A stochastic analysis of greedy routing in a spatially dependent sensor network

TL;DR

This paper models message routing in a sensor network with spatially-dependent node density using stochastic processes, providing analytical tools and simulations to understand hop behavior and the impact of sleep schemes.

Contribution

It introduces a stochastic model using inhomogeneous Poisson processes for spatially-dependent sensor networks and analyzes hop behavior with novel integral and approximation methods.

Findings

Dependence between hops is effectively modeled with only the previous node.

Sleep schemes reduce model complexity and improve efficiency.

Model predictions align closely with simulation results.

Abstract

For a sensor network, a tractable spatially-dependent node deployment model is presented with the property that the density is inversely proportional to the sink distance. A stochastic model is formulated to examine message advancements under greedy routing in such a sensor network. The aim of this work is to demonstrate that an inhomogeneous Poisson process can be used to model a sensor network with spatially-dependent node density. Elliptic integrals and asymptotic approximations are used to describe the random behaviour of hops. Types of dependence that affect hop advancements are examined. We observe that the dependence between successive jumps in a multihop path is captured by including only the previous forwarding node location. We include a simple uncoordinated sleep scheme, and observe that the complexity of the model is reduced when enough nodes are asleep. All expressions involving multidimensional integrals are derived and evaluated with quasi-Monte Carlo integration methods based on Halton sequences and recently-developed lattice rules. An importance sampling function is derived to speed up the quasi-Monte Carlo methods. The ensuing results agree extremely well with simulations.