Salahuddin Mohammad Masum

Asynchronous leader election in mobile ad hoc networks

With the proliferation of portable computing platforms and small wireless devices, the classical dilemma of leader election in mobile ad hoc networks has received attention from the research community in recent years. The problem aims to elect a unique leader among mobile nodes regardless of their physical locations. But, existing distributed leader election algorithms do not cope with highly spontaneous nature of mobile ad hoc networks. This paper presents a consensus-based leader election algorithm that finds a local extrema among the nodes participating in leader election. The algorithm is highly adaptive with ad hoc networks in the sense that it can tolerate intermittent failures, such as link failures, sudden crash or recovery of mobile nodes, network partitions, and merging of connected network components associated with ad hoc networks. The paper also presents proofs of correctness to exhibit the fairness of this algorithm.

A consensus-based l -Exclusion algorithm for mobile ad hoc networks

This paper addresses the @?-Exclusion problem for mobile ad hoc networks. The @?-Exclusion problem, a generalization of distributed mutual exclusion problem, involves a group of processes, each of which intermittently requires access to one of @? identical resources or pieces of code called the critical section (CS). This paper presents a consensus-based mobility-aware @?-Exclusion (LE) algorithm that operates asynchronously and copes explicitly with arbitrary (possibly concurrent) topology changes associated with such networks. The algorithm can tolerate link changes or failures, sudden crashes or recoveries of at most @?-1 mobile nodes. The algorithm is based on collection of enough consensuses for a mobile node intending to enter CS, and uses diffusing computations for this purpose. This paper presents a simulation to demonstrate that the proposed algorithm, as compared to the k-Reverse Link (KRL) algorithm, is quite effective to variety of operating conditions, and is highly adaptive to frequent and unpredictable topology changes due to link changes or failures.

Hierarchical Numbering Based Addressing and Stateless Routing Scheme for Wireless Sensor Networks

Due to energy and other relevant constraints, addressing of nodes and data routing techniques in sensor networks differ significantly from other networks. In this article, we present an energy-efficient addressing and stateless routing paradigm for wireless sensor networks. We propose a dynamic and globally unique address allocation scheme for sensors in such a way that these addresses can later be used for data routing. We build a tree like organization of sensors rooted by the sink node based on their transmission adjacency and then set labels on each sensor with a number according to the preorder traversal of the tree from the root. In this addressing process, each sensor keeps necessary information so that they can later route data packets to the destination depending on these addresses, without keeping the large routing table and running any no route discovery phase. Moreover, the scheme does not use location information as well (as done by geo-routing) and can be used in the indoor environment. We conduct simulations to measure the soundness of our approach and make a comparison with another similar technique TreeCast. Simulation results reveal that our approach performs better than its counterpart in several important performance metrics like address length and communication energy.

Distributed Allocation of Identical Resources in Mobile Ad Hoc Networks

This paper addresses the l—Exclusion problem for mobile ad hoc networks. The l—Exclusion problem, a generalization of distributed mutual exclusion problem, involves a group of processes, each of which intermittently requires access to one of l identical resources or pieces of code called the critical section (CS). In literature, few token—based solutions to this l—Exclusion problem are available. Nevertheless, these solutions suffer from poor failure resiliency, as these do not consider failures associated with mobile ad hoc networks, such as loss or regeneration of tokens, crash or sudden recovery of nodes. This paper presents a consensus—based mobility—aware l—exclusion (LE) algorithm that operates asynchronously and copes explicitly with arbitrary (possibly concurrent) topology changes associated with such networks. The algorithm can tolerate link changes or failures, sudden crashes or recoveries of at most l—1 mobile nodes. The algorithm is based on collection enough consensuses for a mobile node intending to enter CS, and uses diffusing computations for this purpose. The algorithm requires nodes to communicate only with their current neighbors, making it well—suited for use in mobile ad hoc networks. This paper presents a simulation study to demonstrate that the proposed algorithm, as compared to the k—Reverse Link (KRL) algorithm, is quite effective to variety of operating conditions, and is highly adaptive to frequent and unpredictable topology changes due to link changes or failures or formations, sudden crashes or recoveries of at most l—1 mobile nodes, under different mobility settings.

Maintaining a Binary Tree Structure For Mobile Ad Hoc Computing

Mobile ad hoc (multi-hop) wireless network, unlike conventional infrastructured wireless counterparts, operates in the absence of fixed switching stations and thus all networking entities therein can be mobile. In mobile ad hoc network, logical binary tree structure is likely to become volatile or expensive to maintain over time due to changeable network topology. Additional adverse effects take place when a node joins or leaves the computation in the presence of mobility. This paper presents a distributed algorithm that maintains a binary tree among mobile nodes to the network dynamics to reflect overall communication efficiency. This is achieved by modifying the tree structure in a localized, mutual exclusive fashion, thereby allowing for concurrent segment-wise modifications to proceed. Remarkably our proposal operates without global knowledge of the logical structure and can be embodied as an underlying protocol layer that supports transparent deployments of conventional algorithms in mobile environment. Moreover, correctness proofs show that our proposal is promising.

Dynamic Load Balancing for Mining of Molecular Structures using Genetic Algorithm

Graph mining techniques for analyzing large collections of molecules to find regularity or patterns among molecules of a specific class, such as finding common properties in large numbers of drug candidates, finding molecular features that inhibit the desired reaction etc. is an important research issue in bioinformatics as well as molecular informatics. In this context, finding frequent graphs has received increasing attention over the past years. But, the computational complexity of the underlying problem and the large amount of data to be explored essentially render traditional sequential algorithms practically useless. To address such problems a distributed algorithm is adopted to find the frequent sub-graphs and to discover interesting patterns in molecular compounds. However, this problem is characterized by a highly irregular search tree, whereby reliable workload prediction is very hard. Therefore, a genetic algorithm (GA) is proposed to solve the dynamic load-balancing problem of highly irregular search tree.

Asynchronous \ell-Exclusives in Mobile Ad Hoc Networks

A solution to the /spl lscr/-exclusion problem for mobile ad hoc networks aims to provide access and control over /spl lscr/ identical copies of critical resources among mobile nodes. In literature, few token-based solutions to this problem are available. However, these solutions do not consider failures, such as loss or regeneration of tokens, crash or sudden recovery of nodes. This paper presents a consensus-based /spl lscr/-exclusion algorithm that explicitly copes with mobility associated with such networks. The algorithm is fault-resilient in the sense that it can tolerate loss of messages, link failures, sudden crash or recovery of at most /spl lscr/-1 mobile nodes. This paper presents a simulation study considering several performance metrics that can significantly impact the behavior of such an algorithm in different ad hoc settings. This paper also presents a proof of correctness and compares the algorithm with existing ones.

Asynchronous ℓ-exclusion in mobile ad hoc networks

A solution to the /spl lscr/-exclusion problem for mobile ad hoc networks aims to provide access and control over /spl lscr/ identical copies of critical resources among mobile nodes. In literature, few token-based solutions to this problem are available. However, these solutions do not consider failures, such as loss or regeneration of tokens, crash or sudden recovery of nodes. This paper presents a consensus-based /spl lscr/-exclusion algorithm that explicitly copes with mobility associated with such networks. The algorithm is fault-resilient in the sense that it can tolerate loss of messages, link failures, sudden crash or recovery of at most /spl lscr/-1 mobile nodes. This paper presents a simulation study considering several performance metrics that can significantly impact the behavior of such an algorithm in different ad hoc settings. This paper also presents a proof of correctness and compares the algorithm with existing ones.