EAGER: Towards an Open Source Backpressure Protocol Stack for Wireless Networks

虽然理论上已经证明了利用队列回压的随机网络优化技术可以产生高性能的跨层协议，但将其转化为实践的优先工作是有限的。在这个探索性研究项目中，我们在背压收集协议(BCP)的基础上，开发并演示了一种适用于无线传感器网络的新型跨层背压协议栈。这种高度灵活的跨层无线网络方法的可能收益是巨大的，包括即使在存在移动性的情况下也提高了吞吐量，以减少的损失处理突发的外部干扰事件，以及降低协议实施的复杂性。我们结合了MAC层备份优先级排序、传输层实用程序优化、处理各种网络动态的机制以及与异步低功耗休眠计划的互操作性。在该项目中开发的协议实现代码将作为开源公开发布。预计该项目的成功完成将为随机优化理论驱动的协议在实践中取得成功提供重要的概念证明。

While it has been shown theoretically that stochastic networkoptimization techniques utilizing queue backpressure can result inhigh performance cross-layer protocols, there has been limited priorwork on translating them into practice. In this exploratory researchproject, we develop and demonstrate a novel cross-layer backpressureprotocol stack for wireless sensor networks, building on ourimplementation of the backpressure collection protocol (BCP). Thepossible gains of this highly agile approach to cross-layer wirelessnetworking are substantial and include increased throughput even inthe presence of mobility, handling of bursty external interferenceevents with reduced losses, and reduced protocol implementationcomplexity. We incorporate techniques for MAC-layer backoffprioritization, transport layer utility optimization, mechanisms tohandle various kinds of network dynamics, and interoperability withasynchronous low power sleep schedules. Protocol implementation codedeveloped in the project will be released publicly as open-source. Itis expected that the successful completion of this project willprovide an important proof-of-concept that stochastic optimizationtheory-driven protocols can be successful in practice.