation and noncoverage by the sensors is also examined. These results will have applications in planning and design tools for wireless sensor networks, and studies of coverage employing computational geometry.

Wireless sensor networks (WSNs) are being deployed for an ever-widening range of applications, particularly due to their low power consumption and reliability. In most applications, the sensor data must be sent over the Internet or core network, which results in the WSN backhaul problem. Passive optical networks (PONs), a next-generation access network technology, are well suited form a backhaul which interfaces WSNs to the core network. In this paper a new structure is proposed which converges WSNs and PONs. It is modelled and analyzed through queuing theory by employing two M/M/1 queues in tandem. Numerical results show how the WSN and PON dimensions affect the average queue length, thus serving as a guideline for future resource allocation and scheduling of such a converged network.

Ultra high performance digital media applications are now creating the need for new network architectures instead of shared IP links to provide them with dedicated, on-demand high capacity. These demands will be more critical in the future, as applications employing a real time environment become more common. Dynamic optical services at lambda and sub-lambda granularities are proposed because they guarantee the appropriate QoS in terms of bandwidth, jitter and latency. The technologies presented in this paper are merely initial considerations for use when providing the network services under investigation. Initial studies have shown that there is a lot of interest, and there are many advantages, in setting up network services to distribute high-performance media streams via optical networks to one or more locations.

This paper analyses the probability that randomly deployed sensor nodes triangulate any point within the target area. Its major result is the probability of triangulation for any point given the number of nodes lying up to a specific distance (2 units) from it, employing a graph representation where an edge exists between any two nodes close than 2 units from one another. The expected number of un-triangulated coverage holes, i.e. uncovered areas which cannot be triangulated by adjacent nodes, in a finite target area is derived. Simulation results corroborate the probabilistic analysis with low error, for any node density. These results will find applications in triangulation-based or trilateration-based pointing analysis, or any computational geometry application within the context of triangulation.

A distributed algorithm is introduced which detects and recovers holes in the coverage provided by wireless sensor networks. It does not require coordinates, requiring only minimal connectivity information (for example, whether any two nodes are within either the sensing radius or twice the sensing radius.) The radio communications area is assumed to be larger than the sensed area. Two active nodes are called neighbors, and are said to be connected by a link, if their distance lies between these two values. Redundant nodes are likely to exist inside an active node's sensing range. If all connected neighbors of some active node A can form a ring via links between them, there is no large hole inside the ring. Otherwise A is a boundary node of a large hole. All boundary nodes and most holes can be detected with very low probability of error, and simulation results suggest that redundant nodes are selected efficiently for activation when recovering the hole.

Several neural network architectures for computing parity problems are described. Feedforward networks with one hidden layer require N neurons in the hidden layer. If fully connected feedforward networks are considered, the number of neurons in the hidden layer is reduced to N/2. In the case of fully connected networks with neurons connected in cascade, the minimum number of hidden neurons is between log/sub 2/(N+1)-1 and log/sub 2/(N+1). This paper also describes hybrid neuron architectures with linear and threshold-like activation functions. These hybrid architectures require the least number of weights. The described architectures are suitable for hardware implementation since the majority of weights equal +1 and weight multiplication is not required. The simplest network structures are pipeline architectures where all neurons and their weights are identical. All presented architectures and equations were verified with MATLAB code for parity-N problems as large as N=100.

In this paper we present a heuristic tool known as the virtual permanent reconfiguring mechanism (VPRM) which dynamically reconfigures wavelength connections in a WDM transport network with a given physical topology. VPRM seeks to provide an optimised semi-permanent solution of the network configuration or logical topology, and traffic flow of a WDM transport network. The tool is activated in the event of changes within network environment; these may include: (i) normal network environment changes such as a change in traffic pattern, network expansion (introduction of new nodes or increase in wavelength availability) and line shutdown for maintenance purposes and, (ii) abnormal network environment changes such as sudden power failure, and network faults. Detection of the former network changes is achieved using a threshold detection technique. The latter, changes however are detected by means of fault alarm signals from distributed fault restoration mechanisms. The VPRM employs a combination of integer linear programming (ILP) and a heuristic approach. It essentially balances the traffic utilisation throughout the network physical structure and it dynamically ensures that spare connections are available, even after a fault, prior to completion of network maintenance. Unlike other reconfiguration techniques which involve solving ILP problems, the VPRM not only balances the load distribution throughout the physical network but also reduces the overall mean traffic level within each link.

This paper reports on a series of measurements carried out at 2 GHz designed to characterise a wide-band wireless local loop (WLL) in a range of environments. The wide-band channel measurements included data taken in both urban and suburban environments, using directional and omni-directional transmitting antennas, with and without movement near the receiver. The transmitters were building or pole-mounted with receivers inside various rooms in a variety of buildings. In all cases, the receiver was inside a building and no line of sight existed.

In three-divisional photonic switching networks, data is interchanged among channels that are formed in space, time, and wavelength. This paper presents several techniques for designing the architectures of small photonic switching modules that switch in three divisions. In one technique, Marcus' (1970) classical transformation from a space-only Clos network to a time-space-time architecture is generalized to the wavelength division and extended to three divisions. In another technique, Hunter's (see IEEE J. Lightwave Technol., vol.11, no.3, p.495-511, 1993) space-time transformation is also generalized and extended to three divisions. It is illustrated on several classical network architectures, including the most elementary case of a three-division photonic switching module based on the Benes (1965) architecture: an eight-by-eight network in which each side has two fibers and each fiber has two wavelengths and two timeslots.