A Cross-Layer Paradigm for Throughput and Energy Optimization in Distributed and Application-Centric Wireless Sensor Networks

This thesis presents a cross-layer policy for joint throughput and energy optimization in next-generation wireless sensor networks (WSNs). We derive the link suitability of a transmission channel, which determines the optimal physical (PHY) layer resource allocation that jointly maximizes achievable throughput and minimizes the associated power cost of using the link. This differs from current approaches that fit into one of two categories: those that perform rate control but do not model power cost in terms of the achievable throughput on the transmission link, or those that do not separate performance-related benefits and energy-related costs to ensure that we can jointly maximize throughput and minimize energy consumption. Our method ensures that WSNs allocate transmission resources that impact limited sensor energies to operating channels capable of a higher level of throughput and energy conservation.;A Network Planning Protocol (NPP) works in conjunction with our cross-layer policy and contributes the relationship between: expected throughput performance, network lifetime, message load for data and coordination packets, and energy capacity. This relationship enables network planners of next-generation WSNs to study the impact of various network design decisions on performance and network lifetime metrics. This is achieved by deriving NPP in terms of physical resources that affect throughput and energy capacity such that we may allocate resources appropriately to meet network objectives.;We present these contributions for distributed and application-centric WSNs. Application-centric WSNs refer to sensor networks that execute applications with significant performance requirements, while incorporating energy conservation schemes for sensors. In many instances, it shall also be necessary to execute many of these sensory applications simultaneously, with centralized control of geographically distributed sensor clusters. As a result, this thesis explores the use of overlay mesh networks to design a two-tier, application-centric and distributed WSN architecture.;Using link suitability in a cross-layer policy, we contribute a cross-layer optimization protocol that is unique in identifying optimal routing paths with the highest trade-off between throughput benefits and energy costs across multiple transmission channels. Our policy is based on transmit power at the PHY layer, channel allocation at the link layer, and multi-hop path selection at the network layer.