Location information is crucial for the correct interpretation of data collected through wireless sensor networks (WSNs). The de facto system for wireless localization, Global Positioning System (GPS) does not work properly in indoor environment, thus researchers are thriving to find other localization schemes for indoor WSNs. The main goal of this work is to study and design three-dimensional (3D) wireless localization schemes for indoor applications. In this thesis, a new and accurate, efficient and cost-effective algorithm, called parametric loop division (PLD) has been proposed for localizing static nodes within a WSN. In the proposed technique, reference points can help to produce new parametric points by calculating the mid points and by taking step size that falls within the network boundary. The objective of PLD scheme is to estimate the actual localization volume and find the node position in 3D space by using subdivision method. In each step, triangles are subdivided into pairs with the addition of extraordinary nodes in its control ring matrix. Parametric points are generated by using the step size and RSSI is compared with threshold value for localization. The work involves the development of novel solution which utilizes the anchor node position information to calibrate nodes with unknown target, allowing it to work even in a changing environment with increased reliability and accuracy.

Subsequently, PLD is evaluated in presence of different types of noises. Firstly, the localization accuracy was tested without the addition of noise in distance measurement. Like other schemes, PLD is adversely affected by the noise, which reduces the accuracy of the system. A new framework with extended kalman filtering (EKF) is proposed to refine the nodes coordinates affected by the noise. Furthermore, an analytical framework is presented with the detailed study of lower bound of the localization accuracy. The PLD is tested for naive, Gaussian and intelligent noise. The anchor node is modelled by only using the knowledge information of coordinates to redesign the distance vertex from anchor to parametric points.

Finally, we consider the mobile based localization scheme, which has become popular recently with the development of autonomous robots and unmanned aerial vehicles. We designed an extended centroid based localization system that use the weight on distance to compute the signal power. A fuzzy logic approach is adopted for computation. The design is divided in to In the first phase, RSSI is mapped to fuzzy membership function. The mobile anchor exchange beacon and measure distance using RSSI data. The target node position is computed in a circle within the sensing region for a mobile anchor node, which moves on a random walk for broadcasting beacons. RSSI and signal power is used as an input for fuzzy system. In the second phase, for accurate node positioning a perpendicular bisector is drawn from rough estimation to circle drawn previously. Like EKF, fuzzy logic works well in nonlinear estimation of target nodes locations.

Localization problem is evolving with the advance of mobile technologies and this thesis contributes to the fast development of this topic. However, there are still some issued left out as future study, mainly on the effect of anchor node localization error, implementing mobile anchor in a PLD algorithm and energy-aware localization schemes in WSNs.