Choke packets are measures against network overloads. Since the data traffic
between the node computers is subject to strong fluctuations, the technique of choke pockets provides a remedy. In distributed data networks, the unsteady data traffic between the node computers, the Interface Message Processors (IMP), can lead to bottlenecks and data congestion if the buffer storage capacity of the individual node computers cannot cope with the traffic that arises. In the most extreme case, it can lead to a deadlock. Since resolving a deadlock is difficult, attempts are made to prevent it from occurring in the first place. There are several methods to do this, but most of them are effective even with low traffic and thus reduce the data throughput. In contrast to these, the choke packet method only takes effect from a certain upper traffic limit. Technically, this is solved in such a way that an IMP processor that detects the congestion sends a choke packet back to the host, which then reduces its transmission speed. After waiting a certain time, the traffic is checked again and leads either to a further reduction or to an increase in the transmission speed.