Abstract: In general, in one aspect, the disclosure describes an apparatus for selecting a queue from a plurality of queues. The apparatus includes a hierarchal queue occupancy device to indicate an occupancy status of the plurality of queues, a next queue selector to select a queue based on said hierarchal queue occupancy device and a most recently serviced queue, and a queue identification register to identify a most recently serviced queue.

Abstract: Congestion is controlled in an internetwork having at least two segments coupled by a router where at least one connection between communication devices passes through the router. Each connection is assumed to use a window-based flow control protocol between its source and destination. On receiving an acknowledgment from a connection in the router, where the acknowledgment contains a window size set by the destination, the router adaptively determines a second window size for the connection based on the router's average buffer occupancy and its instantaneous buffer occupancy. If the window size in the acknowledgment exceeds this second window size, the window size in the acknowledgment is overwritten to select the second window size. The router then forwards the acknowledgment to the source, thereby controlling the window size available to the source as a function of the congestion in the router.

Abstract: The resources required to support transmission of a traffic source, such as video or multimedia, over a packet-switched network are heavily dependent on the burstiness of the traffic source. The present invention provides methods for accurately and rapidly determining burstiness curves for traffic sources, such as elementary video and MPEG-2 transport streams. The described deterministic computation methods provide space and time efficiency in relation to the level of accuracy selected. The methods teach the exploitation of the piecewise linearity within the burstiness curves such that computations are only performed at points subject to slope change. The efficiency of the described methods suggest wide applicability to a range of traffic characterization situations, including QoS provisioning as well as real-time applications, such as TV distribution.

Abstract: A method for rate allocation within the individual switches of a communication network implementing a rate-based congestion control approach for best-effort traffic. The method enables a guaranteed minimum bandwidth to be allocated to each communication session or connection, in addition to fairly dividing the available bandwidth among the competing connections. The method also estimates the transmission rate of each connection, instead of relying on explicit indications provided by the connection, and uses this information in the computation of its fair share of bandwidth. The method also calculates the available bandwidth on the link where the bandwidth is allocated, by measuring the utilization of the link periodically. Finally, the method incorporates a mechanism to detect connections that remain idle and withdraw allocations from them so as to avoid under-utilization of the link bandwidth.

Abstract: A traffic scheduling system and method are disclosed for providing quality-of-service (QoS) guarantees in a packet-switched communication network where multiple traffic sessions share an outgoing communication link of a network switch in such a way that a certain minimum rate is reserved for each session sharing the link. The method can be applied to both general communication networks supporting variable packet sizes and Asynchronous Transfer Mode (ATM) networks with a fixed packet (cell) size. A packet arriving for transmission on an outgoing link of the switch is assigned a timestamp and placed in a priority queue in the order of increasing timestamps so that the packet with the smallest timestamp is placed at the head of the queue. The timestamp is computed based on a system potential maintained by the system, that tracks the progress of work in the system. The system potential is maintained at zero when no packets are queued at the outgoing link, and increases progressively as packets are transmitted.

Abstract: A traffic scheduling system and method for packet-switched communications networks where multiple sessions share an outgoing communications link. Prior to transmission on the outgoing link, packets are assigned a time-stamp and placed into a priority queue in order of ascending time-stamps so that the packet with the smallest time-stamp is placed at the head of the queue. The time-stamp assigned to a particular packet is calculated as the estimated time at which the last bit of the packet is transmitted in an ideal system, using a global variable called the system potential which tracks the progress of work in the scheduling system. The system potential is recalibrated periodically to avoid any long-term unfairness in service offered to the sessions sharing the outgoing link.

Abstract: A method is disclosed for rate allocation within the individual switches of a communications network implementing a rate-based congestion control approach for best-effort traffic. The methodology of the invention centers on a new rate allocation algorithm which performs its allocation functions independently of the number of connections sharing a network link and therefore performs an allocation in .THETA.(1) complexity. With that implementation simplicity, the algorithm is particularly advantageous for implementation in ATM switches carrying a large number of virtual channels. The algorithm operates on bandwidth information supplied from the source of a connection in special cells or packet headers, such as ATM Resource Management cells. By storing parameter values for other connections sharing a network link, the algorithm requires a constant number of simple computations for each request from a connection for a bandwidth allocation.

Abstract: A network and method for routing optical signals through wavelength-coding of routing tags belonging to the optical signals. The routing tag preferably consists of one or more header pulses S.sub.i which are chosen from among header wavelengths .lambda..sub.hi. Preferably, the header wavelengths .lambda..sub.hi are different from wavelengths used by the data. The optical signal also has reset tag containing preferably one reset pulse R preferably having a unique reset wavelength .lambda..sub.r. The optical data is contained between the routing tag and the reset tag. The network has a splitter for dividing the optical signal into two or more split optical signals, which are copies of the original optical signal. A wavelength differentiating element is positioned in the path of one of the split optical signals to differentiate and preferably spatially resolve the header wavelengths of the header pulses.