Sanjib Sadhu

Modeling broadcasting using omnidirectional and directional antenna in delay tolerant networks as an epidemic dynamics

We study broadcasting of information in a system of moving agents equipped with omnidirectional as well as directional antenna. The agent communication protocol is inspired by the classical SIRS epidemics dynamics. We assume that the antennas of all agents have a fixed transmitting power, while signal reception only occurs when the receivers sense signals with power exceeding a certain threshold. Thus, information exchange is a local phenomenon which depends on the relative distance and antenna orientation between the transmitting and the receiving agent. We derive an expression for the mean broadcasting time and study the information dissemination robustness of the system using elements of classical epidemiology and physics. In particular, we show that the mean broadcasting time depends upon ¿ which quantifies the area the radiation pattern of the antenna sweeps as it moves. We report three important observations (a) directional antennas perform better than omnidirectional antennas, (b) directional antennas whose beam-width is narrower perform even better, and (c) the performance enhances a lot if directional antennas rotate. These behaviors can be understood in the light of the reported analytical findings.

A predictive location management scheme based on semi-random mobility model

Predictive location management schemes based on the user mobility patterns are preferred over non-predictive ones as the paging cost drastically decreases. In existing predictive schemes a MT is expected to be intelligent as it is assigned a lot of computation load which may otherwise be easily shared by the MSC. Furthermore some schemes require the regional route maps. We propose a predictive scheme which assigns more load to the MSC than MT. The MT stores only current day movement log and sends this log to MSC during night. The MSC computes the unique sub-patterns, and identifies hot cell, i.e. the last cell of the sub-patterns which is sent to the MT. The MT sends update when it leaves a hot cell which enables the MSC to identify the sub-pattern to be followed next. The MSC pages only a cell corresponding to the sub-pattern when call arrives. If not found pages a set of probable cells. An analytical study considering semi-random mobility model gives the total location management cost using the proposed scheme

GRP_CH Heuristic for Generating Random Simple Polygon

A heuristic ‘GRP_CH’ has been proposed to generate a random simple polygon from a given set of ‘n’ points in 2-Dimensional plane. The “2-Opt Move” heuristic with time complexity \(\mathcal{O}(n^4)\) is the best known (referred in [1]) among the existing heuristics to generate a simple polygon. The proposed heuristics, ‘GRP_CH’ first computes the convex hull of the point set and then generates a random simple polygon from that convex hull. The ‘GRP_CH’ heuristic takes \(\mathcal{O}(n^3)\) time which is less than that of “2-opt Move” heuristic. We have compared our results with “2-Opt Move” and it shows that the randomness behaviour of ‘GRP_CH’ heuristic is better than that of “2-Opt Move” heuristic.

A linear time algorithm to cover and hit a set of line segments optimally by two axis-parallel squares.

This paper discusses the problem of covering and hitting a set of line segments

Computing the Triangle Maximizing the Length of Its Smallest Side Inside a Convex Polygon

\cal L

Optimal Covering and Hitting of Line Segments by Two Axis-Parallel Squares

in

An alternative approach for computing monotone polygon

{\mathbb R}^2

Generating Empty Convex Polygon Randomly from a Subset of Given Point Set

by a pair of axis-parallel squares such that the side length of the larger of the two squares is minimized. We also discuss the restricted version of covering, where each line segment in

A predictive location management scheme by extracting the unique sub-patterns from the mobility logs

\cal L

Random Polygon Generation through Convex Layers

is to be covered completely by at least one square. The proposed algorithm for the covering problem reports the optimum result by executing only two passes of reading the input data sequentially. The algorithm proposed for the hitting and restricted covering problems produces optimum result in