摘要： Dealing with video streams, transmitting various pictures or even exchanging ?les with a mobile device is common practice in today’s world. Such services produce a large amount of data which has to be transferred over a network. For the current Internet infrastructure, these large amounts of data do not pose a problem. MANETs in contrast struggle with high network load due to their speci?c features, namely high mobility and decentralized infrastructure. However, in some use cases those features are advantageous for civil as well as military applications. Public demonstrations are likely an example for a civil use case, while in the military world soldiers transmit video feeds of observations or share pictures of suspected persons with their command and fellow troops. Transmissions with high data rates provoke a high network load with over-stressed routes. Those situations often result in congestion as nodes that act as forwarding devices are usually part of several routes. Consequently, higher delays and packet loss rates as well as Denial of Service situations become an issue. To encounter decreasing performance, we propose an extension to a path discovery algorithm that we call AODV-CBR, which selects a path between a source and a destination depending on the node’s available capacity and the required resources for the requested transmission. The objective is to increase the overall performance by avoiding nodes that are already overloaded. We show that our extension yields better results in MANETs with high utilization.