Abstract: An important problem in wireless sensor networks is to find the minimal number of randomly deployed sensors making a network connected with a given probability. In practice sensors are often deployed one by one along a trajectory of a vehicle, so it is natural to assume that arbitrary probability density functions of distances between successive sensors in a segment are given. The paper computes the probability of connectivity and coverage of 1-dimensional networks and gives estimates for a minimal number of sensors for important distributions.
Introduction
Wireless sensor networks (WSNs) rely on two fundamental metrics: connectivity and coverage. Connectivity ensures every node can communicate with a central sink, while coverage ensures the entire target area is monitored. In 1D scenarios, such as monitoring roads, rivers, or borders, sensors are typically deployed sequentially from a moving vehicle.
Existing models often assume uniform distributions or focus on asymptotic behavior as the number of sensors $n \to \infty$. However, real-world deployment involves fluctuations in vehicle speed or manual dropping, leading to complex probability distributions for the distances between successive sensors. This paper addresses the practical need for exact calculations of the number of sensors required to achieve a target probability of connectivity and coverage for arbitrary distance distributions.
Methodology
The core of the proposed framework is modeling the distances between successive sensors ($y_i$) rather than their absolute positions.
- Recursive Probability Function: The authors define a recursive function representing the probability that $n$ sensors are "properly" deployed within a segment of length $L$ and are connected within a transmission radius $R$.
- Laplace Transforms: To handle the convolutions resulting from summing random variables, Laplace transforms are used to convert these operations into simpler algebraic products, enabling the derivation of exact formulas for connectivity probability ($P_n$).
- The Universal Approach: By utilizing piecewise constant densities, the authors can approximate any arbitrary probability distribution. This allows the framework to handle virtually any real-world deployment pattern while still providing exact probability calculations.
- Distribution-Specific Analysis: The paper derives specific formulas for common distributions, including Uniform, Truncated Exponential, and Truncated Normal.
Key Results and Insights
The research provides the first exact analytical formulas for the probability of connectivity in finite 1D networks. A significant finding is the distinction between connectivity and coverage; a network may be fully connected but fail to cover the entire segment if the final sensor is too far from the boundary.
Numerical simulations validate these theorems. For a 1km segment with a 50m transmission radius, a perfectly regular deployment requires only 21 sensors. In contrast, a uniform random deployment requires approximately 157 sensors to reach a 95% probability of connectivity. The results demonstrate that higher variance in sensor placement significantly increases the required sensor count. Furthermore, the authors show that 2D "road" scenarios can be reduced to 1D problems using the Pythagorean theorem, extending the applicability of their results.
Conclusion and Future Work
While the 1D model is highly effective for linear infrastructure, it does not directly apply to open 2D or 3D environments. Additionally, the current model assumes static sensors and fixed transmission radii. Future research will explore extending this "universal approach" to non-straight trajectories and identifying optimal distance distributions to minimize sensor count while maintaining high network reliability.
Summary updated