Achieving Two Orders of Magnitude Reduction in Cellular Network Energy Inefficiency

我们的目标是开发一套算法，以提高蜂窝无线网络的能源效率。这些网络最初设计时没有充分考虑它们的效率，据计算效率不到10%。随着蜂窝通信量以指数速度增长，专家们相信，在一个小区内，通信量将是今天的1000倍。在提出的研究中，将开发一些算法来提高能源效率。例如，新的调制技术、新的链路自适应算法或蜂窝网络的算法，以适应不断变化的业务条件，使得业务量小的小区仅使用所需的能量，或者小区关闭，其覆盖范围由其相邻基站接管。我们分析了多种不同技术的组成，得出结论：将蜂窝网络基站的能耗降低到现有水平的1.5-2%是可能的，这一建议的智力价值在于重新设计蜂窝网络，并将其能量效率提高几个数量级。这将通过一组新的工具、方法和算法来实现，这些工具、方法和算法采用了一组不同的权衡，如部署、带宽和可能的延迟。更广泛的影响：由于该项目依赖于了解用户的位置，因此该项目将启用基于位置的服务。这些都是高度期望的，但不能在今天的网络中实现。这个项目将建立在我们现有的研究，原型的努力，和现有的测试平台。

The goal is to develop a set of algorithms to increase the energyefficiency of cellular wireless networks. These networks wereoriginally designed without paying sufficient attention to theirefficiency, calculated to be less than 10%. With cellular trafficincreasing at exponential rates, the experts believe that in onedecade the traffic will be 1000 times today's traffic. In the proposedresearch, a number of algorithms will be developed to improve energyefficiency. Examples are new modulation techniques, new link adaptationalgorithms, or algorithms for the cellular network to adapt tochanging traffic conditions such that cells with small traffic useonly as much energy as needed, or cells shut down, their coverage tobe taken over by their neighboring base stations. We analyze thecontributions of a number of different techniques, and reach theconclusion that it is possible to reduce the energy consumption in thecellular network base stations to the 1.5-2% of its current levels.The intellectual merit of this proposal is in redesigning the cellularnetwork and improving its energy efficiency by several orders ofmagnitude. This will be achieved by a new set of tools, methods, andalgorithms employing a different set of tradeoffs such as deployment,bandwidth, and possibly delay.Broader Impact: Because the project relies on knowing a user'slocation, the project will enable location-based services. These arehighly desired but cannot be implemented in today's networks. Thisproject will build on our existing research, prototyping efforts, andan existing testbed.