Abstract: A method for operating a calculation system including a neural network, in particular a convolutional neural network, the calculation system including a processing unit for the sequential calculation of the neural network and a memory external thereto for buffering intermediate results of the calculations in the processing unit, including: incrementally calculating data sections, which each represent a group of intermediate results, with the aid of a neural network; lossy compression of one or multiple of the data sections to obtain compressed intermediate results; and transmitting the compressed intermediate results to the external memory.

Abstract: A hardware architecture for an artificial neural network ANN. The ANN includes a consecutive series made up of an input layer, multiple processing layers, and an output layer. Each layer maps a set of input variables onto a set of output variables, and output variables of the input layer and of each processing layer are input variables of the particular layer that follows in the series. The hardware architecture includes a plurality of processing units. The implementation of each layer is split among at least two of the processing units, and at least one resettable switch-off device is provided via which at least one processing unit is selectively deactivatable, independently of the input variables supplied to it, in such a way that at least one further processing unit remains activated in all layers whose implementation is contributed to by this processing unit.

Abstract: A method for verifying a calculation of a neuron value of multiple neurons of a neural network, including: carrying out or triggering a calculation of neuron functions of the multiple neurons, in each case to obtain a neuron value, the neuron functions being determined by individual weightings for each neuron input; calculating a first comparison value as the sum of the neuron values of the multiple neurons; carrying out or triggering a control calculation with one or multiple control neuron functions and with all neuron inputs of the multiple neurons, to obtain a second comparison value as a function of the neuron inputs of the multiple neurons and of the sum of the weightings of the multiple neurons assigned to the respective neuron input; and recognizing an error as a function of the first comparison value and of the second comparison value.

Abstract: A hardware architecture for an artificial neural network ANN. The ANN includes a consecutive series made up of an input layer, multiple processing layers, and an output layer. Each layer maps a set of input variables onto a set of output variables, and output variables of the input layer and of each processing layer are input variables of the particular layer that follows in the series. The hardware architecture includes a plurality of processing units. The implementation of each layer is split among at least two of the processing units, and at least one resettable switch-off device is provided via which at least one processing unit is selectively deactivatable, independently of the input variables supplied to it, in such a way that at least one further processing unit remains activated in all layers whose implementation is contributed to by this processing unit.

Abstract: A method for operating a calculation system including a neural network, in particular a convolutional neural network, the calculation system including a processing unit for the sequential calculation of the neural network and a memory external thereto for buffering intermediate results of the calculations in the processing unit, including: incrementally calculating data sections, which each represent a group of intermediate results, with the aid of a neural network; lossy compression of one or multiple of the data sections to obtain compressed intermediate results; and transmitting the compressed intermediate results to the external memory.

Abstract: A method for verifying a calculation of a neuron value of multiple neurons of a neural network, including: carrying out or triggering a calculation of neuron functions of the multiple neurons, in each case to obtain a neuron value, the neuron functions being determined by individual weightings for each neuron input; calculating a first comparison value as the sum of the neuron values of the multiple neurons; carrying out or triggering a control calculation with one or multiple control neuron functions and with all neuron inputs of the multiple neurons, to obtain a second comparison value as a function of the neuron inputs of the multiple neurons and of the sum of the weightings of the multiple neurons assigned to the respective neuron input; and recognizing an error as a function of the first comparison value and of the second comparison value.

Abstract: In a having a plurality of levels, each level has at least one information node; a number of information nodes of a lower k+1-th level are assigned to an information node of a k-th level configured thereabove and linked to this one information node; one information node of the lower k+1-th level is designed to inform the information node of the k-th level configured thereabove, which pieces of information the information node of the k+1-th level needs, and which pieces of information the information node of the k+1-th level is making available.

Abstract: In a method for processing messages with the aid of at least one Ethernet controller having a message store, the at least one Ethernet controller checks a received message for the presence of at least one property, and the received message is stored in the message store if this at least one property applies to the message.

Abstract: A method for acoustically sensing an area is described. An acoustic transmit pulse is transmitted into the area by an acoustic transducer and a received signal is acquired with the aid of the transducer, in order to receive the transmit pulse reflected back from the area. Within a post-pulse oscillation time of the transducer, which directly follows the emission of the transmit pulse, the actual phase response of the received signal is determined. In the event the actual phase response deviates from a predefined setpoint phase response, an object is detected within the area. Also described is a device for acoustically sensing an area, which is designed to carry out the method.

