Resource and performance tradeoffs in delay-tolerant wireless networks
Resource and performance tradeoffs in delay-tolerant wireless networks
复制标题
DOI：
10.1145/1080139.1080144
复制
发表时间：
2005-08
期刊：
影响因子：
--
通讯作者：
T. Small;Z. Haas
中科院分区：
文献类型：
--
作者：
T. Small;Z. Haas
无线和移动的网络技术通常对资源的可用性施加严重的限制，导致通常使用的无线联网协议的性能差并且通常不能令人满意。例如，小型化网络节点的功率和存储器/存储约束降低了吞吐量并增加了网络延迟。通过各种方法和技术进步，研究人员试图以某种方式弥补这种硬件限制。然而，这并不总是必要的。有时，这种网络所需的性能不需要遵守性能关键型应用程序所需的服务级别。例如，对于传感器网络的一些应用，最小延迟不是关键因素，并且它可以被折衷为更有限的资源，例如能量或吞吐量。这种网络被称为延迟容忍网络。因此，为了减少能量消耗，这种传感器节点的传输范围将非常短，导致网络拓扑中网络节点的邻居的平均数量非常小。如果传感器节点是移动的，那么大多数时候一个节点没有邻居;只有很少的另一个节点迁移到它的邻居。这意味着经典的存储转发网络方法将无法很好地工作，因为在源和目的地之间几乎没有完整的路径。对于这种类型的网络环境已经提出了几种路由协议，一个示例是共享无线信息站模型（SWIM），其中分组通过从节点复制（而不是转发）到节点来通过网络传播，因为链路是零星创建的。目标是数据包的一个副本到达目的地。SWIM是一个非关键性能可以用来换取资源不足的例子，例如在能量、延迟、存储、容量和处理复杂性之间进行权衡。在本文中，我们研究了其中的一些权衡，暴露的方式，可以节省资源的性能水平上的妥协，以满足特定的网络技术的限制。
Wireless and mobile network technologies often impose severe limitations on the availability of resources, resulting in poor and often unsatisfactory performance of the commonly used wireless networking protocols. For instance, power and memory/storage constraints of miniaturized network nodes reduce the throughput capacity and increase the network latency. Through various approaches and technological advances, researchers attempt to somehow compensate for such hardware limitations. However, this is not always necessary. Sometimes, the required performance of such networks does not need to adhere to the level of services that would be required for performance-critical applications. For example, for some applications of sensor networks, minimal latency is not a critical factor and it could be traded off for a more limited resource, such as energy or throughput. Such networks are termed delay-tolerant networks. Thus, to reduce the energy expenditure, transmission range of such sensor nodes would be quite short, leading to network topologies in which the average number of neighbors of the network nodes is very small. If the sensor nodes are mobile, then most of the time a node has no neighbors; only infrequently another node migrates into its neighborhood. This means that the classical networking approach of store-and-forward would not work well, as there is nearly never an intact path between a source and a destination. Several routing protocols have been proposed for this type of networking environment, one example is the Shared Wireless Infostation Model (SWIM), where a packet propagates through the network by being copied (rather than forwarded) from a node to a node, as links are sporadically created. The goal is that one of the copies of the packet reaches the destination. SWIM is an example of the way that non-critical performance could be traded off for insufficient resources, such as the tradeoffs between energy, delay, storage, capacity, and processing complexity. In this paper, we examine some of these tradeoffs, exposing the ways in which resources could be saved by compromising on the level of performance, as to satisfy the particular limitations of network technologies.