Jean Frédéric Myoupo

ALL-TO-ALL BROADCASTING ALGORITHMS ON HONEYCOMB NETWORKS AND APPLICATIONS

This paper presents two simple all-to-all broadcasting algorithms on honeycomb mesh. Consider a network with n processors, one has personalized routing strategy at each node and it requires a 3n communication time complexity. This communication time can be reduced to n because the computation time is always assumed to be much lower than the communication time. The other is based on a Hamiltonian path and has a 2n communication time complexity. We show how they can be used to get parallel solutions to a class of problems on honeycomb networks, among others Prefix Sums, Maximal Vectors, Maximal Sum Subsegment, Parenthesis Matching, Decoding Binary Tree, and Sorting. In our knowledge, these all-to-all broadcast algorithms are the only ones so far exhibited on a honeycomb.

A basis for 3-D cellular networks

We propose to extend the standard concept of planar cellular networks into space. Indeed, in cellular networks the trend is to have a smaller cells to meet the growing number of communication services. The smaller the cells, the more important is the third dimension because this model is more efficient. So, it is better to take the height into account. For instance, to create a cellular network in a building, it makes no sense to have a planar cellular network. It would be better to have antennas placed in the three dimensions. This paper presents some explanations about the reason why hexagons tessellation are used in the theory of cellular networks. It describes the 3-D cellular networks used for this work and also discusses the frequency reuse mechanism and channel allocation schemes.

A coarse-grained multicomputer algorithm for the longest common subsequence problem

The paper presents a coarse-grained multicomputer algorithm that solves the Longest Common Subsequence Problem. This algorithm can be implemented in the CGM with P processors in O(N/sup 2//P) in time and O(P) communication steps. It is the first CGM algorithm for this problem. We present also experimental results showing that the CGM algorithm is very efficient.

An Algorithm for a Constraint Optimization Problem in Mobile Ad-hoc Networks

A mobile ad-hoc network is considered as a dynamic autonomous system composed of mobile devices interconnected by links without wire, without the use of a fixed infrastructure and without centralized administration. The absence of a centralized infrastructure forces each device to work in a peer to peer distributed environment, and to act as a router to relay communications, or to generate its own data. The management of the network thus is strongly distributed on all elements of the network. In this paper, we present a modelling of the mobile ad-hoc network (MANET) problem in form of a constraint satisfaction/optimization problem called CSPADhoc. Then, to minimize the consumption of batteries for devices, we describe an approach based on an adaptation of the A star algorithm to the MANET problem called (MANET-Astar). Finally, we present some experimental results using our approach

Time-Efficient Parallel Algorithms for the Longest Common Subsequence and Related Problems

Recently Aklet al. introduced a new model of parallel computation, called broadcasting with selective reduction (BSR), and showed that it is more powerful than any CRCW PRAM and yet requires no more resources for implementation than even EREW PRAM. The model allows constant time solutions to sorting, parallel prefix, and other problems. In this paper, we describe constant time solutions to the longest common subsequence problem and the sequence alignment problem using the BSR model. These are the first constant time solutions to these problems for any model of computation.

HIGHER DIMENSIONAL HONEYCOMB NETWORKS

We define the higher dimensional honeycomb graphs as a generalization of hexagonal plane tessellation, and consider it as a multiprocessor interconnection network. A 3-D honeycomb mesh network with n nodes has degree 4 and diameter approximately 3.63n. The network cost, defined as the product of degree and diameter, is about 20 percents better for the 3-D honeycomb than for the 3-D mesh. We describe the addressing scheme, the routing and broadcasting algorithms for three-dimensional and higher dimensional honeycombs. Furthermore, a formula for the diameter of a higher dimensional honeycomb network of given size is determined.

An Energy-Efficient Clique-Based Geocast Algorithm for Dense Sensor Networks

This paper proposes an energy-efficient geocast algorithm for wireless sensor networks with guaranteed de-livery of packets from the sink to all nodes located in several geocast regions. Our approach is different from those existing in the literature. We first propose a hybrid clustering scheme: in the first phase we partition the network in cliques using an existing energy-efficient clustering protocol. Next the set of clusterheads of cliques are in their turn partitioned using an energy-efficient hierarchical clustering. Our approach to con-sume less energy falls into the category of energy-efficient clustering algorithm in which the clusterhead is located in the central area of the cluster. Since each cluster is a clique, each sensor is at one hop to the cluster head. This contributes to use less energy for transmission to and from the clusterhead, comparatively to multi hop clustering. Moreover we use the strategy of asleep-awake to minimize energy consumption during extra clique broadcasts.

A faster linear systolic algorithm for recovering a longest common subsequence

Abstract We present a new linear systolic array architecture of m cells which outputs a longest common subsequence (LCS) of two input strings A and B in time n + 2 m , where n and m denote the lengths of A and B respectively ( m ⩽ n ). Our approach improves the time of execution required by previous linear systolic arrays for this purpose. Furthermore, a design combining a tree with the linear array provides an LCS and its length in n + m + log m clock cycles only.

A fault-tolerant permutation routing algorithm in mobile ad-hoc networks

Mobile ad-hoc networks (MANET) is a distributed systems formed by wireless mobiles nodes, that can dynamically self organize into arbitrary and temporary networks with no pre-existing communication infrastructure. After initialization, one of a crucial issue in ad-hoc network, is how to transmit items held by the stations to their destinations called Permutation Routing. As mobiles devises are dependant on battery power, it is important that this permutation routing minimize their energy consumption. But, it can occur during the permutation routing that some stations develop faults and stop working. The existence of these faulty nodes can significantly affect the packets delivery rate. If a faulty node, participating in a permutation routing operations, drops packets, all these packets will be lost. Hence, the goal of a fault tolerant permutation routing is to provide certain packet delivery guarantee in spite of the presence of faulty stations. The main contribution of this paper is to present a protocol that provided an adequate level of fault tolerant and minimized energy during the permutation routing. Unlike in [2], in this work, in presence of faulty nodes, items can be transmitted to their destinations without loss.

An Application of an Initialization Protocol to Permutation Routing in a Single-Hop Mobile Ad Hoc Networks

In 1999 Nakano, Olariu, and Schwing in [20], they showed that the permutation routing of n items pretitled on a mobile ad hoc network (MANET for short) of p stations (p known) and k channels (MANET{(n, p, k)) with k < p, can be carried out in