Abstract: A method for acoustically sensing an area is described. An acoustic transmit pulse is transmitted into the area by an acoustic transducer and a received signal is acquired with the aid of the transducer, in order to receive the transmit pulse reflected back from the area. Within a post-pulse oscillation time of the transducer, which directly follows the emission of the transmit pulse, the actual phase response of the received signal is determined. In the event the actual phase response deviates from a predefined setpoint phase response, an object is detected within the area. Also described is a device for acoustically sensing an area, which is designed to carry out the method.

Abstract: In a having a plurality of levels, each level has at least one information node; a number of information nodes of a lower k+1-th level are assigned to an information node of a k-th level configured thereabove and linked to this one information node; one information node of the lower k+1-th level is designed to inform the information node of the k-th level configured thereabove, which pieces of information the information node of the k+1-th level needs, and which pieces of information the information node of the k+1-th level is making available.

Abstract: A method for synchronizing two communication networks of an electronic data-processing system, each of the networks including one or more respective nodes, may include establishing for each of the networks a respective time schedule that establishes at least one respective time slot for a respective synchronization message. The synchronization message time slots are established to coincide. The synchronization message of the first communication network is generated by one of the nodes of the first communication network. The synchronization message of the second communication network is generated as a function of the synchronization message of the first communication network.

Abstract: In a method for processing messages with the aid of at least one Ethernet controller having a message store, the at least one Ethernet controller checks a received message for the presence of at least one property, and the received message is stored in the message store if this at least one property applies to the message.

Abstract: A method for changing an operating mode of at least one device, multiple compare registers, each having a bit pattern, being associated with the at least one device, is provided in which a data packet transmitted via a network is received by the at least one device, a content of the data packet being compared to the bit patterns of the associated compare registers, and in which a change in the operating mode of the at least one device takes place when the at least one device detects conformity of a bit pattern of the associated compare registers with the content of the data packet.

Abstract: A method for providing a communication for at least one device that is provided for a motor vehicle and is linked via a network of the motor vehicle to at least one further device, where data between the at least one device and the at least one further device are exchanged via an Internet Protocol. Also a device and a network are provided.

Abstract: A method for transferring data between a data transmission system and a processor of a participant in the data transmission system. All components of the data transmission system are synchronized to a common global time base. The operating system time base of the participant processor is synchronized to the global time base of the data transmission system at least prior to a data transfer, and, to this end, a synchronization clock pulse that is synchronous to the global time base of the data transmission system is provided for synchronizing the operating system time base. This synchronization clock pulse may be provided by a hardware and/or software arrangement. In a data transmission system for implementing the method, the synchronization clock pulse is provided by suitable arrangement(s) of the communications controller and applied to the host processor via a synchronization line.

Abstract: A communications system for a motor vehicle, including a plurality of electrical components, a data bus structure to which the components are connected in order to transmit information among the components, and a power line structure to which the components are connected in order to be supplied with power. The information is transmitted in successive cycles over the data bus structure, each cycle including at least one time window for transmitting information at specific points in time and at least one event window for transmitting information in response to specific events. The communications system includes an arrangement for redundantly transmitting information which merely transmits the information transmitted in the at least one time window over the data bus structure at least partially over the power line structure as well.

Abstract: A device for simulating a data transmission via a data transmission network, especially in a motor vehicle, for evaluating the transmission quality via the network as a function of an interference environment and of a channel characteristic of the network.

Abstract: A method for synchronizing two communication networks of an electronic data-processing system. One or a plurality of nodes is connected to the two communication networks in each case. Each of the two communication networks has a schedule, which specifies at least one time slot for a synchronization message. One of the nodes of the first communication network, and one of the nodes of the second communication network are linked to one another by a shared arithmetic unit. A deviation between the time slots for the synchronization messages on the two communication networks is determined as a function of the schedules of the two communication networks. The deviation is used to determine correction values, which are forwarded to at least one other node. The occurrence of the time slots for the synchronization messages is modified by the other node as a function of the correction values.

Abstract: A method for diagnosing a synchronization of two communications networks of an electronic data-processing system, each communications network including at least one node and having a respective schedule that specifies at least one respective time slot for a respective synchronization message, may include detecting, by each of the two communications networks, an individual time-based variable, comparing the two detected time-based variables to each other, and determining a quality of the synchronization of the two communications networks based on the comparison.