Banani Saha

Abstract: Mobile Ad-hoc Networks are a collection of two or more devices equipped with wireless communications and multi-hop networking capability. We compare the performance of two prominent on demand routing protocols for mobile ad hoc networks--Dynamic Source Routing (DSR) and Ad Hoc On-Demand Distance Vector Routing (AODV). We exhibit that though DSR and AODV share an inherent on-demand behavior, the nuances in their protocol mechanics lead to performance differentials which are analyzed by varying the network size, network load and mobility. The simulation is carried out in NS2 where we employ the new trace format [1] to base our comparisons. We present a novel approach to analyze the protocols by varying parameters and making comparison in multiple dimensions to facilitate accurate observations. Additionally, we anatomize AODV which performs admirably in all but low mobility situations and make recommendations to enhance performance.

Abstract: The major aim of this paper is possible analysis of fraud activities prevalent in the telecommunication industry particularly regarding subscription fraud problems. A state variable model has been designed for analysis. This model is the basis and explains the necessity of profiling customer's details and classifying and slabbing flat security deposits. This model is logically explained and is evaluated on the basis of a sample data set. Moreover predictability and interpretability of different graphs drawn on the basis of the data is explained with approximate formula. Finally the question of synthetic data generation for construction of a suitable learning rule has also been addressed.

Abstract: This paper proposes a new parallel algorithm for the maximal elements problem with no constraints. It is proposed for a linear array with reconfigurable pipelined bus system (LARPBS) model and on its latest variant, LARPBS(p) model also. It runs in O(log log n ldr log n) time with O(n) processors. Its significance is that it works for any instance of the problem with no constraint laid in [2].

Abstract: This paper presents a novel approach to analyse the traffic efficiency of infrastructureless networks. Wireless communication is rapidly moving towards more infrastructureless routing schemes, where the devices encounter frequent topology changes and communicate with each other anonymously and sporadically. Nodes do not have the opportunity to choose an optimum routing path, they do rather obtain a valid routing path on the fly during the data exchange by cooperating with the nearest neighbours. Packets emanated from an infrastructureless node are disseminated through each node found in the vicinity. This produces a vast number of packet duplicates within the entire communication environment, causing the problem of increased congestion, increased packet delay, more packet drops, and undesirable energy consumption compared to the infrastructured networks. Data exchange and packet routing operations in large networks are highly stochastic. Thus, the performance analysis of routing operations is modeled as a stochastic queueing system, where the input and output traffics and throughput operations of routing systems are precisely formulated in the form of stochastic processes. A hypothetical routing algorithm is also defined here as a reference model in order to perform a comparative analysis of the infrastructureless routing scheme.

Abstract: This paper presents the formal modelling and verification of the Ad-hoc On-demand Distance Vector (AODV) routing protocol. Our study focuses on the quantitative aspects of AODV, in particular the influence of uncertainty (such as packet loss rates, collisions) on the probability to establish short routes. We present a compositional model of AODV’s functionality using probabilistic timed automata. The strength of this model is that it combines hard real-time constraints with randomised protocol behaviour and can deal with non-determinism (due to e.g., queue behaviours at network nodes). An automated analysis by probabilistic model checking provides useful insights on the sensitivity of AODV’s ability to establish shortest/longest routes and deliver data packets via such routes.