A multi-objective optimization approach for sensor network design

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering and Computer Science


Pramod K. Varshney


Sensor network design, Sensor networks, Routing, Distributed detection, Sensor placement

Subject Categories

Electrical and Computer Engineering | Engineering


Many sensor network design problems are characterized by the need to optimize multiple conflicting objectives. However, existing approaches generally focus on a single objective (ignoring the others), or combine multiple objectives into a single function to be optimized, to facilitate the application of classical optimization algorithms. This restricts their ability and constrains their usefulness to the network designer. A much more appropriate and natural approach is to address multiple objectives simultaneously, applying recently developed multi-objective evolutionary algorithms ( MOEAs ) in solving sensor network design problems. This dissertation describes and illustrates this approach by modeling two sensor network design problems (mobile agent routing and sensor placement), as multi-objective optimization problems, developing the appropriate objective functions and discussing the tradeoffs between them.

One of the main contributions of this dissertation is the development of a new MOEA called evolutionary multi-objective crowding algorithm ( EMOCA ). The key new approach in this algorithm is to use a diversity-emphasizing probabilistic approach in determining whether an offspring individual is considered in the replacement selection phase, along with the use of a non-domination ranking scheme. EMOCA is evaluated using nine benchmark multi-objective optimization problems, and shown to produce non-dominated solutions with significant diversity, outperforming three state-of-the-art multi-objective evolutionary algorithms on most of the test problems.

The first part of the application section of the dissertation formulates and solves the multi-objective mobile agent routing problem in wireless sensor networks. A recent approach for data fusion in wireless sensor networks involves the use of mobile agents that selectively visit the sensors and incrementally fuse the data, thereby eliminating the unnecessary transmission of irrelevant or non-critical data. The order of sensors visited along the route determines the quality of the fused data and the communication cost. We model the mobile agent routing problem as a multi-objective optimization problem, maximizing the total detected signal energy while minimizing the energy consumption and path loss. Simulation results show that this problem can be solved successfully using multi-objective evolutionary algorithms such as EMOCA and non dominated sorting genetic algorithm - II ( NSGA-II ). We also demonstrate that EMOCA and NSGA-II outperform the classical multi-objective optimization approach called the weight based genetic algorithm (WGA). In WGA, a genetic algorithm is employed to optimize a weighted combination of the normalized values of the objectives.

In the second part of the application section of the dissertation, we formulate and solve the sensor placement problem for two specific application scenarios: (1) Energy efficient target detection; (2) Distributed detection of air pollutants in indoor environments subject to a constraint on the sensor network lifetime. The sensor placement problem for energy efficient target detection is modeled as a multiobjective optimization problem that addresses multiple optimization criteria including the probability of detection, energy dissipated in the network, and the optimal number of sensors to be deployed. We consider data-level fusion and decision fusion models for evaluating the multiple objectives. We solve the sensor placement problem using EMOCA and NSGA-II. Simulation results show that EMOCA and NSGA-II outperform WGA.

Many sensor network design problems have constraints on energy consumption, data accuracy, and data latency. Hence, it would be useful to develop a constrained multiobjective optimization framework to solve such problems. Such an approach is developed in this dissertation. To illustrate this approach, we model the sensor placement problem for distributed detection of air pollutants as a constrained multi-objective optimization problem. Our goal is to determine sensor locations for maximizing the detection probability and minimizing the number of sensors subject to a constraint on the sensor network lifetime. We develop mathematical models for the system level detection probability based on the diffusion of the air pollutant. We propose a new constrained multi-objective evolutionary algorithm called constrained evolutionary multi-objective crowding algorithm (C-EMOCA) for solving constrained multi-objective optimization problems. C-EMOCA and another well-known constrained multi-objective evolutionary algorithm (NSGA-II) are applied for solving the sensor placement problem. Simulation results show that the detection probability is influenced by the pollutant parameters and the constraint on the network lifetime. We also observe that C-EMOCA performs better than NSGA-II.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.