Wireless Sensor Networks (WSN) are playing a key role in the efficient operation of Cyber Physical Systems (CPS). They provide cost efficient solutions to current and future CPS re- quirements such as real-time structural awareness, faster event localization, cost reduction due to condition based maintenance rather than periodic maintenance, increased opportunities for real-time preventive or corrective control action and fine grained diagnostic analysis. However, there are several critical challenges in the real world applicability of WSN. The low power, low data rate characteristics of WSNs coupled with constraints such as application specified latency and wireless interference present challenges to their efficient integration in CPSs. The existing state of the art solutions lack methods to address these challenges that impediment the easy integration of WSN in CPS.

This dissertation develops efficient resource management algorithms enabling WSNs to perform reliable, real-time, cost efficient monitoring. This research addresses three important problems in resource management in the presence of different constraints such as latency, precedence and wireless interference constraints. Additionally, the dissertation proposes a solution to deploy WSNs based real-time monitoring of critical infrastructure such as electrical overhead transmission lines.

Firstly, design and analysis of an energy-aware scheduling algorithm encompassing both computation and communication subsystems in the presence of deadline, precedence and in- terference constraints is presented. The energy-delay tradeoff presented by the energy saving technologies such as Dynamic Voltage Scaling (DVS) and Dynamic modulation Scaling (DMS) is studied and methods to leverage it by way of efficient schedule construction is proposed. Performance results show that the proposed polynomial-time heuristic scheduling algorithm offers comparable energy savings to that of the analytically derived optimal solution.

Secondly, design, analysis and evaluation of adaptive online algorithms leveraging run- time variations is presented. Specifically, two widely used medium access control schemes are considered and online algorithms are proposed for each. For one, temporal correlation in sensor measurements is exploited and three heuristics with varying complexities are proposed to perform energy minimization using DMS. For another, an adaptive algorithm is proposed addressing channel and load conditions at a node by influencing the selection of either low energy or low delay transmission option. In both cases, the simulation results show that the proposed schemes provide much better energy savings as compared to the existing algorithms.

The third component presents design and evaluation of a WSN based framework to mon- itor a CPS namely, electrical overhead transmission line infrastructure. The cost optimized hybrid hierarchical network architecture is composed of a combination of wired, wireless and cellular technologies. The proposed formulation is generic and addresses constraints such as bandwidth and latency; and real world scenarios such as asymmetric sensor data generation, unreliable wireless link behavior, non-uniform cellular coverage and is suitable for cost minimized incremental future deployment.

In conclusion, this dissertation addresses several challenging research questions in the area of resource management in WSNs and their applicability in future CPSs through associated algorithms and analyses. The proposed research opens up new avenues for future research such as energy management through network coding and fault diagnosis for reliable monitoring.