Abstract:
According to the present invention a method and a device is provided for combining at least two data signals having a first data rate into a single data stream having a second data rate being higher than the first data rate for transmission on a shared medium or vice versa, said device comprises at least one port for receiving said at least two data signals and a port addressing unit for extracting data from the data signals received by said ports, wherein said port addressing unit is configured to place the extracted data at predetermined positions in said single data stream to be transmitted on said shared medium and at least one control data insertion unit is provided for placing control data in said single data stream. Thus, a method and device proposed implementing a multiplexing structure in which the data from M parallel low-speed channels are multiplexed onto a data bus operating with M times the data rate of the M low-speed channels. A multiple-stage process where, e.g., 28 T-1 (DS-1) channels are re-mapped into one T-3 (DS-3) channel and 3 T-3 (DS-3) channels are then mapped into an STM-1 frame is advantageously avoided.
Abstract:
PROBLEM TO BE SOLVED: To obtain a new concept for a variable-length packet process at a very high data transmission speed by providing a packet processor with a distributor which distributes packets to several parallel identical process paths, supplying the packets to the respective process paths, and enabling the packet information of only one packet to be processed for the packet at an arbitrary moment. SOLUTION: A packet stream of variable-length packets which arrive and are received and arrayed is supplied to parallel process paths of the packet processor 40, and demultiplexed into N substreams of variable-length packets or virtual process paths. Demultiplexing is carried out without altering the packets contents. The number N depends upon the data transmission speed or data throughput, the number of priority levels, the number of cycles that the following processes require, the length of the packets, and other parameters. As the data throughput is higher and higher, more and more substreams or virtual process paths are needed.
Abstract:
PURPOSE: To realize femtosecond-order to picosecond-order time resolution and atomic-level space resolution by holding a tunnel chip at a tunnel distance above a sample and applying a tunnel voltage across the gap between the chip and sample, and then, supplying at least one pulsed laser. CONSTITUTION: A tunnel chip 1 is placed at a tunnel distance (namely, at about 1nm) above the surface 2 of a sample 3 to be investigated. A pulsed laser beam (pump beam) 4 is projected upon the surface 2 of the sample 3 from the chip 1 which is separated from the surface 2 by a distance by performing stroboscopic scanning. At every exciting pulse, thermoelectrons are generated at the place of collision and move in the sample 3. The dynamic response of a system including the material of the sample 3 and the electrons moving in the sample 3 is investigated by stroboscopic extraction. When another laser beam 5 is projected upon the surface 2 of the sample 3 near the tunnel chip 1 synchronously to the pump beam 4, the moving thermoelectrons are excited to a higher energy level.
Abstract:
The surface of the IC substrate and the pads on the surface which are connected to the conducting paths to be tested are covered with three layers. The first (13) is of passivating oxide, a second layer of electroluminescent material (15) and a third layer (17) of a conductive material which is so thin as to be transparent to the passage of photons. The focused laser beam is directed to a pad or pads to generate charges by photon-assisted tunnelling of electrons from each selected pad. Consequently, a specific voltage level is induced in each selectedpad and in each non-selected pad which is connected electrically with a selected pad. The whole substrate surface is flooded with a laser beam to excite voltage-dependent photon-assisted electron-tunnelling from pads, the tunnelling intensity being dependent on the potential of each pad. The excitation pattern is detected in the electroluminescent layer, viewed by a TV camera (21).
Abstract:
Verfahren zum eingabe- und belegungsabhängigen Erzeugen von Meldenachrichten, die Überlastung eines Puffers anzeigend, aufweisend:- Erhöhen (502) eines Zählers der letzten Eingaben, der einem Datenstrom von einer von mehreren Quellen zugehörig ist, in Reaktion auf das Empfangen eines Datenpakets von der einen der mehreren Quellen durch eine Datenverarbeitungse i n he it;- Speichern des Datenpakets in dem Puffer, der mehrere Speichereinheiten aufweist, wobei der Puffer die Datenpakete zugehörig zu den Datenströmen von den mehreren Quellen abspeichert;- Feststellen (504), ob der Puffer überlastet ist;- in Reaktion auf das Feststellen, dass der Puffer überlastet ist, willkürliches Auswählen (508) einer belegten Einheit des Puffers und Ermitteln (510) einer Quelle eines Pakets, das in der belegten Einheit des Puffers gespeichert ist;- Erzeugen (506) einer Überlastungs-Meldenachricht;- Senden (514) der Überlastungs-Meldenachricht zu der Quelle des Pakets, das in der belegten Einheit des Puffers gespeichert ist, wenn der Zähler der letzten Eingaben einen Schwellenwert übersteigt, und Verringern des Zählers der letzten Eingaben, der dem Paket, das in der belegten Einheit des Puffers gespeichert ist, zugehörig ist; und- Verwerfen (516) der Überlastungs-Meldenachricht, wenn der Zähler der letzten Eingaben des entsprechenden Stroms gleich einem Schwellenwert oder kleiner als dieser ist.
Abstract:
Ein Verfahren zum Hybrid-Management von eingaben- und belegungsabhängigen Überlastungen enthält das Erhöhen eines Zählers der letzten Eingaben, der einem Datenstrom von einer oder mehreren Quellen zugehörig ist, in Reaktion auf das Empfangen eines Datenpakets von einer der Quellen und Speichern des Datenpakets in einem Puffer, der mehrere Speichereinheiten enthält. Das Verfahren enthält das Feststellen, ob ein Puffer überlastet ist, das willkürliche Auswählen einer belegten Einheit des Puffers und das Ermitteln der Quelle des in der belegten Einheit des Puffers gespeicherten Pakets, das Erzeugen einer Überlastungs-Meldenachricht, das Senden der Überlastungs-Meldenachricht zu der Quelle des in der belegten Einheit des Puffers gespeicherten Pakets, wenn der Zähler der letzten Eingaben einen Schwellenwert übersteigt, und das Vermindern des Zählers der letzten Eingaben, der der Quelle des in der belegten Einheit des Puffers gespeicherten Pakets zugehörig ist, und das Verwerfen der Überlastungs-Meldenachricht, wenn der Zähler der letzten Eingaben einen Wert null aufweist.
Abstract:
The present invention concerns a packet-processing apparatus and a packet switch adapter for the processing of variable-length packets comprising packet data and packet information. The adapter and/or the apparatus comprises a distributor for distributing the packets to several parallel, identical processing paths, each comprising at least one processing unit, whereto the packets are fed and which is able to process only for one of the packets its packet information at any moment in time. The feeding is interruptable for feeding a different of the packets to another of the processing paths. Furthermore, a packet-processing method for distributing received variable-length packets to several parallel, identical processing paths is disclosed.
Abstract:
A method and a device for combining at least two data signals having a first data rate into a single data stream having a second data rate higher than the first data rate for transmission on a shared medium or vice versa. The device has at least one port for receiving at least two data signals and a port addressing unit for extracting data from the data signals received by the ports. The port addressing unit is configured to place the extracted data at predetermined positions in the single data stream to be transmitted on the shared medium and at least one control data insertion unit is provided for placing control data in the single data stream.