Network delays strongly impact time-sensitive distributed systems that communicate over the network. Today, many distributed systems are still using wired network connections. However, with larger installations, wired connections become less feasible, and the need for wireless communication over WIFI or 5G/6G systems becomes greater. These communication methods bring new challenges as messages experience higher network delay, jitter, and packet errors. Therefore, it is necessary to test these systems and their performance reliably. For these tests, network emulation is a promising method. With network emulation, these systems can be tested on wired hardware that behaves like a wireless one. While there are tools available that can emulate networks, there is a lack of flexibility in defining and following stochastic delay distributions, for example, to emulate the unpredictable behavior of wireless connections. This bachelor thesis presents the design and implementation of a tool that specializes in emulating network delays, which can be adapted flexibly and dynamically. This tool is implemented as a Linux Queueing Discipline, which allows it to be used on all Linux-based systems. To achieve the mentioned flexibility, we present a design that separates the generation of new delays from the enforcement. While the Queueing Discipline runs in the restricted kernel space and delays packets, the delays are calculated outside Kernel Space in User Space and then sent as a message to the Queueing Discipline. This separation allows developers to customize how delays are generated more easily or even write their own applications that interact with the Queueing Discipline. To prove the viability of our tool and explore its limitations, we tested and evaluated it. This evaluation showed that, within the limits of software network emulation, our tool can accurately and reliably delay packages both statically and following stochastic distributions.