José Luis Jodrá

Service discovery mechanism over OLSR for mobile ad-hoc networks

This paper considers the subject of provisioning services in ad-hoc networks. Service discovery in these networks is challenging due to the lack of central infrastructure, limited resources and high mobility. We propose integrating this mechanism on the proactive OLSR (optimized link state routing) protocol. A new message type, called service discovery message (SDM), is added. This message will announce (servers) and request (clients) for services. Available local services will only be advertised once, with a specific lifetime, in order to avoid increasing the network overhead. These services will be stored in a local service cache that is maintained on each node. When a node wants to make use of a service, it will look on the cache. If it is not yet stored, it will send a query message. Simulation results are presented to demonstrate that the introduced packet overhead is insignificant compared with the standard OLSR protocol. These mechanisms must be efficient: every node should know the maximum number of services as possible. In networks with few nodes, the percentage of discovered services is around 90%. On the other hand, in bigger networks, was higher than the 80%. We believe that due to the obtained results, the following step consists on proving the protocol here presented in real scenarios using prototypes.

Efficient 3D Transpositions in Graphics Processing Units

Matrix transposition is a basic operation for several computing tasks. Hence, transposing a matrix in a computer’s main memory has been well studied since many years ago. More recently, the out-of-place matrix transposition has been performed efficiently in graphical processing units (GPU), which are broadly used today for general purpose computing. However, due to the particular architecture of GPUs, the adaptation of the matrix transposition operation to 3D arrays is not straightforward. In this paper, we describe efficient implementations for graphical processing units of the 5 possible out-of-place 3D transpositions. Moreover, we also include the transposition of the most basic in-place 3D transpositions. The results show that the achieved bandwidth is close to a simple array copy and is similar to the 2D transposition.

Implementation and Comparison of AODV and OLSR Routing Protocols in an Ad-Hoc Network over Bluetooth

Ad-Hoc networking is one of the most impacting architectures in the development of the Future Internet, where the nodes are expected to be wireless and mobile. Unlike the traditional IP-based and simulation-based approaches inside the Ad-Hoc research field, this work presents a real implementation of proactive (OLSR) and reactive protocols (AODV) over real Bluetooth devices (non-IP approach) in static and dynamic scenarios. Along the paper we point out many handicaps that have been solved in order to make Bluetooth work as supporting technology (the need for a previously established connection to, the lack of broadcast messages, the overhead...). We have carried out a set of experiments to compare the potential and limitations of both protocols. This work was developed inside the Future Internet project supported by the Basque Government within the ETORTEK Programme and the FuSeN project supported by the Spanish Ministerio de Ciencia e Innovacion.

A Study of Memory Consumption and Execution Performance of the cuFFT Library

The Fast Fourier Transform (FFT) is an essential primitive that has been applied in various fields of science and engineering. In this paper, we present a study of the Nvidias cuFFT library - a proprietary FFT implementation for Nvidias Graphics Processing Units - to identify the impact that two configuration parameters have in its execution. One useful feature of the cuFFT library is that it can be used to efficiently calculate several FFTs at once. In this work we analyse the effect this feature has on memory consumption and execution time in order to find a useful trade-off. Another important feature of the library is that it supports sophisticated input and output data layouts. This feature allows, for instance, to perform multidimensional FFT decomposition with no need of data transpositions. We have identified some patterns which may help to decide the parameters and values that are the key for achieving increased performance in a FFT calculation. We believe that this study will help researchers who wish to use the cuFFT library to decide what parameters values are best suited to achieve higher performance in their execution, both in time and memory consumption.

Service Discovery Protocol in Proactive Mobile Ad Hoc Networks

This paper proposes a service discovery protocol for discovering and advertising services in a proactive ad hoc network. The protocol we have defined is piggybacked into the OLSR protocol. We define a new message type into OLSR called Service Discovery Message (SDM) for both advertisement and discovery of services. The advertisement frequency and advertisement lifetime are user-controlled parameters, so that they can be modified depending on the user requirements. Each node maintains a service cache to store information about its own services, and the services each device discovers in the network. We also present simulation results of our protocol and show that the service discovery protocol defined here achieves much efficiency in discovering services, while it introduces practically no packet overhead compared to the basic OLSR protocol.

Analysis of several decision fusion strategies for clustering validation. Strategy definition, experiments and validation

We analysed the efficiency of CVIs decision fusion strategies for cluster validation.Decision fusion strategies over the CVIs perform better than single CVIs.Only the voting strategies that weight the CVIs beat the individual CVIs.The best CVI voting approach weights the impact of the experimental factors. In the unsupervised learning environment, the correct partition of data is not available, making it difficult to evaluate the performance of clustering algorithms. Therefore, one of the biggest challenges in this area is the validation of the results obtained by the algorithms. Amongst the various proposals currently under discussion, one of the most popular approaches is the one based on internal cluster validity indexes (CVIs). Comparative studies of such indexes show that there is no optimal CVI able to cope successfully with all the contexts. The aim of this work is to implement and analyse several decision fusion strategies over the CVIs studied in an extensive comparative work published in the bibliography, motivated by the success achieved by voting strategies in supervised learning. Thus, this experimental work consists of designing and implementing different CVI decision fusion strategies and then evaluating their performance in order to discover which of them are promising and eventually select the best one. Experiments with several strategies showed that the majority of the decision fusion approaches designed cope with the diversity of contexts more effectively than single CVIs. Display Omitted

Solving Poisson's equation using FFT in a GPU cluster

Poissons equation is present in many scientific computations and its efficient solution is achieved by means of several methods. One of the most efficient methods is the Fast Fourier Transform (FFT), which is very widely used in lots of computational problems. In this work we implement a Poisson solver that uses FFT as base method and runs in a cluster of Graphics Processing Units (GPU). We analyze the execution of our implementation to find the main bottlenecks and we compare the results to a CPU based solver. The results show a good scalability up to 16 GPUs, the number of GPUs we have available. Moreover, although the main time consuming part of the process is the network communication, the GPU implementation is about 2.5 times faster than the CPU implementation. We implement the solution of Poissons equation using FFT in a cluster of GPUs.We overlap computation, inter-node communication and CPU/GPU communication.The results show good scalability up to 16 GPUs.Our implementation is about 2.5 times faster than an optimized CPU based solver.