Given the potential benefits offered by wireless sensor networks(WSNs), they are becoming an appealing technology for process,manufacturing, and industrial control applications. In thisthesis, we propose a novel approach to WSN protocol design forcontrol applications. The protocols are designed to minimize theenergy consumption of the network, while meeting reliability andpacket delay requirements. The parameters of the protocol areselected by solving a constrained optimization problem, where theobjective is to minimize the energy consumption and theconstraints are the probability of successful packet reception andthe communication delay. The proposed design methodology allowsone to perform a systematic tradeoff between the controlrequirements of the application and the network energyconsumption. An important step in the design process is thedevelopment of analytical expressions of the performanceindicators. We apply the proposed approach to optimize the networkfor various communication protocols.

In Paper A, we present an adaptive IEEE 802.15.4 for energyefficient, reliable, and low latency packet transmission. Thebackoff mechanisms and retry limits of the standard are adapted tothe estimated channel conditions. Numerical results show that theproposed protocol enhancement is efficient and ensures a longerlifetime of the network under different conditions. Furthermore,we investigate the robustness and sensitivity of the protocol topossible errors during the estimation process.

In Paper B, we investigate the design and optimization ofduty-cycled WSNs with preamble sampling over IEEE 802.15.4. Theanalytical expressions of performance indicators are developed andused to optimize the duty-cycle of the nodes to minimize energyconsumption while ensuring low latency and reliable packettransmissions. The optimization results in a significant reductionof the energy consumption compared to existing solutions.

The cross-layer protocol called Breath is proposed in Paper C. Theprotocol is suitable for control applications by using theconstrained optimization framework proposed in the thesis. It isbased on randomized routing, CSMA/CA MAC, and duty-cycling. Theprotocol is implemented and experimentally evaluated on a testbed,and it is compared with a standard IEEE 802.15.4 solution. Breathexhibits a good distribution of the work load among the networknodes, and ensures a long network lifetime.