While it has been shown theoretically that stochastic network optimization techniques utilizing queue backpressure can result in high performance cross-layer protocols, there has been limited prior work on translating them into practice. In this exploratory research project, we develop and demonstrate a novel cross-layer backpressure protocol stack for wireless sensor networks, building on our implementation of the backpressure collection protocol (BCP). The possible gains of this highly agile approach to cross-layer wireless networking are substantial and include increased throughput even in the presence of mobility, handling of bursty external interference events with reduced losses, and reduced protocol implementation complexity. We incorporate techniques for MAC-layer backoff prioritization, transport layer utility optimization, mechanisms to handle various kinds of network dynamics, and interoperability with asynchronous low power sleep schedules. Protocol implementation code developed in the project will be released publicly as open-source. It is expected that the successful completion of this project will provide an important proof-of-concept that stochastic optimization theory-driven protocols can be successful in practice.
Wireless sensor networks allow smart objects embedded in some environment to measure and communicate meaningful information about that environment. Potential applications of wireless sensor networks that can benefit society range from precision agriculture (where sensors can measure soil moisture content to enable farmers to make informed decisions to maximize the quality and yield of produce) to smart buildings (which can operate energy efficiently while maximizing the comfort experienced by their occupants) to networks of vehicles equipped with radios (communicating road safety information to each other so that drivers can be warned in time to be better prepared to deal with adverse conditions). In this project, we developed and implemented a novel class of data communication protocols for wireless sensor networks based on a data packet scheduling technique known as backpressure scheduling. In backpressure scheduling, packet transmission decisions are made on the basis of packet queue-size differences between neighboring nodes in a network. While it has been known for some time through mathematical investigations that backpressure scheduling is efficient in terms of maximizing data throughput, there has been a gap between theory and practice because there have been very few real software system implementations of this approach to network protocol design. We have worked towards bridging this gap by developing and experimentally evaluating implementations of backpressure protocols for wireless sensor networks on two commonly used software platforms, namely TinyOS and Contiki OS. They serve as a proof of concept that backpressure scheduling can be implemented in practice feasibly and efficiently. Our implementations are open source, which will allow other researchers and practitioners to view and modify the code to their needs. In this project, besides developing and implementing backpressure protocols that can be deployed on wireless sensor networks consisting of static nodes, we also developed novel extensions of backpressure techniques so that they can be used efficiently in relatively sparse networks of mobile nodes, and even to control the motion of robots to ferry messages between distant wireless devices.
